The Natural History of the Present, Chapter 14

Part II: Future Landscapes: Atmospheric Agriculture and the Upland Ocean

Chapter 14: The Buffalo Plains

As one goes west from the Mississippi River, which up to the toe of Illinois, 600 miles from the Gulf, is little above sea level, the land rises and dries. It becomes higher and drier as one approaches the rain shadow of the Rockies. The plains west of the 100th meridan (the line where the West starts) range from 2000 to 4000 feet in elevation and are quite flat. The soils, subhumid and little leached by rainfall, are fertile. Winters in the northern plains can be extremely cold, with severe blizzards. The native grasses are short and the grass cover is discontinuous. Sunshine and wind are abundant. Rainfall is low and variable. The plains experience droughts of up to a decade, during which the grass cover may retreat by 70%, exposing much bare soil. For farmers the long-term probability of rainfall in a given place is more important than the rainfall average over a larger area: that is, in some specific places, because of topography and airflow, rain is more likely. Light rain showers are common but a third of the annual rainfall can fall in an hour. Without irrrigation, farming on much of the plains is a gamble (some say impossible); but the Great Plains are huge, and until recently produced perhaps half of the wheat traded on world markets (much of it irrigated, some dryfarmed). Escarpments to the east and south lead down to the tall grass prairies and perennial streams of the lands nearer the Mississippi Valley.

Earnest Thompson Seton, a Canadian naturalist, estimated the extent of the plains biome in North America at 750,000 square miles. The so-called buffalo commons, a buffalo ranching scheme proposed a decade ago for the depopulating, poorer quality farmlands of the Dakotas, Texas, Wyoming and Nebraska was much smaller, 147,000 square miles. From Nebraska, including a bit of southern South Dakota, south through Texas and New Mexico, approximately 220,000 square miles of the plains are underlain by an aquifer of fossil water, the High Plains Regional Aquifer. This is water stored from a previously rainy time, some of it glacial meltwater. Much this aquifer consists of the better-known Ogallala Aquifer. The depth to the water varies. Before pumping began, it lay within reach of hand-dug wells in parts of Nebraska: home-made windmills, of many different designs, were characteristic of that part of the plains. Some of the water irrigated truck crops. (A decade after Custer’s defeat on the Little Bighorn, one truck farmer commented that if his eggplants sold as well in the Lincoln market as they grew, he’d be rich.) Even now, with much of it pumped down (water pumped for agriculture is about half of what it was in the 1970s), groundwater flow from the aquifer provides substantial parts of the dry season flow of the rivers that cut across the plains. Groundwater pumping for irrigation agriculture has eliminated much of the watery habitat for fish, amphibians and mammals on the plains. Sections of some Kansas rivers now dry up in summer. A hundred and fifty years ago when small streams and seasonal wetlands were more common, the land was more easily habitable. Vernal and autumnal pools provided habitat for shorebirds and amphibians; buffalo dug out tiny seeps for wallows; and prairie dog towns, with their avian, mammalian and reptilian hangers-on, occupied one fifth of the landscape.

The buffalo on the North American Plains constituted the greatest concentration of large herbivores since the Pleistocene; greater than that of any single species on the African plains. Seton estimated the North American buffalo population at European contact at approximately 75 million animals. Before the Europeans arrived buffalo ranged over the plains and the prairies and into the eastern woodlands, and were found in small numbers in the Great Basin, whose grasses are not adapted to continuous grazing, and on the southern parts of the Atlantic Coastal Plain. Seton’s estimate was based on the carrying capacity of the different regions for cattle and horses. The prairie, which Seton called the best habitat, is now all farmland. (Whether it was in fact the best habitat for the buffalo is uncertain.) Prairie was said to support one animal per 7 or 8 acres. One acre of prairie was said to equal 4 acres of plains or woodland; and the estimate followed from these calculations. Seton’s estimate was undoubtedly too high, but the modern rule of thumb is that grassland that will support 2 cattle will support 3 buffalo, despite their similarity in size. Buffalo are browsers as well as grazers and stand cold much better than cattle. During storms, they browse into the wind, rather than downwind like cattle, and by moving this way, historically for long distances, shorten the time the storm lasts. Unlike cattle, buffalo go into a state of reduced metabolism in winter, which reduces their need for food. Most writers now think the short, protein-rich grasses of the High Plains supported a population of 25 to 30 million animals; and was the center of the buffalo’s distribution. The short grasses of the plains are 15-20% protein when dry, higher in protein than the taller grasses of the prairie. They reproduce vegetatively and withstand grazing better. Buffalo need a minimum of 1 part protein to 6 parts carbohydrate, so the plains in winter, with their cured high-protein grasses exposed on windy slopes, might not have been such a bad wintering ground. Most northern mammals are adapted to cold; food is their limiting factor. Buffalo are also migratory. Their hump acts as a suspension system for their legs and lets them maintain a steady, rapid gait (a canter-gallop) for hours at a time. Their capacity for rapid long-distance movement is an adaptation for a steppe environment, where grassses grow thinly and the forage in any one place is limited in amount, where the pattern of high-quality grasses forms a mosaic, and peak forage quality occurs at different times in different places. So buffalo, like other migratory grazers, were here today, gone tomorrow. Animals that eat fibrous forage also need water daily; buffalo will eat snow and drink from springs, but water is sometimes distant on the plains. Buffalo hold their heads low, a sign of an adaptation to a short sward. They probably moved east to the better watered prairies in times of drought. (During long droughts the short grasses also moved east, taking over ground from the taller grasses, among which they live.) Buffalo dig out seeps in damp depressions to drink from (several per acre in good environments; this waters the landscape for other creatures) and can smell surface water from several miles away. After drinking from a stream, they move away from it to feed and rest, being much less an animal of stream margins than cattle, who need more water and tend to hang around on riverbanks. (Cattle are naturally animals of forest edges and streams.) Buffalo probably stopped breeding during droughts, and death rates would have gone up, so the population would have fallen. But like the short grasses, buffalo populations are capable of responding quickly to improved conditions. Various lineages of buffalo have probably grazed the North American Plains (which as a habitat also moved around) for the last 200,000 years.

Perhaps 1.5 million wolves lived with the buffalo and killed up to a third of the calves yearly. (In parts of Alaska grizzly bears kill 50-90% of newborn moose.) So buffalo were limited by predation by wolves on calves; by predation by people, with whom the modern buffalo developed; and by weather. Drowning, falls, grizzly bears and fires also killed some buffalo. Before the Plains Indians had horses, their take of buffalo is thought to have been 300,000 animals a year. These animals were mostly killed in the summer with drives or fire surrounds. A healthy population of buffalo can withstand predation of 7% of its breeding females annually without being reduced. Out of 30 million buffalo, this is about a million animals, if half the population was female. Probably the majority of buffalo killed were young females, whose meat was preferred and whose more pliable hides were easier to process. Early observers on the plains thought the buffalo calved only every other year, since only half the cows had calves; but healthy populations of ungulates calve every year, and this observation was probably the result of a natural calving rate near 80%, with a third of those calves being eaten by wolves.

Where buffalo numbers were limited, as in the eastern woodlands, they may have been overhunted, as they were near the agriculturalists of the Mississippi Valley. Areas inaccessible to hunters because of tribal warfare served as refuges, whose surplus animals drifted into the more heavily hunted areas. Buffalo were not alone on the plains; they shared the landscape with several other herbivores: white-tailed deer in the river valleys; mule deer in the more broken landscapes; the plains elk, a summer grazer like the bison (elk are probably a plains animal that now live in the mountains); bighorn sheep, now also an inhabitant of the hills; pronghorn antelope in the drier sagebrush habitats; various species of grouse, including sage grouse and sharptails; jackrabbits; and in the lighter soils, prairie dogs. Prairie dog towns occupied 20% of the short and mixed grass plains. So the pre-contact plains, so empty to the European farmer’s eye in the 1880s, were marked out by their populations of plants (250 species), and the animals that ate them; by buffalo trails and wallows; by prairie dog towns; by buffalo traps; by the rings of stones that marked the encampments of the Hidatsa or Arapahoe; by Indian trails; by waterways. On another level they were marked by long erosional processes; by the airsheds and flows of the atmosphere that determined rainfall; by sunlight; and by stored carbon. Settlers arriving when the animals were gone commented on their expansiveness and silence. Buffalo trails were the basis of the Wilderness Trail from Virginia through the Cumberland Gap to Kentucky; of the rail line from the Potomac River to the Ohio; of the route of the Union Pacific Railroad up the Platte River in Nebraska.

By the early nineteenth century, the prairies, oak savannahs and woodlands of the Old Northwest, part of the Louisiana Purchase, were being surveyed by teams employed by the government. Unlike the animals, or the Native Americans who depended on them, and whose stories and foot paths interwove with the pastures and trails of the animals, the surveyors marked out the landscape into rectangles, a mile on a side. The survey created Jefferson’s enlightenment landscape, the useable, marketable landscape whose pale squares cover the Middle West and plains as seen from the air today. The survey was another one of those breath-takingly simple schemes by which the West has taken over the world. The survey lines ignored topography and led to many difficulties for the men. Country considered home by the natives, who knew what times of day and what seasons to use different places, and how to get around, was often described as a hell by the surveyors, who went in a straight line from sun-up to sun-down regardless of the season, or of the country’s abundant insect life. By marking out the territory as saleable land, the surveyors changed the lives of large animals like the buffalo. The survey meant the Plains would be settled up to the limit of rainfall necessary to grow wheat: 15 to 18 inches annually, or approximately to the 100th meridian; that is, a line from Bismarck, South Dakota to Amarillo, Texas; a line that divides the humid continental steppe (subtropical in some parts of Texas) from the subhumid dry one; and the so-called eastern birds and mammals from the western ones. The exact position of the line can shift a few hundred miles east or west depending on rainfall. This line is also notable now in that west of it, on the high, dry, depopulated lands between the wheat fields and the Rockies, exists the only landscape in which it is possible to imagine the return of the buffalo.

The buffalo was one of the keystone animals of the plains. Large herbivores raise the productivity of grasslands and manipulate its structure. They vary the habitat. A prairie, like a forest, consists of a shifting mosaic of habitat patches, the result of disturbance (by grazing herbivores, by digging rodents, by fires). Many prairie birds are adapted to the different habitat patches. Buffalo grazing and trampling created bald spots, and (as the vegetation recovered) patches of herbs. By digging wallows and pawing out springs they made surface water available for other animals. Grasses produce biomass more quickly than the dead matter can decompose; herbivores and fire (the megaherbivore) increase a grassland’s productivity by speeding up the recycling of nutrients. By reducing the amount of flammable material, grazing makes fire on the plains less severe. (Most lightening fires on the plains occur during the growing season and burn less than a acre.) Another plains herbivore was the Rocky Mountain locust. Rocky Mountain locusts are probably now extinct thanks to the plowing of the high river bottomlands where the insects bred. They are now studied in part by collecting the animals that thaw out of mountain glaciers. The overall effect of the locusts isn’t known, but may have been substantial. A swarm 1800 miles long and 110 miles wide, consisting of some 6000 tons of insects was reported in 1875 in Nebraska. A typical swarm ate 50 tons of vegetation a day. (Five acres of a good corn crop, if the insects ate the corn to the ground, more like 25 acres in 1880, or—since the vegetation was not totally consumed—tens to hundreds of acres of grassland.) Another grazer, one whose range corresponds rather closely with the High Plains Aquifer and the short grass prairie, is the white-tailed prairie dog. Prairie dogs were another keystone animal of the plains. In general prairie dogs prefer sandy soils and short grass, though the black-tailed prairie dog lives amid the taller mixed-grass prairies to the east. Their strategy for avoiding predators is to keep the grass about their burrows nibbled short and thus be able to see approaching prairie falcons, golden eagles, ferruginous hawks, foxes and coyotes; short grass did not impede the black-footed ferrets, which pursued prairie dogs down into their burrows. The constantly re-growing grass was also nutritious. Nibbling prairie dogs controlled invasive shrubs, such as mesquite. Prairie dog burrows reached densities of 50 per acre in prairie dog towns. The burrows went down 10 feet. It is thought that 100-250 million acres, one fifth of the plains, had such towns, so prairie dogs were as abundant as passenger pigeons, with something like 5 billion animals. The animals moved tens of millions of cubic yards of earth annually; this reshuffling raised soil fertility. The towns with their short grasses and burrows also served as aquifer recharge areas. Water penetrated through to underground more easily than in undisturbed grassland. The towns’ nutritious grasses, fertilized by prairie dog droppings, urine, new soil, and insect parts, was preferentially grazed by buffalo. Where they remain, prairie dog towns are now preferentially grazed by cattle, though cattle gain slightly less weight grazing on prairie dog towns than on adjacent pastures. Grasses growing in the towns are reduced in quantity by a third to a half by the constant grazing by the prairie dogs, but are more digestible and nutritious. The prairie dogs needed a large grazer like the buffalo to help keep the grass low. Both they and the buffalo benefitted from the constantly mown, fertilized grasses of the towns. The urine-soaked grass at the entrances to the burrows held salts needed by the buffalo. Prairie dog towns were refuges from grass fires, as the short, green grasses wouldn’t burn. And prairie dog towns, like burned-over areas, would green up early in the spring.

The twentieth century has seen prairie dogs rendered virtually exitinct on the American plains. Prairie dogs were removed because they were thought to compete with cattle for grass. Removing the prairie dogs greatly simplified the biota of the plains, by doing away with the animals that depended on the prairie dogs. These included not only their predators, but the box turtles, cottontail rabbits, skunks, burrowing owls, snakes, insects and amphibians that used prairie dog burrows. The yearly temperature 4 feet underground in a prairie dog burrow in the Oklahoma plains ranges from 50° F. to 80° F., compared to -10º F. to 120º F. at the soil’s surface. Ten species of amphibians, 15 of reptiles, 101 of birds, and 37 mammals found food or shelter in prairie dog towns. The burrows were a sort of built environment on the grassland: a constructed oasis. C. Hart Merriam, an influential nineteenth century biologist, pointed out that 256 prairie dogs eat the same amount of grass as 1 cow. What Merriam didn’t know was that prairie dogs and buffalo grazing together result in more total biomass (more grass-fed animal and more available grass) than the buffalo or prairie dogs grazing alone. (Prairie dogs reduce the forage effectively available to cattle by only 4-7%. Eliminating them was never economic.) The biomass of the prairie dogs was substantial. At 1 buffalo per 30 acres, an acre of grassland supported approximately 33 pounds of buffalo. At 50 animals per acre, an acre of prairie dog town supported 100 pounds of prairie dog. By comparison, the richest parts of the Serengeti in Africa, which have more rainfall, support 100 pounds of mammalian herbivore per acre (plus 50 pounds of termites). The sensible thing in this case would have been to manage the prairie dogs for leather and meat; or for the skins of their predators (coyotes, foxes, ferrets). Prairie dogs were probably not limited by predation but by competition with each other and by bubonic plague, a microbial predator, which was endemic in the colonies. Now of course they are limited by federally-funded trapping and poisoning programs.

So it is unlikely that buffalo and prairie dogs limited each other’s populations. Grazers that share a landscape tend to partition the environment, concentrating on different parts of it. Large grazers can move great distances (away from drought and prairie dog towns) and tend to raise the potential productivity of grassland, partly by grazing and trampling the grass, partly by temporarily storing grassland nutrients in their bodies and recycling them in faeces and urine. In general, once rainfall reaches 800 millimeters a year (about 31 inches), soil fertility declines because of leaching by rainfall. On African savannahs with large grazers soil nutrients peak at rainfalls of 900 to 1100 millimeters per year. The nutrients herbivores store in their bodies are released slowly in urine and faeces (which after release take some time to break down). Those that remain in their bodies end up stored in the bodies of their predators and scavengers, and released in their faeces and urine. A writer has called the plains piss-driven. Grazing and trampling also promote regrowth and more rapid recycling of nutrients (and thus more nutrients held in the grasses, the soil and the animals). Moderate grazing can double the productivity of grasses and increase their nutrient content and digestibility—which is why prairie dogs didn’t limit the numbers of the buffalo. The situation is a bit more complicated than this: many plant-eaters leave growth-promoting substances on the cut surfaces of the plants; cow saliva contains thiamine that promotes the growth of grass, mouse saliva contains epidermal growth factor, moose browsing encourages branching in the browsed plants. Thus animals and their food plants dance around each other.

The buffalo shared the plains with several other large herbivores whose total biomass was a not-inconsiderable fraction of theirs: elk, with an estimated pre-contact continental population of 10 million, perhaps 2 to 4 million on the plains (perhaps more), and an average body weight of 700 pounds; 9 million pronghorned antelope (out of a continental population of 30 to 40 million), with an average weight of about 100 pounds; 2 to 4 million bighorn sheep (no one really knows), at 200 pounds each; 3.5 million mule deer, abundant in lowlands, rough country, and river valleys; an unknown number of white-tailed deer, an eastern species that with other eastern species like the oppossum colonized the plains up the river valleys; and millions of black-tailed jackrabbits, an abundant small herbivore with a strongly cyclical population, important in some Native American diets, especially west of the plains in the Great Basin. Since most of these animals were shot out relatively quickly and some (especially the sheep) were susceptible to the diseases and parasites of domestic livestock, all these population figures are estimates. Perhaps the biomass of the prairie dogs was 30% that of the buffalo, and that of the large grazers also 25-30% of the buffalo’s biomass. The elk and bighorn sheep may have made altitudinal migrations in summer up to the pastures of the Rockies, or into the Black Hills of South Dakota, as they do today, thus effectively expanding the acreage of the plains (modern bighorn sheep grow larger if they are able to make such migrations); or they may not have and their modern movements may be an attempt to avoid humans. Some populations of both probably stayed down on the plains in winter, browsing in bad weather in the gulleys and the timbered river valleys. Browse is important in the diets of both. Mule deer prefer rough ground, where they can escape wolves, and white-tails were restricted to the river valleys. These rivers also had beaver until the 1840s, and some of the Indian tribes netted fish when they ran upstream in the spring. Antelope are primarily browsers, eat almost no grass and thus compete little with cattle or buffalo. A large part of the diet of the other large herbivores is also browse. Buffalo and elk prefer grass, but will eat browse. (Half of elks’ winter diet is browse.) Pronghorn eat sagebrush (it is the basis of their winter diet), as well as various prairie forbs. Cattle will not eat sage, though buffalo will: sage, a common component of American dryland ecosystems, has a higher protein content and more carbohydrates and fat than alfalfa. Many smaller prairie herbivores, such as sage grouse and black-tailed jackrabbits, also eat it. With their different eating habits, migration patterns, and environmental preferences, the large herbivores partitioned the plains habitat. Some direct competition existed, for instance between elk and buffalo for the standing dead grass of late winter and early spring. Such habitat partitioning used to be taken advantage of by old-fashioned farmers, who would pasture a sheep with each several head of cattle. The sheep would clean up the growth the cattle wouldn’t eat and were thus “free.” This was at a time when the income from 10 or 20 sheep mattered.

In general, a single grazer cannot use the natural biomass of an area efficiently. For this reason, innovative grass farmers follow a herd of cattle with a flock of chickens, which eat the grass the cattle didn’t, and also the insect larvae in the cattle’s dung. A mix of browsers and grazers would provide a higher yield of hides and protein in the dry rangelands of the Sonoran desert than cattle; on this basis some have suggested re-introducing the elephant, Chacoan peccary and llama (as well as the panther, lion and jaguar) to the Southwest. While buffalo by themselves are not efficient users of the plains flora, they are better at it than cattle. They have a more efficient digestive system and eat a wider range of plants. Buffalo are a dryland species, like gazelle, which are 3 times more productive on African savannahs than cattle. The lack of adaptation of cattle to the climate (they withstand drought and cold poorly), together with their effect on streams, where they browse down the willows, trample the banks, and near which they leave much of their daily 20 pounds of urine and 50 pounds of manure, makes them a poor choice of animal to raise on the plains. Several writers have pointed out that without agricultural subsidies for cattle and wheat, farmers would start raising buffalo: nothing else would make money. (In North Dakota, a cattle and wheat state, agricultural subsidies constitute 80% of farm income.) A decade or more ago, two sociologists from New Jersey proposed a communal buffalo range on the drier, poorer parts of the plains, which are now more empty than when Native Americans hunted buffalo there. An idea of what a managed buffalo range would produce can be got from modern examples. The productivity of wild animal populations is commonly manipulated by controlling their sex and age ratios. In Custer State Park in Montana in the 1970s a herd of 1000 buffalo (the winter count of the population, following the fall round-up and slaughter) was kept on 73,000 acres. The ratio of cows to bulls was kept at 10:1. Buffalo are serial monogamists; that is, a bull stays with a cow until she is bred and then approaches another; this takes time and sets a lower limit on the number of bulls. So 10% of the adult animals are bulls, 60% are cows of breeding age, the remaining 30% are calves, yearlings and two-year olds. The breeding cows are slaughtered at 10 years. Some breeding bulls are also kept this long and sold to hunters as trophy animals. Bulls are reliable breeders at 3 years and still edible if slaughtered at 4 years. The manager at Custer State Park tried to keep 70 to 80 animals in each age class up to 10 years. At this stocking rate and under this management regime, the calving rate was 90%, with some twins; some calves were born to two-year old heifers (the usual age of first calf in buffalo is 3 years). Ranchers call this a healthy herd; that is, one in which the animals are capable of rapid growth and reliable reproduction. The herd produced 500 calves annually. With a 5% death rate, and slaughter at 2 years, the herd was producing 450 slaughterable animals a year (90% of its yearly increase), or about 1 animal every 162 acres. (Not all the two-year olds were slaughtered, some went into replacements of older animals that were slaughtered.) When at one point Custer State Park let its herd get to 2600 buffalo, there was some deterioration in the range and the calving rate fell to 60-80%. A calving rate of 80% isn’t bad, and with so many more breeders the yearly slaughter would have been considerably higher: such numbers explain why pastures get overgrazed. With the 1000 wintering buffalo in the Park are 500 elk, 150 white-tailed deer, 120 bighorn sheep, and 120 pronghorn antelope. These animals constitute a total biomass of 35-40% that of the buffalo. So if we take these numbers as a rule of thumb, the plains could produce 1 buffalo carcass for every hundred plus acres, plus an equivalent number, at something less than half the weight, of the carcasses of other large herbivores. These numbers are based on a stocking rate higher than that of Custer State Park, but Custer State Park, in south-central Montana, is very dry. (The carrying capacity of a habitat varies with rainfall. The National Bison Range in Flathead, Montana, is also dry, and runs 1 buffalo for 56 acres and also has a 90% calving rate.) Current estimates for a bison population on the plains are one animal per 24 to 30 acres (high for the drier parts of the plains, where one animal per 50 acres is probably more reasonable), or 1 carcass for double that amount. These animals are produced with no additional feed and very little manipulation of the landscape. The sex, age structure, and size of the herd is manipulated to increase its productivity.

The animals taken from the manipulated herd amount to 45% percent of its fall count (its “population”), compared to a sustainable take of 7% of breeding females (and say 10-15% of two-year old males) in a natural herd. Most of the animals slaughtered in buffalo ranching schemes are young males. Such manipulations of wild herds can create problems. When Canadian wildlife biologists limited the take of muskox on Baffin Island to mature bulls, the survival rate of the younger animals was reduced. The bulls were necessary for some other function than breeding, perhaps for exposing snow-covered lichens for the other animals to reach, perhaps to defend against wolves. The Inuit had warned the biologists that killing the mature bulls wasn’t a good idea, but they didn’t express themselves in a biologically acceptable way; they said the large males encouraged the other animals. More generally, slaughtering males before they reach breeding age is a genetically risky strategy. One tends to remove the more aggressive animals. The result is that the animals become more tame, but it isn’t clear what else is being removed. More intensive management of buffalo also destroys a herd’s social structure and prevents buffalo from learning to be buffalo, something buffalo calves learn from their mothers, aunts and uncles. Elk managed for antlers and meat rapidly become domesticated and exhibit physiological, morphological and behavioral changes (the “tameness”) associated with domestication. Silver foxes selected for neonatal characteristics such as floppy ears and round heads, as well as for behavioral tameness, become tame within several generations.

The modern bison was probably partly created by human predation. Killing large numbers of adult animals results in populations that breed at earlier ages (when their sizes are smaller). One reason the modern bison is smaller than the steppe bison of glacial times is thought to be due to human predation. (Another is the warmer modern climate. In fact, the modern bison is more closely related to the European wisent than to the steppe bison and may have come, like the moose, from Eurasia over the Bering Steppe with human hunters.) There are modern examples of this evolutionary effect. Hunters taking trophy rams from an isolated population of bighorn sheep in Canada (57 animals over 30 years) reduced the mean body weight of four year old rams from 200 pounds to 160 pounds and their horn length from 28 to 20 inches. (Big rams are hunted for their trophy horns.) Fish are extremely susceptible to size-selective predation; experimental populations can rapidly be made smaller by removing the larger breeding individuals. (This also reduces the population’s total biomass.) The effect is reversed by removing the smaller breeders. The point is that human predation on wild animals differs from that of animals like wolves. Wolves are cautious and take the young, the old, the starving, the sick. Buffalo mothers whose calves were not eaten by wolves may have been more wary or more fierce, letting these traits increase in the population. Wolf predation is thought to increase the size of elk, as wolves tend to take the smaller individuals, leaving the big ones which are more difficult for them to handle. Golden eagles take old, young, sick, and pregnant prairie dogs (pregnant females are probably slightly slower), as well as males obsessed with sex. Humans take healthy animals of breeding age, which creates a different selection pressure. Ideally, culling managed herds of buffalo for food should be random. If the point is to manage the buffalo as a genetically fit wild animal, one would have to see how the herds reacted to a scheme that artificially reduced the number of bulls, that was probably biased toward removing the more agressive animals, and that eliminated old animals altogether. Older females are important in a herd’s social structure. (Buffalo can live 20 to 30 years but are usually culled at 9 or 10.) Elk and buffalo bands are traditionally led by older females, who have good spatial memories for food and water. Bighorn sheep have trouble colonizing new territories without an older female to lead them on their yearly migrations. In a new territory, no one knows the way around. So sheep management schemes must allow older “boss” females to remain as leaders.

Seton deduced from fur traders’ records that the buffalo herds made yearly migratory circuits of 300 to 400 miles; the large herds were formed of small groups of 10 to 20 cows, calves, heifers and bulls to 3 years of age (the age of sexual maturity). Current thinking is that many small herds were non-migratory, while large herds (tens of thousands of animals), which were also composed of small family groups, moved all year. The plains are a very variable habitat, with rainfall varying from year to year and from place to place, and movement, even a more or less regular circuit, is a way to deal with this. (African pastoralist systems tend to be more productive than fenced ranches in similar dry climates.) Different environments had different uses. Places with early growth thanks to an early season fire, or many prairie dog towns, were useful in the spring. (Early spring fires were set by Indians to lure buffalo; summer fires were less desireable but usually small, especially if the area had been grazed, and fall fires were a disaster because they destroyed the winter feed.) In summer, buffalo need shade and mud, in which they roll to coat their skin against biting insects. Their wallows, located in low ground, constituted another “built” area on the plains. They were sources of water and mud for birds and insects, and also aquifer recharge areas: breaking up the sod helped water sink into the ground. In winter, buffalo favored broken ground with gullies and brush, partly to get out of the wind, partly because slopes are likely to have their forage exposed by the wind, partly because brushy browse is more accessible than grass in winter. River valleys constituted 7% of the Plains habitat and were important for the buffalo’s survival. With their massive heads and forequarters buffalo are built to move snow out of the way. Their compact bodies and thick coats help them survive cold weather. While they go into a state of reduced energy demand in cold weather, saving energy by taking advantage of what topography and weather offer improves their likelihood of survival.

So what would a buffalo ranching scheme produce? If we take a reasonable estimate of the grassland’s carrying capacity, from somewhere west of Bismarck, North Dakota, to the Rockies, of say, 1 animal (winter count) per 30 acres, one could harvest 1 two-year old animal or its equivalent every 75 acres, more or less. (In drier areas the harvest would be less.) Buffalo, being scarce, are worth more than cattle nowadays, but an animal should be worth $1250 (meat, head, hide, bones). From the same acreage one could take 1 other large herbivore. Since they vary so much in size, their value is harder to estimate—perhaps $500 for an average carcass. A gross return of $23 per acre isn’t far from the net for dryland wheat, in the years in which it will grow in this environment (perhaps 1 year out of 3), and compares well with the returns from cattle. Of course trophy hunters will pay more for an animal: $2500 to $5000 for a large buffalo, elk, or bighorn sheep.

A semi-natural plains means giving up some control over the animals on it. Buffalo are difficult and dangerous to handle, and also likely to injure themselves in close confinement (their skin tears easily and they fight to escape). Fences that contain buffalo are expensive, costing from $5000 to $20,000 a mile, as are confinement facilities such as corrals. If the animals could be marked from a distance with marks that could be recognized at a distance (paint, implanted chips), they could be slaughtered (shot) and dressed in the field. Much of the offal could be abandoned to feed wolves, foxes and other scavengers and to return some of the buffalo’s nutrients to the grassland. (By comparison little is wasted from cattle; of the 35 million cattle slaughtered annually in the United States, the equivalent of 11 million animals goes into products other than human food, much of it animal feed.) Slaughter would take place over several months in the fall and early winter. The offal (like the 5% of the animals that die naturally from accidents every year) would help support populations of grizzly bears, wolves, coyotes, foxes, badgers, bobcats, skunks, all animals once abundant on the plains. Some of these animals could also be trapped or shot in a limited fashion for their fur and flesh. Letting a part of the buffalo return to the soil better mimics the natural situation and also supports less obvious cycles of dependence and opportunism, for instance, early-returning mountain bluebirds that feed on maggots and flies in the decaying flesh of winter-killed buffalo. Prairie dogs and their predators would be left alone; perhaps lightly harvested. One might spread some mineral nutrients to replace what is removed with the buffalo and grow some irrigated hay, or cut some wild hay, and leave the large bales about in favored situations for supplemental winter feed. (Large round bales remain usable for several years.) Like most herbivores, buffalo live as long as their teeth, and reducing the wear on their teeth from eating coarse browse can double their productive life. The productivity of the grassland, and perhaps to an extent, the movement of the animals, could be manipulated by burning. Burning every second or third year maximizes grassland productivity. Early spring, when the ground is wet, and the dead grass eaten down, is the favored time to burn. Any attempt to maximize yields of herbivores always brings one into competition with their predators and the question of whether to harvest predator or herbivore (wolves or buffalo). On a buffalo plains people would replace wolves as the predator limiting buffalo numbers; and people would also try to ameliorate the worst effects of the weather. The number of wolves would have to be kept low (say 50,000 to 100,000 rather than 1.5 million), so as to keep calf mortality from wolf predation low (5%?). A healthy wolf population would be one reason for letting some older buffalo survive. Wolf predation in general falls on old animals, as well as on young animals, or those weakened by injury, disease, or starvation. Such predation is not “waste” but a way of maintaining the buffalo’s (and the wolf’s) fitness. Like a regenerative wheat field, a plains with buffalo is a world in which economic growth, if possible, has to accomodate the productivity of the underlying ecosystem: the grass, soil, rain, the behavior of the buffalo and that of the other herbivores and their predators and parasites.

Tourism would provide another income base, comparable to the $20 to $30 an acre the buffalo, elk, antelope, bighorn sheep, mule deer, prairie dogs, wolves, bears, golden eagles, prairie falcons, sage grouse and the other animals and plants would bring. The plains are now largely empty, cleared of everything but cattle, irrigated hay, and grain, and some native herbivores, whose populations are perhaps 1% of pre-contact levels, and whose behaviors have been altered by fences, loss of habitat, predator control and hunting. (More cattle and sheep die from the poisons spread to control predators than from the predation itself, another example of morality — the moral idea of what a landscape should be, that is, a landscape should be free of predators — trumping economics.) Some native herbivores, especially elk and bighorn sheep, suffer severely from diseases and parasites spread by domestic stock. Suppose we set a goal of $10 an acre for tourism. If the average tourist spends $100 a day for room, food, gas, and so on, then a value of $10 an acre means 10 million people a year visiting for 1 night each, or more likely 2 or 3 million visitors for several nights each, probably concentrated in the better weather of spring, summer, and fall. Three million people visit Yellowstone National Park in Wyoming each year. They don’t all stay there; but they stay somewhere nearby. Ten million yearly visits at double occupancy, with the rooms occupied 150 days a year, would require 30,000 rooms (8000 to 10,000 small bed-and-breakfasts) on the 100 million acres of the dry plains. Something like 70,000 extra people would be visiting the plains at any one time (2 per square mile). The stress so many people would put on the landscape would depend largely on how and where they went into it. A network of seasonal roads would be necessary for bringing out the slaughtered animals; such roads would also provide trails for walking or riding. Keeping people on the roads would keep most of the area out of the way of people.

A buffalo commons was proposed as a solution for the poorest areas on the plains, whose agricultural economies and population are collapsing. While such a scheme would never cover the plains as a whole, it might work for large sections of them. The reason for putting large sections of the landscape together to raise buffalo is partly because the enclosing fences are expensive, and one perimeter fence is cheaper than several smaller ones (internal fences aren’t necessary for buffalo, who practice a sort of rotational grazing on their own), partly because fences that keep buffalo in will also keep other wild animals, such as wolves, grizzlies and elk, in (or out), and partly to let buffalo be buffalo. A small herd leading a more or less natural life requires 4000 to 5000 acres. One can raise buffalo like cattle, on 100 to 200 acres, and make a profit, but then one has a ranching operation with a fairly dangerous wild animal. Some people do this successfully. Raising wild buffalo along with their natural commensuals (wolves, prairie dogs, elk, grizzly bears) requires tracts of several hundred thousand acres. A hundred million acres amounts to about 156,000 square miles, or a square a little under 400 miles on a side. This comes to 1600 miles of fence (actually more, because the land won’t be a square), or $32,000,000 at $20,000 per mile. Thirty-two million dollars is within the range of a year’s agricultural support payments
Grasslands now constitute one-fifth of terrestrial environments. The advance of grasslands is a rather recent development. Grasslands, like peat bogs, store carbon; it has been suggested that the co-evolution of grasslands and grazers is one reason we live in a glacial age: together they remove carbon dioxide from the atmosphere and store it in the soil, continually lowering the atmosphere’s level of carbon dioxide. Undisturbed grassland soils are 10% organic matter a meter down; forest soils are that rich to 10 centimeters. Most of the biomass of perennial grasses is below ground and the annual root turnover (the death and regrowth of small roots) continuously stores carbon in the soil. Grasslands store between 0.5 and 1 ton of carbon per acre per year. Storage of a ton of carbon (paid for by power plants, auto manufacturers, and other industries fueled by combustion) is currently valued at $10 to $50 a year. (Many writers say $20 a ton; the Norwegian government in 2000 was charging its offshore oil industry $38 per metric ton of carbon dioxide released. This charge made it worthwhile to sequester the gas.) The buffalo bring about $20 an acre. So a buffalo rancher who doesn’t have a bed-and-breakfast or a wind farm (wind power being a major potential source of energy on the plains, but one that depends on the construction of new power lines), might, in a regenerative world, expect another $10 to $20 an acre from carbon storage, for a total return of $30 to $40 an acre, a yield considerably above cattle or dryland wheat.
Perhaps in a late-industrial world, such a scheme would work. Similar ones to raise wildlife, harvest sustainable timber, or collect wild fruits in Africa, Peru or Brazil rarely work. Part of the problem is ownership—who owns which buffalo, which wolf pelt? One can see a corporation made up of current landowners managing their section of the plains. Reluctant landowners whose land lay amidst others’ would have to join, or be bought out. Land would have to be converted to shares, goals established, compensation for labor agreed upon. The corporation could invest in tourist facilities or license that development to members. The place of federal land in the scheme, if any, would have to be negotiated. As part of modern society, the gross return from the land should be sufficient to pay for a modern life: heated houses, higher education, medical care, pick-up trucks. The infra-structure demanded by such a scheme, which reflects its integration into and value to the larger economy, is small: gravel roads, mobile meat-packing units, some sort of four wheel drive vehicles to haul the field-dressed animals to the packing units, trucks to haul away the meat, freezing plants to store the surplus; tourist accomodations. While all this falls far below the capital demands of farmland, it is comparable to those of raising cattle on grass, which is considered part of a modern economy. The buffalo would have to be kept to the plains, kept out of towns and cities, off major highways, and away from grain-growing areas; this could be done with fences and cattle guards, or perhaps other, smarter barriers would work: living hedges, ditches, highway overpasses, the incorporation of natural barriers like mountains, rivers or bluffs.
With the technology of the 1860s or 1870s such a scheme was also possible. The Native Americans were already managing the buffalo, but for a low rate of return, in modern economic terms: the feeding and housing of 120,000 people, plus a few hundred dollars income per household per year in robes. (A modern buffalo scheme would support a similar number of people, a reflection of how the value of buffalo has risen.) Sustainable management of animals, forests and fisheries was never the point of capitalism. Capitalist development broke apart the more or less sustainable systems of agriculture, fishery and forestry that were limiting European economic development. If they grow, sustainable systems grow slowly, too slowly to make a large profit or to meet demand. One reaches a point where one can’t raise the productivity of the biological system without compromising its long-term stability. Biological systems are capable of flipping up or down, to greater or lesser levels of productivity, and then remaining stable at those levels. North American cod stocks for instance, may remain stable at their current very low levels for a long time (decades to centuries), until something (several good years for cod recruitment, several bad years for the fish that prey on small cod, temperature changes in the ocean favorable for food resources for larval cod, ocean currents favorable for spawning) raises levels once again. Or cod may never recover; their abundance may have been due to a combination of events that happened once, and won’t happen again (the climate, the ocean, the currents may be too different). Once abundant, the cod could maintain themselves, but once reduced they may stay that way. Populations of moose are also capable of flipping in this way. Lack of sustainability is the downside of modern agriculture: under modern management, crop yields rise, but at the expense of soils, waterways, ground water, the health of the farmers and their neighbors, the long-term productivity of the landscape. Generating electricity from wind and sun presents new possibilities for the plains, but any sort of human industrial use, with its associated demands for water, power lines, roads, factories and people, would have to work around the existing ecosystem in order not to destroy it; the buffalo would have to be seen as part of an ecosystem that mattered. The natural inflation in land values that accompanies industrialization would tend to drive a buffalo-raising scheme out of business, as property taxes rose to reflect the land’s potential economic value. Restricting land-use so as to maintain an ecosystem is not impossible (it is done in some cases now for agricultural land), but it means seeing the value of whole ecosystems as greater than that of their destruction for economic development. (Luckily enough, in this case buffalo and prairie dogs coexist well with windmills, though birds of prey and bats may not.) The economic costs of destroying an ecosystem are always large, but are remote in time and space, and so not a current worry. In an ideal world, terrestrial landscapes (prairies, agricultural lands, forestlands) would be valued at what they can produce in a biologically appropriate way; and the amount of land developed in a non-biologically appropriate way would be limited, and the effects of its development mitigated. Such schemes would not destroy the current economic landscape but re-arrange it; and probably require compensating monies to flow in many directions. At this point our accumulation of capital and fossil-fuelled power allows us to accomplish many things; we could still create an intellectually satisfactory, sustainable, modern material life; but with the end of cheap energy and the economic disruptions caused by climate change, this opportunity is likely to close.
The end of the buffalo came swiftly. The plains tribes had lived as nomadic hunters on horseback for something over a hundred years. Many of the famous tribes were immigrants to the plains; the Cheyenne and Dakota had been corn-growers and maple syrup makers in Minnesota and Wisconsin, until pushed west by the Ojibwa, who were being pushed by other tribes, by Iroquois expansion and by the Americans. Horses began to be used in the 1680s and by the late 1700s all the plains tribes had horses. Twenty-five separate Indian nations, of 7 language groups, lived on the plains in 1800, their existence there (and to an extent their culture) a creation of Euro-American settlement. The take of buffalo by the Indians before the horse has been estimated at 300,000 animals a year. Many of these animals were killed in jumps or fire drives, so little selection by age or sex was involved. (One can guess from the normal composition of the herds, that cows of all ages, calves and young bulls would have been the animals killed.) Fire drives and jumps were sometimes wasteful; not all the animals killed could be used. But all those trapped were killed, since people feared that escaped animals would warn other buffalo about the trap. The take of 300,000 animals
The acquisition of the horse changed the life of the tribes and opened up the plains away from the large river valleys for habitation. Many tribes abandoned agriculture. Horses were owned individually, so the social structure of the tribes became less egalitarian. Wealth in general increased. Tipis became larger. On average, 4 times the weight in possessions could be transported with horses than with dogs, the former beast of burden. There were 120,000 Indians on the Plains in 1800. At a ration of 5 pounds of meat per person per day (the Hudson’s Bay ration was 7 to 8 pounds), the tribes required between 720,000 and 840,000 buffalo. Hunts now took place from horseback and the animals killed were selected. Most of those killed were young cows. This reduced the population of breeders. The robe trade with Americans traders became important in the 1840s and 1850s and probably added 100,000 buffalo to these numbers. Robes were prime in winter, when the animals had their thickest coats. Young bulls were acceptable for robes, but cows, with thicker fur and more pliable skin, were preferred. The winter-killed animals usually weren’t needed for food, and the carcasses were often abandoned, except for the hump and tongue. More available carrion from the winter kill may have increased the survival of wolves, increasing the pressure on buffalo calves. The robe trade further changed Indian social structures. While women prepared the robes (10 to 20 per winter per person), men kept the proceeds. A good hunter with several wives could accumulate wealth and influence. Indian numbers on the plains during the nineteenth century rose from immigration as more tribes were pushed west by the Americans, but fell because of disease, mostly smallpox. There was a net fall of 50% in the population by the end of the century.

During much of the nineteenth century the buffalo herds were stressed by droughts. There was a drought on the southern plains from 1822 to 1832, and one on the northern plains from 1836 to 1851. The lack of grass would have interferred with the animals congregating in late summer to breed. Cows would have been in poor condition, fewer would have become pregnant, and fewer would have carried their fetuses to term. During the droughts one can assume buffalo numbers fell. The native take of perhaps 900,000 animals, mostly cows of breeding age, would have further reduced them, and their potential for renewal. At the same time buffalo had to compete with another large grazer, the horse, whose numbers were increasing on the plains (there were 2.5 million on the southern plains by 1850). Buffalo’s winter habitat along the valley of the Missouri River was being degraded by cutting fuel for steamboats, and by the Native Americans collecting cottonwood branches and bark for winter feed for their horses. The spring grass along the Platte was being eaten by the oxen and cattle of the emigrant wagon-trains, whose members gathered in early spring in Missouri awaiting greenup further west. Buffalo were also susceptible to the diseases of cattle. All this, plus the shooting that accompanied the westward movement of the Americans meant that by the end of the 1850s buffalo were considerably reduced; perhaps the population on the plains amounted to 20,000,000 animals, perhaps fewer. The buffalo had once drifted east to the better watered areas in times of drought, when their favored short grasses began to compete well with the taller grasses of the eastern prairies. As settlement progressed onto the prairies, animals that followed the grass east (once an adaptive behavior) were killed.

Buffalo then became part of the industrial world. Buffalo leather was tougher than that from cattle or deer, but in the late 1860s new tanning techniques made bison leather workable. The British army favored the long-wearing skins for footwear. The Americans found it ideal for the belting needed to connect steam engines to the machinery in their factories. There was a growing demand for belting in the economic expansion that followed the Civil War. A drought in southern South America in the 1860s shut off the usual supply of cow hides from Argentina. Finally, as a matter of public policy, killing the buffalo was seen as a strategy that would eliminate the plains tribes once and for all. When railroads reached the plains in 1867, it became possible to ship out the hides. Then the commercial hunt was inevitable. The old growth hemlock forests of the Adirondacks and northeastern Pennsylvania provided the tanbark. The logs, too heavy to float out, were abandoned in the woods. Only hides and tongues were removed from the dead animals. By some accounts 80% of the hides were thrown away because of poor handling. The hunt lasted about 15 years. Soon afterwards rubber from Brazil came into use for belting, and soon after that electric motors made steam engines obsolete. Less than 1000 wild buffalo were left by 1889. According to an Indian story, the buffalo in their final days, distraught at their treatment by men, went mad, and raced round and round bawling in despair and pain, before returning to the center of the earth from which they had come. In the 1890s, thousands of tons of buffalo bones were collected from the plains and sent east to be ground up for carbon black (a filtering agent, used for purifying sugar) and fertiliser. The bones (by estimates based on their weight) amounted to those of 30 million buffalo.

With the buffalo gone and the Native Americans on reservations, the plains were opened to American settlement. Rainfall on the plains tends to go in 20 year cycles. The 1820s and 1840s were times of drought. In the 1870s, with buffalo becoming scarce, ranchers began to drive herds of cattle north from Texas to take advantage of the summer grass. Wolves, which, if abundant, can make cattle ranching difficult, and sheep raising impossible, were killed for a bounty (80,000 between 1883 and 1918 in Montana). Good rainfall years from 1878 to 1887 led investors from the eastern United States and Britain to invest in cattle ranching on the northern ranges. Then the blizzard of 1886 killed most of the cattle. Cattle are not adapted to winter on the northern plains. They froze to death in the deep snow, or suffocated as it drifted over their heads in the draws. They drifted before the wind and froze standing up. Unable to eat because of the deep snow, they starved.

At the same time the open range was being broken up as homesteading followed the railroads into the Dakotas, Nebraska, Colorado and eastern Montana. Much of this arid land was fertile and would grow wheat in a wet year. Drylands accumulate nitrogen since they are little leached by rain, and their vegetation, lacking water, is unable to make use of the accumulating nutrients. The first crops of the homesteaders were good ones. Dryland wheat in Oklahoma, in a year with sufficient rainfall, yielded up to 75 bushels an acre when the sod was first broken. After 80 years of cultivation, such lands now yield 45 bushels an acre under irrigation or 8 under dryland farming. (Eight bushels an acre means a gross income of about $24, or a net of perhaps $5 an acre; but $5 an acre over 10,000 acres comes to $50,000.) Rain was said to follow the plow, but the wet years were followed by dry ones. Dryland farming on the plains results in soil loss (up to a foot or more a year) through wind erosion. Plowing and dryness tend to destroy the ability of the plains soils to flocculate and absorb water; unless kept vegetated, they turn to dust and blow away. For the most part wheat farming west of the 104th meridian was a failure except in river valleys or on irrigated ground. The drier land in the Dakotas, Montana, Wyoming that didn’t have sufficient water to grow crops was slowly consolidated into large holdings of several square miles, used mostly for cattle pasture, and to raise some hay, and wheat in favorable locations if the year allowed. River water was used for irrigation and farmers who lived over the High Plains Regional Aquifer began experimenting with using the water below them.

Ground water pumping had begun in the 1880s with small acreages irrigated by home-made windmills, where the aquifer was near the surface. Such windmills were common in Nebraska, under whose soil lies two-thirds of the ground water in the plains. Irrigation systems for commodity crops like corn using wells and pumps were experimented with throughout the early part of the twentieth century but it wasn’t until the late 1930s that well-drilling technology, with pumps powered by automobile engines, and guaranteed government loans (a gift of the Depression), put groundwater irrigation from the Ogallala within reach of the average farmer. The invention of center pivot irrigation in the 1950s made irrigation much less laborious. A single irrigation of 150 acres by hand from a central wellhead with irrigation pipe and ditches took 2 weeks of hand labor; the center pivot system, once set up, moved continuously on its own. Center pivot systems required the removal of the windbreaks and hedgerows that had been planted with government support after the droughts and duststorms of the 1930s, when dirt from the plains had fallen on Washington and New York and had drifted over the Atlantic to fall on France. The trees broke the plains wind and helped stop the soil from blowing away. The equipment’s high capital cost ($50,000 to $60,000 for a system that would irrrigate approximately 130 acres, a quarter section, the corners being left out by the circular run of the rig) also made farmers that much more vulnerable to changes in the price of grain, or in the tax system. (For a while in the 1970s, the tax laws were all that made new irrigation rigs in Nebraska profitable.) Regular irrigation also leached pollutants down into the aquifer. Such pollutants include manufactured nitrogen fertiliser, herbicides and pesticides; gasoline and diesel oil; nutrients and toxic chemicals from feedlots and septic systems. Polluted runoff seeped out of abandoned water or oil wells into the aquifer directly. Wells throughout the plains have tested above the level of nitrates supposed to be toxic to babies. Nitrates in rather low amounts are also toxic for cattle and hogs. The Agricultural Extension Service recommends that farmers pumping irrigation water out of the aquifer count its nitrogen content as part of their fertiliser applications. Life on the plains depends on underground water. Aquifers like the Ogallala are essentially impossible to clean, once polluted.

But irrigation of the plains worked. The crops were profitable, and so the plains were mined for water. Where there was water, the plains prospered. Towns gained population. Property values rose. Different states had different laws governing the use of groundwaters. In Texas, a landowner could pump a well until it ran dry, whether the well was drawing water from under someone’s else’s property or not. So a Texas farmer could sink wells along his fencelines and pump the aquifer under his and the surrounding lands dry. Pumping in much of Texas, in eastern New Mexico and in Colorado has already depleted much of the recoverable water. On the plains as a whole the irrigated area peaked in 1980, when the water pumped from underground exceeded the flow of the Colorado River. Only about 15% of the water stored in the Ogallala is economically recoverable for agriculture. The rest is too deep, too salty, or too difficult to extract. So a great deal of water will be left when irrigated agriculture on the plains is over in 20 to 30 years. It will be more expensive to recover, but water for household use (and perhaps for some industrial uses) is worth orders of magnitude more than water for agriculture, which must be literally as cheap as dirt, and thus is always provided with a substantial government subsidy, either directly or through the tax system. Groundwater pumping has already dried up much of the surface water habitat on the plains. Hundreds of miles of once perennial streams are now seasonally or permanently dry; springs are gone. (The Ogallala accumulates water from rainfall at a fraction of an inch a year, while pumping depletes it by feet a year. Over the whole aquifer, use, at 22.2 cubic kilometers per year in 2000, is 3-4 times recharge, at 6-8 cubic kilometers per year.) Recently, local water management committees were formed to regulate depletion. Their rules vary mostly in the length of time they favor depleting the resource to zero. (Unregulated ground water pumping has gone out of favor because your investment in a well might be affected by your neighbor sinking several wells nearby and pumping out all the water too quickly for you to recover your investment.) The time scales for depletion range from a 1% loss a year in Kansas to a 20 year pumpdown in Oklahoma. One Kansas committe is planning to stop use shortly before economic depletion of the water.

When the water is gone, some areas of the plains will still be suitable for dryland farming. Current operators claim the use of deep plowing, which brings up heavy clays from the subsoil (and is economic because of cheap fossil fuels) will prevent another Dust Bowl in the next droughts. Perhaps the changes in patterns of rainfall with global warming will bring more rain to the plains; but this is doubtful. As far as I am aware, no rules exist to require revegetation of irrigated lands as the water disappears. Revegetating the landscape with native grasses is neither simple nor cheap. The two major plains species do not reproduce well from seeds and require irrigation to get started. They are more difficult to start in wind-eroded, silt-depleted soils, and do most of their propagation vegetatively, which allows them to withstand grazing better than most American bunch grasses, and to expand rapidly after droughts. (An introduced Eurasian grass, crested wheatgrass, stands summer grazing better than any of the native species.) Perennial grasses, native or not, turn the soil into a carbon store. Soils store three-quarters of the carbon in terrestrial ecosystems, more than in the plants that grow on them. Cultivated grassland soils in the United States are thought to have lost 30-50% of their carbon to date. Turning the plains once again into a carbon-hungry ecosystem of perennial grasses and grazers, their restoration and maintenance partly paid for by the burners of fossil fuels, is a happy possibility. And using windmills to harvest some of the constant wind doesn’t interfere with uses like grazing (buffalo and cattle, unlike birds and bats, ignore windmills). A considerable proportion of the United States’ coal reserves lies under the plains and thousands of acres of regraded strip-mined land will be available to site wind machines. Irrigation agriculture and strip-mining currently return the greatest income per acre on the plains; sustainable uses (grass, buffalo and wind power) would return the greatest total income on the plains over time, and a better income than dryland farming. As global warming worsens, and the place of people, soils, and vegetation in the global context becomes clearer, perhaps attitudes will change, and industrial combustion will begin to subsidize more biologically appropriate uses of this landscape.