Jeffrey S. Dean
Dr. Dean is Professor Emeritus in the Laboratory of Tree-Ring Research at the University of Arizona. His research interests include dendroarchaeology, the archaeology of the southern Colorado Plateau, Navajo ethnoarchaeology, paleoenvironmental reconstruction, human-environment interaction, and agent-based modeling of human behavior. He has served as President of the Arizona Archaeological and Historical Society, Treasurer of the Tree-Ring Society, President of the Society for Archaeological Sciences, Treasurer of the Society for American Archaeology, and is a Fellow with the Arizona-Nevada Academy of Science and the American Association for the Advancement of Science.
Paleoenvironment and Arroyo Hondo
28 August 2015
(Revised 24 November 2015)
Archaeological interest in past environmental conditions and modifications derives from efforts to understand the effects over time of these phenomena on human populations. Generally, emphasis is placed on the environment as the major resource of individual and group livelihoods and secondarily on other effects such as damage from catastrophic events like earthquakes and volcanic eruptions. The former aspect is focused on human subsistence systems as the main interface between survival and the natural world. For the last four millennia, a large segment of the southwestern United States has relied heavily on farming for its sustenance. For this and other reasons, the natural and social sciences in the Southwest have contributed disproportionately to advancing knowledge of human agricultural systems in semiarid environments. The region has provided a productive test bed for examining the complex interactions among human behavior, demography, and the environment that typify traditional Southwestern farming.
All agriculture involves the creation of artificial ecosystems that are subject to greater human control than natural ecosystems and thereby constrain some of the uncertainty inherent in the latter. Such control is manifested in many ways including the domestication and genetic modification of plants and animals, field locations, exploitation of productive soils, monitoring environmental predictors of future yields, protection of crops from early and late frosts, timing of planting and harvesting, water augmentation (runoff control, terracing, gridded fields, irrigation), and natural and artificial fertilization. The greater economic stability afforded by agricultural systems underlies numerous sociocultural developments that characterize agricultural societies. Among these developments are surplus production, long-term food storage, accumulation and redistribution of “wealth,” enhanced trade, population growth, increased competition for more rapidly depleted resources, greater sociocultural complexity, and many others.
Given the importance of food production in the course of human history and the widespread temporal and spatial distribution of agriculture around the world, knowledge of such subsistence systems is crucial to understanding the survival, expansion, and development of human societies through time and space. Because agricultural systems involve networks of intimate and complex relationships among human behavior, technology, social organization, demography, and environmental conditions, they can be adequately understood only with reference to detailed knowledge of past and present environmental conditions and fluctuations.
Archaeological investigation of human ecology involves assessing changing relationships among environment, culture, and demography (Dean 1988a; 1996). A large array of paleoenvironmental, ecological, agricultural, archaeological, ethnographic, historical, and demographic data has been entrained to better understand the role of agriculture in these interactions. Such efforts have not always met with unqualified success. Varying success rates are due to a variety of theoretical and methodological factors that must be made explicit in order to better understand such “failures,” recognize and rectify weaknesses in such studies, and to devise more appropriate and productive research strategies for investigating these issues.
The southwestern United States has long been a focus of efforts to understand interactions among environment, behavior, and demography. Bitter experience led to the early realization that many Southwestern environments are marginal for both native and European agricultural practices and that environment is a critical factor in the human occupation of the region. Historical and ethnographic records document principles, practices, technology, adaptations, successes, and failures of both indigenous and introduced agricultural systems in the region. Detailed archaeological records provide a long-term perspective on the full range of efforts to agriculturally support a variety of human societies across four millennia in a region characterized by enormous spatial and temporal variability in crop-raising potential. Increasingly sophisticated reconstructions of past environmental dynamics illuminate the conditions under which farming has been conducted in different times and places by diverse societies. This felicitous combination of circumstances makes the Southwest a particularly fruitful venue for exploring the archaeological study of agricultural adaptation to environmental stability, variation, and change across broad temporal and spatial scales. The following topics are particularly important in this regard: general approach; paleoenvironmental reconstruction; reconciling the amplitude, temporal, and spatial dimensions of paleoenvironmental techniques; relating human behavior to environmental conditions and variations; using ethnographic and historical data; and apparent conflicts between “scientific” expectations and real-world outcomes.
Dozens of techniques exist for reconstructing various aspects of past environments. The relevance of the reconstructions produced by these techniques to understanding agricultural systems varies tremendously from almost none (fluctuations in coral reefs) to a great deal (climate, hydrography, soils, etc.). Therefore, careful selection is necessary to identify and use techniques pertinent to each individual study. To make such selections, ways in which attributes reconstructed by particular techniques relate to the human beliefs, behavior, technology (agricultural methods and crops), and population that characterize local situations must be understood. For example local precipitation reconstructions are not directly relevant to irrigation systems that tap the flow of large through-flowing rivers, just as streamflow reconstructions are of limited relevance to dry farming systems that depend on local precipitation. Both, however, may be pertinent to mixed farming systems that exploit both rainfall and streamflow. Judicious assessment of specific reconstructions is critical to making such informed decisions.
Different paleoenvironmental techniques have certain attributes that render their reconstructions suitable for some archaeological purposes but not others. Therefore, definitional clarity and specificity are necessary to identify reconstructions pertinent to particular agricultural situations. A few examples illustrate this point. First, by focusing attention on variables such as precipitation and temperature, the common practice of conflating climate with environment obfuscates a wide range of potentially relevant variability. While climate is an environmental variable, it does not encompass the full range of environmental variation. Second, the temporal structure of environmental systems is important in this regard. Stability exists as long as the boundary conditions that regulate a system remain fairly constant. Thus, stable systems are characterized by means and variances that vary little over long time spans. Variability defines fluctuations within limits established by the boundary conditions. Change occurs when one stable state is transformed into another by changes in the boundary conditions; that is, by “permanent” changes in the means and variances that “define” the systems. Third, an important set of concepts defines the kinds of variability evident in environmental systems and clarifies the nature of reconstructions of these factors (Dean 1988b, 1996, 2000). Stable aspects (such as bedrock geology and large-scale topography) have not changed appreciably over the time period of interest (for agriculture in the Southwest, the last 4,200 years). Climate type and the elevational zonation of plant communities originally were included in this category, but, due to the recent advent of global climate change, these assignments may have to be revised. Stable aspects need not be reconstructed because they are accurately represented by present conditions. Low frequency (fluvial processes) and high frequency (climate) aspects of the environment have periodicities shorter than the length of the study period and, therefore, must be reconstructed. Episodic factors (earthquakes and volcanic eruptions) exhibit no known periodicity and must be identified in paleoenvironmental or archaeological records. Fourth, archaeological use of environmental information usually focuses on the amplitude (warm vs cool, dry vs wet, erosion vs. deposition, etc.) of the variations. Attributes not related to magnitude, however, are equally important. Temporal variability is the rate (gradual or rapid) of change from maximum to minimum values. Spatial variability reflects how conditions differ from one place to another. Coherence is the spatial patterning of associations (similarities vs. differences) among recording stations through time. Fifth, two different kinds of paleoenvironmental reconstruction are employed to capture variation through time and space in one or more of the foregoing attributes. Qualitative reconstructions employ measures of relative variability (positive vs. negative, longer vs. shorter, higher vs. lower, departures from mean values, etc.) in environmental factors, while quantitative reconstructions utilize standard units of measurement such as inches of precipitation and degrees of temperature. Consistent application of these or equivalent agreed-upon concepts and terms significantly clarifies archaeological dialogue on environmental matters.
In any effort to apply paleoenvironmental reconstructions in particular archaeological contexts, serious consideration must be given to the strengths and limitations inherent in the reconstruction techniques employed. Each technique is sensitive to only a portion of the total range of potentially relevant variability in any environmental factor. Alluvial chronostratigraphy, for example, informs on the rise and fall of floodplain water tables and the deposition and erosion of floodplain sediments but reveals little about climate. Palynology and packrat midden studies illuminate fluctuations in the composition and distribution of plant communities but not floodplain processes. Dendroclimatology reconstructs several climate variables but tells nothing about floodplain processes or plant communities.
Different paleoenvironmental techniques also are sensitive to only a small portion of the total possible range of different temporal and spatial scales of variation and resolution. Temporal sensitivity relates directly to the ability of dating techniques to resolve specific units of time. For example, alluvial stratigraphy reflects order of deposition but provides little information on the associated time scale, radiocarbon dates have a resolution of ±25 years, and tree-ring dates are accurate to the calendar year. Furthermore, the application of different types of dating varies with the different contexts (natural and human) in which the dated materials are found. As a result, alluvial, palynological, and packrat midden studies, which depend heavily on radiocarbon dating, normally do not resolve time periods shorter than about 50 years, while dendroclimatology detects annual and, sometimes, seasonal variations. Finally, the original purposes for which reconstructions were made must be factored into the analysis. Most reconstructions are designed to illuminate topics other than human-environment interactions, and such emphases may make them less than ideal for studying agricultural systems. For example, the vast majority of dendroclimatic reconstructions are intended to illuminate climate behavior, and some are ill-suited for investigating aspects of human behavior.
Once the foregoing issues have been addressed to the extent possible, the thorny problem of reconciling the various reconstructions with one another must be confronted. Such a step is necessary in order to derive hypotheses as to how reconstructed environmental conditions, variations, and changes might affect targeted agricultural systems. Because different paleoenvironmental techniques are sensitive to different agriculturally relevant aspects of the environment and to different temporal and spatial scales of variation and because reconstructions can be qualitative or quantitative, they cannot be combined simply by superimposing, averaging, summing, or subtracting values (Dean 1988b). Plotting dendroclimatic time series on top of the trace of fluctuations in alluvial groundwater levels (Euler et al. 1979: Figure 5) produces implausible representations of environmental variability simply because the former represents high-frequency variability in climate and the latter represents low-frequency fluvial processes. Experimentation reveals that the most productive procedure is to plot different reconstructions separately against a single time scale for visual inspection (Figure 1). This procedure identifies periods of environmental deterioration or amelioration that could adversely or favorably impact extant farming systems (Dean 1996).
A problem similar to the above afflicts attempts to relate environmental reconstructions to past human behavior. A major problem is rectifying the time scales of archaeological and paleoenvironmental data sets so that meaningful comparisons can be made and testable expectations can be derived. For example, the quarter-century to multicentury resolution of ceramics-based archaeological chronologies falls so short of the annual resolution of dendroclimatic reconstructions that it impedes relating observed behavioral changes to high-frequency climate fluctuations. Paleoenvironmental reconstructions with periodicities congruent with the resolution of archaeological chronologies (fluvial processes, plant community fluctuations) may be more suitable for this purpose. Possible solutions to this impasse include improving the resolution of archaeological chronologies, extracting low-frequency environmental variability from tree-ring data (Meko et al. 2007), and comparing a wide range of environmental reconstructions to derive expectations about human behavioral responses that can be evaluated archaeologically (Dean 1996). Similarly, spatial scales of paleoenvironmental and archaeological resolution must be reconciled. Thus, although continental-scale dendroclimatic reconstructions (Herweijer et al. 2007) can identify “megadroughts” that affected much of North America, they cannot reveal small-scale food production differentials that may have influenced human events in the Southwest.
Another factor that contributes significantly to our imperfect understanding of prehistoric agriculture in the Southwest is incongruencies among modern ideas about farming derived from scientific, historical, and ethnographic data on the one hand and archaeological evidence for farming in the distant past on the other. Recent attempts to use modern agricultural knowledge to assess the prehistoric farming potential of various localities have concluded that farming could not have supported the archaeologically documented population levels achieved in the past. While suggestive, these studies involve variables that may not be strictly analogous to those affecting prehistoric farming in the same areas, particularly as regards soil water-holding capacity and soil nutrient depletion. The nutrient requirements of prehistoric maize may have been lower than those of their modern counterparts resulting in slower soil depletion. The nutrient content and depletion rates of modern soils also may not reflect those of prehistoric fields, many of which now lie buried at considerable depths beneath the modern ground surface. The modern studies also do not consider the probability of soil enrichment by deposition due to periodic flooding, especially when and where floodwater farming was practiced. The use of modern records to characterize prehistoric farming suffers from the fact that historical and ethnographic records do not encompass the full range of past farming technology. One of the major conundrums of Southwestern anthropology is that indigenous farming seems to exceed expectations based on modern science. Both Navajo and Pueblo fields are known to continue in productive service many years beyond their theoretical durations. These limitations can be resolved by focused, diachronic research on native crop production systems that combine rigorous environmental and crop yield measurements with direct observation of farmers in action throughout the agricultural cycle. Recently, several innovative studies along these lines have significantly enhanced knowledge of modern indigenous farming that augment understanding of prehistoric agriculture in the region (Dominguez and Kolm; Muenchrath and Salvador 1995; Russell 1983; Vlasich 2005).
Related to the foregoing problem is the application of historically recorded environmental excursions in other parts of the world to the Southwest. Chief among these are the European Medieval Warm Period (WMP) and Little Ice Age (LIA). Such applications have serious flaws that limit their relevance to the Southwest. First, I am aware of no studies that define the effects that the atmospheric causes of these distant events would be expected to have in the Southwest. As a result, there is no way to characterize MWP or LIA variability in terms applicable to the Southwest. Second, lacking such studies, there is no reason to expect that either excursion would reflect similar conditions in other regions, such as aberrant warmth (MWP) or cold (LIA) in the Southwest. To invoke a more recent analogy, the atmospheric concomitants of El Niño have different effects in different regions, e.g., excessive rainfall in California contemporaneous with severe drought in Australia. Note also that Salzer’s (Salzer and Kipfmueller 2005) Flagstaff-area low-frequency temperature reconstruction (Figure 2) shows no evidence of either the MWP or LIA. Third, the ambiguous dating of both phenomena creates chronological elasticities that commonly are indiscriminately stretched or compressed to fit analytical convenience. The MWP is most precisely placed in the eleventh century A.D. and the LIA in the A.D. 1550-1850 interval. Fourth, the spatial distributions of LIA-related climatic phenomena are irregular, and “the notion of the Little Ice Age as a globally synchronous cold period has all but been dismissed” (Mann 2002). Because of these uncertainties, it is preferable to use the higher resolution, climate sensitive, locally specific paleoenvironmental reconstructions available in the Southwest rather than nebulous concepts imported from other parts of the world. Such local reconstructions would capture any climate variations due to the MWP or LIA with far more accuracy, precision, and relevance than imposing these remote occurrences on the Southwest.
Archaeologists interested in human ecology are faced with the multifaceted problem of assessing the potential effects of paleoenvironmental variability on human subsistence systems. This effort involves both evaluating the likely impact of different environmental factors on human behavior and identifying what ranges of variation in the factors can be expected to have limited subsistence systems. The first task is the easier of the two. It is not difficult to visualize how different environmental factors – such as floodplain and groundwater fluctuations, precipitation, temperature, growing season, and many others – may affect human societies. More difficult to establish are the levels at which changes in these variables impact humans. For example, at what depth of incision does arroyo cutting effectively hinder floodplain farming or at what precipitation threshold does drought significantly impact crop production. Subjective estimates of the potential impact on crop production can be made for qualitative reconstructions such as floodplain erosion-deposition (geomorphology), effective moisture (palynology), and plant community changes (pack rat midden studies). Other more objective limits may be drawn on the basis of statistical characteristics of qualitative reconstructions such as standard deviations of annual tree-growth departures (Dean and Robinson 1977; Dean et al 1985). More realistic estimates can be made for quantitative reconstructions by referencing ethnographic data on the environmental limitations on native crop production.
PALEOENVIRONMENTAL RECONSTRUCTION IN THE SOUTHWEST
Historically, the Southwest has been a focal point of paleoenvironmental research relevant to understanding the human ecology of the region over the last 4,000 years. Geomorphological and chronostratigraphic studies illuminate the relationship between farming populations, the land surfaces that supported their fields, and important sources of water for crops. Palynological analyses and packrat midden studies document low frequency changes in plant community distributions and compositions as well as human impacts on biotic environments. Dendroclimatology has produced a plethora of high-frequency qualitative and quantitative reconstructions of precipitation, temperature, streamflow, and other measures of high frequency variability in climate across the Southwest. Ethnographic studies of Indian farming practices (Bradfield 1971; Dominguez and Kolm 2005; Hack 1942; Russell 1983; Vlasich 2005) elaborate the human component of indigenous agriculture. As enlightening as these studies have been for understanding the long-term, agriculture-based human occupation of the Southwest, much still remains to be learned about the human-environment interface in the region.
LFP PALEOENVIRONMENTAL VARIABILITY IN THE RIO GRANDE VALLEY
Archaeologically oriented LFP paleoenvironmental research in the northern Southwest has been concentrated on the Colorado Plateau. For example, research into fluctuations in floodplain sedimentation and hydrology on the Plateau dates back to the 1920s and continues unabated to the present. This geographical emphasis appears to be the result of several factors including opportunity, the number of interested institutions and individuals, the clarity of the stratigraphic record in numerous exposures caused by widespread arroyo cutting since 1880, an abundance of large-scale contract archaeological projects, and the high quality chronological control afforded by radiocarbon and tree-ring dates. A fairly consistent regional pattern characteristic of many of the tributaries of the Colorado River has emerged from these studies. Considering only the last 2,000 years, alluvial deposition and rising groundwater levels prevailed (with variations) from approximately A.D. 1-200, 400-750, 925-1250, and 1500-1850. Erosion and falling water tables prevailed from 200-400, 750-925, 1250-1500, and 1850-present. Fewer alluvial studies suggest that this pattern cannot be projected into the Rio Grande drainage, where comparable depositional/erosional units have not yet been recognized.
HFP DENDROCLIMATIC RECONSTRUCTION IN THE RIO GRANDE VALLEY
The Arroyo Hondo Project’s seminal role in Southwestern archaeological dendroclimatology began in the late 1970s when Doug Schwartz broached the possibility of LTRR’s undertaking dendrochronological research focused specifically on Arroyo Hondo Pueblo. His query occurred at a particularly propitious time. In the early 1960s, the Archaeology Program of the Laboratory of Tree-Ring Research (LTRR) reanalyzed of all the Southwestern archaeological tree-ring samples in its possession. One outcome of this so-called Synthesis Project was using measured ring widths of dated archaeological samples to construct climate sensitive numerical tree-ring index chronologies for several localities across the region. At the same time, the Archaeological Program’s Southwest Paleoclimate Project was extending the archaeological tree-ring chronologies out to the present by incorporating tree-measurements from living trees to produce long composite chronologies to reconstruct aspects of past annual variability in climate. Initially, this research was focused on qualitative reconstructions that documented relative fluctuations (high vs. low) in climate. By the late 1970s, however, dendroclimatic interest had shifted to the quantitative reconstruction of conventional measures of precipitation (inches or millimeters) and, when possible, temperature (degrees). The time was ripe for the Southwest Paleoclimate Project to switch to quantitative reconstruction.
Enter Doug Schwartz. With his encouragement and support, the Arroyo Hondo Project became the test bed for the first use of archaeological tree-ring chronologies to retrodict inches of precipitation on annual and subannual scales of resolution. The result was quantitative reconstructions of Santa Fe area precipitation for the tree year (prior August through current July) and spring (current March through June). This research was fully described in Volume 4 of the School of American Research Press Arroyo Hondo Archaeological Series: The Past Climate of Arroyo Hondo, New Mexico, Reconstructed from Tree Rings by Rose, Dean, and Robinson (1981). Based largely on what was learned with the Arroyo Hondo study, the Southwest Paleoclimate Project went on to produce quantitative reconstructions for nearly 30 locations in the Southwest and numerous related applications.
The Arroyo Hondo volume also provided the opportunity to describe quantitative dendroclimatic reconstruction as applied in an archaeological context. Thanks to Jane Kepp’s skillful editing and insistence on clarity, the volume succeeded in this regard. Despite many significant changes in dendroclimatology since 1981, the process described in the volume remains fundamental (Figure 3). The statistical methods associated with each of the six analytical stages, however, have undergone continual modification as new theoretical approaches and mathematical techniques were developed. The first stage involves the construction of climate sensitive ring-width chronologies based on ring series from both archaeological contexts and living trees. These chronologies serve as the tree-ring data base for the analyses. The second stage (the first stage discussed in the book) includes the statistical evaluation of the tree-ring chronologies for dendroclimatic analysis. The third stage entails evaluating the relevant climate data to determine their suitability for dendroclimatic processing. The fourth stage includes the two- step calibration of the tree-ring and climatic time series. Calculating a regression-based response function establishes the effects of climate variability on tree growth. A transfer function derived from the response function produces regression equations used to reconstruct climate variables from the time series of tree ring width indices. The fifth stage, verification, statistically determines which potential reconstructions are valid. Finally, the accepted transfer equations are used to reconstruct climatic variability in the study area. In the Arroyo Hondo case, the tree-year (prior August through current July) and spring (current March through June) precipitation reconstructions in inches per year for the A.D. 850 to 1970 interval survived the testing process and are presented in Figures 33 and 34 and Tables 21 and 22 in the report. Since then, these reconstructions have been used assess the relationships between precipitation and human behavior at Arroyo Hondo Pueblo in particular and the Santa Fe area in general.
Since 1981, numerous additional quantitative dendroclimatic reconstructions have been made for various locations and various purposes in the central Rio Grande region of New Mexico. These reconstructions are an inadequately tapped resource for understanding past climate of the region and human interaction with the natural environment over the last 1500 years. Augmented by the incorporation of additional samples, the original Southwest Paleoclimate Project dendroclimatic chronologies were used to retrodict annual tree-year precipitation and current June Palmer Drought Severity Indices for four localities relevant to the northern Rio Grande area. These stations are Chama River Valley, Rio Grande North (the Taos area), Jemez Mountains, and Santa Fe. These reconstructions reveal a strong relationship between climate and the occupational history of Arroyo Hondo Pueblo. Establishment and initial growth of the pueblo occurred during an early 14th century wet period following the regional drought of the late 1200s. Between 1335 and 1345, population decline and near abandonment of the pueblo occurred during a severe drought. A reoccupation of the site during a wet interval in the late 14th-early 15th centuries was terminated around 1420 by the fourth worst drought in the Arroyo Hondo dendroclimatic record.
Recently Ron Towner and Matt Salzer (2013) undertook a more advanced dendroclimatic reconstruction of previous September through current June precipitation in the Chama, Jemez, and Arroyo Hondo localities for the A.D. 985-2002 period (Figure 4). Despite the use of new data selection criteria and analysis techniques, their Arroyo Hondo reconstruction is quite similar to the Rose et al. (1981) effort. This result testifies to the robust nature of climate-sensitive tree-ring chronologies in the Southwest. The major difference between the two reconstructions is mean annual precipitation values of thirteen inches for the 1981 reconstruction and 10 inches for the 2013 reconstruction, respectively. This discrepancy likely reflects the use of the Santa Fe and Las Vegas weather records for the 1981 analysis and the Las Vegas weather records for the 2013 study. In addition to the dendroclimatic reconstructions themselves, the use of three locations allow Towner and Salzer to address spatial variability across the Santa Fe area. Their study provides a firm foundation for refining the relationship between climate and the population of Arroyo Hondo Pueblo.
Two other recently-derived dendroclimatic reconstructions that are not archaeologically oriented also may be relevant to understanding the history of Arroyo Hondo Pueblo. Touchan et al. (2011) provide an A.D. 824-2005 reconstruction of prior October-current June precipitation on the Pajarito Plateau and discuss its wider ramifications for drought occurrence and atmospheric circulation patterns. Margolis et al. (2011) reconstructed calendar-year streamflow in millions of cubic meters for the Santa Fe River from 1305 to 2007 that could prove invaluable in assessing the potential contribution of irrigation agriculture to the survival of prehistoric populations in the Santa Fe area.
This review would not be complete without discussing current and future research projects that will have considerable relevance to the Santa Fe region. First, LTRR and SMU are wrapping up a five-year NSF Coupled Human and Natural Systems Program sponsored study of the human ecology of the Pajarito Plateau with numerous publications immanent. In a addition to dendroclimatic reconstructions, this project will produce forest fire histories that extend into the prehistoric period, analyses of changing agricultural productivity and human population dynamics on the Plateau, studies of the environmental impact of the large population influx in reaction to the Pueblo Revolt, calculations of the rate of forest recovery after the Post-Reconquest human abandonment of the area, dating the abandonment of archaeological sites through tree-ring dated forest dynamics studies, assessing the effects of human occupation on forest fire regimes, and others. Finally, an ongoing large scale project undertaken by Tim Kohler and colleagues will produce paleoenvironmental and demographic data pertinent to the Rio Grande Valley.
SUMMARY AND CONCLUSIONS
The Arroyo Hondo Project played an important role in dendroclimatic research in the central Rio Grande Valley. Initially, the Project provided the incentive for the first quantitative dendroclimatic reconstruction to use ring-width data from archaeological contexts, in this case sites (including Arroyo Hondo Pueblo) in the Santa Fe area. The worth of the Arroyo Hondo study is indicated by the degree of correspondence between climatic variability between 1300 and 1450 and the occupational history of Arroyo Hondo Pueblo. This work laid the foundation for comparable quantitative reconstructions for nearly thirty localities and numerous special projects in the Southwest. Subsequent dendroclimatic research projects – some connected with archaeology some not – in the area have generated high quality HFP data on numerous environmental factors including climate, streamflow, forest fire history, human disturbance of the environment, and environmental recovery from such disruptions. All this information will improve understanding of the human occupation of the middle Rio Grande Valley in general and the Arroyo Hondo Pueblo locality in particular.
Bradfield, Maitland 1971. The Changing Pattern of Hopi Agriculture. Royal Anthropological Institute of Great Britain and Ireland Occasional Paper, No. 30. London.
Dean, Jeffrey S. 1988a. A Model of Anasazi Behavioral Adaptation. In The Anasazi in a Changing Environment, edited by George J. Gumerman, pp. 25-44. Cambridge University Press, Cambridge.
Dean, Jeffrey S. 1988b. Dendrochronology and Peleoenvironmental Reconstruction on the Colorado Plateaus. In The Anasazi in a Changing Environment, edited by George J. Gumerman, pp. 119-167. Cambridge University Press, Cambridge.
Dean, Jeffrey S. 1996. Demography, Environment, and Subsistence Stress. In Evolving Complexity and Environmental Risk in the Prehistoric Southwest, edited by Joseph A. Tainter and Bonnie Bagley Tainter, pp. 25-56. Santa Fe Institute Studies in the Sciences of Complexity, Proceedings Volume 24. Addison-Wesley, Reading.
Dean, Jeffrey S. 2000. Complexity Theory and Sociocultural Change in the American Southwest. In The Way the Wind Blows: Climate, History, and Human Action, edited by Roderick J. McIntosh, Joseph A. Tainter, and Susan Keech McIntosh, pp. 89-118. Columbia University Press, New York.
Dean, Jeffrey S., Robert C. Euler, George J. Gumerman, Fred Plog, Richard H. Hevly, and Thor N. V. Karlstrom 1985. Human Behavior, Demography, and Paleoenvironment on the Colorado Plateaus. American Antiquity 50:537-554.
Dean, Jeffrey S. and William J. Robinson 1977. Dendroclimatic Variability in the American Southwest, A.D. 680-1970. Laboratory of Tree-Ring Research, University of Arizona, Tucson.
Dominguez, Steven, and Kenneth E. Kolm 2005. Beyond Water Harvesting: A Soil Hydrology Perspective ion Traditional Southwestern Agricultural Technology. American Antiquity 70:732-765.
Euler, Robert C., George J. Gumerman, Thor N. V. Karlstrom, Jeffrey S. Dean, and Richard H. Hevly 1979. The Colorado Plateaus: Cultural Dynamics and Paleoenvironment. Science 205:1089-1101.
Hack, John T. 1942. The Changing Physical Environment of the Hopi Indians of Arizona. Reports of the Awatovi Expedition, Peabody Museum, Harvard University, Report No. 1. Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, Vol. 35, No. 1. Cambridge
Herweijer, Celine, Richard Seeger, Edward R. Cook, and Julian Emile-Geay 2007. North American Droughts of the Last Millennium from a Gridded Network of Tree Ring Data. Journal of Climate 20:1353-1376.
Mann, Michael E. 2002. Little Ice Age. Encyclopedia of Global Environmental Change, Vol. 1, Earth System: Physical and Chemical Dimensions of Global Environmental Change, edited by Michael C. MacCracken and John S. Perry, pp. 504-509. John Wiley and Sons, Chichester, Great Britain.
Margolis, Ellis Q, David M. Meko, and Ramzi Touchan 2011. A Tree-Ring Reconstruction of Streamflow in the Santa Fe River, New Mexico. Journal of Hydrology 397:118-127.
Meko, David M., Connie A. Woodhouse, Christopher H. Baisan, Troy Knight, Jeffrey J. Lukas, Malcolm K. Hughes, and Matthew W. Salzer 2007. Medieval Drought in the Upper Colorado River Basin. Geophysical Research Letters 34.
Muenchrath, Deborah A., and Ricardo J. Salvador 1995. Maize Productivity and Agroecology: Effects of Environment and Agricultural Practices on the Biology of Maize. In Soil, Water, Biology, and Belief in Prehistoric and Traditional Southwestern Agriculture, edited by H. Wolcott Toll, pp. 303-333. New Mexico Archaeological Council Special Publication 2. Albuquerque.
Rose, Martin R., Jeffrey S. Dean, and William J. Robinson. 1981. The Past Climate of Arroyo Hondo, New Mexico, Reconstructed from Tree Rings. Arroyo Hondo Archaeological Series, Vol. 6. School of American Research Press, Santa Fe.
Russell, Scott C. 1983. Factors Affecting Agricultural Production in a Western Navajo Community. Ph.D dissertation, Department of Anthropology, Arizona State University, Tempe. University Microfilms International, Ann Arbor.
Salzer, Matthew W., and Kurt F. Kipfmueller. 2005. Reconstructed Temperature and Precipitation on a Millennial Timescale from Tree Rings-on the Southern Colorado Plateau, U.S.A. Climatic Change 70:465-487.
Touchan, Ramzi, Connie A. Woodhouse, David M. Meko, and Craig Allen 2011. Millennial Precipitation Reconstruction for the Jemez Mountains, New Mexico, Reveals Changing Drought Signal. International Journal of Climatology 31:896-906.
Towner, Ronald H., and Matthew W. Salzer 2013. Dendroclimatic Reconstructions of Precipitation for the Northern Rio Grande. In From Mountaintop to Valley Bottom: Understanding Past Land Use in the Northern Rio Grande Valley, New Mexico, edited by Bradley J. Vierra, pp. 54-65. University of Utah Press, Salt Lake City.
Vlasich, James A. 2005. Pueblo Indian Agriculture. University of New Mexico Press, Albuquerque.