Chapter 4: Central Arizona from 1200 to 1450 Return to Table of Contents
In this chapter, I describe
the climatic, environmental, and cultural diversity of central Arizona during the 1200
to 1450 period. Climatic and
environmental diversity--from the dry and hot desert to the wet and cool
mountains--provides a range of conditions to evaluate the influence of
environmental characteristics on vulnerability to dry periods. Cultural diversity--from
sedentary irrigation agriculturalists to newly arriving immigrants--suggests
that the findings of this study may be broadly applicable to other culturally
diverse regions and not limited by the subsistence strategies or vulnerabilities
of a particular society. It is important
to consider vulnerability to dry periods during the 13th through 15th
centuries so that we can advance our understanding of the complex social issues
occurring during this time.
This chapter is organized into four sections. First, I delineate the spatial boundaries of the study area and the watersheds therein that are an important analytical unit of this study. Second, I describe the climatic diversity within central Arizona and focus on aspects of this climate that are shared across the study area. Third, I describe the cultural and environmental diversity within the study area. Fourth, I discuss why the 1200 to 1450 period is an important period to consider vulnerability to dry periods.
Study Area and Scales of Analysis
The central Arizona study area (Figure 1, all shaded polygons) includes the low and hot Sonoran desert in the south, a transition zone north to the Colorado Plateau, and the cooler and wetter high mountains of eastern Arizona (Fish and Nabhan 1991; Turner and Brown 1982; Whittlesey and Ciolek-Torrello 1997). Populations throughout the study area were widely distributed throughout this landscape along perennial rivers, intermittent streams, ephemeral washes, and among mountains and mesas distant from perennial rivers.
The shaded polygons in Figure 4.1 are the watersheds that comprise the study area. A watershed is an area of land that drains water, sediment, and dissolved materials to a common outlet at some point along a stream channel (Dunne and Leopold 1978). Watersheds are also referred to as drainage basins or catchment areas and they occur at multiple scales. I identify the smallest watershed units ("cataloging units" or "sub-basins") identified by the U.S. Geological Survey (Seaber et al. 1987). Watersheds are a common spatial unit of analysis for archaeologists focused on central Arizona prehistory because these watersheds roughly correspond to differences in the material indicators used to infer cultural identity (as illustrated later in this chapter). Watersheds also delineate a reasonable spatial boundary that may approximate actual resource acquisition zones.
See Figure 4.1
The primary unit of analysis of this study is settlement-scale residential abandonment examined at the watershed scale and at the scale of the entire central Arizona study area. The "central Arizona” scale includes the nine identified watersheds in Figure 4.1: Agua Fria, Big Chino-Williamson Valley, Carrizo, Lower Verde, Lower Salt, Tonto, Upper Salt, and Upper Verde, and White watersheds. I selected these watersheds for study because there have been thorough and targeted efforts to compile records of all identified settlements in these areas (Wilcox et al. 2001a, 2001b, 2003) and dry periods in these watersheds can be effectively identified by available tree-ring precipitation reconstructions (as discussed in Chapter Five). I evaluate the models, when possible, at the watershed scale as this allows each model to be tested in multiple locations. If the relationship between dry-period severity and residential abandonment differs among watersheds, then differences in watershed characteristics are examined to understand factors that may have contributed to differences in relationships (discussed further in Chapter Five).
I use the central Arizona scale to evaluate a few aspects of the models. For example, settlements in the Agua Fria watershed were located only in areas of low and moderate precipitation while settlements in the Upper Salt were located only in areas of moderate and high precipitation. Thus, neither watershed offers a comprehensive opportunity to evaluate the influence of precipitation levels on vulnerability to dry periods. When watersheds are aggregated at the central Arizona scale, however, problems of low numbers of settlements in particular classifications are avoided.
I focus this analysis on the Agua Fria, Lower Verde, Lower Salt, Tonto, Upper Salt, and Upper Verde because each of these watersheds has a long-term settlement history and sufficient numbers of settlements to represent a range of demographic and environmental conditions necessary to evaluate the vulnerability models. I exclude the Big Chino-Williamson Valley, Carrizo, and White watersheds from the watershed-scale analysis due to low settlement and/or room numbers and the limited duration of occupation in these watersheds. Settlements within these watersheds are, however, included in the all-central-Arizona scale analyses.
Climatic Diversity
The climatic diversity of central Arizona provides a range of environmental conditions to evaluate models of vulnerability to dry periods. I use this diversity to identify differences in potential productivity among settlements and watersheds and to evaluate models of vulnerability to dry periods that emphasize these differences to explain spatial variation in vulnerability to dry periods (as discussed in Chapter Five and in the Supply models). I focus exclusively on precipitation and streamflow conditions because water is the primary limiting factor on resource productivity in the region (Muenchrath and Salvador 1995). The precipitation and streamflow levels and ranges noted in this section are all modern average annual values calculated from historical climate records (United States Geological Survey 2010; Western Regional Climate Center (2010). Modern long-term precipitation averages are appropriate for characterizing average precipitation levels during the 1200 to 1450 period because the atmospheric and physiographic controls on Southwest climate have not changed since the period of study (Sheppard et al. 2002). Studies of pollen, plant and animal distributions, geology (Schoenwetter 1962), and the tree-growth response to climate over time also demonstrate that there has been no change in the type of climate prevalent for at least the past 2,000 years (Dean and Robinson 1982). [It is important to understand that I use modern climate data with relatively fine spatial resolution to characterize settlement-scale potential resource productivity and tree-ring precipitation reconstructions to identify central Arizona scale dry periods during the 1200 to 1450 period. These data are thoroughly discussed in Chapter Five.]
Precipitation levels varied among settlements and watersheds based largely on differences in elevation, topography, and location. Settlements during the period of study were located in areas that historically (ca. 1900 to 2000) receive an annual average of 8" to 35"of precipitation (Western Regional Climate Center 2010). Figure 4.2 displays the average annual precipitation levels of settlement locations within each watershed. In general, higher elevations receive more precipitation than lower elevations. While these spatial differences in precipitation levels are relatively constant, year-to-year changes in regional-scale precipitation levels will occur that change the absolute levels received at each location. That is, if a dry-period decreases precipitation across the region, the relative rank-ordering of settlements by precipitation levels will remain constant (I demonstrate this spatial consistency in Chapter Five). This is because the absolute precipitation levels are controlled by hemispheric atmospheric circulation patterns and will not vary substantially within the study area (McPhee 2004). The implication of the synchronicity is, if people used population movement from areas of lesser to greater productivity to manage dry-period shortfall risks, then these movements were likely between areas of inherently different productivity rather than to areas with a short-lived anomalous precipitation advantage. I use locations near and far from perennial rivers and areas receiving on-average low, moderate, and high precipitation as indicators of inherently greater or lesser inherent productivity.
See Figure 4.2.
Streamflow discharge levels also varied substantially throughout the study area (Table 4.1). Precipitation levels are a reasonable indicator of relative changes in these discharges because the source flows are entirely within the study area watersheds. The watershed of the Lower Salt River, however, extends outside of the study area and into northern and eastern Arizona. As a result, its flows will sometimes be out-of-sync with local precipitation patterns within the rest of the study area. Perennial rivers in the region trend north to south beginning in the high elevation mountain and plateau country then join the east to west flowing Lower Salt River in the Sonoran Desert. All rivers are subject to short and intense flooding possibly with changes in channel morphology (Graf 1988; Graybill et al. 2006; Nials et al. 1989). In addition to the irrigation potential of some perennial rivers, riparian vegetation adjacent to each river was probably an important source of food (Fish and Nabhan 1991).
See Table 4.1.
Although people lived in locations with widely different precipitation and streamflow levels, all watersheds shared a common biseasonal precipitation pattern. This biseasonal pattern allows for a greater structural diversity of the flora than in other North American deserts (Brown 1994:182). The wettest periods are winter (November through April) and summer (July through September) (Sellers and Hill 1974; Sheppard et al. 2002). Winter precipitation is strongly affected by westerly storm tracks originating over the Pacific Ocean. Summer precipitation is the product of moisture from several oceanic sources moving into the region in July. Summer convective storms occur when local conditions cause these moist air masses to ascend (Sheppard et al. 2002). The proximity of Arizona to the Pacific Ocean, the Gulf of California, and the Gulf of Mexico also subject the region to atmospheric processes affected by changes in sea surface temperatures (i.e., El Nino, La Nina, Pacific Decadal Oscillation). The result of these complex processes interacting with the diverse topography of the region is high intra-annual and inter-annual precipitation variability. It seems likely that this variability challenged the successful scheduling of crop planting and made harvest success continually uncertain throughout the region (e.g., Dean 1988, 1996; Van West and Dean 2000).
Temperatures across the study area are, like precipitation, mostly a function of differences in elevation, topography, and location. In general, temperatures decrease with increasing elevation. Maximums are mid-summer and minimums are mid-winter. Temperature variations affect resource productivity by influencing water requirements of plants, growing season durations, and the timing and magnitude of snow-fed stream discharges relied on for irrigated agriculture. Throughout the study area the frost-free period exceeds 120 days, the approximate length of time necessary for a successful maize harvest (Muenchrath and Salvador 1995). I do not consider temperature variation in this study.
Archaeologically focused studies of climatic influences on human behavior in central Arizona have been dominated by a focus on changes in annual streamflow discharge volumes along the Lower Salt River (Graybill 1989; Graybill et al. 2006; Nials 1989), the Verde River (Van West and Altshcul 1997), and Tonto Creeks (Waters 1998; 2006). Only two studies have considered precipitation changes. Rose (1994) reconstructed Palmer Drought Severity Indices for Climate Divisions 3, 4, and 6, which cover the central Arizona study area. His retrodictions stop in A.D. 1370 as the data were not available at the time for further retrodiction. Weaver (1972) used precipitation and environmental reconstructions developed for the Black Mesa area of northern Arizona to articulate a model of Hohokam collapse that relied on decreases in effective moisture beginning in the 1200s. Neither of these studies examined long-term patterns of population movement in relation to changes in dry-period severity. Regional-scale patterns have not emerged from these studies and there is no empirically driven consensus on the impact of climate extremes on population movement in central Arizona.
Cultural and Environmental Diversity
Archaeologically defined cultural traditions referred to as Hohokam, Salado, Mogollon, Sinagua, and ancestral Puebloan have been identified within the study area and throughout prehistoric Arizona (Cordell 1997; Reid and Whittlesey 1997). Distinct boundaries separating these traditions, however, did not exist and most material assemblages from particular locations share a number of common traits. As a result, archaeologists typically conceptualize the region by watersheds and associated river valleys. These watersheds contain roughly similar constellations of material indicators used to identify the cultural traditions of the region. This suggests that these watersheds, and particularly the perennial rivers that define them, may have been socially meaningful. I follow this watershed-scale approach in this selective summary of these traditions, some of their material indicators, and culture-historical events. I also note some specific environmental and climatic characteristics that distinguish each watershed.
Lower Salt Watershed
The Lower Salt River watershed was home to the "Hohokam" tradition for more than a thousand years (Bayman 2001; Crown and Judge 1991; Doyel 1987; Doyel et al. 2000; Gumerman 1991). People of this tradition transformed their hot and dry Sonoran Desert homeland into the "most populous and agriculturally productive valley in the [U.S.] West before 1500 CE" (Fish and Fish 2007:1). This transformation was the result of the development of large-scale irrigation systems and effective social strategies for managing these systems (Howard 1993; Howard 2006). Eight canal networks with separate intakes off of the Lower Salt River have been identified (Howard 1993). During the period of this study, these networks irrigated about 100,000 acres (Howard 1993:296) of maize, beans, squash, and cotton (Gasser and Kwiatkowski 1991). Residential stability, likely due to these extensive canal systems, was greater in the Hohokam area than in other parts of the U.S. Southwest (Dean et al. 1994:70).
In addition to irrigation agriculture, characteristic Hohokam cultural features include monumental architecture (ball courts, platform mounds, and big houses), marine shell ornament production and circulation, cremation and inhumation mortuary practices, sedentary village-based communities, and red-on-buff ceramics (Gumerman 1991). Many of these features are thought to signal Mesoamerican origins or influences (McGuire et al. 1994; McGuire and Villalpando 2007). The extent of influence of the Hohokam tradition throughout the study area during the period of interest varied over time and is a matter of continuing inquiry (Van West and Altschul 1997:391-392; Whittlesey 1997b).
The 13th through 15th centuries, identified by archaeologists as the Classic Period, is characterized by significant sociocultural changes in the Lower Salt watershed. Following a retraction of the Hohokam interaction sphere in the 1100s, the people of the Lower Salt River valley developed new forms of residential and public architecture, modified existing mortuary practices, and developed new pottery styles and ceremonial objects, and severely restricted the extent of exchange relationships throughout the watershed (Abbott et al. 2003:8; Bayman 2001:280-283 and references contained therein). The transformation of a previously diffuse distribution of settlements across the watershed to walled, multi-family compounds after about 1100 suggests a more exclusionary pattern had developed (Fish 1989:50-51). The development of walled platform mounds placed at regular spatial intervals within the watershed also suggests increasing hierarchical social organization (Fish 1989:50-51). Sometime during the 15th century, the archaeological visibility of Hohokam (Dean 1991) and the other cultural traditions of central Arizona cease.
Based on retrodicted annual streamflow discharge volumes from tree-rings and analogies with historic irrigators in the watershed, Graybill et al. (2006) and Nials et al. (1989) have argued that catastrophic floods and associated geomorphic channel changes during the late 1300s contributed to settlement and population changes and ultimately the depopulation of the lower Salt River valley after A.D. 1400. Abbott and colleagues (Abbott 2003 ed.), however, develop a strong case for a change in Hohokam society around 1100 that initiated a gradual decline in sociopolitical conditions that ultimately led to the depopulation. Studies of dry-period impacts in the watershed have been limited to my (Ingram 2008) analysis of the relationship between droughts and changes in population growth rates in the most well documented canal system in the valley. I found that as dry periods increased in severity, population growth rates increased during a 700 year period. This relationship suggests that the Lower Salt may have been a refuge for people moving away from other areas.
The Lower Salt watershed is in the northern portion of the Sonoran Desert, a vast arid region that extends south and west into Mexico and California. The Sonoran Desert is classified as a "tropical-subtropical desertland" climatic zone (Brown 1994) similar to the Kalahari Desert of southern Africa, the Namib Desert of Saudi Arabia, and the Patagonian Desert of South America. Fish (1989:22) characterizes the Sonoran desert as "one of the major food-rich areas for a gathering economy in North America." The desert is within the Basin and Range physiographic province and consists of north-south trending faulted mountains and flat valley floors (Fenneman 1931). Precipitation, temperature, and streamflow conditions in the Lower Salt are more extreme than throughout the rest of the study area. Precipitation along the Lower Salt where settlement was concentrated is the lowest among all the watersheds--8.37" annually. Temperatures are high with daytime-high averages over 100 F during the summer (Western Regional Climate Center 2010).
Agua Fria Watershed
Peoples living in the northern portion of the Agua Fria watershed have been difficult to assign to traditional Southwestern cultural traditions. Instead, they are sometimes referred to as the "Central Arizona tradition" (see Wilcox and Holmlund 2007:122, note 23, for a discussion of the possible origins of this concept) or "Perry Mesa tradition" (Stone 2000). Puebloan-style architecture and decorated ceramics in the northern portion of the watershed suggests residents were not closely affiliated with the Hohokam tradition. Perry Mesa is the locus of settlement in the watershed during the period of study and settlement patterns on and around the mesa are an integral part of the Verde Confederacy model, a prominent case study of endemic warfare and alliance formation in the late prehistoric Southwest (Wilcox 2005; Wilcox et al. 2001b). Further south along the intermittent Agua Fria River, the watershed is sparsely populated after 1300 and the people there were probably most strongly affiliated with the Hohokam of the Lower Salt River watershed.
The Agua Fria watershed is a transition zone that begins with the Sonoran Desert in the south and ends with high mesas covered with grasslands cut by canyons in the north. Settlements in the northern portion of the watershed were located in areas receiving an average of 15" to 17" of precipitation annually, twice as much as those living in the Lower Salt. Settlement patterns in the watershed include hill-top sites, dispersed small sites near field systems, and settlements on mesas (Wilcox et al. 2001a, 2001b). The Agua Fria watershed is distinct among the watersheds considered in this study because is has the least extensive perennial riverine resources. The Agua Fria River and its tributaries are perennial in only a few places.
Upper and Lower Verde Watersheds
The peoples of the Upper Verde watershed are archaeological known as the "Sinagua" (Colton 1939; Plog 1989). Material correlates include Alameda Brown Ware, paddle-and-anvil ceramic techniques, extended inhumation, alcove houses, deep pit houses, and masonry pueblos and cliff dwellings. Colton (1946) argued that the Sinagua were influenced by Mogollon, Hohokam, and Pueblo people who were drawn to the Flagstaff area after the eruption of the Sunset Volcano. Although the current status of Colton's concept of a distinct culture is unclear (Plog 1989:264), research supports the diversity and variety of influences in the area. Schroeder (1957, 1979) interprets the diverse material patterns of the upper Agua Fria, Verde, and Tonto watersheds as part of a single "Hakataya" culture. The Upper Verde includes the well-known archaeological sites (cliff dwellings) of Montezuma's Castle and Tuzigoot, both national monuments.
In the Lower Verde watershed, ceramics, architecture, and other material traits are diverse but share similar patterns with the Sinagua of the Upper Verde (Pilles 1976; Whittlesey 1997a). The Lower Verde watershed also has strong evidence of Hohokam influence and is often considered a part of the Hohokam periphery (Whittlesey and Ciolek-Torrello 1997). The Lower Verde is culturally similar to the Tonto Basin (Pilles 1976). Settlements in the watershed are mostly located along river terraces, alluvial fans, and along tributaries of the Verde. There is evidence of prehistoric irrigation along the Verde although much of the valley is too steep and narrow for extensive irrigation (Van West and Altschul 1997; Whittlesey and Ciolek-Torrello 1997). Roughly separating the Upper from the Lower Verde watersheds is the Verde Valley, the greatest extent of arable land and biotic diversity in the watershed (Whittlesey 1997a).
Settlement patterns in the Verde watersheds, like those in the northern portion of the Agua Fria, are thought to be strongly influenced by the intentional creation of defensive clusters of settlements and buffer zones. A "Verde Confederacy" is argued to have existed that united Perry Mesa residents with residents of the Upper and middle Verde against populations located in the Tonto Basin (Wilcox et al. 2001b). In this model, the depopulation of the Lower Verde in the 12th century is thought to be the result of the intentional creation of a buffer zone against enemies in the Lower Salt or Tonto watersheds (Wilcox et al. 2001b).
The Lower and Upper Verde River watersheds are a mountainous transition zone between the Sonoran Desert in the south and the Colorado Plateau in the north. Unlike the wide and flat expanse of land along the Lower Salt River, the Verde River is constricted by a narrow valley throughout much of its length. The Verde River is one of the largest perennial rivers in Arizona (USGS 2010) and would have provided a reliable water supply for prehistoric irrigators. Settlements in the Lower Verde were located in areas receiving 9" to 23" inches of precipitation annually (Western Regional Climate Center 2010). Settlements in the Upper Verde received between 13" and 23" annually. Vegetation is dense along the floodplains and the river is bordered by a rich riparian zone. Plant communities in the watershed include semi-desert Grassland, Great Basin Conifer Woodland, and Sonoran Desertscrub (U.S. Fish and Wildlife Service 2010).
Tonto
The Tonto watershed and portions of the Upper Salt watersheds are considered "Salado" (Dean 2000). The material correlates of this pattern, particularly polychrome pottery, extend into western New Mexico and elsewhere in the Southwest (Crown 1994, 1995). Debate continues on the distinctiveness of a Salado culture or whether the traits are more of a "horizon" of styles shared by an amalgamation of peoples of varying backgrounds (e.g., Dean 2000; Doyel 1981; Rice 1998). In addition to polychrome pottery, some of these traits include walled residential compounds, monumental architecture, and in some cases irrigated agriculture. In the Tonto watershed, existing populations were strongly influenced by an influx of migrants during the late 13th century (Stark et al. 1995). These migrants, probably pushed by social and environmental changes in the north (Van West et al. 2000), formed multi-ethnic communities that ultimately proved relatively short lived. The watershed experienced substantial population loss in the mid to late 14th century, perhaps due to regional-scale changes in precipitation variability and more attractive social conditions elsewhere (Van West et al. 2000). The Tonto watershed was the first to be depopulated among the watersheds I focus on in this study.
The Tonto watershed is a transition zone from the Sonoran desert in the south to the mid-elevation mountains and uplands of the northern portion of the watershed. Vegetation ranges from Saguaro cactus to pine-forested mountains. Whittlesey et al. (2000:242) argue that a diversity of resources in the watershed created an island of resource advantage that attracted people of different ethnic and cultural traditions to the basin. The Tonto Creek is mostly perennial through the watershed and settlements were located in areas receiving from 15" to 25" of precipitation annually (Western Regional Climate Data Center 2010).
Upper Salt
The Upper Salt River watershed included people archaeologically referred to as Mogollon and Salado. Both traditions in the watershed were influenced by ancestral Puebloan (Anasazi) and Hohokam cultural traditions (Cordell 1997). Mogollon traits are widely distributed throughout the watershed and Salado characteristics are mostly along the Salt River in the southwestern portion of the watershed. The Mogollon, while sharing many similar characteristics with ancestral Puebloan (Anasazi) populations, maintained distinctive methods of pottery manufacture, architectural construction, residence location, and mortuary treatment (Cordell 1997; Haury 1936; Reid and Whittlesey 1997). Subsistence systems were initially focused on the use of wild plants and animals and the cultivation of small garden plots of corn, beans, and squash although an increasingly sedentary and agriculturally focused lifestyle developed during the period of study.
During the late 1200s and early 1300s, people living in the Upper Salt became increasing influenced by ancestral Puebloan immigrants fleeing the effects of deteriorating conditions in the northern Southwest (Reid and Whittlesey 1997). These immigrants substantially increased population levels and density in the watershed. Rapid population growth in the watershed has also been attributed to increases in opportunities for inter-community exchange (Graves et al. 1982). Rising social and economic tensions in the 1300s, as in other areas of the Southwest, are suggested by the selection of defensible locations for major and minor settlements and the use of cliff shelters for secure food storage (Reid and Whittlesey 1997:164). Following a period of population aggregation, population decline began sometime in the mid to late 1300s. Based on detailed study at a large pueblo in the watershed (Grasshopper), these declines may have been caused by dry-period decreases in resource productivity, increased population, reduced soil fertility, and depleted resources (Reid et al. 2006; Reid and Whittlesey 1997:164). Factors that may have influenced the decline of Salado populations include rising population levels in the context of climatic conditions unfavorable for irrigated agriculture (Waters 2006). Archaeological visibility of human occupation of the watershed ceases sometime during the early 1400s.
The Upper Salt River watershed is a mountainous environment characterized by extreme changes in elevation. Settlements occupied during the period of study are located between 2,000 and 7,000 feet in elevation. Average annual precipitation received at settlement locations ranges from 15" to 35" and the watershed average is about 18.7", greater than any other populated watershed in Arizona (Western Regional Climate Center 2010). Perennial rivers include the Salt with three tributaries (Cherry Creek, Canyon Creek, and Cibecue). Based on my GIS analysis of the slopes of land throughout the watershed, opportunities for floodplain and/or irrigated farming were probably restricted and minimal in most places except along the Salt River near the western edge of the watershed (slopes of between 0 and 5% are considered optimal for irrigation; Walker 1989). Assessing opportunities for floodplain and irrigated farming along perennial rivers of the watershed, however, requires further on-site evaluation.
Contributions to the Prehistory of Central Arizona
It is important to consider vulnerability to dry periods during the 13th through 15th centuries so that we can advance our understanding of the complex social issues occurring during this time. These issues include rising warfare and the 15th century regional depopulation. Each involves population movement on a large scale and over long-time periods. Models of increasing conflict and warfare explain movements out of particular places on the landscape as defensive responses to the real or perceived threat of violence (LeBlanc and Rice 2001). When settled areas are abandoned and unoccupied zones around clusters of settlements are observed, these patterns are interpreted as an effort to create defensive open spaces between socially distant peoples (DeBoer 1981; LeBlanc 1999; Wilcox et al. 2001b; Wilcox and Haas 1994). Models of regional depopulation need to explain large-scale population movements out of a region (unless in situ demographic decline is argued, e.g., Hill et al. 2004). Our current understanding of the process in central Arizona is limited to rising warfare (Wilcox et al. 2001b), demographic decline associated with community coalescence (Hill et al. 2004), and general notions of breaches in regional carrying capacity (e.g., LeBlanc 1999, 2006). Along the Lower Salt River, a prevailing depopulation hypothesis is technological and social challenges to irrigated agriculture related to streamflow variability, flooding, and channel change in the late 1300s (Graybill et al. 2006; Nials et al. 1989). In contrast to environmentally focused hypotheses, Abbott and colleagues (Abbott 2003 ed.) have made a strong case for a gradual, centuries-long decline of populations along the Lower Salt. Causes for this decline are complex and include demographic instability, truncated trade networks, political strife, environmental impacts and ultimately ineffective responses to these challenges (Abbott 2003).
Advancing our understanding of both warfare and depopulation in central Arizona requires a better understanding of the factors influencing population movements. This study focuses on a frequently considered explanation for population movement--dry periods--and asks whether or not and under what conditions there is long-term evidence that dry periods were related to population movements out of settlements and watersheds in central Arizona. If dry-period influences on movements are detected over long time periods, then explanations of warfare and depopulation must accommodate climatic contributions to these phenomena. It should not be sufficient that we gloss climatic influences as simply context in explanations of the major events during this period in prehistory. For example, Wilcox et al. (2001b:165) argue the role of conflict in the depopulation of central Arizona and suggest that "Environmental perturbations could have further exacerbated such a process." No further suggestions of what these perturbations might be were provided. I am not suggesting dry periods as the sole cause of any social phenomenon. Rather, I am suggesting that if population movements constitute an integral part of a social phenomenon being explained and if these movements are strongly related dry periods, then explanations that do not consider potential dry-period impacts on movement are incomplete.
In sum, the cultural, environmental, and climatic diversity of central Arizona during the 13th through 15th centuries provide an important context for evaluating models of vulnerability to dry periods and furthering our understanding of dry-period influences on population movement. Population movements were an integral part of some of the critical social issues of this period. At the conclusion of this study (Chapter Ten), I discuss the implications of the findings of this research for understanding both the depopulation and models of increasing warfare in the region.
This chapter is organized into four sections. First, I delineate the spatial boundaries of the study area and the watersheds therein that are an important analytical unit of this study. Second, I describe the climatic diversity within central Arizona and focus on aspects of this climate that are shared across the study area. Third, I describe the cultural and environmental diversity within the study area. Fourth, I discuss why the 1200 to 1450 period is an important period to consider vulnerability to dry periods.
Study Area and Scales of Analysis
The central Arizona study area (Figure 1, all shaded polygons) includes the low and hot Sonoran desert in the south, a transition zone north to the Colorado Plateau, and the cooler and wetter high mountains of eastern Arizona (Fish and Nabhan 1991; Turner and Brown 1982; Whittlesey and Ciolek-Torrello 1997). Populations throughout the study area were widely distributed throughout this landscape along perennial rivers, intermittent streams, ephemeral washes, and among mountains and mesas distant from perennial rivers.
The shaded polygons in Figure 4.1 are the watersheds that comprise the study area. A watershed is an area of land that drains water, sediment, and dissolved materials to a common outlet at some point along a stream channel (Dunne and Leopold 1978). Watersheds are also referred to as drainage basins or catchment areas and they occur at multiple scales. I identify the smallest watershed units ("cataloging units" or "sub-basins") identified by the U.S. Geological Survey (Seaber et al. 1987). Watersheds are a common spatial unit of analysis for archaeologists focused on central Arizona prehistory because these watersheds roughly correspond to differences in the material indicators used to infer cultural identity (as illustrated later in this chapter). Watersheds also delineate a reasonable spatial boundary that may approximate actual resource acquisition zones.
See Figure 4.1
The primary unit of analysis of this study is settlement-scale residential abandonment examined at the watershed scale and at the scale of the entire central Arizona study area. The "central Arizona” scale includes the nine identified watersheds in Figure 4.1: Agua Fria, Big Chino-Williamson Valley, Carrizo, Lower Verde, Lower Salt, Tonto, Upper Salt, and Upper Verde, and White watersheds. I selected these watersheds for study because there have been thorough and targeted efforts to compile records of all identified settlements in these areas (Wilcox et al. 2001a, 2001b, 2003) and dry periods in these watersheds can be effectively identified by available tree-ring precipitation reconstructions (as discussed in Chapter Five). I evaluate the models, when possible, at the watershed scale as this allows each model to be tested in multiple locations. If the relationship between dry-period severity and residential abandonment differs among watersheds, then differences in watershed characteristics are examined to understand factors that may have contributed to differences in relationships (discussed further in Chapter Five).
I use the central Arizona scale to evaluate a few aspects of the models. For example, settlements in the Agua Fria watershed were located only in areas of low and moderate precipitation while settlements in the Upper Salt were located only in areas of moderate and high precipitation. Thus, neither watershed offers a comprehensive opportunity to evaluate the influence of precipitation levels on vulnerability to dry periods. When watersheds are aggregated at the central Arizona scale, however, problems of low numbers of settlements in particular classifications are avoided.
I focus this analysis on the Agua Fria, Lower Verde, Lower Salt, Tonto, Upper Salt, and Upper Verde because each of these watersheds has a long-term settlement history and sufficient numbers of settlements to represent a range of demographic and environmental conditions necessary to evaluate the vulnerability models. I exclude the Big Chino-Williamson Valley, Carrizo, and White watersheds from the watershed-scale analysis due to low settlement and/or room numbers and the limited duration of occupation in these watersheds. Settlements within these watersheds are, however, included in the all-central-Arizona scale analyses.
Climatic Diversity
The climatic diversity of central Arizona provides a range of environmental conditions to evaluate models of vulnerability to dry periods. I use this diversity to identify differences in potential productivity among settlements and watersheds and to evaluate models of vulnerability to dry periods that emphasize these differences to explain spatial variation in vulnerability to dry periods (as discussed in Chapter Five and in the Supply models). I focus exclusively on precipitation and streamflow conditions because water is the primary limiting factor on resource productivity in the region (Muenchrath and Salvador 1995). The precipitation and streamflow levels and ranges noted in this section are all modern average annual values calculated from historical climate records (United States Geological Survey 2010; Western Regional Climate Center (2010). Modern long-term precipitation averages are appropriate for characterizing average precipitation levels during the 1200 to 1450 period because the atmospheric and physiographic controls on Southwest climate have not changed since the period of study (Sheppard et al. 2002). Studies of pollen, plant and animal distributions, geology (Schoenwetter 1962), and the tree-growth response to climate over time also demonstrate that there has been no change in the type of climate prevalent for at least the past 2,000 years (Dean and Robinson 1982). [It is important to understand that I use modern climate data with relatively fine spatial resolution to characterize settlement-scale potential resource productivity and tree-ring precipitation reconstructions to identify central Arizona scale dry periods during the 1200 to 1450 period. These data are thoroughly discussed in Chapter Five.]
Precipitation levels varied among settlements and watersheds based largely on differences in elevation, topography, and location. Settlements during the period of study were located in areas that historically (ca. 1900 to 2000) receive an annual average of 8" to 35"of precipitation (Western Regional Climate Center 2010). Figure 4.2 displays the average annual precipitation levels of settlement locations within each watershed. In general, higher elevations receive more precipitation than lower elevations. While these spatial differences in precipitation levels are relatively constant, year-to-year changes in regional-scale precipitation levels will occur that change the absolute levels received at each location. That is, if a dry-period decreases precipitation across the region, the relative rank-ordering of settlements by precipitation levels will remain constant (I demonstrate this spatial consistency in Chapter Five). This is because the absolute precipitation levels are controlled by hemispheric atmospheric circulation patterns and will not vary substantially within the study area (McPhee 2004). The implication of the synchronicity is, if people used population movement from areas of lesser to greater productivity to manage dry-period shortfall risks, then these movements were likely between areas of inherently different productivity rather than to areas with a short-lived anomalous precipitation advantage. I use locations near and far from perennial rivers and areas receiving on-average low, moderate, and high precipitation as indicators of inherently greater or lesser inherent productivity.
See Figure 4.2.
Streamflow discharge levels also varied substantially throughout the study area (Table 4.1). Precipitation levels are a reasonable indicator of relative changes in these discharges because the source flows are entirely within the study area watersheds. The watershed of the Lower Salt River, however, extends outside of the study area and into northern and eastern Arizona. As a result, its flows will sometimes be out-of-sync with local precipitation patterns within the rest of the study area. Perennial rivers in the region trend north to south beginning in the high elevation mountain and plateau country then join the east to west flowing Lower Salt River in the Sonoran Desert. All rivers are subject to short and intense flooding possibly with changes in channel morphology (Graf 1988; Graybill et al. 2006; Nials et al. 1989). In addition to the irrigation potential of some perennial rivers, riparian vegetation adjacent to each river was probably an important source of food (Fish and Nabhan 1991).
See Table 4.1.
Although people lived in locations with widely different precipitation and streamflow levels, all watersheds shared a common biseasonal precipitation pattern. This biseasonal pattern allows for a greater structural diversity of the flora than in other North American deserts (Brown 1994:182). The wettest periods are winter (November through April) and summer (July through September) (Sellers and Hill 1974; Sheppard et al. 2002). Winter precipitation is strongly affected by westerly storm tracks originating over the Pacific Ocean. Summer precipitation is the product of moisture from several oceanic sources moving into the region in July. Summer convective storms occur when local conditions cause these moist air masses to ascend (Sheppard et al. 2002). The proximity of Arizona to the Pacific Ocean, the Gulf of California, and the Gulf of Mexico also subject the region to atmospheric processes affected by changes in sea surface temperatures (i.e., El Nino, La Nina, Pacific Decadal Oscillation). The result of these complex processes interacting with the diverse topography of the region is high intra-annual and inter-annual precipitation variability. It seems likely that this variability challenged the successful scheduling of crop planting and made harvest success continually uncertain throughout the region (e.g., Dean 1988, 1996; Van West and Dean 2000).
Temperatures across the study area are, like precipitation, mostly a function of differences in elevation, topography, and location. In general, temperatures decrease with increasing elevation. Maximums are mid-summer and minimums are mid-winter. Temperature variations affect resource productivity by influencing water requirements of plants, growing season durations, and the timing and magnitude of snow-fed stream discharges relied on for irrigated agriculture. Throughout the study area the frost-free period exceeds 120 days, the approximate length of time necessary for a successful maize harvest (Muenchrath and Salvador 1995). I do not consider temperature variation in this study.
Archaeologically focused studies of climatic influences on human behavior in central Arizona have been dominated by a focus on changes in annual streamflow discharge volumes along the Lower Salt River (Graybill 1989; Graybill et al. 2006; Nials 1989), the Verde River (Van West and Altshcul 1997), and Tonto Creeks (Waters 1998; 2006). Only two studies have considered precipitation changes. Rose (1994) reconstructed Palmer Drought Severity Indices for Climate Divisions 3, 4, and 6, which cover the central Arizona study area. His retrodictions stop in A.D. 1370 as the data were not available at the time for further retrodiction. Weaver (1972) used precipitation and environmental reconstructions developed for the Black Mesa area of northern Arizona to articulate a model of Hohokam collapse that relied on decreases in effective moisture beginning in the 1200s. Neither of these studies examined long-term patterns of population movement in relation to changes in dry-period severity. Regional-scale patterns have not emerged from these studies and there is no empirically driven consensus on the impact of climate extremes on population movement in central Arizona.
Cultural and Environmental Diversity
Archaeologically defined cultural traditions referred to as Hohokam, Salado, Mogollon, Sinagua, and ancestral Puebloan have been identified within the study area and throughout prehistoric Arizona (Cordell 1997; Reid and Whittlesey 1997). Distinct boundaries separating these traditions, however, did not exist and most material assemblages from particular locations share a number of common traits. As a result, archaeologists typically conceptualize the region by watersheds and associated river valleys. These watersheds contain roughly similar constellations of material indicators used to identify the cultural traditions of the region. This suggests that these watersheds, and particularly the perennial rivers that define them, may have been socially meaningful. I follow this watershed-scale approach in this selective summary of these traditions, some of their material indicators, and culture-historical events. I also note some specific environmental and climatic characteristics that distinguish each watershed.
Lower Salt Watershed
The Lower Salt River watershed was home to the "Hohokam" tradition for more than a thousand years (Bayman 2001; Crown and Judge 1991; Doyel 1987; Doyel et al. 2000; Gumerman 1991). People of this tradition transformed their hot and dry Sonoran Desert homeland into the "most populous and agriculturally productive valley in the [U.S.] West before 1500 CE" (Fish and Fish 2007:1). This transformation was the result of the development of large-scale irrigation systems and effective social strategies for managing these systems (Howard 1993; Howard 2006). Eight canal networks with separate intakes off of the Lower Salt River have been identified (Howard 1993). During the period of this study, these networks irrigated about 100,000 acres (Howard 1993:296) of maize, beans, squash, and cotton (Gasser and Kwiatkowski 1991). Residential stability, likely due to these extensive canal systems, was greater in the Hohokam area than in other parts of the U.S. Southwest (Dean et al. 1994:70).
In addition to irrigation agriculture, characteristic Hohokam cultural features include monumental architecture (ball courts, platform mounds, and big houses), marine shell ornament production and circulation, cremation and inhumation mortuary practices, sedentary village-based communities, and red-on-buff ceramics (Gumerman 1991). Many of these features are thought to signal Mesoamerican origins or influences (McGuire et al. 1994; McGuire and Villalpando 2007). The extent of influence of the Hohokam tradition throughout the study area during the period of interest varied over time and is a matter of continuing inquiry (Van West and Altschul 1997:391-392; Whittlesey 1997b).
The 13th through 15th centuries, identified by archaeologists as the Classic Period, is characterized by significant sociocultural changes in the Lower Salt watershed. Following a retraction of the Hohokam interaction sphere in the 1100s, the people of the Lower Salt River valley developed new forms of residential and public architecture, modified existing mortuary practices, and developed new pottery styles and ceremonial objects, and severely restricted the extent of exchange relationships throughout the watershed (Abbott et al. 2003:8; Bayman 2001:280-283 and references contained therein). The transformation of a previously diffuse distribution of settlements across the watershed to walled, multi-family compounds after about 1100 suggests a more exclusionary pattern had developed (Fish 1989:50-51). The development of walled platform mounds placed at regular spatial intervals within the watershed also suggests increasing hierarchical social organization (Fish 1989:50-51). Sometime during the 15th century, the archaeological visibility of Hohokam (Dean 1991) and the other cultural traditions of central Arizona cease.
Based on retrodicted annual streamflow discharge volumes from tree-rings and analogies with historic irrigators in the watershed, Graybill et al. (2006) and Nials et al. (1989) have argued that catastrophic floods and associated geomorphic channel changes during the late 1300s contributed to settlement and population changes and ultimately the depopulation of the lower Salt River valley after A.D. 1400. Abbott and colleagues (Abbott 2003 ed.), however, develop a strong case for a change in Hohokam society around 1100 that initiated a gradual decline in sociopolitical conditions that ultimately led to the depopulation. Studies of dry-period impacts in the watershed have been limited to my (Ingram 2008) analysis of the relationship between droughts and changes in population growth rates in the most well documented canal system in the valley. I found that as dry periods increased in severity, population growth rates increased during a 700 year period. This relationship suggests that the Lower Salt may have been a refuge for people moving away from other areas.
The Lower Salt watershed is in the northern portion of the Sonoran Desert, a vast arid region that extends south and west into Mexico and California. The Sonoran Desert is classified as a "tropical-subtropical desertland" climatic zone (Brown 1994) similar to the Kalahari Desert of southern Africa, the Namib Desert of Saudi Arabia, and the Patagonian Desert of South America. Fish (1989:22) characterizes the Sonoran desert as "one of the major food-rich areas for a gathering economy in North America." The desert is within the Basin and Range physiographic province and consists of north-south trending faulted mountains and flat valley floors (Fenneman 1931). Precipitation, temperature, and streamflow conditions in the Lower Salt are more extreme than throughout the rest of the study area. Precipitation along the Lower Salt where settlement was concentrated is the lowest among all the watersheds--8.37" annually. Temperatures are high with daytime-high averages over 100 F during the summer (Western Regional Climate Center 2010).
Agua Fria Watershed
Peoples living in the northern portion of the Agua Fria watershed have been difficult to assign to traditional Southwestern cultural traditions. Instead, they are sometimes referred to as the "Central Arizona tradition" (see Wilcox and Holmlund 2007:122, note 23, for a discussion of the possible origins of this concept) or "Perry Mesa tradition" (Stone 2000). Puebloan-style architecture and decorated ceramics in the northern portion of the watershed suggests residents were not closely affiliated with the Hohokam tradition. Perry Mesa is the locus of settlement in the watershed during the period of study and settlement patterns on and around the mesa are an integral part of the Verde Confederacy model, a prominent case study of endemic warfare and alliance formation in the late prehistoric Southwest (Wilcox 2005; Wilcox et al. 2001b). Further south along the intermittent Agua Fria River, the watershed is sparsely populated after 1300 and the people there were probably most strongly affiliated with the Hohokam of the Lower Salt River watershed.
The Agua Fria watershed is a transition zone that begins with the Sonoran Desert in the south and ends with high mesas covered with grasslands cut by canyons in the north. Settlements in the northern portion of the watershed were located in areas receiving an average of 15" to 17" of precipitation annually, twice as much as those living in the Lower Salt. Settlement patterns in the watershed include hill-top sites, dispersed small sites near field systems, and settlements on mesas (Wilcox et al. 2001a, 2001b). The Agua Fria watershed is distinct among the watersheds considered in this study because is has the least extensive perennial riverine resources. The Agua Fria River and its tributaries are perennial in only a few places.
Upper and Lower Verde Watersheds
The peoples of the Upper Verde watershed are archaeological known as the "Sinagua" (Colton 1939; Plog 1989). Material correlates include Alameda Brown Ware, paddle-and-anvil ceramic techniques, extended inhumation, alcove houses, deep pit houses, and masonry pueblos and cliff dwellings. Colton (1946) argued that the Sinagua were influenced by Mogollon, Hohokam, and Pueblo people who were drawn to the Flagstaff area after the eruption of the Sunset Volcano. Although the current status of Colton's concept of a distinct culture is unclear (Plog 1989:264), research supports the diversity and variety of influences in the area. Schroeder (1957, 1979) interprets the diverse material patterns of the upper Agua Fria, Verde, and Tonto watersheds as part of a single "Hakataya" culture. The Upper Verde includes the well-known archaeological sites (cliff dwellings) of Montezuma's Castle and Tuzigoot, both national monuments.
In the Lower Verde watershed, ceramics, architecture, and other material traits are diverse but share similar patterns with the Sinagua of the Upper Verde (Pilles 1976; Whittlesey 1997a). The Lower Verde watershed also has strong evidence of Hohokam influence and is often considered a part of the Hohokam periphery (Whittlesey and Ciolek-Torrello 1997). The Lower Verde is culturally similar to the Tonto Basin (Pilles 1976). Settlements in the watershed are mostly located along river terraces, alluvial fans, and along tributaries of the Verde. There is evidence of prehistoric irrigation along the Verde although much of the valley is too steep and narrow for extensive irrigation (Van West and Altschul 1997; Whittlesey and Ciolek-Torrello 1997). Roughly separating the Upper from the Lower Verde watersheds is the Verde Valley, the greatest extent of arable land and biotic diversity in the watershed (Whittlesey 1997a).
Settlement patterns in the Verde watersheds, like those in the northern portion of the Agua Fria, are thought to be strongly influenced by the intentional creation of defensive clusters of settlements and buffer zones. A "Verde Confederacy" is argued to have existed that united Perry Mesa residents with residents of the Upper and middle Verde against populations located in the Tonto Basin (Wilcox et al. 2001b). In this model, the depopulation of the Lower Verde in the 12th century is thought to be the result of the intentional creation of a buffer zone against enemies in the Lower Salt or Tonto watersheds (Wilcox et al. 2001b).
The Lower and Upper Verde River watersheds are a mountainous transition zone between the Sonoran Desert in the south and the Colorado Plateau in the north. Unlike the wide and flat expanse of land along the Lower Salt River, the Verde River is constricted by a narrow valley throughout much of its length. The Verde River is one of the largest perennial rivers in Arizona (USGS 2010) and would have provided a reliable water supply for prehistoric irrigators. Settlements in the Lower Verde were located in areas receiving 9" to 23" inches of precipitation annually (Western Regional Climate Center 2010). Settlements in the Upper Verde received between 13" and 23" annually. Vegetation is dense along the floodplains and the river is bordered by a rich riparian zone. Plant communities in the watershed include semi-desert Grassland, Great Basin Conifer Woodland, and Sonoran Desertscrub (U.S. Fish and Wildlife Service 2010).
Tonto
The Tonto watershed and portions of the Upper Salt watersheds are considered "Salado" (Dean 2000). The material correlates of this pattern, particularly polychrome pottery, extend into western New Mexico and elsewhere in the Southwest (Crown 1994, 1995). Debate continues on the distinctiveness of a Salado culture or whether the traits are more of a "horizon" of styles shared by an amalgamation of peoples of varying backgrounds (e.g., Dean 2000; Doyel 1981; Rice 1998). In addition to polychrome pottery, some of these traits include walled residential compounds, monumental architecture, and in some cases irrigated agriculture. In the Tonto watershed, existing populations were strongly influenced by an influx of migrants during the late 13th century (Stark et al. 1995). These migrants, probably pushed by social and environmental changes in the north (Van West et al. 2000), formed multi-ethnic communities that ultimately proved relatively short lived. The watershed experienced substantial population loss in the mid to late 14th century, perhaps due to regional-scale changes in precipitation variability and more attractive social conditions elsewhere (Van West et al. 2000). The Tonto watershed was the first to be depopulated among the watersheds I focus on in this study.
The Tonto watershed is a transition zone from the Sonoran desert in the south to the mid-elevation mountains and uplands of the northern portion of the watershed. Vegetation ranges from Saguaro cactus to pine-forested mountains. Whittlesey et al. (2000:242) argue that a diversity of resources in the watershed created an island of resource advantage that attracted people of different ethnic and cultural traditions to the basin. The Tonto Creek is mostly perennial through the watershed and settlements were located in areas receiving from 15" to 25" of precipitation annually (Western Regional Climate Data Center 2010).
Upper Salt
The Upper Salt River watershed included people archaeologically referred to as Mogollon and Salado. Both traditions in the watershed were influenced by ancestral Puebloan (Anasazi) and Hohokam cultural traditions (Cordell 1997). Mogollon traits are widely distributed throughout the watershed and Salado characteristics are mostly along the Salt River in the southwestern portion of the watershed. The Mogollon, while sharing many similar characteristics with ancestral Puebloan (Anasazi) populations, maintained distinctive methods of pottery manufacture, architectural construction, residence location, and mortuary treatment (Cordell 1997; Haury 1936; Reid and Whittlesey 1997). Subsistence systems were initially focused on the use of wild plants and animals and the cultivation of small garden plots of corn, beans, and squash although an increasingly sedentary and agriculturally focused lifestyle developed during the period of study.
During the late 1200s and early 1300s, people living in the Upper Salt became increasing influenced by ancestral Puebloan immigrants fleeing the effects of deteriorating conditions in the northern Southwest (Reid and Whittlesey 1997). These immigrants substantially increased population levels and density in the watershed. Rapid population growth in the watershed has also been attributed to increases in opportunities for inter-community exchange (Graves et al. 1982). Rising social and economic tensions in the 1300s, as in other areas of the Southwest, are suggested by the selection of defensible locations for major and minor settlements and the use of cliff shelters for secure food storage (Reid and Whittlesey 1997:164). Following a period of population aggregation, population decline began sometime in the mid to late 1300s. Based on detailed study at a large pueblo in the watershed (Grasshopper), these declines may have been caused by dry-period decreases in resource productivity, increased population, reduced soil fertility, and depleted resources (Reid et al. 2006; Reid and Whittlesey 1997:164). Factors that may have influenced the decline of Salado populations include rising population levels in the context of climatic conditions unfavorable for irrigated agriculture (Waters 2006). Archaeological visibility of human occupation of the watershed ceases sometime during the early 1400s.
The Upper Salt River watershed is a mountainous environment characterized by extreme changes in elevation. Settlements occupied during the period of study are located between 2,000 and 7,000 feet in elevation. Average annual precipitation received at settlement locations ranges from 15" to 35" and the watershed average is about 18.7", greater than any other populated watershed in Arizona (Western Regional Climate Center 2010). Perennial rivers include the Salt with three tributaries (Cherry Creek, Canyon Creek, and Cibecue). Based on my GIS analysis of the slopes of land throughout the watershed, opportunities for floodplain and/or irrigated farming were probably restricted and minimal in most places except along the Salt River near the western edge of the watershed (slopes of between 0 and 5% are considered optimal for irrigation; Walker 1989). Assessing opportunities for floodplain and irrigated farming along perennial rivers of the watershed, however, requires further on-site evaluation.
Contributions to the Prehistory of Central Arizona
It is important to consider vulnerability to dry periods during the 13th through 15th centuries so that we can advance our understanding of the complex social issues occurring during this time. These issues include rising warfare and the 15th century regional depopulation. Each involves population movement on a large scale and over long-time periods. Models of increasing conflict and warfare explain movements out of particular places on the landscape as defensive responses to the real or perceived threat of violence (LeBlanc and Rice 2001). When settled areas are abandoned and unoccupied zones around clusters of settlements are observed, these patterns are interpreted as an effort to create defensive open spaces between socially distant peoples (DeBoer 1981; LeBlanc 1999; Wilcox et al. 2001b; Wilcox and Haas 1994). Models of regional depopulation need to explain large-scale population movements out of a region (unless in situ demographic decline is argued, e.g., Hill et al. 2004). Our current understanding of the process in central Arizona is limited to rising warfare (Wilcox et al. 2001b), demographic decline associated with community coalescence (Hill et al. 2004), and general notions of breaches in regional carrying capacity (e.g., LeBlanc 1999, 2006). Along the Lower Salt River, a prevailing depopulation hypothesis is technological and social challenges to irrigated agriculture related to streamflow variability, flooding, and channel change in the late 1300s (Graybill et al. 2006; Nials et al. 1989). In contrast to environmentally focused hypotheses, Abbott and colleagues (Abbott 2003 ed.) have made a strong case for a gradual, centuries-long decline of populations along the Lower Salt. Causes for this decline are complex and include demographic instability, truncated trade networks, political strife, environmental impacts and ultimately ineffective responses to these challenges (Abbott 2003).
Advancing our understanding of both warfare and depopulation in central Arizona requires a better understanding of the factors influencing population movements. This study focuses on a frequently considered explanation for population movement--dry periods--and asks whether or not and under what conditions there is long-term evidence that dry periods were related to population movements out of settlements and watersheds in central Arizona. If dry-period influences on movements are detected over long time periods, then explanations of warfare and depopulation must accommodate climatic contributions to these phenomena. It should not be sufficient that we gloss climatic influences as simply context in explanations of the major events during this period in prehistory. For example, Wilcox et al. (2001b:165) argue the role of conflict in the depopulation of central Arizona and suggest that "Environmental perturbations could have further exacerbated such a process." No further suggestions of what these perturbations might be were provided. I am not suggesting dry periods as the sole cause of any social phenomenon. Rather, I am suggesting that if population movements constitute an integral part of a social phenomenon being explained and if these movements are strongly related dry periods, then explanations that do not consider potential dry-period impacts on movement are incomplete.
In sum, the cultural, environmental, and climatic diversity of central Arizona during the 13th through 15th centuries provide an important context for evaluating models of vulnerability to dry periods and furthering our understanding of dry-period influences on population movement. Population movements were an integral part of some of the critical social issues of this period. At the conclusion of this study (Chapter Ten), I discuss the implications of the findings of this research for understanding both the depopulation and models of increasing warfare in the region.