Soil Nitrogen Flows and Balances

Of the sixteen elements essential for plant growth, three are limiting in most natural ecosystems on earth: nitrogen (N), phosphorous (P), and potassium (K).Footnote ² A shortage of one or more of these elements often serves as the practical limitation on the amount of plant biomass that can grow in any given place, given local climate conditions. And in the natural world, N is by far in the shortest supply. In natural as well as human-managed ecosystems, nitrogen availability often determines the amount of plant matter present and constrains additional biomass growth. Although the earth’s atmosphere is nearly 78 percent nitrogen, it exists in an inert form unusable by plants. Only two natural forces convert (or “fix”) inert atmospheric N into compounds (such as ammonia) that plants can absorb through their roots: lightning and selected bacteria. Both forces deploy energy to split apart and recombine atmospheric N, making it available to plants. Lightning’s sudden jolt of energy does the job in an instant, releasing plant-useable nitrogen that drifts down onto soils (dry deposition) or falls attached to raindrops (wet deposition). Alternatively, some bacteria in soils ingest atmospheric N and excrete plant-useable N as part of their metabolism, slowly depositing the compounds in soils where plant roots can reach them. This “slow burn” also requires an input of energy, but in miniscule amounts distributed widely across time and space. All living plants on earth access their necessary nitrogen from one of these two sources, and the distribution of bacterial digestion and lightning storms establishes soil N concentrations and sets the pace of plant biomass productivity at a global scale. At least that was the case until the early twentieth century, when scientists created a third mechanism for nitrogen fertilization by deploying fossil fuel energy to fix nitrogen in the lab. The resulting flood of synthetic fertilizers has doubled the supply of plant-available N at the global scale, one of the most significant human interventions in natural systems (Smil Reference Smil2001).

Once fixed nitrogen is in the soil, plants can use it over and over. They draw nitrogen from soil into their roots and use it to build stems, leaves, flowers, and fruit. When a plant dies its constituent nitrogen decomposes back into the soil, becoming available for new growth; this is the basis of composting. Or perhaps a cow or other animal eats the plant, ingesting nitrogen for its own protein-building needs. After digestion, a considerable amount of the original N returns to the soil through manure, providing fertilizer for new plant growth.

If such natural recycling continues over thousands of years, soils can accumulate rich stockpiles of plant-available N that support luxuriant vegetation year after year. But ammonia and other forms of plant-useable nitrogen are inherently volatile, always at risk of dispersal from soils. It may dissolve in water and run off the surface or leach deep into the subsoil. It may volatilize and return to the atmosphere directly. It may feed a different group of bacteria whose metabolic processes send soil N back into the air (denitrification). In these cases, N returns to the atmosphere in inert form, no longer able to support plant growth. Natural ecosystems tend toward an unstable equilibrium, as N flows into soils, recycles repeatedly, and then flows out again. Farmers, as ecosystem managers, attempt to manage these processes in support of their preferred crop plants. They encourage processes that fix nitrogen in soils, for example by planting legume crops that come with their own herd of N-fixing bacteria. They invest considerable labor to support nitrogen recycling, by plowing down stubble or piling up compost heaps. And they transport nitrogen from one part of the farm (such as pastures) to another (such as cropland), often by managing livestock who collect N by grazing during the day, walk back to the stable at night, and deposit N-rich manure for application on cropland. A large part of agriculture—both crop planting and livestock herding—might be considered, in fact, a system of nitrogen cultivation, processing, and transport.

When Euro-American farmers began plowing prairie sod in the mid-nineteenth century they discovered rich soils in most places. In Kansas, deep soils contained decomposed nutrients of grass roots that penetrated 6 feet (2 meters) into the ground. Soil nitrogen had been accumulating here at least since the end of the last ice age, more than 10,000 years earlier. Compared to long-farmed croplands of Europe and even those of North America’s forested eastern seaboard, grassland soils contained abundant nitrogen—between 10 and 12 tons of nitrogen per hectare in the locations considered here (US Department of Agriculture 1978, 1982, 2003). Crop productivity was prodigious. In one comparison, Austro-Hungarian immigrants to Kansas produced double the grain yield they grew back home with only 5 percent of the labor (Cunfer and Krausmann Reference Cunfer and Fridolin2009). It is no wonder that Americans and Europeans flooded into the plains when the opportunity arose.

All that was possible when there was enough rain, which was far from certain, and throughout the pioneer era farmers worried more about soil moisture than about soil fertility (Miner Reference Miner1986, Reference Miner2006). Still, soil nutrient dynamics began changing immediately after sodbreaking, whether farmers attended to them or not (Haas et al. Reference Haas, Evans and Miles1957). Cropland began each spring with a stockpile of nitrogen carried over winter from the previous year. Through the growing season, soils received inputs of nitrogen from both natural and human-managed sources, which added to nutrient stocks. Soils also lost nitrogen through natural and human-controlled pathways. This study estimates inputs and outputs of nitrogen to calculate annual balances (figure 2). It begins with the known nitrogen content of unplowed soil, then adds annual inputs and deducts outputs to calculate the amount of soil nitrogen carried forward into the coming year (Garcia-Ruiz et al. Reference Garcia-Ruiz, Manuel, Gloria, David and Juan2012). Because soil science operates in the metric system, nitrogen flows appear in kilograms per hectare (kg/ha), with 1 kg/ha equivalent to slightly less than 1 pound/acre. Tracking soil nitrogen dynamics at 19 time points between 1880 and 1997 (corresponding with US Agricultural Census years) reveals farmers’ interaction with one of the planet’s most important natural systems through 120 years of agricultural evolution.

Figure 2. Nitrogen inputs and outputs in agroecosystems. White arrows represent primarily natural processes, while gray arrows indicate human-managed pathways (Garcia-Ruiz et al. Reference Garcia-Ruiz, Manuel, Gloria, David and Juan2012).

Note: While leaching and erosion are natural processes, land management practices affect their magnitude and might be considered hybrid natural and human processes.

Natural nitrogen inputs include two forms of atmospheric deposition plus free bacterial fixation. Atmospheric nitrogen drifted down onto cropland attached to dust particles (dry deposition) and to raindrops (wet deposition). In the Great Plains, atmospheric deposition was quite low. Wet deposition is estimated at 0.57 grams N per cubic meter of rainfall (Allen Reference Allen2008; Vasquez et al. Reference Vázquez, Costoya, Peña, García and Herrero2003). For each year that amount is multiplied by the total annual precipitation (estimated from historical weather station data available for 1895 to 1993). For example, in 1950 Nemaha County, Kansas received 845 mm of precipitation bringing an estimated 5 kg N per hectare to cropland through wet deposition. For 1880, 1890, and 1997, when weather station data are not available, the 99-year average precipitation substitutes. Dry deposition is typically an order of magnitude smaller, so negligible that it is not estimated here (Holland et al. Reference Holland, Braswell, Sulzman and Lamarque2005). Free-living bacteria in soils also generate plant-useable nitrogen as part of their own metabolic processes (free fixation), but they are more active in humid climates than in the semiarid grasslands. Free fixation adds about 4 kg N per hectare in similar environments, reduced somewhat by each cultivation event in a year (Boring et al. Reference Boring, Swank, Waide and Henderson1988; Unkovich and Baldock Reference Unkovich and Baldock2008). This analysis assumes two cultivations per year for most crops and thus estimates free fixation inputs at around 2 kg N per hectare. Adding roughly 5 kg/N/ha from rainfall plus 2 kg from free fixation means that natural processes contributed about 7 kg N/ha to Kansas soils in most years.

Several nitrogen inputs are wholly or primarily human-managed. A traditional way to replenish cropland fertility was collecting livestock manure for application on fields. The census reports the numbers of domestic animals on farms, allowing an estimation of their manure production and its nitrogen content. Manure from animals living in the barnyard (horses, mules, milk cattle, swine, poultry) was available for cropland application. Beef cattle, however, spent much of the year on dispersed pastures, meaning that their manure was not easily collected and therefore must be excluded from calculations. Another human-managed N input depended on symbiotic fixation, a natural process by which bacteria on root nodules of legumes draw nitrogen from the air and transform it into plant-useable nitrogen in the soil. Symbiotic fixation happens in nature, but on farms it is a human-controlled process because farmers decide how much or how little of their cropland to plant in legumes. In the Great Plains, key legume crops were alfalfa hay and, after 1930, soybeans. Multiplying known rates of symbiotic nitrogen fixation for each crop by its acreage estimates the legume contribution to soil fertility. Seeding contributed a small amount of nitrogen to soils each year, estimated here on the basis of recommended seed application rates in historical agriculture manuals (Hutcheson et al. Reference Hutcheson, Wolfe and Kipps1936). Finally, synthetic fertilizer gained importance in Great Plains agriculture after World War II. US Department of Agriculture data allow crop-based estimates of nitrogen fertilizer application beginning in 1959 (US Department of Agriculture 2013, 2017). The sum of these natural and human-managed vectors represent the total annual nitrogen inputs to agroecosystems.

On the other side of the balance sheet were nitrogen outputs, many of them losses resulting from natural processes. The major human-managed output, of course, was the harvest, which removed nitrogen in the form of grain, hay, and fodder. The estimates used here calculate the nitrogen content of these products, and assume that for grain crops, the stubble and other residues returned to the soil, either directly in the field or as livestock bedding straw mixed with manure. (While farmers sometimes burned off wheat stubble, thus losing its nitrogen content, they more often collected it for livestock bedding or else left it in the ground as a wind erosion prevention.) Hay and fodder crops (like corn and sorghum for silage or forage), by contrast, consumed the entire above-ground portion of the plant, including all its nitrogen content. Census production data multiplied by the nitrogen content of various crops allows an estimate of harvest-based nitrogen outputs from cropland soils.

Because plant-useable nitrogen is volatile, there are a number of natural pathways by which it leaves soil. Agricultural land use could affect these vectors—for example erosion rates—but they were outputs farmers were unaware of or, at least, did not actively manage. Rainfall (and irrigation) leach nitrogen deep into the ground as water trickles through soils, removing it from plants’ root zone. Erosion physically washes away or blows away soil particles, including the nitrogen attached to them. Denitrification is a function of bacterial metabolism—the reverse of fixation—that decomposes plant-available nitrogen into an inert form plants cannot use. And ammonia volatilization is a chemical process by which nitrogen in unstable ammonia escapes into the atmosphere. Scientific studies of these processes make it possible to estimate the amount of natural nitrogen losses based on cropland area (Garcia Ruiz et al. Reference Garcia-Ruiz, Manuel, Gloria, David and Juan2012: 670–75). With the emergence of synthetic N fertilizers from the 1950s, leaching and volatilization increased because as much as half of applied fertilizer never made it into the targeted crop plants.

The sum of all the estimated nitrogen inputs and outputs over a year results in an annual balance, reported here per hectare of cropland (1 hectare = 2.47 acres). If outputs and inputs were equal, then soil nitrogen was in balance, and each new crop year benefitted from the same soil fertility as the previous year, a circumstance that might appear sustainable, at least for this one important agricultural component. When inputs exceeded outputs, soils stockpiled additional nitrogen, building fertility over time, as had been the case for thousands of years before sodbusting. When outputs were higher than inputs, however, the annual nitrogen balance was negative, meaning that next year’s crop inherited less available nitrogen than previously, resulting in soil fertility depletion or “soil mining.” These methods make it possible to track the relative sustainability of soil fertility in one place over an extended period. If applied uniformly, they also allow comparison of soil sustainability from one place to another. This article compares three Kansas counties (figure 1), and this special issue of Social Science History extends the comparison to other locations on either side of the Atlantic.

Nitrogen Flows through a Pioneer Agroecosystem, 1880–1930

Nemaha County, Kansas was the ideal to which nineteenth-century Great Plains settlers aspired. Euro-American pioneers began to make farms there in 1854, but made little headway before the end of the Civil War; by 1875 there were about 10,000 residents who had plowed 12 percent of the land for crops.Footnote ³ The county covers 460,000 acres (186,000 hectares) of low-rolling hills, rich soils, and tallgrass prairie, with woodland along rivers and on steep hillsides. Population density in Nemaha County rose to nearly four people per square mile (10 per km²) by 1889, as the grassland filled with farms and small towns during the first thirty years of agricultural colonization.

Settlers slowly altered their environment, creating an agricultural mosaic of crop fields, pastures, woodlands, and farm yards, crisscrossed by dirt roads and railway lines. The most ecologically important activities were the plow-up of prairie grasses for cropland and the suppression of landscape fire. In Nemaha County, cropland area rose from 15 percent of all land in the 1870s to 60 percent by 1900. Crop diversity was low (figure 3). In its early years, Nemaha County farmers planted corn on 60–70 percent of cropland and cut native prairie hay on most of the remainder. Cropland was in continuous use, with virtually no fallow. The dominant corn crop provided feed for livestock, and grain surpluses flowed into national markets via railroads. This early agricultural system was highly productive. Rich soil, adequate rainfall, and a plenitude of sunshine fueled crop growth, and farmers’ ingenuity and hard work brought in abundant harvests. By the end of the plow-up, settlers had completely transformed the ecological system on a bit more than half of the region’s land. It was the most dramatic environmental impact of the agricultural occupation. The remaining land, the 40 percent not plowed for crops, did not sit idle. There settlers grazed livestock, mostly beef cattle. Domestic cattle meant fences, rearranged water supplies, fire suppression, and intensified grazing and trampling. In Nemaha County, livestock density rose rapidly during the late nineteenth century, from 4 animals per square mile (11 per km²) in 1875 to 12 (30/km²) by the late 1880s.

Figure 3. Area of major crops (hectares), Nemaha County, Kansas, 1880–1997.

After 1900 crop diversity increased. Hay continued roughly stable and corn remained the dominant crop but declined in area as wheat and sorghum gained ground. By the 1920s, corn still represented about half of cropland, and hay, wheat, and sorghum split the other half.

As Nemaha families built new rural communities they depleted soil fertility. Between 1880 and 1910, nitrogen outputs significantly exceeded inputs, creating sharply negative annual nitrogen balances (figure 4). Outputs of 50 to 60 kilograms of nitrogen per hectare were matched by inputs of around 20 kg/ha. For the first several decades of agricultural occupation, Nemaha cropland lost approximately 35 kg of nitrogen per hectare every year. This was a soil mining process, but by world standards crop yields were high (figure 5). Through the 1880s and 1890s gross cropland productivity—the aggregate biomass of all harvested crops—approached 3 tons per hectare. However, by the early 1900s it dropped below 2.5 and in 1930, some 60 years after sod-breaking, yield drifted down to 1.9 tons per hectare. Productivity was still good, but noticeably lower than in the early years.

Figure 4. N inputs and outputs (kilograms per hectare), Nemaha County, Kansas, 1880–1997.

Figure 5. Soil nitrogen balance (kilograms N per hectare on the left axis) and gross crop productivity (tons per hectare on the right axis), Nemaha County, Kansas, 1880–1997.

As yields declined, the gap between nitrogen outputs and inputs narrowed. From 1910 onward, farmers extracted a bit less nitrogen and contributed more than they had done in the late nineteenth century. Annual negative balances continued, but improved steadily after 1890’s low point of –39 kg N/ha. By the 1920s annual balances hovered around –20 kg/ha. Even six decades after initial agricultural colonization, famers mined nitrogen from their soils year in and year out, but more slowly each decade.

Soil mining is a universal process on agricultural frontiers. Human and livestock populations are typically low in the early years, meaning chronic shortages of labor and limited manure availability. Initially rich land relieved farmers of the necessity of replenishing soil fertility. It is not surprising that Kansas settlers mined soil nitrogen, but nutrient balances reveal the extent of those losses and their duration in the context of the local environment and the contemporary economy, technology, and culture. Why did negative nitrogen balances improve somewhat between 1900 and 1930? Part of the answer lies in the slowly declining yields that were a direct consequence of soil mining. Lower yields also meant lower nitrogen extractions through harvest. The livestock dynamics in Nemaha County did not change much. From initial inputs of about 15 kg nitrogen per hectare, manure contributions dropped to about 10 kg/ha thereafter and remained stable (figure 6). The early decline resulted not from fewer animals, but rather from rapidly increasing cropland area as farmers plowed new sod into the twentieth century. One important addition was alfalfa, a legume hay. Virtually nonexistent in 1890, nitrogen inputs from alfalfa’s symbiotic fixation rose to 20 kg/ha by 1920. Even a decline in the next decade left legume contributions about equal to manure. Through these vectors farmers adjusted toward lower annual nitrogen losses by the end of the pioneer era.

Figure 6. Manure, legume, and synthetic fertilizer inputs (kilograms N per hectare), Nemaha County, Kansas, 1880–1997.

It is tempting to seek conscious adaptation toward soil nitrogen sustainability in these changes, but that is probably not what happened. Even as late as 1930, soil nitrogen content was high compared to other long-farmed regions of the world and yields were quite good, if not as good as 50 years earlier. Rather, these improvements in nitrogen dynamics appear to be a lucky by-product of adaptations to more pressing agricultural problems: shortfalls of soil moisture and horse power. Alfalfa was superb feed for horses and cattle, clearly superior to native hay, and farmers grew it to give their animals a boost of protein along with necessary roughage. That was probably their main incentive to increase legume acreage. And crop diversification was primarily a response to low and unreliable rainfall. Corn is a thirsty crop, prone to failure during drought years; as farmers learned that painful lesson they reduced corn acreage to lower risk in dry times. The crops that replaced it—wheat and sorghum—were more drought-hardy than corn. The fact that they were also less demanding of soil nutrients was a bonus, but probably not the driving consideration. Even after 60 years of soil mining, nutrient maintenance was a small concern and a low priority for farmers struggling to keep farm systems productive and sustainable on multiple fronts.

To the west, in Rooks County, Kansas, agricultural colonization began in the 1870s, two decades later than in Nemaha (Cunfer Reference Cunfer2005; Rooks County Historical Society 1981). Considerably dryer climate reduced cropland productivity compared to Nemaha County, altered farmers’ crop choice, and shifted soil nutrient dynamics, in magnitude if not in kind (Miner Reference Miner1986). As with most American pioneer communities, corn was the first crop settlers planted (figure 7). It was the quintessential American subsistence crop, useful for feeding milk cows, hogs, and chickens that sustained the family, and for feeding horses and mules that powered plow-up. But corn was even more vulnerable to drought and hot, drying winds in Rooks County than in wetter eastern Kansas (Miner Reference Miner2006). Corn was risky and did not become the cash crop here. Corn persisted and even increased in Rooks County as population grew through 1930, but wheat became the dominant grain (Malin Reference Malin1944). Wheat was a valuable cash crop for export by rail; it was also relatively drought resistant and less demanding of soil nutrients compared to corn. By 1910, crop choice had diversified, split three ways between wheat, corn, and hay. Twenty years later sorghum replaced hay, but there were still three significant crops in the county. Until the 1920s there was virtually no fallow land, with all cropland in continuous production.

Figure 7. Area of major crops (hectares), Rooks County, Kansas, 1880–1997.

Nitrogen outputs in Rooks County significantly exceeded inputs throughout the pioneer era (figure 8). Outputs were virtually flat between 1880 and 1930, at about 40 kg N/ha per year, while inputs hovered around 18 kg/ha. The result was an annual soil nitrogen balance of about –24 kg N/ha. This was soil mining, but the magnitude of nitrogen flows was smaller than in wetter Nemaha County. Annual outputs and inputs were each about 8–10 kg N/ha lower than in the east, as was the nitrogen balance (figure 9). Crop yields were also lower in the dry west, at about 2 tons of gross biomass per hectare across all crop types. Such productivity was still quite good, and it remained stable for more than five decades.

Figure 8. N inputs and outputs (kilograms per hectare), Rooks County, Kansas, 1880–1997.

Figure 9. Soil nitrogen balance (kilograms N per hectare on the left axis) and gross crop productivity (tons per hectare on the right axis), Rooks County, Kansas, 1880–1997.

Initial soil nitrogen content was even richer in Rooks County than in Nemaha County, around 11 tons N/ha. But farmers there did not narrow the gap between nitrogen outputs and inputs as happened in Nemaha. Nitrogen balances in Rooks County were about –23 kg/ha in the 1880s, and remained at –25 kg/ha in 1930. Here there was no evidence of soil nutrient adaptation, whether intentional or incidental, through half a century of agricultural colonization. The county started at a somewhat higher level, mined soils at a slower pace, and as late as 1930 had not yet depleted soil fertility enough to reduce yields or prompt significant adjustments. Farmers applied what manure they had available, but took no other action to address soil fertility, neither planting legumes nor rotating fallow. Even with significantly reduced soil nitrogen by 1930, there was still enough available to support abundant wheat harvests each year, if rainfall was sufficient.

Manure availability declined throughout the era, dropping from about 10 kg N/ha in 1880 to below 4 kg/ha in 1930, by which time the transition from horse power to tractors was well under way (figure 10) (Cunfer Reference Cunfer2005; Fitzgerald Reference Fitzgerald2003). A brief introduction of alfalfa brought a legume crop to Rooks County between 1900 and 1925. For a decade legume nitrogen contributions matched declining inputs from manure, but then fell away again by 1930 and were never more than a small component of Rooks County’s nitrogen equation.

Figure 10. Manure, legume, and synthetic fertilizer inputs (kilograms N per hectare), Rooks County, Kansas, 1880–1997.

Crisis and Recovery, 1930–50

By 1930 the pioneer era of agricultural colonization on the Great Plains was complete. Nearly all land that would ever be plowed was already in cropland, and population peaked in many places in the 1920s (Cunfer Reference Cunfer2005; Sylvester et al. Reference Sylvester, Myron and Brown2016). Extensification drew to a close by 1930. Intensification, however, gathered speed after 1950 and accelerated through the second half of the twentieth century (Cunfer et al. Reference Cunfer, Andrew and Joshua2018). But between those years were two decades of crisis, turmoil, and recovery in the grasslands. The decade-long Dust Bowl drought that began in 1932 is well known. Accompanied by high temperatures that evaporated scarce soil moisture, the weather downturn caused repeated crop failures and wind erosion, from routine field-level blowing to massive dust storms that tracked across the continent. These environmental conditions caused bankruptcy, land abandonment, and emigration (Egan Reference Egan2006; Hurt Reference Hurt1981; Worster Reference Worster1979). The environmental crisis coincided with an international economic collapse, the Great Depression. That combination of regional disaster and global depression hit plains residents hard. It also opened a political opportunity for revolutionary changes in the US agricultural economy.

For the first time in American history the federal government took an active role in managing the farm economy, and even made a limited (and ultimately incomplete) effort to direct land use (Duffin Reference Duffin2007; Gregg Reference Gregg2010; Kirkendall Reference Kirkendall1966). Beginning in 1933, new federal government agencies offered farmers annual commodity subsidies (Agricultural Adjustment Administration), price controls (Commodity Credit Corporation), low-cost credit (Farm Credit Administration), and occasional compensation for crop failures (Cunfer Reference Cunfer2005; Grant Reference Grant2002). The New Deal agencies resettled some farmers, converted some cropland to pasture, and attempted to manage annual production through land set-aside programs (Maher Reference Maher2008; Phillips Reference Phillips2007). Beginning as emergency relief programs designed to keep farmers from starving or losing their land, they evolved into economic recovery programs and, by the late 1940s, became permanent components of the nation’s agricultural economy (Danbom Reference Danbom1995). Fixed in outline, if constantly tweaked and adjusted by Congress, federal agencies created a reliable safety net that by turns raised farm incomes, brought urban standards of living to rural places, increased farm sizes, and depopulated small communities (Hurt Reference Hurt2011).

World War II demand for food and fiber coincided with the end of the long drought in 1941. Both events made the 1940s a time of rebuilding from the deep losses of the Depression. Farmers reinvested in their properties, rebuilt livestock herds, and became prosperous again (Grant Reference Grant2002; Hurt Reference Hurt2011). The two decades between 1930 and 1950 were a topsy-turvy time for plains farmers on almost every count, from weather to environment to economy to politics to society. The disruptions of the period affected soil nutrient dynamics in a variety of ways, causing marked fluctuations in nitrogen indicators. For example, drought-induced crop failure in 1935 and 1940 in Rooks County dramatically reduced yields and thus harvest-related nitrogen outputs, spiking the nitrogen balance dramatically upward (figure 9).

Early in the 1930s, livestock populations collapsed in many plains communities (Cunfer Reference Cunfer2001). When hay and fodder crops failed and pasture grasses dried up, livestock faced starvation. Farmers sold or slaughtered cattle by the millions because they could not feed them. The federal Drought Relief Service paid farmers a few dollars per head for emaciated cattle and simply buried their carcasses in trenches (Worster Reference Worster1979). One consequence of livestock reduction was lower manure nitrogen inputs to soils in subsequent years. In many places cattle herds did not recover for a decade or more. As manure contributions fell, legumes could not make up the difference. Both alfalfa and recently introduced soybeans—the two significant plains legume crops—needed considerable soil moisture to thrive. The drought hit legume crops hard, and farmers planted fewer of them during the 1930s, then increased acreage again in the 1940s (figures 6 and 10). By 1941 rains finally returned, but it took the rest of the decade to rebuild agroecosystems and rural society. Farmers restocked pastures, replanted legume crops, began to invest in synthetic N fertilizers, and developed new irrigation systems. Crisis and catharsis led on to recovery and restoration. By the end of the 1940s the Great Plains was poised for a dramatically new direction. Grassland agriculture would look very different after 1950 than it had before 1930.

The Socioecological Transition to Modern Industrial Agriculture, 1950–97

Beginning about 1950, and continuing through the end of the twentieth century, farming underwent a “socioecological transition” (Cunfer et al. Reference Cunfer, Andrew and Joshua2018; Fischer-Kowalski and Haberl Reference Fischer-Kowalski and Helmut2007). This change encompassed soil nutrient management, energy, machinery, pest control, seed breeding, landscape structure, rural demography, and economics. A restructuring of the biophysical basis of agriculture revolutionized the fundamental relationship between people and nature in agroecosystems (González de Molina and Toledo Reference Gonzalez de Molina and Toledo2014). While this transformation affected multiple components of agriculture, soil fertility serves as a representative indicator of sustainability. How did farmers restructure nitrogen outputs and inputs as this rural revolution took off and accelerated during the second half of the twentieth century?

In 1950, Great Plains soils remained nutrient-rich, though 70+ years of soil mining left nitrogen contents far below their initial levels (Haas et al. Reference Haas, Evans and Miles1957). With sufficient rain, farmers produced yields of around 2 tons per hectare—only slightly lower than the bumper crops of the late nineteenth century and still quite reasonable. But new technology was about to change the nitrogen equation for farmers. In 1909, German chemist Fritz Haber discovered how to synthesize ammonia and soon thereafter engineer Carl Bosch manufactured it at industrial scales (Smil Reference Smil2001). The technology depended on fossil fuels to supply both the raw material of nitrogen fertilizer and the high temperatures and pressures required to produce necessary chemical reactions. For the first time, people could fix nitrogen in a plant-useable form; previously only lightning and a handful of bacteria species could do so. Now human beings intervened in the global nitrogen cycle, with dramatic consequences. With the onset of World War I, synthetic nitrogen fixation first fed the munitions industry, not agriculture, generating ammonia for high explosives. The disruptions of the early twentieth century—economic Depression and another world war—delayed full deployment of ammonia synthesis for fertilizer use. But by the late 1940s, low cost and highly effective fertilizer products began to spread through the US agricultural economy, not least in the Great Plains (Cunfer Reference Cunfer2004; Duffin Reference Duffin2007).

Data from the US Department of Agriculture make it possible to track fertilizer use beginning in 1959 (US Department of Agriculture 2013, 2017). By then Nemaha County farmers were already applying an average of 15 kg N/ha on their cropland, matching legume inputs and tripling contributions from manure (figure 6). Fertilizer use skyrocketed thereafter, climbing nearly every year for the next 40 years. By the end of the century Nemaha County farmers applied 65 kg N/ha to their land, nearly triple their total nitrogen inputs in 1880. Legume contributions rose in tandem, primarily from expanding acreages of soybeans, but also from alfalfa. Combined, nitrogen inputs nearly quintupled, from 30 kg N/ha in 1950 to 145 kg/ha in 1997 (figure 4). Never in history had farmers been able to deliver nutrients to their crops in these quantities—it was truly revolutionary.

Nitrogen outputs rose in tandem with inputs, primarily due to dramatically increased crop yields plus losses from synthetic fertilizer that leached or volatilized N into the environment (figure 4). By 1997, Nemaha croplands that had done well to produce 3 tons per hectare a century earlier were producing more than 8 tons/ha of total crop biomass (figure 5). The gap between nitrogen outputs and inputs narrowed and then disappeared. Soil nitrogen balances remained negative, but ever less so, and in the 1990s approached positive territory for the first time in 150 years of agricultural occupation (figure 5). In 1997, farmers harvested 8 tons per hectare in all crops combined and still managed to supply virtually as much nitrogen as they extracted, in roughly equal parts from synthetic fertilizer and legumes (figure 6). From one point of view they had achieved a level of soil nutrient sustainability never before accomplished. From another perspective, they did this by decoupling themselves from any real reliance on soil fertility at all. Rather, they delivered plant nutrients directly as fertilizer or legume residue, then harvested those nutrients in marketable form later in the year, all but bypassing the soil.

In dry Rooks County synthetic fertilizer inputs were 10 kg N/ha in 1959, a third lower than in Nemaha County because of the smaller nutrient demands of wheat compared to corn (figure 10). The late-twentieth-century increase in fertilizer application was dramatic, but only reached 35 kg N/ha in 1997, just more than half the level in Nemaha. Legumes remained unimportant in the county, unlike in the east. Total nitrogen inputs quintupled, from 10 kg N/ha in 1950 to 50 kg/ha in 1997, but never approached the dizzying heights of the 145 kg/ha in Nemaha County (figure 8). Even in the fertilizer era, dry central Kansas was no match for the gross biomass productivity of wetter eastern Kansas. Still, the transformation was remarkable. Yields of all crops combined rose from 1.3 tons per hectare to more than 4 during the half century, and the gap between nitrogen outputs and inputs narrowed, but did not close (figures 8 and 9). Annual nitrogen balances, that had hovered around –24 kg N/ha prior to 1930, moved into a range between –15 and –10 kg/ha after 1959 (figure 9). Soil mining continued, but at a slower pace.

One more comparison bears scrutiny. South of Rooks County lies Pawnee County, Kansas, in the same semiarid rainfall zone. Farmers there had a virtually identical land use and soil nutrient history as Rooks between 1880 and 1930. Initial settlers grew corn, but wheat quickly dominated (figure 11). Between 1910 and the early 1950s, sorghum increased as a drought-resistant alternative grain that could feed livestock in place of corn.

Figure 11. Area of major crops (hectares), Pawnee County, Kansas, 1880–1997.

But there were some noticeable differences between the two western Kansas counties in their soil fertility dynamics. First, legume nitrogen contributions rose alongside fertilizer inputs after 1959, much like in wetter Nemaha County, and in clear contrast to similarly dry Rooks County (figure 12). Nitrogen outputs and inputs rose through the second half of the twentieth century and the gap between them narrowed (figure 13). As in Nemaha County (but not in Rooks), Pawnee County’s soil nitrogen balance closed the gap and moved briefly toward positive territory for the first time in the 1990s (figure 14). Nitrogen outputs and inputs rose quickly, not as high as in Nemaha but still to levels some 40 percent higher than those in nearby Rooks County. Yields spiked too, up to nearly 6 tons per hectare in the 1990s, when Rooks could only muster 4 tons/ha. Why did Pawnee County’s nitrogen profile look more like Rooks County before 1930 but more like Nemaha County after 1950? How did Pawnee County outperform Rooks so noticeably through the late-twentieth-century socioecological transition, when they shared virtually identical climate and soil conditions and had been so similar before?

Figure 12. Manure, legume, and synthetic fertilizer N inputs (kilograms N per hectare), Pawnee County, Kansas, 1880–1997.

Figure 13. N inputs and outputs (kilograms per hectare), Pawnee County, Kansas, 1880–1997.

Figure 14. Soil nitrogen balance (kilograms N per hectare on the left axis) and gross crop productivity (tons per hectare on the right axis), Pawnee County, Kansas, 1880–1997.

The answer relates to another form of intensification that emerged after 1950 in selected parts of the Great Plains: irrigation (Green Reference Green1973; Opie Reference Opie1993; Watson Reference Watson2020). Rooks County remained a dryland crop area right through the twentieth century, without significant access to either surface water from nearby rivers or to groundwater from aquifers. Pawnee County farmers, however, developed irrigation from river-associated shallow aquifers after 1950, when they irrigated fewer than 3,800 acres (1,500 hectares). In 1978 farmers irrigated 57,000 acres (23,000 hectares); by 1997 it was 70,000 (28,000 hectares). Rooks County, in the latter year, irrigated only 1,700 acres (670 hectares). Irrigating farmers in Pawnee County spread their water across several crops, including corn, alfalfa, soybeans, and wheat (US Department of Agriculture 1978). One consequence of irrigation was an increase in legumes; another was crop diversification. Corn and soybeans area rose from near 0 to 25,000 acres (10,000 hectares) each, virtually all of it irrigated, and hay also increased in area. In the meantime, wheat acreage fell. Irrigation allowed farmers to reduce wheat monoculture and diversify. Irrigation also raised crop yields and reduced the risk of crop failure during droughts. But applying water to cropland subsequently required higher inputs of synthetic nitrogen fertilizer because the extra water running through soils increased leaching (Cunfer Reference Cunfer2005). Irrigated land had higher nitrogen losses and therefore needed higher nitrogen inputs compared to dry cropland. The advent of irrigation explains why Pawnee and Rooks Counties diverged after 1950, despite tracking one another closely during the previous 70 years.