INTRODUCTION

In the known histories of the impact of human intervention on the landscape, that of the Great Plains of the United States is among the most frequently described. In the 1930s, clouds of dust rose off the recently plowed land to catch the attention of media and politicians as far away as Washington, DC. The Dust Bowl permanently focused attention on the ways that human interaction with the environment could have consequences for both people and the environment (Worster, 1979; Hurt, 1981; Gutmann and Cunfer, 1999; Cunfer, 2002).

The Great Plains of North America is a large region spanning the area from the end of the Midwest mesophytic forests to the front range of the Rocky Mountains (east to west), and from northern Canada to Central Texas (north to south) (Riebsame, 1990). The climate of the Great Plains is one of dry winters and wet summers. It is shaped by three air masses, each leading to its own seasonal dynamics (Lauenroth and Burke, 1995). A Pacific air mass originating in the Gulf of Alaska is dominant in the winter, its air dried by crossing several mountain ranges. An additional polar air mass also shapes winter weather, creating a strong north-south gradient in air temperature and snow cover. In summer, the westerly flow weakens and polar air retreats to the north, allowing an air mass that comes from the subtropical Atlantic Ocean to bring moisture into the region.

The distribution of both natural ecosystems and land use management is controlled in large part by two major climatic gradients: an east-west gradient of increasing precipitation and a north-south gradient of increasing temperature (Figures 4-1 and 4-2). Mean annual precipitation ranges from more than 1,200 mm/yr to less than 300 mm/yr, and mean annual temperature from less than 0°C to greater than 20°C. Plant species composition varies from tallgrass prairie to shortgrass steppe, with decreasing precipitation. In addition to influencing ecosystem type, these gradients have large influences on net primary production and soil organic carbon (Sala et al., 1988; Burke et al., 1989).

FIGURE 4-1

Average annual precipitation, Great Plains counties, 1961-1990.

FIGURE 4-2

Average annual temperature, Great Plains counties, 1961-1990.

There is a conventional storytelling of the development of the Great Plains.¹ Originally lightly settled by native people, the region was colonized by the European-origin population of the United States in the decades following the Civil War (Powell, 1878; Webb, 1931). This settlement was encouraged by the institutional context of the United States in the nineteenth century. At that time a combination of economic, social, and political processes made land and transportation available for individuals and families who wanted to move out of more densely populated parts of the country (Opie, 1987). These settlers discovered a semiarid grassland landscape with agricultural possibilities that ranged from consistent arable cropping in the east to limited cropping and steady pasturing in the west. Some of the implications of aridity (lack of wood for building and unfamiliar prospects for agriculture, for example) delayed settlement, but demographic pressure in the more eastern United States pushed people west. With what seemed like good years of rainfall and with increasing agricultural prices in the 1910s and 1920s, an aggressive stream of would-be farmers and ranchers acquired land and attempted to farm it (Gutmann and Sample, 1995; Worster, 1992). In this conventional history, the plowing up of the grasslands for wheat, combined with the drought of the 1930s, provoked disastrous dust storms and social dislocation. While we may not agree that those were the only causes, or that the greatest areas of wheat farming suffered the worst drought and dust storms, there was a causal relationship. Plowed land is a much greater source of blowing dust than uncultivated grassland.

Despite the social and economic disruptions of the 1930s, land use changed little as a result of the Dust Bowl. Acreage planted in crops dipped during the worst of the 1930s drought, but by 1945 had recovered fully to predrought levels. The balance between cropland and pasture remained virtually stable from the 1920s through the 1990s in most plains counties (Cunfer, 2005). People responded to the problems of the 1930s by changing some of their cultivation practices (Hargreaves, 1992). Some of them migrated out of the region, although there may have been less out-migration during the drought of the 1930s than during the drought of the 1950s, when postwar economic development in the United States provided places to go (Gutmann et al., 2002). Those who remained changed their farming practices between the 1930s and 1970s, by introducing techniques that reduced the risk of erosion and made better use of soil moisture, implementing improved crop varieties, enhancing nutrients through fertilization and by increasing the use of irrigation where groundwater was available (Green, 1973). The drought of the 1950s did not last as long as that of the 1930s, and improved agricultural practices made the impact of the later drought less severe. Nonetheless, the adoption of more intense irrigation since the 1940s has led to groundwater depletion and other land use changes that provoked further damage by the 1980s. The Ogallala aquifer has been the largest source of groundwater for irrigation, and it has suffered steady declines in level since the 1970s and 1980s (Opie, 1993; Riebsame, 1991; Cunfer, 2005).

Whether consistently provoked by environmental conditions or spurred by broader social change, the Great Plains region has experienced significant demographic changes since its colonization by people of European descent beginning in the mid-nineteenth century. The region has grown in population overall, but that growth masks two defining population patterns. The first pattern applies to rural areas, where population grew with settlement until the 1930s and has declined since. While the story is more complex than that told by the media to general audiences, the overall rural pattern since the 1930s has been one of a declining and aging population, with less and less agricultural employment and growing poverty.² In the second pattern, most of the long-term growth occurred in a small number of metropolitan areas, led by Denver, Colorado. Since the 1960s, counties close to the front range of the Rocky Mountains have grown, spurred by the growth of recreation and by industrial development in the same areas.

This history of land use and demographic transformation in a context of environmental change animates our study of population and environment in the Great Plains. Population change has acted through land use to influence the environment. Conversely, environmental conditions have influenced demographic change. All this has taken place over more than a century of rapid economic and social change in the United States. The challenge is to measure changes in population, land use, and environment, as well as to disentangle their impacts from the underlying social changes that would have happened anyway.

CONCEPTUALIZING THE CONNECTIONS AND THINKING ABOUT THE DATA

This project has three main research components, and in this chapter we will emphasize two of them. Those three components are: (1) an analysis of county-scale processes based on historical population, land use, and environmental and other data; (2) a series of historical studies of individual localities, of the experiences of farmers and their families, and of agricultural practices throughout the region; and (3) interviews of about 180 farm families in six different parts of the region. This chapter is primarily about the first component, with some information from the second and only brief references to the third.

The Great Plains region as we define it covers much of 10 states of the United States, comprising nearly one in seven U.S. counties: Colorado, Kansas, Montana, Nebraska, New Mexico, North Dakota, Oklahoma, South Dakota, Texas, and Wyoming. Our goal has been to represent a large portion of that region, and our strategy has been to make use of readily available aggregate data about counties. We recognize that we are unable to analyze the behavior and experiences of individual persons, families, farms, or environmental settings. This emphasis on county-scale processes determines how we conceptualize the connections between population, land use, and environment. It led us to relate population to land use through a set of hypotheses, some brought to the project from the beginning, and others formed as our research evolved. These are the hypotheses that are important to the work reported in this chapter.

1. We began with the hypothesis that the growth of population density through European-origin settlement led to changes in land use—and therefore environment—with a more or less linear relationship: the more people there were, the more land use changed. We also hypothesized that different kinds of population had differential impacts on land use, with urban impacts greater than rural, and some ethnic groups having different impacts than others.

2. We also began with the hypothesis that population density could itself be a function of the resource attributes of the land, so that population density could become greater in areas with more precipitation than in those with less, or where there was oil or coal, because of greater opportunities for employment. This manifests itself as a limitation on the first hypothesis, so that in areas with fewer resources there would be less population, but also less land use change.

3. A third main hypothesis operates more in the realm of methods than results, asserting that it is possible to gauge the impact of population change on the environment by estimating biogeochemical models linked to accurate historical data about population and land use at the county scale, and then draw conclusions based on the modeled results.

Our approach is thoroughly interdisciplinary but still anchored in disciplinary strengths (Riebsame et al., 1994). Our team has a core of historians, ecologists, and sociologists, plus geographers and anthropologists. As we venture further, we build on the cross-disciplinary tools that enable a historian's deep knowledge of the past and the agricultural practices of the past century, coupled with the demographer's clear understanding of demographic processes to inform the ecosystem modeler's task of model specification and interpretation.

DATA, DESIGN, AND STATISTICAL TECHNIQUES

While we make use of many varieties of data, our primary sources come from county tabulations drawn from the U.S. censuses of population and agriculture. We have collected those data for the decennial population censuses from 1880 through 2000, as well as for the agricultural censuses (which were decennial until 1920 and then more frequent thereafter) from 1880 through 1997.³ In addition to census-based sources, we have collected other county-level tabulations of social characteristics. We use the population and social indicators data to understand population structure and change, and the agricultural census data to understand agricultural land use. Their consistency, as well as the effectiveness and long-term quality of the U.S. census, have made this part of our project straightforward. Some of these data were available to us in digital form, and others we collected in print form and then hand-keyed into our database. All of these data are described in Gutmann et al. (1998). Since that document was published, we have added data from recent censuses (1997 agriculture and 2000 population), while maintaining their content and structure. Although our study area is not coterminous with the 10 states, we have collected data that covered the entire area of the 10 Great Plains states, and often neighboring states, especially Iowa and Minnesota.

While data about population size and structure are remarkably similar from 1880 to 2000, we are well aware of the changes that have occurred in the tabulated record during such a long time. On the population side, only in 1910 and since 1980 are good data about ethnicity available. In another example, population size divided into 5-year age categories was not tabulated from 1880 through 1920. On the agricultural side, the definition of major categories of farmland into cropland, pasture, and other types changed significantly in the 1920s, and the yields of irrigated crops have always been difficult to estimate. One challenge of our research has been to find ways to overcome these limitations.

Environmental data are not as easy to acquire as those about land use and population, because they are not normally tabulated at the scale of the county (Gutmann, 2000). We have sought temperature and precipitation data at reasonable intervals of time and space. We have done this by interpolating VEMAP Project data (http://www.cgd.ucar.edu/vemap/) to historical county boundaries. We have also acquired recent soil structure and elevation data, as well as data giving locations of bodies of water and streams, interpolating them to historical county boundaries. Finally, we have acquired other weather data (dust storms, for example), when possible (see Gutmann and Cunfer, 1999; Deane and Gutmann, 2003).

No single method has worked in all our analyses. Some of what we have done is descriptive, based on examining tabular and graphical series of results that display change over time or variation through space. Yet we have consistently attempted to make use of statistical approaches that are appropriate for the data and the hypotheses involved. One challenge has been to take account of the role of time and space in our analysis, with appropriate study of temporal and spatial autocorrelation. We have also pursued innovative ideas about the role of causality (Deane and Gutmann, 2003). Finally, one of our important goals has been to integrate what we can learn from the historical data with ecosystem modeling techniques.

Despite the widespread recognition that the Great Plains represent a perfect test case for the study of population and environment in grassland settings, no earlier study has been able to amass the large body of systematic, quantitative, and region-wide data that we have brought to bear on the problem. These data allow us to start from the beginning, first describing the impact of settlement by European-origin populations, and then systematically analyzing elements of that impact. In this chapter we highlight our findings thus far and point to new directions of research to follow.

MAJOR FINDINGS

In writing about our major findings, we begin with the broadest view of what we have learned about the impact of population on environment over the last 100 years, and then turn to two additional questions. Our presentation echoes the major hypotheses stated earlier, with one major section reporting on our exploration of each hypothesis. From this wide-angle view, we show that a growing European-origin population has massively changed the environment of about a third of the Great Plains region, both through the replacement of open grasslands with fenced and heavily managed croplands and through the creation of significant areas of urban and suburban development. While we cannot always directly measure environmental outcomes, this analysis, based largely on land use change, is dramatic. The other two-thirds of the Great Plains, used primarily for extensive livestock grazing, has seen only limited environmental change, as cattle mimicked the ecological impacts of the native bison they replaced. This finding is important, because, while it confirms that settlement changed land use, it demonstrates that there was less land use change than previous researchers have believed.

Our second main point shows the complexity of gauging the relative importance of population and the environment in shaping land use decisions. Paradoxically, although population change has shaped the environment through land use change, we also see that the environment limits human action. Humans can decide to plant crops, for example, but they will not succeed in raising crops in most of the Great Plains. This conclusion is a necessary antidote to the thinking that asserts that human domination of environmental systems is always complete and without limits.

Our final discussion highlights the progress we have made in finding new ways to measure the environmental impact of population on the environment of the Great Plains. In our most recent research, we have run the century ecosystem model with parameters based on historical land uses in a small number of Great Plains counties. The century ecosystem model is a generalized ecosystem model that simulates the dynamics of carbon, nitrogen, and phosphorus in grassland, forest, savanna, and crop systems (Parton et al., 1993; Metherell et al., 1993 Metherell et al., 1995; Kelly et al., 1997; Paustian, Parton, and Persson, 1992). The results of these model runs allow us to begin the process of estimating historical and current soil composition and greenhouse gas scenarios for the region. This detailed modeling of environmental outcomes based on accurate historical data is the most precise effort yet undertaken to gauge the impact of human agricultural activity on the grassland environment.

HOW POPULATION CHANGED THE GREAT PLAINS

Our first hypothesis asserts that, as the population of the Great Plains grew due to the influx of the European-origin population, environmental impacts through agricultural (and later urban) land uses accelerated, closely coupling demographic change with land use and environmental change. The exploration of that process has informed much of our work, with conclusions that show that the process is very generally as we expected, but with a number of interesting complications. While it is important to see the broad generalization confirmed, the most valuable findings may come from the complications and the insights they give us for future research.

The Great Plains, by our definition, consist of approximately 390 million acres of land. In 1880, U.S. farmers reported tht they had about 19 million of those acres in farms (see Figure 4-3). By 1910 they had 10 times as much, and by 1930 they reported 288 million acres of farms, nearly three-fourths of the region's land area. At its peak in 1959, nearly eight out of every nine acres of land in the region was reportedly in a farm. For census purposes, pasture and ranch land is included in farmland, so a considerable majority of this farmland was not plowed. The conversion of native grassland to cropland happened somewhat later. Although farmers began plowing out sod in the 1870s, it was a Herculean effort, and as late as 1900 only 8 percent of the region, some 31 million acres, were used for crops. Most of the plowing up of the Great Plains happened in the first three decades of the twentieth century, when farmers brought about 88 million additional acres into crop production, peaking at 31 percent of total land area in 1935 (Cunfer, 2005).

FIGURE 4-3

Population, farmland, and cropland, Great Plains counties, 1880-1990. The figures for cropland represent the sum of cropland harvested for major crops from 1880 through 1920, (more...)

The demographic change that accompanied the transformation of land use is just as dramatic. The region experienced steady population growth as land use changed through 1930, followed by a rapid transformation from overall population growth to urban population growth. The region's rural population has been shrinking since the 1930s, in some decades quite rapidly. We follow the U.S. Census Bureau's definition of an urban place as one having a population of 2,500 inhabitants, a relatively low threshold that moves small towns with populations over 2,500 into the urban category.

The comparisons between population and land use trends displayed in Figure 4-3 show that the growth in farmland generally increased with overall population size, but it continued to increase long after the rural population stopped growing in 1930. Cropland, however, stopped increasing after 1940, only a decade after the rural population peak. Put another way, the link between rural population change and land use change that led to conversion to cropland was tighter than that between rural population and overall conversion of land to farm uses, or between the overall population change and either farmland or cropland conversions. The reverse process was not true, however. As rural population declined after 1930, the amount of land in crops remained stable. Depopulation did not equate to land abandonment in the second half of the twentieth century, as remaining farmers continued to plant acreage relinquished by emigrant neighbors.

Figure 4-3 shows that most of the direct impact of rural population change on land use and consequently on the environment took place during the era of rapid settlement from 1880 through 1930 and then diminished. This is true in other areas as well (Moran, Brondizio, and VanWey, Chapter 5). Yet the demographic impact on an agricultural region such as the U.S. Great Plains is not limited to the direct results of local population growth and decline. Environmental changes are as much the result of large-scale shifts in the market for agricultural products (driven by populations and tastes elsewhere) as they are the result of local or regional population change (Cronon, 1991). Local demographic change is usually the symptom that spurs immediate changes in land use, for example, because one needs farmers to make the change from native grassland to crops. That is what we can measure. We can also measure the role of local population changes in the conversion of land from agricultural to urban and suburban uses. This is a topic not dealt with directly in this chapter, but it is part of our research agenda for the future.

Not all the conversion to farmland produced dramatic change in land character with environmental consequences. The amount of cropland reported, however, is a strong indication of the intensity of environmental change that took place in the Great Plains between 1880 and 1992. At its peak in the late 1930s, between 31 and 38 percent of the total land in the region had been converted from native grassland to cropland.⁴ The geographical distribution of that cropland varies from subregion to subregion because the eastern Great Plains has more rainfall than the west and is better suited for cropping.

Figure 4-4 shows the growth of total area cropped in the Great Plains from 1880 through 1992, dividing the region into an eastern tier of states (North Dakota, South Dakota, Nebraska, Kansas, Oklahoma, and Texas) and a western tier of states (Montana, Wyoming, Colorado, and New Mexico). Figure 4-5 shows the spatial distribution of cropping in the Great Plains in 1930; a few counties in the east had as much as three-fourths of their land in cropland, while those in the west had 10 percent or less. While not displayed in the figures, there was also a significant increase in the amount of land that is irrigated in small parts of the Great Plains, reaching a peak of more than 17 million acres in the late 1970s. While this represents only about 4 percent of the entire Great Plains, the increase in irrigation has had environmental consequences of local importance. Putting all these findings another way, we see little support for a simple linear relationship between population growth and land use change.

FIGURE 4-4

Differences between eastern tier and western tier states, 1880-1990.

FIGURE 4-5

Percentage of county area in crops, 1930.

The conversion from native grassland to cropland and other farm uses is as dramatic over the long term as that reported by others in this volume (see Chapters 5, 6, and 8) for regions—such as the Amazon—where deforestation and other kinds of land use conversion has taken place. While it can be difficult to measure the environmental consequences of this change at the large scale that we have chosen to study, we know in general terms what happens when native grasses are plowed and replaced with crops:

• Soil texture changes and topsoil may be lost to wind and water erosion.
• Carbon, nitrogen, and other minerals change composition as plowing and cropping alter soil chemistry.
• Natural processes related to fire are suppressed.
• Hydrological systems are disrupted or altered.
• Species diversity diminishes as introduced crop species replace native plants.
• Wildlife distribution patterns change as farmers replace habitat and build roads, fences, and other barriers to migration.

In the two-thirds of the grassland that farmers could not successfully plow for crops, environmental change has been much less dramatic. While nearly all of that land has been used for extensive grazing by cattle, and in much smaller proportions by sheep and horses, most of it remains in native vegetative cover. Grazing livestock can have the following environmental impacts on grasslands:

• Interference with wildlife, especially competing large grazers (bison, pronghorn) and predators (wolves, grizzly bears).
• Changes in plant diversity, species composition, and ground cover, especially increases in invasive species.
• Disruption of riparian and aquatic habitats along rivers and streams.

Yet unlike with cropping, grazing has little impact on soil texture and depth or on carbon or nitrogen systems, and it is less disruptive of plant and animal biodiversity (Lauenroth et al., 1994; Cunfer, 2005). Whereas nearly all forest ecosystems in the continental United States have been logged in the past 400 years, nearly two-thirds of the Great Plains remains in unplowed native vegetation. These consequences are by no means all that has happened to the environment of the Great Plains as a consequence of the region's growing population and its long-term conversion from native grassland to cropland, managed rangeland, and later urban and suburban development. While we have only partially quantified them, their impact is important and they constitute the starting point for our understanding.

THE PARADOX OF POPULATION AND ENVIRONMENT

The message in the previous section is clear: the change in population that took place in the Great Plains when the European-origin population grew transformed the environment in ways that have had local consequences and global consequences. They range from the local disruption of faunal wildlife to the global alteration of the carbon cycle, and probably every scale in between.

Despite this simplicity and certainty, our results show that environmental constraints limit the impact of the human population on the environment. These constraints may be in keeping with our second hypothesis, and they are worthy of note. They operate by limiting the flexibility that farmers had to choose how they used the land. Put simply, farmers on the Great Plains were unable to convert all their land to cropland—or to any other single use that they desired—because the land was not environmentally suited to every possible use. In a straightforward way, we see this limitation through the range of variation in cropland in Figures 4-4 and 4-5, and in the parallel knowledge that few Great Plains farms or counties were ever transformed into a single-crop monoculture. This is one important finding in the research of Cunfer (2005).

While change in population is the most important determinant of the overall likelihood of any change in land use in the historical time period, a limited group of environmental characteristics are the most important determinants of the specific kind of agricultural land use adopted by farmers. If we measure variation in land use as the choice to use the land for cropping or pasture, almost all the variation in agricultural land use in the Great Plains is explained by environmental variables, especially precipitation, temperature, soil texture, and slope (Burke et al., 1994; Burke, Lauenroth, and Parton, 1997; Gutmann et al., 2004; Lauenroth, Burke, and Paruelo, 2000; Sala et al., 1988; Cunfer, 2005). Not much room is left for human intervention beyond deciding whether to irrigate and which crops to plant, and, even then, irrigation and cropping choices are themselves largely determined by a mix of environmental and market factors over which the farmer has relatively little control. This is the result we report in Gutmann et al. (2004) and Cunfer (2005), confirming earlier work by others and making clear that even when there are ethnic and cultural preferences for certain crops or land uses, the environmental determinants are very strong.

REFINING OUR UNDERSTANDING OF ENVIRONMENTAL IMPACTS

Thus far the story we have told is very simple, limited to the big picture of the impact of population change on environment and the constraints of environment on the exact nature of those impacts. Put another way, the influx of European-origin people to the United States and especially the Great Plains caused a dramatic change in the way land was used in the region, driven largely by the introduction of crop-based and livestock-based agriculture. At the same time, we show that the introduction and continuation of agriculture has limits imposed by the environment.

To refine our measurement of the environmental impact we have undertaken a new series of analyses (Burke et al., 2002; Parton et al., in press; Cunfer, 2004 Cunfer, 2005) that make use of the history of nineteenth- and twentieth-century land use change to estimate the impact of population-induced agricultural land use on soil biogeochemistry. These analyses are designed to confirm the assertion of our third hypothesis, that model-based work on environmental outcomes has value. Burke et al. (2002) and Cunfer (2005) show that, at both the regional scale and at the level of individual counties, crop farming resulted in significant losses of soil nitrogen. Nitrogen declined most in the northeastern plains, where higher rainfall supported more vegetation and cooler temperatures slowed decomposition of plant matter. There soils lost an estimated 1,080 kg nitrogen per hectare as a result of plowing and cultivation over 75 to 100 years. Losses were much smaller in the western and southern plains because of dryer and warmer conditions and significantly less cropping. Soil nitrogen in grazing systems is roughly in balance, so much of the western plains have lost virtually no nitrogen since the European-origin settlement. Across the entire Great Plains, soil nitrogen declined by about 20 percent from original levels (Burke et al., 2002).

Since farmers began using synthetic fertilizer after World War II, nitrogen dynamics have changed. While soil nitrogen is now roughly stable at about 20 percent below presettlement levels region-wide, farmers annually apply an average of 35 kg nitrogen fertilizer per hectare of cropland per year. About half of that goes into crop plants, increasing their growth. The other half is lost to the system, either leaching into waterways or volatilizing into the atmosphere, in either case becoming an environmental pollutant. In the eastern parts of the region, some of the excess fertilizer nitrogen may accumulate in soils, restoring some of the soil nitrogen lost due to a century of cropping, but the extent to which this may be happening is unknown.

At the scale of the individual county, Cunfer (2004 Cunfer (2005) shows that before 1940 Great Plains farm systems produced enough livestock manure to fertilize only about 20 percent of their cropland each year. Traditional, organic, small family farms mined soil fertility, extracting more nitrogen each year than they returned, and crop yields fell during the first 50 years of cultivation. Like many previous American agricultural frontiers, the Great Plains may have been on a path toward widespread land abandonment due to depleted soil fertility, but the development of synthetic fertilizers after 1945 allowed farmers to artificially replenish the nitrogen they removed each year. Crop farming has continued, year in and year out, for more than 130 years in the Great Plains, longer than most other American agricultural regions, mainly because of energy-intensive inputs of synthetic nitrogen. Widespread farming in New England and in the South, for example, lasted only about a century before land abandonment or reversion to forest became widespread. This process has not yet happened in the Great Plains, as crop acreage has remained roughly stable since the 1920s (Cunfer, 2005).

In another approach that showcases carbon as well as nitrogen, Parton et al. (in press) focus their analysis on four counties in the Great Plains (Hamilton, Nebraska; Ramsey, North Dakota; Pawnee, Kansas; and Hockley, Texas) that have different mixes of agriculture involving dryland and irrigation, grains, grasses, and cotton. This analysis makes use of the century ecosystem model. Figure 4-6 summarizes the model results for carbon by reporting a general pattern of large losses (approximately 50 percent) of soil carbon during the first 50 years following the plowing up of native grasslands, with most of the carbon loss occurring during the first 20 to 30 years. Soil nitrogen mineralization followed a general pattern of increased nitrogen mineralization for 10 to 20 years following the plowing up of grassland, and a sharp decrease in nitrogen mineralization 20 to 50 years after plowing up, with nitrogen mineralization rates approaching 20 percent of grassland levels after 50 years of cultivation. These simulated patterns in soil carbon and nitrogen mineralization are consistent with other studies (Schimel et al., 2000) showing that rapid losses of soil carbon following the plowing up of grassland soils, stabilization of soil carbon levels at 50 percent of initial values after 50 years of cultivation, and substantial decreases in soil mineralization after 50 years of dryland cultivation. The high nitrogen mineralization rates following plowing up of grassland soils are consistent with the observation that nitrogen fertilizer responses are minimal for wheat fields after 30 years of dryland cultivation (Metherell et al., 1995) in the Great Plains, and data showing that wheat yields in response to fertilizer increase with time since cultivation (Greb et al., 1974).

FIGURE 4-6

Simulated soil carbon in four Great Plains counties.

The locally dramatic expansion of irrigated agriculture is one of the major land use changes that has taken place during the past 50 years in some parts of the Great Plains, with corn and alfalfa grown in the northern and central Great Plains, and cotton in the southern Great Plains (Texas and Oklahoma). The land use data suggest that most of the irrigated land in the northern and central Great Plains had previously been cultivated using dryland techniques. Model results from Pawnee and Hamilton counties show that irrigated corn-alfalfa rotations begun in the 1960s produced substantial increases in crop yields, soil carbon levels, and soil nitrogen mineralization rates. Most of the increases in soil carbon and nitrogen mineralization occurred from 1970 to 1990 because of the large increases in the amount of carbon (300 to 400 grams of carbon per square meter per year added as corn stover) and nitrogen (100 to 150 kilograms per hectare per year of fertilizer) added to the system with irrigated agriculture. Model results suggest that soil carbon levels increased by more than 800 grams of carbon per square meter for irrigated land in Pawnee and Hamilton counties from 1970 to 2000. Extrapolating these carbon accumulation rates to the 4 million hectares of irrigated land added from 1960 to 1980 for the central and northern Great Plains would result in 56.0 trillion grams of carbon sequestered in the soil.

CONCLUSIONS

In this chapter we have discussed an approach to the study of population-environment relationships that focuses on changes over a large region that makes use of data at the scale of the U.S. county. Taken to its fullest extent, our approach yields estimates of the consequences for soil chemistry of population-driven changes in land use. As we develop estimates for the specific agricultural practices of more counties, we are able to gauge the large-scale and long-term impact of the transition into and out of agriculture for the region as a whole, which will provide valuable data to everyone studying the past experience and the future prospects of a major northern temperate region, one in which potential carbon storage is a significant question.

The results we have presented do not yet show a tight connection between population change and environmental change. There is a good reason for that, because most of the land use changes that we have signaled are not closely tied to population change, except perhaps during the early years of European-origin settlement. In more recent years, population has changed land use patterns in small parts of the Great Plains by forcing the conversion of land from agricultural uses to residential and other uses. This conversion to residential uses has produced other ecosystem consequences that remain to be studied. Our next approach will be to run century ecosystem model estimates of ecosystem processes for counties that are converting to those uses, measuring the impact of uses, such as lawns, that differ from native grasses, irrigated cropland, and dryland crops. While these land uses are small in the region as a whole (see Parton, Gutmann, and Travis, 2003), recent data suggest that their impacts are large with respect to the region as a whole (Kaye et al., 2004).

Finally, we wish to address the theme of integrating social and natural science. In our experience, the great challenge is to spend enough time working together so that all the parties begin to understand the questions asked by the othr disciplines and begin to understand the range of acceptable answers. Concepts that seem as trivial as “which is the dependent variable in this analysis, population or environment?” can lead to very fruitful insights, and indeed, new scientific questions. In our case, we have developed new perspectives on carbon and nitrogen for the region as a whole, and for a period longer than a century, as they are anchored in a rich historical and demographic record. Perhaps the most difficult challenge is to take our improved understanding of the linkages between human and ecological processes and begin to make predictions for the future of the people and the environment of the Great Plains.

ACKNOWLEDGMENTS

This project has benefited from the hard work of many individuals, too many to be authors of a single chapter. We are especially grateful to Glenn Deane, Lenora Bohren, Denis Ojima, Steve Williams, Mark Easter, Kathleen Galvin, William Travis, Susan H. Leonard, Kenneth Sylvester, and Sara Pullum-Piñón, as well as to a host of others.

REFERENCES

1. Bonnifield MP. The Dust Bowl: Men, Dirt, and Depression. Albuquerque: University of New Mexico Press; 1979.
2. Burke IC, Yonker CM, Parton WJ, Cole CV, Flach K, Schimel DS. Texture, climate, and cultivation effects on soil organic matter content in U.S. grassland soils. Soil Science Society of America Journal. 1989;53:800–805.
3. Burke IC, Lauenroth WK, Parton WJ, Cole CV. Interactions of landuse and ecosystem function: A case study in the central Great Plains. In: Groffman PM, Likens GE, editors. Integrated Regional Models: Interactions Between Humans and Their Environment. New York: Chapman Hall; 1994. pp. 79–95.
4. Burke IC, Lauenroth WK, Parton WJ. Regional and temporal variability in aboveground net primary productivity and net N mineralization in grasslands. Ecology. 1997;78:1330–1340.
5. Burke IC, Lauenroth WK, Cunfer G, Barrett JE, Mosier A, Lowe P. Nitrogen in the central grasslands region of the United States. BioScience. 2002;52:813–823.
6. Cronon W. Nature's Metropolis: Chicago and the Great West. New York: W.W. Norton and Co.; 1991.
7. Cunfer G. Causes of the Dust Bowl. In: Knowles AK, editor. Past Time, Past Place: GIS for History. Chapter 7. Redlands, CA: ESRI Press; 2002.
8. Cunfer G. Manure matters on the Great Plains frontier. Journal of Interdisciplinary History. 2004;34:539–567.
9. Cunfer G. College Station: Texas A&M University Press; 2005. On the Great Plains: Agriculture and Environment.
10. Deane GD, Gutmann MP. Blowin' down the road: Investigating bilateral causality between dust storms and population change in the Great Plains. Population Research and Policy Review. 2003;22:297–331.
11. Egan T. Amid dying towns of rural plains, one makes a stand. New York Times. 2003 December 1;Section A1
12. Greb BW, Smika DE, Woodruff NP, Whitfield CJ. Conservation Research Report. Vol. 17. Washington, DC: U.S. Department of Agriculture; 1974. Summer fallow in the central Great Plains; pp. 51–85.
13. Green D. Land of the Underground Rain: Irrigation on the Texas High Plains, 1910-1970. Austin: University of Texas Press; 1973.
14. Gutmann MP. Scaling and demographic issues in global change research. Climatic Change. 2000;44:377–391.
15. Gutmann MP, Cunfer G. A New Look at the Causes of the Dust Bowl. Lubbock: International Center for Arid and Semiarid Land Studies, Texas Tech University; 1999. Publication 99-1.
16. Gutmann MP, Deane GD, Lauster N, Peri A. Two population-environment regimes in the Great Plains of the United States, 1930-1990. Submitted to Population and Environment. 2005
17. Gutmann MP, Sample CG. Land, climate, and settlement on the Texas frontier. Southwestern Historical Quarterly. 1995;99:137–172.
18. Gutmann MP, Pullum S, Cunfer GA, Hagen D. The Great Plains Population and Environment Database: Sources and User's Guide, Version 1.0. Austin: Texas Population Research Center; 1998.
19. Gutmann MP, Pullum-Piñón S, Baker SG, Burke IC. German-origin settlement and agricultural land use in the twentieth century Great Plains. In: Kamphoefner W, Helbich W, editors. German-American Immigration and Ethnicity in Comparative Perspective. Madison: University of Wisconsin Press; 2004.
20. Hargreaves MMW. Dry Farming in the Northern Great Plains: Years of Readjustment, 1920-1990. Lawrence: University Press of Kansas; 1992.
21. Hurt RD. The Dust Bowl: An Agricultural and Social History. Chicago, IL: Nelson-Hall; 1981.
22. Kaye JP, Burke IC, Mosier AR, Guerchman JP. Methane and nitrous oxide fluxes from urban soils to the atmosphere. Ecological Applications. 2004;14(4):975–981.
23. Kelly RH, Parton WJ, Crocker GJ, Grace PR, Klir J, Körschens M, Poulton PR, Richter DD. Simulating trends in soil organic carbon in long-term experiments using the century model. Geoderma. 1997;81:75–90.
24. Kilborn PT. Bucking trend, they stay on plains, held by family and friends. New York Times. 2003 December 2;Section A1
25. Lauenroth WK, Burke IC. Encyclopedia of Environmental Biology. Vol. 2. San Diego, CA: Academic Press; 1995. Great Plains, climate variability; pp. 237–249.
26. Lauenroth WK, Milchunas DG, Dodd JL, Hart RH, Heitschmidt RK, Rittenhouse LR. Effects of grazing on ecosystems of the Great Plains. In: Vavra M, Laycock WA, Pieper RD, editors. Ecological Implications of Livestock Herbivory in the West. Lakewood, CO: Society for Range Management; 1994. pp. 69–100.
27. Lauenroth WK, Burke IC, Paruelo JM. Patterns of production and precipitation-use efficiency of winter wheat and native grasslands in the central Great Plains of the United States. Ecosystems. 2000;3:344–351.
28. Malin JC. Dust storms, 1850-1900. Kansas Historical Quarterly. 1946;14:129–144. 265–296, 391–413.
29. Metherell AK, Harding LA, Cole CV, Parton WJ. CENTURY—Soil Organic Matter Model Technical Document, Agroecosystems Version 4.0, Great Plains Systems Research Unit, Technical Report No. 4). Ft. Collins, CO: U.S. Department of Agriculture, Agricultural Research Service; 1993.
30. Metherell AK, Cambardella CA, Parton WJ, Peterson GA, Harding LA, Cole CV. Simulation of soil organic matter dynamics in dryland wheat-fallow cropping systems. In: Lal R, Kimball J, Levine E, Stewart BA, editors. Soil Management and Greenhouse Effect. Boca Raton, FL: CRC Press Inc.; 1995. pp. 259–270.
31. Opie J. The Law of the Land. Two Hundred Years of American Farmland Policy. Lincoln: University of Nebraska Press; 1987.
32. Opie J. Ogallala: Water for a Dry Land. Lincoln: University of Nebraska Press; 1993.
33. Parton WJ, Scurlock JMO, Ojima DS, Gilmanov TG, Scholes RJ, Schimel DS, Kirchner T, Menaut J-C, Seastedt T, Garcia Moya E, Kamnalrut A, Kinyamario JL. Observations and modeling of biomass and soil organic matter dynamics for the grassland biome worldwide. Global Biogeochemical Cycles. 1993;7:785–809.
34. Parton WJ, Gutmann MP, Travis WR. Historical land use change in eastern Colorado. Great Plains Research. 2003;13:97–125.
35. Parton WJ, Gutmann MP, Williams SA, Easter M, Ojima D. The Ecological Impact of Historical Land Use Patterns in the Great Plains: A Methodological Assessment. Submitted to Ecological Applications. in press.
36. Paustian K, Parton WJ, Persson J. Modeling soil organic matter in organic-amended and nitrogen-fertilized long-term plots. Soil Science Society of America Journal. 1992;56:476–488.
37. Powell JW. Report on the Lands of the Arid Region (45th Congress, 2nd session, House Executive Document No. 73.). Washington, DC: U.S. Government Printing Office; 1878.
38. Riebsame WE. The United States Great Plains. In: Turner BL, Clark WC, Kates RW, Richards JF, Mathews JT, Myers WB, editors. The Earth as Transformed by Human Action. Cambridge, England: Cambridge University Press; 1990. pp. 561–575.
39. Riebsame WE. Sustainability of the Great Plains in an uncertain climate. Great Plains Research. 1991;1:132–1150.
40. Riebsame WE, Parton WJ, Galvin KA, Burke IC, Bohren L, Young R, Knop E. Integrated modeling of land use and cover change: A conceptual scheme for applying an integration strategy to agricultural land use on the U.S. Great Plains. Bioscience. 1994;44:350–356.
41. Riney-Kehrberg R. Rooted in Dust: Surviving Drought and Depression in Southwestern Kansas. Lawrence: University Press of Kansas; 1994.
42. Sala OE, Parton WJ, Joyce LA, Lauenroth WK. Primary production of the central grassland region of the United States. Ecology. 1988;69:40–45.
43. Schimel D, Melillo J, Tian H, McGuire AD, Kicklighter D, Kittel T, Rosenbloom N, Running S, Thornton P, Ojima D, Parton W, Kelly R, Sykes M, Neilson R, Rizzo B. Contribution of increasing CO2 and climate to carbon storage by ecosystems in the United States. Science. 2000;287:2004–2006. [PubMed: 10720324]
44. Webb WP. The Great Plains. Dallas, TX: Ginn and Company; 1931. (Reprint: University of Nebraska Press, 1981)
45. Worster D. Dust Bowl: The Southern Plains in the 1930s. New York: Oxford University Press; 1979.
46. Worster D. Under Western Skies: Nature and History in the American West. New York: Oxford University Press; 1992. New West, true West; pp. 19–33.

Footnotes

1

For our interpretation of this history, see Gutmann and Cunfer (1999) and Cunfer (2002 and Cunfer (2005). For aspects of the conventionally written history of the Great Plains narrative, see Webb (1931), Bonnifield (1979), Hurt (1981), Malin (1946), Riney-Kehrberg (1994), and Worster (1979).

2

See, for example, the New York Times stories by Egan (2003) and Kilborn (2003). They are only a few among many.

3

Since 1920, the agricultural census was taken every five years until 1950. Beginning in 1954, the censuses were enumerated for 1954, 1959, 1964, 1969, 1974, 1978, 1982, 1987, 1992, 1997, and 2002, usually at the beginning of the next year. The 2002 Agricultural Census forms were due on February 3, 2003.

4

The difference between the 31 and 38 percent depends on the interpretation of land that farmers classified as “cropland used as pasture.” The sources are not clear (and probably not uniform in their classification) about the extent to which that land was transformed through plowing and planting of nonnative species.