87Sr/86Sr sourcing of ponderosa pine used in Anasazi great house construction at Chaco Canyon, New Mexico
Introduction
In the southwestern U.S.A., strontium (Sr) isotopic analyses of archeological bone and plant tissue are being used increasingly to decipher prehistoric migration patterns, residential shifts in population [29], and long-distance procurement of timber [17] and foodstuffs [4]. Strontium exists in predictable and measurable quantities in rocks and sediment, as well as in plant matter, bones, and teeth recoverable from archeological sites. The ratio of 87Sr to 86Sr isotopes is maintained from soil water, through plants, and up the trophic chain, making it ideal for provenance studies [8].
Architectural timber common in southwestern ruins is particularly well suited for 87Sr/86Sr analysis. In most cases, the abundant and well-preserved logs have been identified to species, their architectural function and placement have been established, and their cutting dates have been determined to the exact year, depending on the condition of the timber. This dated and well-provenienced wood has been archived and is readily accessible for geochemical studies. Probably the richest such archive is associated with more than a century of excavations at Chaco Canyon, New Mexico [14], [23], [35], [36]. Construction of the great houses at Chaco Canyon required over 200,000 highland conifers, including spruce (Picea spp.), fir (Abies spp.), Douglas-fir (Pseudotsuga menziesii), and ponderosa pine (Pinus ponderosa) [14].
Paradoxically, Chaco Canyon is set in desertscrub-grassland in the middle of the San Juan Basin and removed at least 75 km from the nearest potential timber sources (Fig. 1, Fig. 2). The packrat midden record indicates that highland conifers were either absent locally, as in the case of spruce and fir, or too rare, as in the case of Douglas-fir and ponderosa pine, to have provisioned the Anasazi (prehistoric Puebloans) builders of the Chacoan great houses [6]. Most of this wood had to be felled, processed, harvested and hauled from one or more distant highland areas [5]. The distances and directions of sources for timber procurement are one measure of economic, political and social relationships across the San Juan Basin. Location of the source stands can be narrowed somewhat by species (i.e., spruce and fir grow only on some mountaintops), but more specific locations require application of geochemical tracers.
Diverse rock types with diverse chemical signatures characterize the San Juan Basin and surrounding highlands, particularly the mountaintops (Fig. 3). The Chuska Mountains west of Chaco Canyon are a north–south trending range capped with Tertiary sandstone, locally overlain by Tertiary basalt. The San Mateo Mountains (also referred to as Mt. Taylor) to the south represent a succession of lava and ash flows erupted 2–4 million years ago. The San Pedro/Nacimiento Mountains to the east are composed of highly faulted blocks of Precambrian crystalline rock, of diverse composition, overlain in places by Paleozoic sedimentary rocks. The La Plata Mountains to the north are composed of a Cretaceous granitic batholith capped by Tertiary sedimentary basin deposits. Sandstones and shales of various ages and composition flank each of the four ranges.
A preliminary study by Hatch [20] used inductively coupled plasma-atomic emission spectrometry to characterize 77 modern ponderosa pine from Chimney Rock and 52 architectural samples from Chimney Rock, Aztec, and Salmon Ruins. Hatch [20] found higher elemental concentrations in the prehistoric samples suggesting elemental contamination with burial. Hatch [20] recognized that differences in elemental concentrations between different wood species and different wood compositions (sapwood versus heartwood and living versus dead wood) needed to be constrained in her results before secondary contamination could be quantified in the architectural wood samples. Also, the authors did not address the effects of pretreatment and drying methods on wood elemental concentrations.
Durand et al. [16] conducted a study of 28 major and trace elements in wood of 62 ponderosa pine and Douglas-fir trees growing on three bedrock types (basalt, sandstone and shale) in the San Juan Basin. The authors recognized heartwood elemental concentrations to be more reliable than sapwood for trace element studies of wood. Durand et al. [16] suggested that barium concentrations in architectural wood tend to vary with relative mobility of barium for different lithologies, with high mobility in sandstones and low in basalts. Low barium concentrations in 13 ponderosa pine beams from Chacoan great houses suggested a source in either San Mateo or the Chuska Mountains, both capped at least in part by Neogene basalts (Fig. 3). The authors admitted that more research needed to be done to conclusively source the architectural timbers using elemental concentrations due to the lateral, differential movement of most elements in sapwood [3], [9], [15], [26], [31]. The study shows, however, that less mobile elements such as barium may be useful in sourcing studies.
English et al. [17] narrowed the search for Chacoan wood sources by reasoning that trees assimilate Sr from local soils and atmospheric dust and incorporate it into wood without fractionating the isotopes 87Sr and 86Sr. The 87Sr/86Sr ratio in the wood should be independent of the vital effects controlling fluctuations in elemental concentrations and solely reflect the ratio in local soil water, which is some combination of local weathering products and atmospheric dust [18]. English et al. [17] also narrowed the search by focusing on spruce and fir, which are restricted today to only a few widely separated and geologically distinct mountaintops.
English et al. [17] analyzed 87Sr/86Sr ratios in wood of 73 living spruce and fir trees from three mountain ranges (Chuska, San Mateo, and San Pedro/Nacimiento Mountains) within 100 km of Chaco Canyon. Despite the presence of a “Great North Road” [34], they excluded sampling the La Plata and San Juan Mountains to the north because those spruce–fir forests were most distant (>150 km) and least accessible, requiring transport of large beams across deep canyons and flowing rivers. Trees from the Chuska, San Mateo and San Pedro/Nacimiento Mountains each exhibited distinguishable 87Sr/86Sr ratios.
English et al. [17] also measured 52 architectural specimens of spruce and fir from several great houses in Chaco Canyon (Pueblo Bonito, Chetro Ketl, Pueblo del Arroyo, Wijiji, Hungo Pavi and Una Vida). Two-thirds of the architectural specimens could be traced to the Chuska Mountains and one-third to the San Mateo Mountains. The two fingerprinted sources were both logged as early as A.D. 974, suggesting that selection of timber sources was driven more by socioeconomic ties with communities to the west and south than by timber depletion with distance and time. Conversely, the absence of logs from the San Pedro/Nacimiento Mountains is consistent with the lack of roads and paucity of outlying Chacoan communities to the east. The importance of the Chuska Mountains for resource procurement was recently confirmed by 87Sr/86Sr analysis of prehistoric corn from Chaco Canyon [4]. This study, however, points to the San Juan floodplain as an additional source of corn.
Here we extend the Chaco Sr isotope work to ponderosa pine. Ponderosa pine is more widely distributed than spruce and fir and stands occur within 100 km of Chaco Canyon in all directions. Therefore, we expanded the original geographic scope of English et al. [17] to encompass sampling of live trees in the La Plata and San Juan Mountains, and previously unsampled mesas (e.g., Hosta Butte and Lobo, Cuba and Sisnathyel Mesas; Fig. 1) and mountain flanks throughout the San Juan Basin. Ponderosa pines grow at lower elevations and in closer proximity to Chaco Canyon than spruce and fir, so ponderosa might have been used earlier and from a wider variety of geologic settings and elevations. Moreover, ponderosa pine represents more than half of the estimated 200,000 timbers employed in great house construction [14], and may further constrain exploitation of different timber sources. For example, the abundance of trees in size classes desirable for Chacoan architecture was restricted by substrate, elevation, slope aspect and grade, which mediate interspecific competition and rates of ecological disturbance, tree mortality, and recruitment. Smaller secondary beams may have been acquired from fast-growing stands in the moister highlands, while larger primary beams could have come from more open and fire-prone parklands on mesas and foothills characterized by a few mature trees and a scarcity of saplings.
Section snippets
Sr isotopes
Sr is an alkaline earth element that has the same charge and a similar atomic radius to calcium and thus commonly substitutes for Ca in chemical reactions. Sr has four stable, naturally occurring isotopes with different abundance: 84Sr (0.6%), 86Sr (9.9%), 87Sr (7.0%), and 88Sr (82.6%). 87Sr/86Sr ratios are commonly reported in geochemical investigations. The 87Sr/86Sr ratio of minerals is mostly a function of the initial 87Rb/86Sr ratio and the age of the rock. 87Sr is radiogenic, derived from
Sampling of living trees to establish baseline for provenance studies
In the summer of 2002, we sampled more than 200 living trees from 20 localities in the La Plata Mountains, San Juan Mountains, Sisnathyel Mesa, Cuba Mesa, San Pedro/Nacimiento Mountains, San Mateo Mountains, Lobo Mesa, Hosta Butte and Chuska Mountains. All living trees were cored at breast height with a ¼-inch increment borer. No lubricants were used to avoid chemical contamination. Individual trees were chosen from stands growing on a diverse array of bedrock types. Five individual tree cores
87Sr/86Sr variations in living trees from potential source areas
Sixty-two living ponderosa, and three subalpine fir (Abies lasiocarpa, Abies concolor), two aspen (Populus tremuloides), and two Engelmann spruce (Picea engelmannii) were analyzed to characterize 87Sr/86Sr variations across 21 potential source stands (Fig. 1). The firs and spruce were sampled from the La Plata and San Juan Mountains, which were previously excluded from the range of sites sampled by English et al. [17]. 87Sr/86Sr values for the living trees ranged from 0.7055 to 0.7192 (Table 1;
Testing for diagenesis in the architectural samples
When analyzing the chemistry of architectural specimens, diagenetic contamination (especially secondary salts) is always a concern, particularly in porous materials such as wood and bone. To test for diagenetic additions in the architectural wood, we (1) plotted Sr concentrations against 87Sr/86Sr values, and (2) compared the Sr concentrations in the architectural wood to those in the living specimens, with the assumption that pristine archaeological samples should yield comparable
Conclusions
Strontium isotopic analyses indicate that the architectural ponderosa used in three of the Chacoan great houses likely came mostly from the Chuska Mountains to the west and the La Plata and San Juan Mountains to the north and possibly Hosta Butte to the south. A previous study showed that spruce and fir in six of the great houses came mostly from the Chuska Mountains, with some contributions from the San Mateo Mountains to the south. Comparison between ponderosa and spruce/fir architectural
Acknowledgments
We thank Taft Blackhorse for collection of samples from Hosta Butte, Nathan English and Clark Isachsen for advice on lab protocols, and Dick Warren for help with retrieving archived samples. We also thank Bruce Allen for discussions on the nuances of New Mexico geology, and Tom Windes and Camille Holmgren, and four anonymous reviewers for comments on the manuscript. This work was funded by the Chaco Sites Protection Program of the Navajo Nation, the U.S. Geological Survey, and an NSF-IGERT
References (37)
- et al.
Strontium isotopes in trees as an indicator of calcium availability
Catena
(1990) - et al.
Impact of forest tree species on feldspar weathering rates
Geoderma
(2000) - et al.
Study of the transfer of radionuclides in trees at a forest site
Sci. Total Environ.
(1995) - et al.
Strontium isotopes as tracers of ecosystem processes: Theory and methods
Geoderma
(1998) - et al.
Trees, chemistry, and prehistory in the American Southwest
J. Archaeol. Sci.
(1999) - et al.
Spatial variation of strontium isotopes (87Sr/86Sr) in the Maya region: a tool for tracking ancient human migration
J. Archaeol. Sci.
(2004) The biological system of the elements (BSE) for terrestrial plants (glycophytes)
Sci. Total Environ.
(1994)- et al.
Isotopic evidence for eolian recycling of pedogenic carbonate and variations in carbonate dust sources throughout the southwest United States
Geochim. Cosmochim. Acta
(2000) - et al.
Residential mobility in the prehistoric southwest United States: A preliminary study using strontium isotope analysis
J. Archaeol. Sci.
(1994) - et al.
Effects of rainfall on weathering rate, base cation provenance, and Sr isotope composition of Hawaiian soils
Geochim. Cosmochim. Acta
(2001)
Tracing spatial and temporal variations in the sources of calcium in pedogenic carbonates in a semi-arid environment
Geoderma
Ancient maize from Chacoan great houses: Where was it grown?
Proc. Natl Acad. Sci.
Prehistoric long distance transport of construction beams, Chaco Canyon, New Mexico
Am. Antiquity
Fossil packrat middens from Chaco Canyon, New Mexico: cultural and ecological significance
Mycorrhizal weathering of apatite as an important calcium source in base-poor forest ecosystems
Nature
Changes in Sr/Ca, Ba/Ca, and 87Sr/86Sr ratios between trophic levels in two forest ecosystems in the northeastern U.S.A.
Biogeochemistry
Radial trends in cation ratios in tree rings as indicators of the impact of atmospheric deposition on forests
Can. J. For. Res.
Changing sources on nutrients during four million years of ecosystem development
Nature
Cited by (84)
Were timbers transported to Chaco using tumplines? A feasibility study
2023, Journal of Archaeological Science: ReportsCombining conventional tree-ring measurements with wood anatomy and strontium isotope analyses enables dendroprovenancing at the local scale
2023, Science of the Total EnvironmentDropped in the Ocean – <sup>87</sup>Sr/<sup>86</sup>Sr as a provenance tool for ice-rafted Arctic driftwood
2022, Palaeogeography, Palaeoclimatology, PalaeoecologyCitation Excerpt :The element strontium is present within all rocks, with relative abundance of the radiogenic isotope (expressed as 87Sr/86Sr) specific to certain rock types (Faure and Powell, 1972; Guinn, 1978; English et al., 2001). Therefore, lithologies can be characterized by a defined 87Sr/86Sr ratio (English et al., 2001; Reynolds et al., 2005). Strontium is not as prone to geological or biological fractionation as lighter elements, meaning that soils and the plants growing on them will uptake bedrock values without fractionation (Capo et al., 1998; English et al., 2001; Dijkstra et al., 2003; Bentley, 2006).
Defining isotopic signatures of potential procurement sources: A case study in the Mesa Verde region of the US Southwest
2022, Journal of Archaeological Science: Reports