The 1930s Dust Bowl: Geoarchaeological lessons from a 20th century environmental crisis

Abstract

The Dust Bowl of the 1930s was an environmental crisis of epic proportions characterized by unprecedented soil erosion, population redistribution, and profound transformation of the landscape in the Great Plains region of the United States. Therefore, one might expect that this event left a mark in the sedimentary and geomorphological record of the region. This study is a field reconnaissance of sedimentary and landscape changes that can be linked to the event. To achieve this objective, the Dust Bowl is analyzed as a climatic event, a geomorphic event, and a socioeconomic event. The geoarchaeological approach taken here is based on that of the archaeology of the contemporary past (Harrison and Schofield, 2010), which incidentally coincides with the Anthropocene, as defined by Crutzen (2002). The results of this survey can be useful to those who search for geomorphic evidence of environmental crises in the more distant past.

Keywords

Anthropocene drought Dust Bowl geoarchaeology Great Plains industrial age

Introduction

Since the 1990s, archaeologists, historians, and geographers have increasingly turned their interest to understanding the processes of decline, collapse, and resilience of past civilizations. In part, this interest was stirred by the increasing awareness of current global environmental changes (Dearing et al., 2010; Tainter, 1988; Santos, 1999). The Classic Mayan civilization, the Norse Greenland settlement, and the Easter Island settlement, all of which have been popularized among the educated public by Jared Diamond (1997, 2005), are among the most sounded examples of collapse. Controversy, however, on the concept of civilizational collapse has risen in terms of how the archaeological, historical, and paleoecological record is interpreted and how complexity of societal and natural systems has been simplified in the popular and academic literature (Butzer and Endfield, 2012; McAnany and Yoffee, 2010; O’Sullivan, 2008).

Environmental crises have not ceased to exist in the Industrial Age. In fact, they occur more frequently and are better known. Examples of such crises include the Dust Bowl of the Great Plains, the Sahel droughts and famines of the late 20th century, the Mississippi Floods of 1993, and the devastation of New Orleans by Hurricane Katrina. Of these, the Dust Bowl of the 1930s is of great relevance in American history because it signals a tremendous environmental, societal, and economic transformation of the Great Plains, the heartland and breadbasket of the country (Figure 1). The event, often referred to as a catastrophe of epic proportions (Egan, 2006), is perhaps a characteristic environmental crisis of the Industrial Age.

Figure 1.

Dust Bowl region.

One can assume that the severity of the Dust Bowl of the 1930s left such an impact on the land that eventually the remains of this crisis will emerge in the archaeological record of the Great Plains. Artifacts and remains of the culture that prevailed previously and during the event can be used for dating the event in the future (Figure 2). Some of them can be associated with blowing sediment (Figure 2c and d), which may permit geoarchaeologists to relate the artifacts and buildings of the early 20th century to the Dust Bowl or another eolian event.

Figure 2.

Artifacts, features, and remains of material culture that could bear archaeological evidence to generations of future archaeologists. (a and b) Ruins of a former farmhouse with post–Dust Bowl additions and later abandoned, Cimarron County, Oklahoma; (c) vehicles and agricultural machinery, Cimarron County, Oklahoma; (d) building by the Works Administration Projects (WPA), dated to 1937, Baca County, Colorado; (e) barbed wire that can be dated using manufacture catalogues (e.g. Clifton, 1970); (f) buried fence by dunes, Woods County, Oklahoma; and (g) abandoned farmhouse with eolian sediments abutted to a wall, Baca County, Colorado.

Certainly, we don’t need an archaeological research project to learn about the Dust Bowl in the Great Plains because there are multiple and detailed records of the event, including meteorological data, photographs, first-hand accounts, government reports, and more. However, the archaeology of the contemporary past can offer the advantage of corroborating our interpretations with these records (Harrison and Schofield, 2010). Furthermore, a study of the Dust Bowl archaeologically and geoarchaeologically should be part of reconstructing the long-term legacies of land use change (Dearing et al., 2010), something that has become important in studies of past and present to build the ‘long now’ (Brand, 1999; Carpenter, 2002; Holling, 2001).

In this study, a reconnaissance of field evidence for the Dust Bowl of the 1930s was carried out in the hardest hit area of the crisis. This reconnaissance was done in the same way a geoarchaeologist would investigate field evidence of a past environmental crisis. The results revealed that climatic, geomorphic, and socioeconomic aspects intermingle to make the geoarchaeological record a very complex one. Furthermore, this exercise demonstrates not only aspects of geoarchaeological research but also contributes to understanding environmental crises and resilience in an increasingly globalized world.

Methods and information sources

This study is based on a research approach called ‘archaeology of the contemporary past’, applied by Harrison and Schofield (2010). However, the matters discussed here are mainly geoarchaeological, that is to say they relate to the evidence of the Dust Bowl event in sedimentary deposits, soils, and geomorphic features. Therefore, we view the Dust Bowl in the same manner geoarchaeologists view environmental crises of the remote past (historic and prehistoric). To achieve this objective, this study examines the evidence of the Dust Bowl environmental crisis from three separate perspectives: as a climatic event, as a geomorphic event, and as a socioeconomic event. The combination of the three views provides an integrated ecological perspective of the crisis.

Climatic data include the combined calculated Palmer Drought Severity Indices (PDSIs) obtained from meteorological records and reconstructed with tree-ring data, (Fye et al., 2003) and from National Oceanic and Atmospheric Administration records published by the High Plains Regional Climate Center (2013). Data used to supplement regional information were obtained from Schubert et al. (2004) and Cook et al. (2009).

Geomorphic data are scarce for the brief period encompassed by the Dust Bowl event, but studies focusing on late-Holocene geomorphic changes, particularly in sand dunes and river systems, that occurred in the 20th century are applicable. Changes in the geomorphic systems that occurred during and after the event have been singled out in numerous studies dealing with eolian deposits (Arbogast, 1996; Cordova et al., 2005; Forman et al., 2008; Halfen and Johnson, 2013; Muhs and Holliday, 1995; Rhinewald, 2005; Rogers, 2007; Werner et al., 2011), and stream behavior in the Dust Bowl region (Schumm and Lichty, 1963; VanLooy and Martin, 2005; Wilson, 1972). Maps of wind erosion by county in the Dust Bowl area compiled by Joel (1937) and digitized and analyzed via geographic information system by Cunfer (2011) were also consulted. Additional unpublished data gathered by the authors have been processed and presented here as support for the study. These include micro-stratigraphic information as well as data obtained from aerial photographs since the 1930s. Many of the photographs suggest aspects of resilience as demonstrated by geomorphological changes. The geomorphic examples discussed in this study refer mainly to localities in western Oklahoma, and a few others in southwest Kansas, and southeastern Colorado.

Socioeconomic information is more readily available than climatic and geomorphic information. The earliest studies to provide socioeconomic insight into the crisis (e.g. Johnson, 1947) draw heavily from the federal government’s document The Future of the Great Plains (Great Plains Committee, 1936). The overriding theme of this work, best illustrated by Worster’s (1979) seminal Dust Bowl book, is human failure to adapt to the limits imposed by nature. In these studies, the disaster manifests socioeconomically in the form of mass farm failures and widespread depopulation of the region. In review of Dust Bowl literature, one must be aware that this ‘declensionist’ narrative (Cronon, 1992) permeates many of the studies that examine the socioeconomic dimensions of the disaster (Porter, 2014). Even-handed, empirically driven assessments (e.g. Cunfer, 2005; Hurt, 1981; Riney-Kehrberg, 1994), although less common, are essential components of a socioeconomic assessment of the crisis. At the other end of the spectrum, authors such as Bonnifield (1979) build upon the themes of Walter Prescott Webb (1931) to report their findings from the perspective of human ‘progress’ (Cronon, 1992) and accomplishment over nature (Porter, 2014). Additionally, anecdotal studies (e.g. Egan, 2006; Svobida, 1986) help document changes in the land and its people.

More recently, however, a series of studies have addressed the Dust Bowl event in its broader extent, including more recent droughts experienced in the region. Porter (2012, 2014) set to evaluate the social impact through perceptions of local residents according to their generation, from those who are considered survivors to those of younger generations, their place of residence, and other demographic characteristics. Studies concerning adaptations, such as land reclamation led by the government, are useful for understanding processes of resilience (e.g. McLeman et al., 2008; Orlove, 2005; Parton et al., 2007). The establishment and administration of the National Grasslands (Hurt, 1985) and the Conservation Reserve Program (Dunn et al., 1993) provide interesting examples of this. Given the limitation in words for this paper, only general sets of data are included here.

The Dust Bowl as a climatic event

From the point of view of defining the 1930s Dust Bowl as a climatic crisis, it is important to address four of the main climatic aspects discussed in the literature: (1) the causes of the 1930s drought, (2) the spatial and temporal extent of the drought, (3) how it relates to other droughts before and after, (4) and the role of the drought in the development of the environmental crisis.

The drought that accompanied the Dust Bowl is often referred to as the 1930s drought or the Dust Bowl drought (Cook et al., 2014; Fye et al., 2003; Stahle et al., 2007; Woodhouse and Brown, 2001). Its relevance, in comparison with other droughts before and after, has been depicted in diagrams using instrumental and reconstructed values using the PDSI in a 10-year time series (Figure 3). The reconstructed PDSI is based on the calibrated values of tree rings, thus facilitating comparison of intensity and length of droughts across centuries. For details on the geographic provenience of the data, the calibration, and cumulative curve for each drought period, see Fye et al. (2003).

Figure 3.

(a) Instrumental and (b) reconstructed summer PDSI for the Dust Bowl Region and Northern Rockies and Plains.

Despite its intensity and length, the 1930s drought is not considered a megadrought, a term used to refer to major, multidecadal droughts in the past millennium (Fye et al., 2003; Stahle et al., 2007; Woodhouse and Overpeck, 1998). The most notable of these megadroughts occurred across the territories of the United States and Mexico in the 16th and 18th centuries (Stahle et al., 2007). Additionally, droughts in the 19th century were longer and more persistent than those occurring during the time of farming settlement (1890s, 1910s, 1930s, and 1950s) (Figure 3).

In the broader regional context of the Great Plains, the 1930s drought extended from 1928 to 1940 (Fye et al., 2003). The driest year of the drought was 1934, which has been singled out as the driest year on record for the region during the past millennium (Cook et al., 2014). This fact is evident in the regional PDSI data and precipitation records (Figures 3 and 4a). The year 1934 also registered the highest temperature in the High Plains in the 20th century (Figure 4). That record was recently surpassed by 2013, which was also a dry year.

Figure 4.

Temperature and precipitation in the High Plains during the 20th century. (a) High Plains annual precipitation departure from average (%). (b) High Plains average annual temperature departure from average (°C).

The causes of the 1930s drought have long been debated, but more recent studies using modeling and instrumental data from that time suggest the ‘low jet’, which powers the flow of moist air from the Gulf of Mexico into the Great Plains, is culpable on a regional scale (Helfand and Schubert, 1995). The reconstruction of circulation patterns for the 1930s suggests that the low jet flow was weaker and flowed farther south than normal (Brönnimann et al., 2009). On a broader, global scale, the 1930s drought has been associated with anomalies in the sea surface temperatures and developments associated with La Nina conditions in the Pacific (Cook et al., 2009; Schubert et al., 2004). Additional factors implicated in the persistence of the drought include land degradation and the effects of aerosol produced by the dust (Cook et al., 2009).

The picture of 1930s drought and its relation to the Dust Bowl event makes more sense when seen in the broader context of the climatic events before and after, namely, those from the 1890s to the present. It is noteworthy that between the end of the 1890s and 1930s, several climatic events had roles in shaping the effects of the 1930s drought. In particular, it is important to note the so-called Early Twentieth Century Pluvial (1905–1928), which provided an extended period of above normal precipitation for the Great Plains, only interrupted by the 4-year drought of the early 1910s. During this period, a massive expansion of farming on the Great Plains occurred.

Of certain importance is the 1950s drought, which in some parts of the Great Plains is also known as the 1950s Dust Bowl. The geographic extent of this drought was arguably larger than that of the 1930s (Fye et al., 2003), with the driest areas shifting from year to year. It mostly affected the southern and southwestern Great Plains and the southwestern United States. The driest years, particularly for the Great Plains, were the mid-1950s when dust was mobilized, although not at the scale seen in the 1930s. Nonetheless, an important distinction lies in the fact that less dust was mobilized during the 1950s, presumably because of the conservation practices that had been implemented in the wake of the 1930s event (Stout and Lee, 2003), but also because of dramatically increased irrigation from the Ogallala aquifer (Opie, 1993).

The Dust Bowl as a geomorphic event

There is no doubt that the Dust Bowl disaster led to a great mobilization of sediment and tremendous erosion in the Great Plains. In theoretical terms, a geoarchaeologist of the future would first look at the geomorphological and soil records of the region, particularly the eolian record. Also, because a great deal of soil erosion occurred on slopes and large amounts of sediment were carried by water, it would be important to look at the fluvial systems, as well.

Eolian dust and sand

There is the perception that the Dust Bowl event might be identified in the records by accumulations of dust. As noted in previous studies, the movement of particles relates to wind velocity, turbulence, and particle size (Pye, 1987). Dust particles which are usually smaller than 50 µm (from medium silt to clay) travel in suspension, while larger particles, mostly sand, travel by saltation (Figure 5).

Figure 5.

Mechanisms of dust and sand entrainment based on Pye (1987) as related to mechanisms of eolian erosion and transport of fine-grained soils and sandy soils during the Dust Bowl event.

In the case of the Dust Bowl event, large amounts of dust were produced from fine-grained soils that had no vegetative or crop cover, as well as from fine grains in soils that had developed on sand dunes. Based on Pye’s (1987) model, fine grains (i.e. dust) travel long distances. Although ‘dust rolled over the plains’ as described by locals who witnessed the event, the finer particles in suspension reached faraway destinations in the northeast and east of the United States during the dust storms in 1933 and 1934 (Mattice, 1935). In northwestern Minnesota, high-resolution varved deposits of Elk Lake do have evidence of increases in aluminum (Al) and sodium (Na) that has been dated to the Dust Bowl of the 1930s (Dean, 1997). Dust was reported to reach ships off the Atlantic coast (Egan, 2006). Even further, particles recorded in the Greenland Ice have been linked to the storms originating in the Great Plains (Donarumo et al., 2003).

Interestingly, however, is the fact that although the Dust Bowl event was recorded in the concentrations of Al and Na in varved sediments of Elk Lake, Minnesota, they appear only as a bump in the curve of concentrations of these minerals within the broader period of 1870 to 1970 (Dean, 1997). The aluminum and sodium concentration curve shows that a steady increase occurred between 1905 and 1919, and there were conspicuous peaks in the 1930s and 1950s (Dean, 1997). A close look at the distribution shows that the concentrations are slightly higher in the 1950s peak. Therefore, it is evident that the 1930s Dust Bowl can be identified in the record, embedded in a series of dust storm events related to the plowing of lands in the first two decades of the 1900s, and other drought events (e.g. the Dust Bowl of the 1950s).

Although dust accumulated locally, even inside houses and closed attics, most of the material that was left behind, burying houses and abandoned agricultural equipment, was sand. This is evident by the angle of repose of sediment accumulations visible in photographs of the 1930s (Figure 5). Some documentaries show people shoveling sand after dust storms (Burns, 2012; Tyrrell, 2008). The source of sand moved by saltation had several origins: the sand fraction of upland soils and/or older, destabilized dunes. In summary, fine particles in the soils would become part of the dust (suspension) while the sand grains would be mobilized by saltation. Therefore, it is important to look at the two main components, dust and sand, separately.

Dust evidently formed from areas with no vegetation protection. In theory, a drought would reduce this protective vegetation cover. However, after sod has been completely removed for crops, the amount of exposed sediments for deflation would increase. There is no doubt that the Dust Bowl event is known for its frequent, overwhelming dust storms. However, dust storms of the magnitude of the ones seen in the 1930s have been recorded earlier. Historical records compiled by James Malin (1946a, 1946b, 1946c) in Kansas provide a picture of the frequency and magnitude of dust storms since the 1850s. In some cases, descriptions of the storms are similar to those described by local residents in the Dust Bowl region in the 1930s (see Cunfer, 2008; Malin, 1946a, 1946b, 1946c).

Strong recurrent dust storm events concentrated in some years, particularly during the 1879–1881 period. This may be considered a Dust Bowl as well, but at the time there were fewer people living in the Great Plains and the events were not advertised because they would discourage immigrants from buying land in the region (Cunfer, 2008). In summary, large events like the 1879–1881 Dust Bowl have been eclipsed by the most recent 1930s Dust Bowl, where more inhabitants were affected.

Sand dune mobilization

In the areas where the Dust Bowl took place, some evidence of dune mobilization in association with human structures still exists (see Figure 2c and d). However, the best evidence left by the Dust Bowl event is in the sand dune fields of the region, some of which were active during the megadroughts of the previous centuries or in the middle Holocene (Cordova et al., 2005; Forman et al., 2008; Halfen and Johnson, 2013; Lepper and Scott, 2005).

Many of these fields had been stabilizing since the megadroughts of previous centuries, as revealed by historical accounts (Cordova et al., 2005). In general terms, the dune fields were not historically cultivated, but droughts and grazing pressure of cattle periodically removed the vegetation on stabilized (i.e. vegetated) dunes. For example, evidence of eolian activity was recorded in some localities of New Mexico, Texas, and Colorado in association with droughts in the 1820s, 1850s, 1890s, and 1930s (Muhs and Holliday, 1995).

The Little Sahara and Beaver dune fields in northwestern Oklahoma became more active (Cordova et al., 2005). The comparison of their extent between the end of the Dust Bowl and 60 years later shows how much the fields stabilized (Figure 6). The trend suggests that by the beginning of the 21st century, the sand sheets would have disappeared if it had not been for human-led destabilization as a result of recreational all-terrain vehicle riding in Beaver Dunes State Park and Little Sahara State Park (Figure 7). Thus, the destabilization of eolian sand landforms during the 1930s drought was only a small aberration in the ongoing process of long-term dune stabilization begun in previous centuries.

Figure 6.

Post-1930s sand dune vegetation resilience in aerial photography Comparison of the extent of sand dune activity in Beaver Dunes, Oklahoma, in 1942 and 1995, and Little Sahara State Park, in 1937 and 1995.

Figure 7.

Post-1930s sand dune vegetation resilience in measured area. Decline of bare sand surfaces after the 1930s Dust Bowl event in three dune fields.

Rather than extensive destabilization of sand dune fields, the Dust Bowl event was characterized more by the formation of blowout dunes (Arbogast, 1996; Rogers, 2007). This mainly occurred in areas where disturbance had occurred (Figure 8a). It is important to point out that many of the blowouts were formed on older, stabilized sand dunes, some of which had been formed during the megadroughts of past centuries. In the locality represented in Figure 8a, the Tivoli generation has been OSL dated to around 213 ± 23 BP and the Eda generation between 831 ± 60 and 936 ± 60 BP (Cordova et al., 2005). In general, the case depicted in Figure 8a epitomizes the stratigraphic complexity of sand deposits of various ages at one place. This shows how difficult it would be to single out sand mobilized by a single uni-decadal event like the Dust Bowl.

Figure 8.

Examples of recent dynamics in eolian and alluvial localities in northwestern Oklahoma. (a) Sand dune dissected showing several generations of eolian activity at one place at a locality north of Woodward, Oklahoma. Stratigraphic study based on Cordova et al. (2005). (b).Upstream incision in Tesesquite Creek. Incision began with an unusual flood event in the lower reaches in 1914 and continues upstream today. Left: Photograph taken in 1968 (Wilson, 1972) and the same location in 2004 (the authors).

Another development that characterized the eolian activity during the Dust Bowl was the entrainment of sand from the dry beds of the large rivers, which was reported by Melton (1940). The rapid sedimentation and widening of the alluvial channel plains of the large rivers provided large amounts of sand to be entrained by the strong winds that characterized the drought of the 1930s. However, as evident in the aerial photographs, the dry channels did not provide sand for the large dune fields on the banks (Figure 6). Instead, as observed by Melton (1940), they formed long dunes bordering the channel. These dunes have become vegetated and recognized in the soil surveys as the Lincoln series (Scott, 1989).

The Dust Bowl in the eolian sand record

Although eolian sand processes were active practically everywhere in the Dust Bowl region, the eolian deposits that can be attributed to the event are hard to find in the stratigraphy of many regional studies. In many cases, the resolution of the record suggests that sand accumulation deposits can be attributed to the Dust Bowl (Forman et al., 2005, 2008; Halfen and Johnson, 2013; Lepper and Scott, 2005). However, some deposition may not necessarily be related to the Dust Bowl event. Each dune field has its own conditions, sometimes occurring under non-climatic controls (Halfen and Johnson, 2013; Werner et al., 2011). Varied methods of research and dating, as well as sampling, are additional variables that can interfere with the correlation of data.

Stratigraphic complexity in the eolian record, particularly of later historical periods, can be seen practically at any locality within the fields. Figure 8a provides an example of how each generation of dunes cannibalizes older dunes. The 20th century layer, which is likely to have been the result of the 1930s Dust Bowl, occurs because of a local blowout. At the same location, destabilization by road work created another blowout in the past decade.

Fluvial changes

Large rivers in the Great Plains have undergone modifications through the 20th century. Many of these modifications cannot be directly linked to the 1930s drought. In fact, many of the changes predate the Dust Bowl, as is the case of the Cimarron River, as shown in a study of its channel in Kansas by Schumm and Lichty (1963) and a follow-up study by VanLooy and Martin (2005). Reportedly, at the turn of the century, the Cimarron had a single channel meandering on a grassy floodplain, which after a major flood in 1914 changed to an anastomosed (braided) stream (Schumm and Lichty, 1963). Further widening of the anastomosed channel occurred in 1942, after a high-flood event. The year 1914 was one of excessive rainfall that followed a 3-year drought period (the 1910s drought), and similarly, 1942 was an excessively wet year after the decade-long drought (the 1930s Dust Bowl drought). But after the 1940s, the channel began to become more stable as peak floods decreased, allowing riparian vegetation to colonize the anastomosed alluvial plain and leading to channel narrowing (VanLooy and Martin, 2005). The reduced peak floods of this period are attributed to the widespread construction of erosion dams in the watersheds, a conservation measure implemented by the federal government after the Dust Bowl. One can see the differences in a braided stream in the late 1930s and early 1940s on aerial photographs (Figure 6).

Smaller streams underwent changes, but in most cases, the modifications consisted of increased stream incision. In Cimarron County, Oklahoma, in the core of the Dust Bowl region, Tesesquite Creek’s incision related in part to the same weather phenomena that caused the changes in the Cimarron River. Tesesquite Creek is a tributary of the Cimarron (called at that stretch ‘the Dry Cimarron’). A historical and geomorphological reconstruction of the incision process has been documented by Wilson (1972) who was also able to link the beginning of channel incision to the extreme rains of 1914. In the 1960s, the process was still occurring, evidenced by headward erosion between 1964 and 1970. In fact, the Tesesquite Creek knickpoint is about 1 km upstream from its location in the late 1960s. Photographs published by Wilson (1972) can be compared with the modern channel (Figure 8b).

Although the sedimentation and channel anastomosis in the Cimarron River and stream incision in Tesesquite Creek were initiated by extreme weather events in the 1910s, the role of land use in both cases was also important in accelerating the process (Schumm and Lichty, 1963; VanLooy and Martin, 2005; Wilson, 1972). Other cases in Oklahoma, such as streams feeding into the Washita River, seem to have undergone channel entrenchment as a consequence of extensive farming and overgrazing during initial settlement in the late 1800s and early 1900s (Johnson et al., 1980). There is no evidence that the Dust Bowl caused major changes, except for the case of the 1942 flood, which occurred during an excessive-rainfall event over an area where vegetation cover had been reduced by the 1930s drought.

In summary, no clear sedimentary evidence exists in the streams of the Dust Bowl region that can be associated directly to the 1930s drought because processes of change and sedimentation began to occur earlier. The same can be said for channel entrenchment in tributaries; there is no direct link between channel braiding and channel entrenchment triggered by the 1930s drought. There is also no evidence that the drying of the channels of the large rivers contributed sand to the dune fields. Rather, the development of sand mobilization has often been reported as sand being cannibalized from older dunes (Cordova et al., 2005; Lepper and Scott, 2005; Scott, 1989).

The Dust Bowl as a socioeconomic event

Centuries of human interaction with the environment intensified between 1850 and 1930 as farmers’ cultivation of the Great Plains expanded at the expense of the native grasslands. Increased moisture during the Early Twentieth Century Pluvial (1905–1928) even led some inhabitants to believe they had overcome the area’s variable weather and climate. The 1930s disaster provided a harsh rebuttal, and demonstrated once more that the Great Plains is a region of boom and bust, ebb and flow. The 1930s event was simply another chapter, albeit a harsh one, in the cyclical history of Great Plains weather and climate. A new regime of land use (expansive monoculture) combined with the 1930s drought and broader national economic malaise to create a particularly adverse socioeconomic episode for the region. Many people migrated from the Dust Bowl region, but many more survived the dust and the drought by forging new community ties and by embracing new government programs.

Migration and economic transformation in the long run

There is wide perception, influenced by popular readings, movies, and the media, that the 1930s Dust Bowl had a disastrous effect on the population of the Great Plains (Porter, 2014). However, many authors have illustrated that the Dust Bowl population and migration story is far more complex than the stereotypical Okie migrants headed to California, like Tom Joad in Steinbeck’s The Grapes of Wrath. For example, Bonnifield (1979) provides data that show businesses and schools in towns remained open for the most part, and that only banks were reduced in number, a matter that was more the result of the Great Depression. A number of authors (e.g. Clements, 1938; Gregory, 1991 [1989]; Lawson and Baker, 1981; Rathge and Highman, 1998) have shown that outmigration from the Dust Bowl region predated the crisis, as increased mechanization began to replace human labor.

Since the start of the 20th century, the population has decreased by more than 60% in 98 counties of the Great Plains (Figure 9). Within this broader regional transformation lies the more localized outmigration of the 1930s Dust Bowl event. Despite the accelerated migration reflecting the cumulative effects of farm foreclosures, unemployment because of crop failure, and the systematic impacts of the Great Depression, when the long-term depopulation of the region is viewed (1900 to the present), the depopulation of Dust Bowl counties does not stand out. Furthermore, regional population decline can be deceptive, and thus it is better to think of ‘population redistribution’ rather than population collapse (Rathge and Highman, 1998). While most rural counties’ populations have declined considerably, some metropolitan ones have gained considerable numbers (Gutmann et al., 2005; Parton et al., 2007). As illustrated by Figure 9, the overall picture of long-term Great Plains depopulation shows that the environmental crisis of the 1930s did not have a disproportionately negative impact on the population totals of the historic Dust Bowl region.

Figure 9.

Great Plains counties with population decline between 1900 and 2010. The Dust Bowl area is indicated.

The causes of outmigration through the 20th century, outside the 1930s, show no correlation with adverse climatic conditions like droughts and dust storms (Parton et al., 2007). Population data show that it is the young cohorts of the Plains population that have been most likely to leave the depopulating counties (Rathge and Highman, 1998). The highest losses, occurring in later periods, particularly after 1990, are in the more farm-dependent counties where unemployment and a lack of amenities are the main reason for young people to leave (Gutmann et al., 2005; McGranahan and Beale, 2002). Overall, the picture of depopulation before and after the Dust Bowl shows that the environmental crisis of the 1930s had little impact on the population characteristics of the region after the Dust Bowl. A study of the impact of the four drought decades in the Great Plains (1890s, 1910s, 1930s, and 1950s) showed that in terms of impacts on abandonment, the 1890s drought was most effective in depopulating the region, followed by the 1910s, the 1930s, and 1950s (Meyer et al., 1998; Warrick, 1980). In fact, in terms of duration and intensity, the droughts of the 1930s and 1950s were worse than those of the 1890s and 1910s (Figure 2). Meyer et al. (1998) interpret the lesser impact of the droughts on the 1930s and 1950s farmers as an adaptation via technological changes, institutional development, market integration, and an increased governmental role in crisis relief.

On a different approach to the environmental causes of post-1930s outmigration, Deane and Gutmann (2003) ran a time-series model to evaluate the effect of environmental deterioration on population numbers. The output of the model revealed a short lag between dust storm peaks and population decline. Although this trend contrasts with the no-trend between population decline and drought discussed above, it is important to remember that dust storm events are a common appearance in the Great Plains regardless of extended drought periods (see Cunfer, 2008; Lee et al., 1993; Malin, 1946a, 1946b, 1946c).

On the economic side, although productivity decreased considerably during the 1930s, the improved weather of the 1940s combined with new technological and institutional measures to allow agriculture and other industrial production to increase (Bonnifield, 1979; Meyer et al., 1998). In fact, production increase has been disproportional to population increase, as outmigration has not adversely impacted farming because of ever-increasing mechanization and efficiency (Parton et al., 2007).

Discussion

Time-transgressive nature of geomorphic processes

The eolian and alluvial records, as well as landform change recorded in photographs, suggest that geomorphic responses to the 1930s Dust Bowl are ambiguous in part for two reasons. First, the resolution of the sedimentary records to document that period is low. Second, geomorphic systems were still responding to drought events before and after the Dust Bowl and/or to human disturbances of vegetation and soil by farming settlement since the late 19th century.

Direct response to the 1930s drought occurred relatively quickly in terms of production of dust and the development of blowouts, both of which have practically no representation in the stratigraphy. The fluvial systems are still in the process of re-adjustment, as demonstrated by the cases of Tesesquite Creek and the Cimarron River floodplain. Nonetheless, processes of channel transformation are not solely because of climatic shifts, but also a series of other factors such as extraction of water from the upper aquifers and the construction of dams (Schumm and Lichty, 1963; VanLooy and Martin, 2005). In fact, one reason why the channel of the Cimarron River, as well as other rivers, has been modified after the Dust Bowl to a more braided and vegetated stream is related to the construction of flood dams (VanLooy and Martin, 2005), and not particularly to changes in precipitation.

In summary, the time-transgressive nature of the geomorphic responses to climatic and human changes at a centennial and multidecadal scale makes the effects of the 1930s Dust Bowl (barely a decade-long event) difficult to discern in the stratigraphic and geomorphic records of the region. Therefore, it is advantageous to look at the events from a longer term, more holistic perspective to see how general processes in the development of the Dust Bowl crisis manifested (Figure 10). Then one can appreciate that the Dust Bowl is but one event in a series of events of landscape transformation.

Figure 10.

Schematic sequential model of destabilization of sand dunes and fine deposits in the Dust Bowl area in a series of episodes of a major event since the 1850s.

Resilience after the crisis

Measures taken by the government after the disastrous erosion occurred on the soils of the Great Plains during the Dust Bowl included the implementation of less destructive plowing techniques and the restoration of areas damaged by erosion (Worster, 1979). Restoration took place on lands abandoned by farmers, purchased by the government, and turned into grasslands. Most of these lands were marginal and not apt for re-cultivation; they included eolian sands and shallow soils on Ogallala gravels and calcretes. Based on empirical and experimental data, the natural resilience of grasslands after drought and erosion takes a long time (Judd, 1974; Weaver and Albertson, 1956). For this reason, grassland restoration in lands in possession of the federal government took place through re-seeding between 1940 and 1956 (Laycock, 1988). This suggests that many of the restored grasslands are not the result of natural resilience, as is the case of many cases in the archaeological record of the past, but through government-sponsored projects.

Since 1953, these grasslands fell within the institutional management of the US Forest Service, which regulates their uses (limited grazing, tourism, and research). Although government management has faced criticism from ranchers and environmentalists alike, monitoring demonstrates a marked decrease of dust (as dust storms and blowing dust) in the Great Plains. This is most likely the result of improved land management practices (Lee and Tchakerian, 1995; Stout and Lee, 2003). Even years with low PDSI show diminished amounts of blowing dust (Lee et al., 1993).

Federal or institutional intervention played an important role in the resilience process after the environmental crisis of the 1930s Dust Bowl. In the case of other environmental crises of a larger regional magnitude and duration (e.g. the Sahel Droughts since the 1960s) and more localized short-term events like Hurricane Katrina, national and international aid has provided relief, facilitated resilience, and in some cases facilitated mitigation. This is one aspect that contrasts modern environmental crises from those in the past (e.g. Classic Maya, Viking Greenland, or Easter Island environmental crises) where such a degree of national or international cooperation did not exist (Orlove, 2005).

Collapse, decline, or continuity?

The Dust Bowl of the 1930s did not mark the end of farming on the Great Plains. Despite being arguably the worst environmental disaster in the history of the United States, farming did not collapse. In fact, Great Plains farming has seen better days in the post–Dust Bowl era, during which time farming evolved into a more mechanized, irrigated, and state-subsidized enterprise. The extraction of water from the Ogallala aquifer sustains much of the agriculture of the area affected by the Dust Bowl.

Migration, during and after the event, is another aspect that has to be addressed when thinking of the environmental crisis of the 1930s Dust Bowl. Urban and suburban populations in the Dust Bowl region belie the common perception of Great Plains population and have generally increased (Rathge, 2003). However, it is the rural depopulation that has captured the attention of popular culture and academics alike. As data cited above suggest, migration did occur during the Dust Bowl of the 1930s, but rural population decline in the long-run diminishes the acclaimed exodus of the 1930s, since it began earlier and continues to this day.

This rural depopulation is reflected in the material culture of the cultural landscape. Ghost and semi-ghost towns in western Oklahoma, for example, rarely stand as testament to the Dust Bowl. Some had become ghost towns before the 1930s, and many more developed in the post–Dust Bowl decades. One can see the 1950s and 1960s architectural styles in many of the abandoned buildings in towns. It is more common to see abandoned farmhouses that date to that period, particularly in some of the cattle ranches and the National Grasslands. Very commonly, the grassy landscape is interrupted by vestiges of the past human landscape such as groves of honey locust, western elm, and sumac trees.

The Dust Bowl event in the context of the environmental crises of the past

When interpreting the impact of environmental crises on past populations, archaeologists look at the societal dynamics involved in the process of societal collapse or decline. Resilience and continuity of a society after a crisis is sometimes evaluated (Butzer, 2012). In many cases, aspects of population dispersal and land management are also evaluated (Meyer et al., 1999; Orlove, 2005).

When interpreting societal dynamics and environmental crises in the Paleoanthropocene (pre-1850s according to Foley et al., 2013), archaeologists focus on a series of aspects reconstructed from remains of a material culture and, if lucky, of historical records. Depending on the society, archaeologists may search for answers in paleoenvironmental records as well, including eolian and alluvial sediments and landforms.

Correlations to regional or global changes, often contained in isotopic records in deep sea cores and ice sheets, are often clear in a long perspective. For example, the case of the collapse of the Pre-Pottery Neolithic in the Near East, which is often associated with the 8.2 ka event, is something that has been deduced by comparing paleoclimatic and archaeological records in an area, as well as through modeling. A more recent environmental crisis, the decline and abandonment of Norse settlements in Greenland, is a case explained through local records, as well as regional ones from Europe, where the ‘Little Ice Age’ has been better studied.

In the case of the Dust Bowl, the climatic event had a stronger relationship to changes in the sea surface temperatures of the Atlantic and Pacific oceans (Cook et al., 2009; Schubert et al., 2009). At a broader scale, however, the Dust Bowl event does not bear clear relation to the warming that has characterized the Industrial Age. It may be erroneous then to assign the causality of the event to global warming trends in the 20th century. This is a cautionary note for researchers deliberately assigning an event recorded in the archaeological record to particular global climatic events.

As the case of the Dust Bowl shows, population decline associated with environmental crises of the past, and usually reflected in the abandonment of sites, can be deceptive. Without individual records on where people went and why they left, it is difficult to link population decline to collapse. A major shift in population, over a period of years or decades, does not necessarily equate to economic collapse.

However, using the Dust Bowl event as an analogy to explain human ecological dynamics during environmental crisis may not be appropriate. After all, the Industrial Age has provided a different way for coping with crises, namely, state assistance for recovery and conservation, better communication, and many of the other benefits provided by the globalized world. Yet, it can still serve as an example to question the conclusions that we build when we interpret past environmental crises out of archaeological and geoarchaeological records.

Conclusion

Despite being an environmental catastrophe of epic proportions in an economic and historical perspective, the Dust Bowl event in its geomorphic and climatic forms was not really extraordinary. The drought was not the worst in the long-term record; eolian erosion and deposition events have been occurring for centuries before the extensive plowing of the 20th century, but it drew public attention because of its impact on population and agriculture, which is an important aspect in its significance as an environmental crisis.

If one thinks of a hypothetical future when archaeologists dig archaeological sites of the early Industrial Era (i.e. early Anthropogene, sensu Crutzen, 2002) in the Great Plains, they may find remains of material culture (e.g. Figure 2) that may be associated with the historical records of the Dust Bowl, if preserved. Archaeologists may associate some sites (e.g. farms and abandoned towns) with depopulation, but they would not be able to link it directly to the geomorphic and stratigraphic records in which the 1930s Dust Bowl would be discernible and in fact perhaps totally invisible. In essence, wind erosion and sand mobilization that occurred during the 1930s Dust Bowl would be events welded together with other events in a long-term process of geomorphic evolution of the landscape, spanning 1000 years or longer. Furthermore, many of the processes that occurred during the 1930s may have already started earlier, as is the case of dust and sand mobilization, while other processes may still be ongoing, as is the case of headward channel erosion and incision in some arroyos.

The lesson learned for archaeologists and practitioners of geoarchaeology is that many environmental crises of the ancient past may not necessarily be directly linked to a climatic or geomorphic crisis very easily. Even worse, one can wrongly assign socioeconomic events such as farm abandonment to time-transgressive processes of climatic and geomorphic change. Perhaps a look at some of the modern environmental catastrophes like the Dust Bowl or the Sahel Droughts, for which we have more accurate data, should provide a guidance for interpreting past historic and prehistoric environmental crises.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Arbogast

(1996) Stratigraphic evidence for late-Holocene aeolian sand mobilization and soil formation in south-central Kansas, USA. Journal of Arid Environments 34: 403–414.

Bonnifield

(1979) The Dust Bowl: Men, Dirt, and Depression. Albuquerque, NM: University of New Mexico Press.

Brand

(1999) Clock of the Long Now: Time and Responsibility. New York: Basic Books.

Brönnimann

Stickler

Griesser

. (2009) Exceptional atmospheric circulation during the ‘Dust Bowl’. Geophysical Research Letters 36: 1–6.

Burns

(Director) (2012) The Dust Bowl (Documentary). Walpole, NH: Florentine Films.

Butzer

(2012) Collapse, environment, and society. Proceedings of the National Academy of Sciences of the United States of America 109: 3632–3639.

Butzer

Endfield

(2012) Critical perspectives on historical collapse. Proceedings of the National Academy of Sciences of the United States of America 109: 3628–3631.

Carpenter

(2002) Ecological futures: Building an ecology of the long now. Ecology 83: 2069–2083.

Clements

(1938) Climatic cycles and human populations in the Great Plains. The Scientific Monthly 47: 193–210.

10.

Clifton

(1970) Barbs, Prongs, Pricks, Points & Sticks: A Complete and Illustrated Catalogue of Antique Barbed Wire. Norman, OK: University of Oklahoma Press.

11.

Cook

Miller

Seager

(2009) Amplification of the North American ‘Dust Bowl’ drought through human-induced land degradation. Proceedings of the National Academy of Sciences of the United States of America 106: 4997–5001.

12.

Cook

Seager

Smerdon

(2014) The worst North American drought year of the last millennium: 1934. Geophysical Research Letters. Epub ahead of print 17 October. DOI: 10.1002/2014GL061661.

13.

Cordova

Porter

Lepper

. (2005) Preliminary assessment of sand dune stability along a bioclimatic gradient, north central and northwestern Oklahoma. Great Plains Research 15: 227–249.

14.

Cronon

(1992) A place for stories: Nature, history and narrative. The Journal of American History 78: 1347–1376.

15.

Crutzen

(2002) Geology of mankind. Nature 415: 23.

16.

Cunfer

(2005) On the Great Plains: Agriculture and Environment. College Station, TX: Texas A&M University Press.

17.

Cunfer

(2008) Scaling the dust bowl. In: Knowles

(ed.) Placing History on Maps, Spatial Data, and GIS are Changing Historical Scholarship. Redlands, CA: ESRI Press,pp. 95–122.

18.

Cunfer

(2011) The southern Great Plains wind erosion maps of 1936. Agricultural History 85: 540–559.

19.

Dean

(1997) Rates, timing, and cyclicity of Holocene eolian activity in north-central United States: Evidence from varved lake sediments. Geology 25: 31–34.

20.

Deane

Gutmann

(2003) Blowin’ down the road: Investigating bilateral causality between dust storms and population in the Great Plains. Population Research and Policy Review 22: 297–331.

21.

Dearing

Braimoh

Reenberg

. (2010) Complex land systems: The need for long time perspectives to assess their future. Ecology and Society 15: 1–21.

22.

Diamond

(1997) Guns, Germs and Steel: The Fate of Human Societies. New York: W.W. Norton.

23.

Diamond

(2005) Collapse: How Societies Choose to Fail or Succeed. New York: Penguin Books.

24.

Donarumo

Ram

Stoermer

(2003) Possible deposit of soil dust from the 1930’s US dust bowl identified in Greenland ice. Geophysical Research Letters 30: 1–4.

25.

Dunn

Stearns

Guntenspergen

. (1993) Ecological benefits of the conservation reserve program. Conservation Biology 7: 132–139.

26.

Egan

(2006) The Worst Hard Time: The Untold Story of Those Who Survived the Great American Dust Bowl. New York: Houghton Mifflin Harcourt Company.

27.

Foley

Gronenborn

Andreae

. (2013) The Paleoanthropocene – The beginnings of anthropogenic environmental change. Anthropocene 3: 83–88.

28.

Forman

Marin

Gomez

. (2008) Late Quaternary aeolian depositional record for southwestern Kansas: Landscape sensitivity to droughts. Palaeogeography, Palaeoclimatology, Palaeoecology 265: 107–120.

29.

Forman

Marin

Pierson

. (2005) Aeolian sand depositional records from western Nebraska: Landscape response to droughts in the past 1500 years. The Holocene 15: 973–981.

30.

Fye

Stahle

Cook

(2003) Paleoclimatic analogs to twentieth-century moisture regimes across the United States. Bulletin of the American Meteorological Society 84: 901–907.

31.

Great Plains Committee (1936) The Future of the Great Plains. Washington, DC: Government Printing Office.

32.

Gregory

(1991 [1989]) American Exodus: The Dust Bowl Migration and the Okie Culture in Oklahoma. New York: Oxford University Press.

33.

Gutmann

Deane

Lauster

. (2005) Two population-environment regimes in the Great Plains of the United States, 1930. Population and Environment 27: 191–225.

34.

Halfen

Johnson

(2013) A review of Great Plains dune field chronologies. Aeolian Research 10: 135–160.

35.

Harrison

Schofield

(2010) After Modernity: Archaeological Approaches to the Contemporary Past. Oxford: Oxford University Press.

36.

Helfand

Schubert

(1995) Climatology of the simulated Great Plains low-level jet and its contribution to the continental moisture budget of the United States. Journal of Climate 8: 784–806.

37.

High Plains Regional Climate Center (HPRCC) (2013) Climate Change on the Prairie: A Basic Guide to Climate Change in the High Plains Region – Update. Lincoln, NE: HPRCC. Available at: http://www.hprcc.unl.edu/publications/files/HighPlainsClimateChangeGuide.pdf (accessed 17 May 2015).

38.

Holling

(2001) Understanding the complexity of economic, ecological, and social systems. Ecosystems 4: 390–405.

39.

Hurt

(1981) The Dust Bowl: An Agricultural and Social History. Chicago, IL: Nelson-Hall.

40.

Hurt

(1985) The national grasslands: Origin and development in the Dust Bowl. Agricultural History 59: 246–259.

41.

Joel

(1937) Soil Conservation Reconnaissance Survey of the Southern Great Plains Wind-Erosion Area. USDA Technical Volume 556. Washington, DC: United States Department of Agriculture.

42.

Johnson

(1947) Heaven’s Tableland: The Dust Bowl Story. New York: Farrar, Straus and Giroux.

43.

Johnson

Cantwell

Dean

(1980) The impact of agricultural settlement on Canadian Sandy Creek, Oklahoma. Proceedings of the Oklahoma Academy of Sciences 60: 82–88.

44.

Judd

(1974) Plant succession of old fields in the Dust Bowl. The Southwestern Naturalist 19: 227–239.

45.

Lawson

Baker

(1981) The Great Plains: Perspectives and Prospects. Lincoln, NE: University of Nebraska Press.

46.

Laycock

(1988) History of grassland plowing and grass planting on the Great Plains. In: Mitchel

(ed.) Impacts of the Conservation Reserve Program in the Great Plains (USDA Forest Service Technical Report RM-158). Fort Collins, CO: Rocky Mountain Forest and Range Experiment Station, pp. 3–8.

47.

Lee

Tchakerian

(1995) Magnitude and frequency of blowing dust on the Southern High Plains of the United States, 1947–1989. Annals of the Association of American Geographers 85: 684–693.

48.

Lee

Wigner

Gregory

(1993) Drought, wind, and blowing dust on the southern High Plains of the United States. Physical Geography 14: 56–67.

49.

Lepper

Scott

(2005) Late-Holocene aeolian activity in the Cimarron River valley of west–central Oklahoma. Geomorphology 70: 42–52.

50.

McAnany

Yoffee

(2010) Why we question collapse and study human resilience, ecological vulnerability, and the aftermath of empire. In: McAnany

Yofee

(eds) Questioning Collapse: Human Resilience, Ecological Vulnerability and the Aftermath of Empire. Cambridge: Cambridge University Press, pp. 1–20.

51.

McLeman

Mayo

Strebeck

. (2008) Drought adaptation in rural eastern Oklahoma in the 1930’s: Lessons for climate change adaptation research. Mitigation and Adaptation Strategies for Global Change 13: 379–400.

52.

Malin

(1946a) Dust storms, 1850–1900. Kansas Historical Quarterly 14(2): 129–144.

53.

Malin

(1946b) Dust storms, 1850–1900. Kansas Historical Quartery 14(3): 265–296.

54.

Malin

(1946c) Dust storms, 1850–1900. Kansas Historical Quarterly 14(4): 391–413.

55.

Mattice

(1935) Dust storms, November 1933 to May 1934. Monthly Weather Review, February, pp. 53–55.

56.

Melton

(1940) A tentative classification of sand dunes its application to dune history in the southern High Plains. The Journal of Geology 48: 113–174.

57.

Meyer

Butzer

Downing

. (1998) Reasoning by analogy. In: Rayner

Malone

(eds) Human Choice and Climate Change. Columbus, OH: Battelle Press, pp. 217–289.

58.

Muhs

Holliday

(1995) Evidence of active dune sand on the Great Plains in the 19th century from accounts of early explorers. Quaternary Research 43: 198–208.

59.

Opie

(1993) Ogallala: Water for a Dry Land. Lincoln, NE: University of Nebraska Press.

60.

Orlove

(2005) Human adaptation to climate change: A review of three historical cases and some general perspectives. Environmental Science & Policy 8: 589–600.

61.

O’Sullivan

(2008) The ‘collapse’ of civilizations: What paleoenvironmental reconstruction cannot tell us, but anthropology can. The Holocene 18: 45–55.

62.

Parton

Gutmann

Ojima

(2007) Long-term trends in population, farm income, and crop production in the Great Plains. Bioscience 57: 737–747.

63.

Porter

(2012) Lessons from the Dust Bowl: Human-environment education on the Great Plains. Journal of Geography 111: 127–136.

64.

Porter

(2014) What was the Dust Bowl? Assessing contemporary popular knowledge. Population and Environment 35: 391–416.

65.

Pye

(1987) Aeolian Dust and Dust Deposits. London: Academic Press.

66.

Rathge

Highman

(1998) Population change in the Great Plains. Rural Development Perspectives 13: 19–26.

67.

Rathge

(2003) The Changing Population Profile of the Great Plains. Fargo, ND: North Dakota State Data Center, North Dakota State University.

68.

Rhinewald

(2005) Dune stabilization in the Little Sahara and Beaver Dunes State Parks in relation to climate change and recreational activities. Master’s Thesis, Oklahoma State University.

69.

Riney-Kehrberg

(1994) Rooted in Dust. Lawrence, KS: University Press of Kansas.

70.

Rogers

(2007) Morphology of fixed sand dunes for paleowind reconstruction in Northwestern Oklahoma. Master’s Thesis, Oklahoma State University.

71.

Santos

(1999) The Environmental Crisis. Westport, CT: Greenwood Press.

72.

Schubert

Suarez

Pegion

. (2004) On the cause of the 1930s Dust Bowl. Science 303: 1855–1859.

73.

Schumm

Lichty

(1963) Channel Widening and Flood-Plain Construction along Cimarron River in Southern Kansas. Professional paper 352-D. Washington, DC: United States Geological Survey.

74.

Scott

(1989) Aeolian modification of Pleistocene terraces along the Cimarron River in Major County, Oklahoma. Master’s Thesis, Oklahoma State University.

75.

Stahle

Fye

Cook

. (2007) Tree-ring reconstructed megadroughts over North America since A.D. 1300. Climatic Change 83: 133–149.

76.

Stout

Lee

(2003) Indirect evidence of wind erosion trends on the Southern High Plains of North America. Journal of Arid Environments 55: 43–61.

77.

Svobida

(1986) Farming the Dust Bowl. Lawrence, KS: University Press of Kansas.

78.

Tainter

(1988) The Collapse of Complex Societies. New York: Cambridge University Press.

79.

Tyrrell

(Producer) (2008) When Weather Changed History: Dust Bowl (Documentary). Atlanta, GA: The Weather Channel.

80.

US Census Bureau (2015a) Country Population Census Counts, 1900–1910: USA Counties. Washington, DC: US Census Bureau. Available at: http://www.census.gov/population/www/censusdata/cencounts/index.html (accessed 2 May 2015).

81.

US Census Bureau (2015b) USA Counties. Washington, DC: US Census Bureau. Available at: http://censtats.census.gov/usa/usa.shtml (accessed 2 May 2015).

82.

VanLooy

Martin

(2005) Channel and vegetation change on the Cimarron River, Southwestern Kansas, 1953–2001. Annals of the Association of American Geographers 95: 727–739.

83.

Warrick

(1980) Drought in the Great Plains: A case study of research on climate and society in the USA. In: Ausbel

Biswas

(eds) Climatic Constraints and Human Activities. Oxford: Pergamon Press, pp. 93–123.

84.

Weaver

Albertson

(1956) Grasslands of the Great Plains. Lincoln, NE: Johnson Publishing Company.

85.

Webb

(1931) The Great Plains. Boston, MA: Ginn & Company.

86.

Werner

Mason

Hanson

(2011) Non-linear connections between dune activity and climate in the High Plains, Kansas and Oklahoma, USA. Quaternary Research 75: 267–277.

87.

Wilson

(1972) Geomorphology of Tesesquite Creek valley, Cimarron County, Oklahoma. Oklahoma Geology Notes 32: 195–208.

88.

Woodhouse

Brown

(2001) Tree-ring evidence for Great Plains drought. Tree-Ring Research 57: 89–103.

89.

Woodhouse

Overpeck

(1998) 2000 years of drought variability in the Central United States. Bulletin of the American Meteorological Society 79: 2693–2714.

90.

Worster

(1979) Dust Bowl: The Southern Plains in the 1930’s. Oxford: Oxford University Press.