Chapter 12: Some Ecological Implications of the Land Trip; a Short and Anecdotal Essay on the Environmental History of the United States, with a focus on Streams, Especially Those in the Middle of the Country

Now, long after it all happened, a large literature is appearing on the ecological consequences of European settlement of “new lands” in North and South America, Australia, New Zealand, the South Pacific islands. Exploitation of the lands’ resources began with settlement. By the 1670s, Boston merchants were buying timberland in Maine in order to log it and abandon it. (Fernando Gorges’ settlement in Maine preceded that of the Pilgrims, but the natives, not yet affected by European diseases, were too numerous and too unfriendly, and after attempting to set up a sawmill, the English left.) Beaver insured the success of the colony at Plymouth. The fur trade had begun in the 1520s as a sideline of the cod fishery off Newfoundland, then expanded with the Dutch settlements in New York, and with the growing popularity of felt hats, made of beaver fur, in Europe after the 1550s. Two million skins were shipped out of New York during the 1600s, and by 1650 beaver were more or less gone from New England and New York State. Most were gone from streams east of the Mississippi by 1700. By 1680 the French had a trading network from Quebec to the mouth of the Mississippi and west to the Rockies, where they came up against Spanish traders; the British were trading in most of northwestern Canada west of an outpost on Hudson’s Bay. Recurrent epidemics, collapsing populations, the desire for trade goods, and the movement of European and Native populations into others’ lands demoralized the Native Americans and removed their cultural constraints on hunting. In the 1650s the tribes who hunted the immense swamplands between Lake Erie and the Mississippi (wooded swamps of elm and red maple, now drained to grow corn) were forced by the Iroquois to provide skins for the British (a sort of tribute: acting as middlemen for the British, the Iroquois reaped the profits). Their part in the fur trade led to Iroquois domination from Hudson’s Bay to the Carolinas, and from the Atlantic to the Mississippi.

The landscape between the Great Lakes and the upper drainages of the Mississippi and Ohio was one of the centers of the North American beaver population. Some writers put the aboriginal population of beaver in the United States at 200 to 400 million animals. (Seton estimated 50 million, other estimates go as low as 10 million.) If 200 million is correct, at an acre of pond per beaver there were then 200 million acres of beaver wetlands, about 10% of the continental United States. (Five percent is probably a more reasonable number.) Beaver ponds occupied about 20% of the area around pre-contact Detroit. In good habitat, each 1000 meters of stream contains an active beaver lodge. Beaver wetlands supported waterfowl such as wood ducks, sandpipers and teal; small animals like fish, turtles, frogs and salamanders and the mink, otters and raccoons that ate all these; willows and moose; populations of invertebrates and microbes; and the wolves that ate the beaver and the moose. They increased the width of the land-water ecotone, where microbes recycle nutrients, and regulated stream flow. In pre-contact North America, men and beaver shaped the colder, wetter parts of the continent, from the Atlantic to the Rockies, north to the margins of the Arctic. Beaver wetlands produce more methane than other boreal wetlands, and so affected global atmospheric chemistry. Settlers cut hay in abandoned beaver wetlands (“beaver meadows”) and drained large former wetlands for farms. Draining wetlands affects water quality downstream: the filtering effect of the wetland is eliminated, so most of the nutrients and sediment coming from upstream continue on downstream; and the oxidation of wetland soils releases stored nutrients and sediment that also flow downstream. In the 1600s and 1700s, as their populations fell, and their cultures collapsed around them, Native Americans trapped out the beaver. (The trap-out by the mountain men in the 1800s in the streams draining the Rockies, using steel traps and castoreum bait, occurred after the Louisiana Purchase.) Managing beaver sustainably, as the Indians had done for millennia, removes a quarter or less of the population annually (50 million animals out of a population of 200 million).

On the frontier, the settlers fished, shellfished, fowled, cut timber, hunted and trapped at will. In the mid-1700s New Jersey farmers complained to Swedish botanist Peter Kalm about the growing shortage of ducks. As Kalm pointed out, since the farmers shot ducks whenever they pleased, including in spring and summer, when the ducks were raising their young, they had only themselves to blame. New England coastal natives, who had depended on an endless supply of shellfish for thousands of years, complained that the Englishmen’s pigs rooted up their clam beds; but they were liable to punishment if they killed the pigs. Arguments over dams that prevented the movement of migratory fish would finally be lost by riverbank landowners in New England in the 1800s, when Boston capitalistists started to develop New England’s water resources to power textile mills. Dams that shut off fish runs changed the character of whole drainages. Dams flooded riverside wetlands permanently (turning them into shallow lakes) and changed the vegetation along the lengths of regulated streams. (Trees on a floodplain grow where the period of inundation is limited. When the flood lasts too long, the tree roots, deprived of oxygen in the wet ground, stop transporting water to the leaves, which die of drought. Dams that raised water tables along the Mississippi, together with cutting of the streamside forests for fuel and timber, changed a mixed forest oaks, hickory, willow, sycamore, pecans and elm to a monoculture of silver maple, a pioneer species that could stand the higher water levels and shaded everything else out.) Rates of land clearing before the Revolution were half or less the rates of land clearing that came afterwards, but the effects on streams were severe. Development spiraled upwards after wars, which pushed technological development forward; rates of western settlement increased after the French and Indian War, the American Revolution, and the Civil War.

Early forest clearance was for agriculture. Logging overtook agriculture as a reason for cutting the forest only in the 1880s. Forest clearance in New England reached a rate of 0.4% to 0.5% of the land area per year by the late 1600s, a rate that would take about 200 years to bring the whole landscape under cultivation. Despite a speeding up of this process after 1750, cleared land in New York and New England reached its maximum extent in the 1880s, with something like 80% of the landscape cleared. (Some writers say 65%.) New York State, in a typical reversal, is now about 70% forestland. Before freight railroads, transportation constituted 50-75% of the cost of sawn lumber. Without transportation to a market, probably 75% of the timber cut to clear farmland was wasted. If wagon roads and water transportation were available, ashes, turned into potash, were the first crop on a farm. An acre of hardwood produced 60 to 150 bushels of ashes, worth $0.06 to $0.08 each; this would pay for the labor in clearing the land. (In some forests the yield of ashes was higher and would pay for the land and clearing it.) After being “proved” by harvesting a crop, the land could be sold for 10 times its original cost. Some people established several farms during a lifetime. In the years from, say, 1650 until 1935, when agriculture was the country’s dominant business, the notion of the family farm was something of an illusion. In 1910, more than half of all American farmers had been on their land less than 5 years.

Woods that were not cleared for farmland were changed by cutting for timber. First growth forests in the eastern United States have 3 to 6 times the woody biomass of second growth forests (125 to 250 tons of wood per acre for good first growth, 20 to 80 tons for second growth). Early cutting, for prime sawlogs, fencing, and building material, was light, but changed the character of the forest. American chestnut, which sprouts from stumps, was 4-15% of woodland in New Jersey and Connecticut in early surveys, 60% by 1900. Chestnut, which resists rot, was popular as a building material and for fencing and was cut extensively. Softwoods (which do not sprout) were also favored, largely because they could be floated down rivers to market. Old trees of some long-lived species, such as eastern hemlock, cut for tanbark, and white pine and red spruce, cut for dimension lumber, disappeared. Logging, like burning, increases deer populations. White-tailed deer are 4 to 5 times more abundant in a logged forest, than in mature deciduous forest. (Deer are found at a density of 2 per square kilometer in mature deciduous forest, 8 to 10 per square kilometer in a logged forest. Densities as high as 20 to 40 per square kilometer may be carried for a limited time.) Deer suppress the regeneration of hemlock, yew and white cedar, and browse on many more species. (Their winter browse lines are a characteristic of the shorelines of northern lakes.) A density of 4 animals per square kilometer may eliminate species sensitive to browsing. An old estimate of the animals in pre-contact New England forests lists, for every 10 square miles (25 square kilometers, a square a little over 3 miles on a side), 5 black bears, 2 to 3 mountain lions, 2 to 3 wolves, 10,000 to 20,000 gray squirrels, 400 deer, and 200 turkeys. The number of deer is very high (about 16 per square kilometer). The woodlands of southern New England were not mature deciduous forest in an ecological sense, but managed by fire for game. (The numbers may also be wrong.) The numbers for squirrels look high but only come to 2 to 3 per acre. By 1750 most of the predators were gone from the settled parts of New England and by 1900 hunting had made deer extinct in most of New England and New York State, removing both the threat of Lyme disease (colonial records report a disease something like it) and the influence of deer on forests.

The Middle West, settled in the boom after the Revolution, was plowed and cleared at about 1% a year, twice as fast as New England. River transportation, canals, and railroads, rapidly succeeding and supplementing each other, made this a more commercial proposition from the start. Ohio had 600,000 people by the 1820s, the six other states of the Old Northwest—the modern Middle West—200,000. By 1860, Ohio had more than 2 million people, the rest of the Old Northwest 5.5 million. In the first period of settlement of the Middle West, only river bottoms and wooded slopes were considered farmable. Successful grain farming on the wet upland prairies of Illinois and Indiana, landscapes which we now consider quintessentially Corn Belt, had to wait for the practice of tile drainage. The wet prairies were first used as cattle range. Drainage was expensive, but finally took off in the 1880s. One can trace its progress in the records of local manufactories for drainage tile (clay tile is heavy and expensive to ship). Drainage continued through the 1920s in the wet prairies of Illinois, Indiana, Iowa and Ohio. Drained lands now constitute 27% of Illinois, 22% of Iowa, 29% of Indiana: large numbers that imply major changes in the habitat. Drainage helped create the Corn Belt. Until recently, the turning of wild land into land that produced something of marketable value was considered a public good in the United States. The reclamation of seasonally wet “waste” land was held to be a public benefit by the courts in suits that gave drainage ditches the right of eminent domain to cross other peoples’ land. Under some laws, if a majority of the landowners of 60% of an area to be drained favored a drainage district, the other landowners were forced to join it, and pay for the improvements. (This potentially let a minority of landowners rule. Such laws may also have been a way to force capital-poor farmers off their land.) The expense of drainage meant that over half the farmers on drained land were tenants in 1914. If one could afford it, drainage paid, covering its costs in about 5 years, through increased yields. Tile drainage, the lines 40 to 100 feet apart, 4 feet deep, with tiles 2 to 5 inches in diameter, shunted the prairie’s water and nutrients into the headwaters of the Mississippi. The nutrient load was high partly because the plant cover had been removed, partly because drainage let the nutrients in these formerly wet soils oxidize and flow away with the water. By 1964 drainage in southern Minnesota had reduced the area’s once immense waterfowl population (tens of millions of birds) to insignificance. (High soil fertility also means high wildlife populations. Probably 45 million acres of duck habitat were drained in the upper Midwest, reducing populations by at least 90 million birds.)

Grain farming in the late 1800s was a mixed agriculture, with grain, animals, and hay or legume-hay mixtures raised on the same farm; the hay and some of the grain was fed to the animals. Hay was a major crop. Grass and clover were grown for seed, and hay was grown to feed the farm’s cattle and work horses and exported to feed the cab horses of the eastern cities. As late as 1939, Jasper County, Illinois had 47% of its cultivated area in grasses for hay and seed crops; this would fall to 1% by 1974, as soybeans, a row crop like corn, rose from 9% to 69% of the county’s agricultural lands. In Illinois and Iowa in the 1990s, 30% of agricultural land was in continuous corn, 60% in two years of corn followed by one of soybeans. Virtually all the crops were heavily fertilised and herbicided. Farm animals were gone. Only one in six farms spread manure. Such changes were part of the general boom in the farm economy, the shift from horses to tractors, and from mixed farming with its rotations and manures, to specialized cropping with its commercial fertilisers and pesticides after World War II. In general, crop rotations kept 25% of cropland in sod (hays or legume hays; small grains like wheat also form sods and hold soil better then row crops). The land in sod reduced soil erosion, which was enormous on the lighter, hillier prairie lands. Hayfields, especially if used for seed, and so harvested late in the year (or if the first cutting was delayed until the third week of June), provided habitat for grassland birds, which hayfields no longer do so under the intensive cutting of modern management. The fields must look inviting to the returning migrants, but the time between cuttings is not sufficient to raise a brood, so the fields become a trap for breeding birds and after a few years have none.

Such changes in the landscape affected streams. Forest clearance, and cultivation and drainage of the prairie, increased the rate and amount of run-off from rain or melting snow. Soil running off agricultural lands widened and shallowed the headwaters of the Middle Western rivers and raised their beds. Cultivation of rich alluvial floodplains in the Midwest and South turned the bottomland forests that were originally sinks for floodborn nutrients and silt (accumulating silt at a rate of 10-20 tons per acre per year) into—as fields of corn or soybeans—donors of silt to streams (15-60 tons of sediment per acre per year). The growing Mississippi floods in the early twentieth century were known to be caused by agriculture and forest clearance upstream. Agricultural erosion peaked in the 1920s and 1930s along the tributaries of the upper Mississippi. The 1927 flood (after which the Corps of Engineers took over management of the river) destroyed 160,000 structures in the Mississippi Valley and all the bridges 1000 miles upriver of Cairo, Illinois. Soil conserving practices such as strip cropping, contour plowing, and the use of winter cover crops, adopted after the droughts of the 1930s, helped reduce erosion, but erosion would rise again after the Second World War, as rotation into sod crops ceased. During the War prices for grain were high and farmers brought more land under cultivation. With the use of artificial fertilisers, rotations into sod were unnecessary (a waste of good land). Agricultural Extension Agents encouraged farmers to grow the crops “best adapted” to the soils and region. The crop rotations of the mixed farm had reduced soil erosion and helped control water run-off. Crop rotations also maintained soil fertility and helped control weeds and pests. On farms that grew only grain, the land lay open two-thirds of the year, erosion and run-off increased, and fertility was maintained with manufactured fertilisers. The weed and pest populations that built up were controlled with herbicides and pesticides. These chemicals ran off into streams and sank into ground water. Under traditiional crop rotations before World War II, levels of nitrates in Iowa’s Des Moines River were already high: rates of soil and water run-off from bare fields were also high; and cultivating corn for weed control during the summer, up to 7 times, led to high rates of erosion from cornland. Now such rates would rise. Much modern agricultural runoff contains suspended solids and nutrients on the level of sewage wastewater.

After World War II, as some farms specialized in row crops, others began to specialize in cattle and hog raising; that is, they became essentially feedlots. Raising thousands of animals in one place meant that the animals’ manure, rather than being a useful soil amendment, became a disposal problem; there was not enough cropland to spread it on. There are approximately 9 billion domestic animals in the United States at any one time, compared with about 300 million humans; the animals produce several times the urine and faeces of the humans. Most of it is handled badly. Manure becomes a pollutant if it is applied too heavily or at the wrong time of year. Large-scale operations that raise confined animals now handle it as a liquid, since the capital costs of dealing with it this way are less. This lets much of its nitrogen volatilize as ammonia (a waste of the nutrient), makes it difficult to store for long because of the expense of the facilities, and creates horrendous pollution problems when storage ponds fail, as they do regularly during heavy rains. (During Hurricane Floyd in the 1990s, hog manure lagoons failed along 23 of the 26 river systems in North Carolina and flooded Albemarle Sound with a layer of nutrient-rich muck 6 inches to several feet thick.) Composting the manure, while somewhat more expensive, would be a far better solution. Aerobic composting locks up much of the ammonia as usable nitrite and nitrate, and releases considerably less of the greenhouse gases methane, nitrous oxide and carbon dioxide. Composting reduces the bulk of the material by half, so there is less to store or spread. After composting, the material is easily stored under a shed or a plastic tarp and can be spread as needed. Composting makes the nutrients in manure less soluble, and careful composting, say with fly ash, a waste product of coal-burning power stations, greatly reduces its load of pathogens, such as the intestinal bacteria E. coli, and Cryptosporidium, a diarrhea-causing organism whose oocytes, probably derived from manure running off dairy farms, are extremely common in shellfish in river estuaries of the eastern United States. A more capital intensive solution for farms with large numbers of animals (feedlots; or dairy farms with 1000 cows or more) is generating electricity from the methane in the manure. The electricity can be used on the farm or sold and the material left over is useful as a soil amendment (it is essentially compost). Neither composting nor electricity generation will happen without some sort of enticement; the returns on capital aren’t great and methane generators have to be overseen by someone, creating an additional cost. However currently the systems for handling liquid manure are largely paid for by the federal government, which could require more environmentally appropriate solutions. (All this of course ignores the question of the health, well-being and tastiness of animals held in the miseries of close confinement.)

* * *

Before European contact, the rivers of the upper Mississippi Valley ran clear. Their valleys were forested. Along the Wabash, a tributary of the Ohio, sycamores grew 200 feet tall between the river bluffs (50 feet taller than tall eastern trees), and 6 to 8 feet in diameter. Illinois as a whole was about a third forested, Wisconsin largely forested, Ohio more or less completely so. Large parts of the bottoms of the rivers were covered with beds of mussels. The wide, shallow Ohio was known for its mussel beds. As with the shellfish of Chesapeake Bay, their filtering capacity (the time it took for the volume of river water above them to pass through them) is thought to have been measured in days. (Perhaps as long as a week. Within the limitations of their habitat, and of losses caused by their predators and parasites, mussel populations would have expanded to the limits of their food supply, which was plankton and detritus from the water.) The mussels filtered out the primary producers in the river (the bacteria and algae) and kept the water clearer, letting sunlight penetrate further. They thus benefitted the rooted underwater plants, which grew where the bottom was less firm (mussels need a firm bottom to attach) and which grow less vigorously, or die, if algae cut off the light. The underwater plants anchor the bottom against disturbance by bottom-feeding fish (like the introduced European carp); their stems and leaves slow currents, damp waves, and help silt settle out, which helps mussels, which are sensitive to siltation. By reducing turbidity the plants let light penetrate further into the water column and thus maintain a more favorable environment for themselves and for clear water species of fish, such as bass, perch and pike. Their roots oxygenate bottom muds. Their lower stems and roots provide a large oxygenated surface in the anaerobic zone, where microbes living on them convert ammonia (toxic to many aquatic animals) to usable nitrate (which the plants convert to plant tissue) and oxidize toxic metals into harmless forms. Aquatic plants generate dissolved organic matter and detrital particles that support the invertebrates, bacteria, and plankton at the base of the food chain (including the mussels’ food chain); they shelter snails, aquatic insects and juvenile fish; and are eaten directly by waterfowl and muskrats. Enormous flocks of migratory waterfowl (100 to 200 million birds) once visited the rivers of the Mississippi Valley and were probably important in the natural nutrient regimes of the rivers and their floodplain lakes. Waterfowl and fish ate the algae and plants, zooplankton, small fish, invertebrates, fingernail clams and various aquatic insects in the rivers and floodplain lakes. They ate mast from the floodplain forests. The mussels that covered the firmer bottoms were thus part of various mammalian, avian and fishy food chains. Mussels were extremely abundant in all Middle Western rivers including the larger ones (the Illinois, the Ohio, the Tennessee, the Wisconsin, the upper Mississippi).

The discovery of fresh-water pearls in mussels led to their systematic elimination. Since mussels were abundant and the work was done by hand, this took several decades. (Times to economic extinction vary: during the 1820s sealers took only 5 years to eliminate the southern fur seal and elephant seal as economic resources; and modern ocean fisheries take about 15 years to fall to 10% of their former abundance.) During this time the shells were also used for making buttons, an industry that employed 20,000 people in 1920. From 1914 to 1920 the upper Mississippi produced an average of 35,000 tons of shells a year. Button-making might have led to a stable, renewable use of the mussels, and perhaps use of the meat as well as of the pearls and shells, but the pearl fishery was uncontrollable. One year 10,000 tons of shells were harvested from a mile and a half of the Mississippi near Muscadine, Iowa, constituting perhaps 100 million mussels. As wild animals, the mussels were free for the taking. Rivers were public. The shells were steamed open on the riverbank, the pearls picked, the marketable shells sold, the meat and unusable shells abandoned. The stench left by professional mussel-fishers made them disliked.

Other factors were also in operation to doom the mussel fishery. The eastern tributaries of the Mississippi flowed down from the pineries of Wisconsin and Minnesota. From the mid-nineteenth century on, logs were floated down the rivers to sawmills, and the sawn lumber shipped west by wagon or railroad to the growing prairie towns. Michigan’s timber resources were exported via Lake Michigan to Chicago, where depending on demand, much of it rotted in the yards. The pineries of Michigan, Wisconsin and Minnesota comprised about 300 billion board feet of timber, which was cut off in less than 50 years. About 100 billion board feet had already been cut in New York, Pennsylvania and New England. (Because of continued cutting, and because catastrophic fires on the logged-over ground converted much of the pineland of the upper Middle West to aspen and scrubland, the current stand of white pine in the northern states is about 10 billion board feet.) No dam was built on the Mississippi or its tributaries while the timber trade was underway. But the end of the trade was a disaster for the timber towns. Most American towns have arisen for an economic reason: to service newly opened agricultural lands; near a mill dam site; near a water source suitable for brewing beer, or tanning hides; near mines; near good harbors. Most of these functions would sooner or later end: the timber would be used up; the hemlock or oak tanbark gone; mines or fisheries depleted; coal would replace waterpower; wool fall in value. Cheap transportation of materials and goods would make cheaper labor available, that is, more economic, elsewhere, and so (for instance) textile and shoe manufacturers moved from water-powered New England to the cheaper labor of the American South, then to southern and eastern Europe and Brazil, then to Mexico, the Caribbean, Vietnam and China; while cheap, fast transportation made poorer land, in, say, New England (but land close to the market), uneconomic to farm, compared to better land in the Middle West, or to government-financed irrigated land in California. Thus American cities and towns have always had to re-invent themselves. The main motive for this was to protect land values; when the businesses left, land values collapsed, and buildings were worth little or nothing. Peoples’ savings (and banks’ assets, most of them in real estate) were lost. So when the cut was over in the upper Middle West, a clamor arose from the Mississippi towns for dams, and the water-powered electricity and river navigation they would bring. River navigation was an anachronism by then; railroads into Chicago had already largely eliminated St. Louis, an old river port, as a transportation center. Railroads ran year-round and were not dependent on water levels, wind or ice. But boosters pointed out that bulk transportation of prairie grain by barge down to the port of New Orleans would be cheaper than rail if the federal government financed and maintained the roadway (that is, the dam and lock system on the river). Sales of hydroelectricity from the dams would pay for the construction of the lock and dam system and for ongoing maintenance. In 1894 the hydroelectric station at Niagara Falls, New York, had helped bring industry to Buffalo, N.Y. Buffalo milled prairie grain and made steel, thanks to its position on the Great Lakes and on a major railroad corridor. Hydroelectric power was also considered important because in the early twentieth century the United States was thought to be running out of coal.

Keokuk Dam at Alton, Illinois, was the first dam on the Mississippi’s main stem. The site was suggested by Major Stephen Long during a survey of the river in 1817. (Long also called the Mississippi upstream from St. Croix marvelously clear, an echo of Jefferson’s assessment of the Illinois.) Keokuk Dam was built for transportation and electric power. Electric power is generally the only use that will pay for modern dams; irrigation and transportation uses will not pay for ongoing maintenance, much less capital costs (a measure of how cheap food and transport are). Public water supply and flood control provide what are considered essential services, and so a large part of their cost is usually borne by the public treasury. (In general, flood control can be provided more cheaply and effectively by altering land use upstream than by building dams; or by keeping buildings out of reach of the flood; but for the most part this means altering land use on private lands, which means taking politically unpopular positions. Dams also create jobs, are a visible expression of the power of government, and are almost always the more politically palatable solution.) Dams of course prevent fish migrations. Many North American fish and mussel species are co-adapted. Fish carry mussel glochidia (a sort of larvae) in their gills. That is, the mussels use the fish, some of which eat adult mussels (and some of those release their glochidia into the fish’s mouth as their shells are crushed) for distribution of their young. So dams cut off the movement of mussels and their recolonization of new territories. If overfishing hadn’t doomed the mussels, the dams would have. Fish migrated around some dams on the Missouri in the 1993 floods and the levee-breaking high waters of 1993 resulted in the best fish reproduction on the Missouri and lower Mississippi in several decades; but this is not a desirable situation on a controlled river.

The lands alongside large flatland rivers like the Illinois, the Wisconsin, and the Mississippi are overflow lands, with swamps, marshes and floodplain lakes, and require levees and drainage, or both, to function as cropland or buildable land. Such drainage schemes are very expensive. Early private schemes usually failed and were completed, if at all, at public expense. Because of the expense of making them usable, riverside lands along the Mississippi (overflow land and swampland) had remained a forgotten part of the public domain. But it was government policy to convert as much public land to private land as possible and by the middle of the nineteenth century such lands began to be regarded as an opportunity. The Swampland Act turned them over to the states, who sold the land to speculators. The money from the sales was supposed to be used to finance drainage. The speculators, resold the land to farmers, with or without providing drainage. The general notion was that once turned into farmland, the lands would enrich farmers and other landholders, increase local tax revenues, and give another turn to the upward spiral of wealth; but the money raised by land sales was usually not enough to provide drainage. The land available was not small; the Mississippi alluvial valley, all of which was once subject to flooding, varies from 20 to 80 miles wide from its junction with the Ohio River to the Gulf of Mexico, a straightline distance of 600 miles (about 1200 miles by river, before navigation improvements). The land rises about 6 inches per mile from sealevel at the Gulf to 300 feet in southeastern Missouri. After the Second World War, the federal government took over management of the Mississippi River and slowly the wide flatland rivers of the Middle West were dammed and their floodplains were leveed off. Thanks to political pressure, the levees were built much closer to the main stem of the Mississippi than the Army Corps of Engineers wanted. The Corps wanted to set them back a mile on each side, which would have made the river much easier to control. Channeling eventually drained 17 million acres in the alluvial valley (that is, 70% of it) and a total of 120 million acres along all the rivers of the Mississippi drainage.

The annually flooded lands are an important part of the riverine ecosystem. Most of the productivity of large river-floodplain ecosystems is in the floodplain. A floodplain greatly enlarges the area available to fish for feeding and spawning. A floodplain expands a river’s littoral zone, the shallow water along its margins. The littoral zone is usually much more productive than the deeper waters of the river channel. Its water is well oxygenated and warmer. Sunlight reaches to the bottom. In a river-floodplain ecosystem the advancing margin of the floodwater creates a moving littoral zone over the width of the floodplain, the actual width depending on the topography of the floodplain and the height of the flood. (The Illinois had an exceptionally flat floodplain, with an exceptionally long flood, and was a very productive fishery.) The flood pulse controls the productivity of the floodplain; great floods re-arrange the plants and landscape and may result in major spawning and recruitment events for fish. The nutrients that are released from the flooded soils stimulate the growth of green plants, algae, zooplanckton, aquatic insects, and various other invertebrates and the animals that feed on them. The firm terrestrial soils and the vegetation of the floodplain make better spawning sites for many fish than the soft sediments of the permanent lakes and backwaters. As the flood recedes, the fish, including the young fish of that year, return to the permanent backwater lakes and the river. In a river/floodplain ecosystem, the main channel of the river serves as a migratory pathway for fish and as a refuge and feeding area during periods of low flow, while floodplains with their lakes, backwaters and marshes provide new sources of food and spawning and nursery areas for fish and greatly increase the productivity of the main channel. Especially in turbid rivers, where silt reduces the penetration of light, the channel may be least productive part of the ecosystem. Floods are essential to the system, providing large episodes of fish recruitment and refreshing floodplains with nutrients and water; and if large, rearranging floodplain vegetation and the river channel. In the 1920s, before many riverside wetlands were drained, and the physical, chemical and biological integrity of the system began to break down, fish rescue operations were mounted to returned stranded juvenile fish to the river from the drying pools of the floodplain (people competing with the raccoons, minks and herons). For successful fish reproduction, the flood must be high enough, last long enough, and occur when temperatures are favorable for spawning.

The permanent floodplain lakes and marshes along the upper Mississippi and its tributaries like the Illinois were spawning grounds for northern pike, large-mouth bass, and yellow perch, and were used by by big-mouth buffalo and bluegills; they were feeding and resting places for migratory waterfowl; nesting sites for dabbling ducks, flycatchers, rails and herons; home for furbearers like muskrats and mink; habitat for turtles. The fisheries in such natural rivers were extremely productive. In 1900 the Illinois produced approximately 10% of the freshwater commercial fish catch in the United States, or 24 million pounds. It also produced about 500,000 pounds of snapping and softshelled turtles. This comes to about 170 pounds of fish per acre of permanent water (and an unknown fraction of the standing crop of fish). Such numbers are not unusual: the Tippah River in Mississippi had a standing crop of 241pounds of fish per acre before channelization; and 5 pounds per acre after channelization to turn it into a waterway destroyed the riverine habitat. (Similarly, catches in the Illinois fell to 4 pounds per acre by the 1970s, as the effects of pollution and riverworks took firmer hold.) Whether a catch of 170 pounds per acre on the Illinois was sustainable isn’t known.

The banks of these Middle Western rivers were wooded. In the 1820s streamside forests began to be cut for fuel for steamboats. Steamboats required a lot of wood and cutting soon deforested the banks. After deforestation, or a destructive flood (one that killed trees), cottonwoods and willow would resprout, sycamore and silver maple seed into the new mud; other trees would come later in the succession. The tallest trees in the floodplain forest topped out at about about 100 feet and were used as rookeries by herons and as nesting and roosting sites by hawks and eagles. In spring and fall migratory ducks would eat the mast of the nut trees. Trees on the banks would fall into the river, float away, jam against the bottom, collect other trees. Stable jams might become islands. Large logjams were common on Middle Western rivers. They extended the floodplain, raising the height of the flood.

By the 1920s pollution from industry, sewage plants and oil refineries; siltation from agricultural erosion, from bank erosion by barge traffic, from sewage solids; damage from dredging; and over-fishing would eliminate mussels as an economic resource, or a resource of much biological importance to the rivers. One can see this in the history of the Illinois River, a major tributary of the Mississippi. The Illinois drains west central Illinois, rising at a height of land south and west of Chicago and entering the Mississippi above St. Louis. In 1848 a canal was dug to connect the Mississippi to Lake Michigan, via the lower Illinois. The canal was 36 to 48 feet wide and 6 feet deep and doubled the flow in the Illinois, permanently flooding some backwater swamps and killing some riverside vegetation. For a time, the additional water, together with the fertilising effects of the nutrients from Chicago sewage and stockyard waste, probably increased fish production. In 1900 the Chicago Ship and Sanitary Canal, a larger waterway, opened and began the wholesale transfer of sewage and stockyard waste from the City of Chicago to the Illinois River. Algal blooms fertilised by this material grew, died and sank to the bottom of the river, where the algae decayed, using up the oxygen in the water. Navigation dams stopped the river from flowing in low water, turning it into a series of polluted pools, making the condition worse. (As long as the water was moving, the dams, by oxygenating the water that flowed over them, may have helped the situation, but when the flow slowed, the water in the pools stagnated.) By 1910 the river was anoxic for much of its length. The low levels of summer oxygen eliminated a normal bottom fauna (including the mussels) and killed most of the rooted vascular plants in the river. Levels of decayed material (algae, sewage sludge) accumulated on the bottom of the river. The sediments became too soft for aquatic vegetation to root or for mussels to anchor. Both were replaced by tubifed worms, which live in mud. Agricultural runoff peaked in the 1920s and 1930s in northwestern Illinois and added heavy silt loads to the river. (During the droughts of the 1930s, when erosion was enormous, stream beds in the Middle West were raised 10 to 30 feet by soil erosion from farmland, making streams wider and shallower and more prone to flooding.) Dredging for navigation removed the accumulating material, both silt and sludge, but placed it on the floodplain, where floods returned it to the river, and where constant barge traffic kept it in suspension. Resuspending the sediments greatly increased the oxygen demand in the water, supporting the anoxia. Dredging also negatively affected the winged water insects (the naiads), whose larvae inhabit bottom muds. (Naiads, such as mayflies, are insects of clean streams, sometimes extremely abundant, whose larvae and adults are food for waterfowl and fish. The return of mayflies to Lake Erie late in the twentieth century was a sign of the lake’s recovery from massive nutrient pollution.) In 1948 the last factory making buttons from mussel shells on the Illinois closed and diving ducks (ring-necked ducks, canvasbacks, ruddy ducks, lesser scaup: fish and invertebrate eaters) began to disappear from the river. By 1955 the fall population of scaup was zero, probably because organochlorine pollution had eliminated the fingernail clams. By 1965, 1200 metric tons of chlorinated hydrocarbons were being spread on farmland in the Illinois basin every year. Much of this ended up in the river and the floodplain lakes. While many of the persistent organochlorines were phased out in the 1970s, abundant applications of insecticides, herbicides and fungicides continued. In the late 1980s annual applications to cropland in the Mississippi basin reached 100,000 tons of pesticides and 6 million tons of nitrogen fertiliser. These figures don’t include fertilisers and pesticides applied to local lawns. Many modern herbicides are biodegradable, but break down more slowly in water than in soil, and so effect vegetation in the rivers and in the gulf marshlands. They affect riverside trees during floods. Rooted underwater plants more or less completely disappeared from the lower 200 miles of the Illinois and its well-connected backwaters in the 1950s and with them the dabbling ducks (mallards, pintails, widgeon, teal: the plant and insect eaters).

The water quality problems caused by sewage sludge, stockyard waste and toxics slowly improved on the Illinois during the last quarter of the twentieth century, thanks partly to the Clean Water Act of 1972. Stockyards moved from Chicage to the High Plains. The federal government began subsidizing the construction of sewage treatment plants. Some toxics were banned. However the discharge of many toxic chemicals into the river (so-called permissible discharges) continues. Industrial use of Illinois water comes to approximately 1.5 billion gallons a day. Some writers claim the effect of this use is more affected by the quality of the effluent than by the volume of use, but water used industrially is likely to be more or less sterilized of small fish, fish larvae, and the larvae of aquatic invertebrates, returning to the river as a nutritious, but dead, soup. Between 1974 and 1989 there were 350 spills of hazardous materials in the river, that is, about 1 every 2 weeks. The use of the river for shipping continues. (The valley of the lower Illinois ships more grain per mile than any other midwestern river.) Dissolved oxygen is still low, tubifed worms have replaced fingernail clams and naiads in the more polluted stretches of the river, and heavy metals are found in the mussels that survive; some mussels are still collected and their shells shipped to Japan, where they are ground into the grains that seed cultured pearls.

Agricultural erosion strongly affects the rivers of the Mississippi Drainage. The Illinois basin is almost entirely agricultural. (The State of Illinois is 96% farmland.) From 1945 to 1986 row cropland increased by 67% in the Illinois basin. Hedgerows, once an important component of the farm landscape in Illinois, were removed to make fields larger and more easily cultivable. (Most row crops require spring work to be done at the same time, so speed is essential. With 10 miles of hedgerow per square mile of farmland, hedgerows were important habitat for birds, insects and small mammals. They reduced wind erosion, water flow off fields, and assisted in aquifer recharge.) Contour plowing, developed in the 1700s in Virginia by a nephew of Thomas Jefferson to slow erosion on hillsides, and rediscovered in the 1930s, was used less and less after the Second World War. Fall plowing (good farmers plow in the fall so as to have the fields ready to plant in the spring) and growing the summer crops of soybeans and corn left the ground bare for two-thirds of the year. Improved farm machinery let farmers square off fields by channeling the streams that drained them. Channelization, by shortening a stream’s length and increasing its slope, increases its velocity and erosive power. The bare banks of channelized streams, no longer protected by grasses and by bird-planted trees and shrubs, erode. Bank and bed erosion in channelized watercourses produces 50% of the annual sediment yield of Illinois streams. The sediment is carried down into the marshes and forested deltas where feeder streams enter the Illinois, raising their beds and eventually causing problems with flooding. To control the flooding, the marshes are channelized. Channelization destroys their ability to trap sediments and nutrients, sending both into the Illinois. The Illinois floodplain originally occupied 400,000 acres of the 18.5 million acres in the lower Illinois basin, or about 2% of its area. Half of the original floodplain was leveed off for agriculture, including areas below the level of the river, which have to be kept drained by pumps operating much of the year. This meant the increased yield of sediment was concentrated in the remaining oxbows and in the floodplain lakes and backwaters. Some floodplain lakes lost half their depth. From 1958 to 1961 the remaining clear and vegetated backwaters and lakes became turbid and barren. Their use by gamefish and ducks declined drastically. Lakes tend not to cleanse themselves but to collect and recycle nutrients and pollutants. Ecosystems can flip to different productivity states, some of which (usually those lower in the succession) have markedly less biotic regulation of energy flow and biogeochemical cycles. Turbid, shallow, eutrophic lakes, with heavy algal blooms, are a biologically stable alternative to shallow, clear lakes with rooted aquatic vegetation.


Turbidity is a major problem in most Midwestern rivers. Increasing turbidity during the 1950s came partly from the increased amounts of silt in runoff, partly from algal blooms caused by nutrients in the water, partly from rooting in the bottom by the introduced European carp, partly from boat traffic and natural wave action. Rooted aquatic vegetation helps control turbidity, but excessive turbidity reduces the light available to vascular plants and thus reduces photosynthesis. This weakens and kills the plants in the deepest parts of the rivers and their lakes and backwaters. No longer damped by plants, the waves from periodic windstorms and from barge traffic become stronger, further increasing the turbidity, and killing more plants. Rooted vegetation goes into a downward spiral. Once killed, the vegetation is difficult to re-establish. Fine-grained sediments take 7 to 12 days to settle out after a windstorm, but in the Illinois basin strong to moderate winds occur every 7 days. Waves from barge traffic are undamped and rooting in the bottom muds by carp (part of its feeding behavior) also raises sediments. So the excess nitrogen and phosphorus from the fields that might have been turned into useful biomass by functioning riverside swamps and submergent vegetation, by trees in the floodplain, and finally by the vegetation in the wetlands of the Mississippi Delta, turns into algae in the lakes and rivers, and creates the dead zone in the Gulf of Mexico. (Bacteria in the remnants of the floodplain and the still pools of the reservoirs still remove about 35% of the nitrogen that enters the Mississippi.) The lack of aquatic plants and their associated plants and animals means little habitat remains for waterfowl or furbearers. Up through the 1950s furbearers, especially muskrats, but also skunks, racoons and mink, provided extra income for many small Middle Western farms. The fish population shifts from sight predators and nest builders (the traditional gamefish) to fish that locate their prey by scent and scatter their eggs over the bottom (such as the bottom-feeding carp). Northern pike, predatory fish of clear lakes, inhabit the interface between rooted aquatic vegetation and open water. A pike population needs 25% plant cover to maintain sufficient biomass to control bottom-feeding fish, like carp, in a lake, and so reduce their effect on turbidity. (Carp can also be controlled by drawing down a lake at spawning time, exposing their eggs and fry in shallow pools. This is a more drastic solution and a temporary one if the lake is connected to other bodies of water with carp.)

The Illinois is a satisfactory industrial river. It is a highway for barge traffic, a source of industrial water, and a dilution basin for industrial discharges. Its fish are not as unhealthy as the bullheads in the Anacostia River of Washington, D.C., half of whom suffer from liver cancer. Its floodplain lakes, despite their blooms of blue-green algae, are used for water-skiing. The return of a healthy riverine environment would mean eliminating barge traffic and might in some ways be a nuisance: mayflies splatter on windshields during their hatches, collect on sidewalks and make roads slippery. Millions of migrating waterfowl shut down airports, if only for periods of a few hours. (Ninety million ducks headed south in the fall of 1997, when the continental population had recovered from the droughts of the 1980s.) Shipping out grain by truck or rail to some point on the Mississippi or the Great Lakes that can better take barge or deep water ship transport means more truck and rail traffic. (The Mississippi at Vicksburg, a natural deepwater port, is 2000 feet wide and 60 feet deep.) The withdrawal of industrial water is easier to manage satisfactorily: most of it could be recycled. As for toxics, their release could be much reduced, or eliminated by modifying industrial chemistries. Sometime in the 1990s, the Illinois was invaded by two species of Asian carp that been introduced to southern catfish ponds to control algae. The carp subsequently escaped in floods to the Mississippi drainage. Ten years later one of them, silver carp, dominates fish biomass in the Illinois. The introduced carp do so well compared to native fish because the Illinois is a damaged habitat. Silver carp are a common food fish in China and surprisingly tasty. One way to deal with the Illinois is to accept the present situation and clean up the ruined river so that its fish are edible. (Similar justifications could be used to accept falling salmon numbers in the altered rivers of the Pacific Northwest, or falling numbers of native fish in the Colorado.)

* * *

Modern rivers are maintained according to standards developed during the first half of the nineteenth century in Germany (the country that also developed the modern chemical industry, with the production of dyes for cloth). The nineteenth century channelization of the Rhine was the first great modern riverwork. The ideal industrial river is a controlled, single-channelled stream that stays within its banks, and is useful as a waterway, a source of water, and a drain. Gangs of men dug by hand new, narrower, straighter channels for the Rhine (a notoriously meandering river), in order to speed its flow and make it dig itself a deeper bed. The natural drainage this created allowed former riverside marshlands to be cultivated and reduced floods upstream. It lowered water tables in nearby fields (not always an advantage and for this reason some communities resisted riverworks), made navigation easier, and eliminated malaria, salmon, and the gold sifted from its gravelly banks (about 10 pounds per year in the early 1800s). The improvements, by sending more water more quickly down the river, increased flooding downstream in Holland. The Rhine changed from a slow, meandering, silty river with riverside swamps and forests into a fast, silt-starved, erosive stream, bordered by roads and railways, that constantly erodes its bed. Earth and stone have to be continually dumped in some reaches to prevent the river from eroding its bridge abutments. As for salmon, with some further improvements the Rhine could now support a run of 6-12,000 fish, about 2% of the run in the natural river. Similar changes to the Mississippi for navigation above its junction with the Ohio have reduced the river’s surface area by 33%, its island area by 25%, and its riverbed by nearly 25%. The river was narrowed to increase its depth and scour, to make it more fit for barge transportation.

Natural rivers flood, migrate across their flood plains, open two or more channels, dig cross-over flows and cutoffs, and are full of snags (about 1 every 10 feet on large rivers). Riverwater spirals downwards through the watershed, using and acquiring nutrients (sometimes recycled within 100 yards), abruptly forming new habitats, as when pools give way to riffles and shallows. Rivers change as their slope changes, as they flow over different substrates (gaining or losing chemicals), as tributaries enter them, from exchanges of gases with the atmosphere, of nutrients with their connected groundwaters. The wooded, more easterly rivers of the Mississippi drainage, such as the upper Mississippi, the Illinois, the Wisconsin, and the Ohio, were clear in pre-contact times; and under development became more silty (and warmer in some reaches, deeper and colder in others). Many plains rivers, such as the Missouri and the Arkansas, were naturally warm and muddy, with populations of fish that were adapted morphically and behaviorally to such conditions. Channeling changed the Missouri, at whose entrance the Mississippi changed from a clear to a silt-laden river, from a warm-water stream with high turbidity, wide seasonal variations in flow, and a braided channel that constantly changed course, to a narrow, cold, fast-moving stream, with (here and there, below its dams) excellent trout fisheries. The gallery cottonwood forests that had lined the river for 1000 miles died from drought. Ninety-three percent of the emergent wetlands, backwaters and sloughs along the Missouri were converted to agriculture or dredged for channels. Many plains rivers are used for irrigation. Largely because most of its water was taken to grow corn, the North Fork of the Platte River, a major tributary of the Missouri, changed from a meandering, braided channel 2500 to 4000 feet wide to one narrow, well-defined channel 200 feet wide. (The mean annual flood of the Platte fell from 13,000 cubic feet per second to 3,000 cubic feet per second and the mean annual discharge from 2300 cfs to 560 cfs after its development for irrigation: that is, about 75% of the flow was removed.) In dammed rivers, most of the silt and organic material end up being stored in the river, instead of feeding riverine wetlands downstream. Two-thirds of the Missouri’s silt ends up behind dams (considerably shortening their lifetimes: there are 60 dams on the Missouri and its major tributaries). The Ohio, once a wide, clear, shallow stream that flowed through a watershed that was almost entirely forested was changed by dams (for water supply, electricity, and river transport, mostly of coal for power plants) into a deep, cold river, its mussel population a remnant, with new fisheries for walleyed pike below its dams. The silt the Ohio contributed to the Mississippi increased 10 times thanks to agriculture and industrial and urban development in its watershed, but the Ohio’s increased silt loads did not make up for that lost from the plains rivers.

East of the Appalachians, in the Northeast, where rivers flow into the Atlantic Ocean, single channel rivers without extensive wetlands connected to the river, seemed normal to twentieth century observers, though some remarked that rivers in undeveloped areas formed multiple channels and had connected wetlands. It is now thought that eastern rivers had also been transformed by agriculture and dams. By 1800 most eastern streams had mill dams at every suitable site (18,000 in Pennsylvania alone). These were low wooden dams, 6 to 8 feet high. They were set on long pine or hemlock timbers laid into the riverbed in the direction of the current. Braces mortised into the timbers held up two parallel wooden walls that crossed the stream. The space between the walls was filled with clay, with clay and stones in the center. On the upstream side planking angled down from the top of the dam to the streambed to let ice and trees slide over the dam. The weight of the earth fill held the dam in place. These dams flooded any adjacent wetlands or side channels. As the countryside about them developed, their ponds slowly filled with silt eroded off the surrounding watershed. The rivers became a series of long pools. When the dams, abandoned from the late 1800s thanks to other sources of power, eventually failed, the river cut itself a single channel through the accumulated sediment. It was now too far below its floodplain to have any connected wetlands. When in great floods the rivers try to re-arrange themselves over the floodplains (and construct new channels or excavate wetlands) they are put back in their former places since too much human development now occupies what has become dry land.

About 250 miles from the Gulf, the Mississippi starts distributing its water, that is, rivers start to flow out of it rather than into it. Much of the land along the river (and for many miles to each side) was once bottomland hardwood forest that accumulated sediment at 10 to 20 tons, a fraction of an inch, per acre per year. Some was marshland, which was inhabited by alligators. The holes the alligators dug were used as refugia in winter (and during summer droughts) by amphibians and fish. Their nesting mounds (a mix of mud and vegetation 5 to 7 feet wide, and 3 feet high) were egg-laying sites for turtles. Dry spots in the swamp, they were colonized by plants and became nesting sites for birds. Mast-bearing trees like oaks rooted in them. Thus alligators (like the accumulating silt) varied the habitat. Because of the silt, the river constantly changed its course through the delta, creating new cutoffs or abandoning old ones. (In contrast, the Mississippi and its major tributaries upstream of Cairo Illinois, which carry much less silt, have maintained more or less stable channels over the last 2500 years.) Downstream the river met the sea in a maze of channels and islands, in which it dropped its final load of sediment. The main distributaries have shifted several times but the wetlands at the river’s mouth have continued to grow at about 1.5 square miles a year for the last 5000 years. The delta shoreline is a balance between sedimentation, sealevel rise, and subsidence of the land underneath (both natural and from the pumping of water and petroleum). The wetlands become more fresh as one moves upriver, until trees (including cypresses, which live 400 to 1000 years) grow in the marshes, anchoring the islands and providing better protection for developed areas inland from storm surges and hurricane winds (such as New Orleans and the adjacent coasts of Louisiana and Mississippi). The delta wetlands provide nursery areas for much of the marine life in the Gulf of Mexico (perhaps 80% of a $5 billion fishery, which represents 200,000 jobs and 25% of the U.S. fish catch). The plants of the freshwater marshes are eaten by migratory waterfowl.

The Mississippi now carries 35-40% of the silt of 150 years ago, despite more erosion in the watershed (the balance is stored behind dams), but the silt it now carries, if allowed to spread over the delta wetlands, would cover 60 square miles 0.5 inch deep each year. River control structures send the silt (and the nutrients carried with it) out into the gulf, bypassing the marshes, which are now losing land at something like 50 square miles per year. (The rate increased in the 1990s; about one-sixth of the Louisiana marshes have been lost to the sea over the last 150 years and the delta is now subsiding at about 0.5 inch a year.) The silt, biocides and fertilizers pour directly into the gulf, and, partly through direct action (pesticides killing zooplankton or shrimp larvae), but mostly through fertilizing the algal plankton, which grow, die, sink, decay, and so deoxygenate the bottom waters, create a so-called dead zone with lethally depleted levels of dissolved oxygen covering several thousand square miles of the gulf bottom. Currently the dead zone covers an area the size of New Jersey. Fish swimming here in the summer die, as do the shellfish and worms that live on the bottom. An artificial shipping channel (Mr. Go, not much used; most ships use the natural river) has been widened extensively by storms and tidal action and lets salt water intrude upriver into the marshlands toward New Orleans, making them less desireable to waterfowl and killing the cypresses. Ten thousand miles of abandoned oil exploration canals, also constantly enlarged by tidal action, help erode the wetlands. Letting the silt flow into the wetlands through siphons or gates would make it do useful work in rebuilding them. The fertiliser carried with it would grow useful biomass. Mr. Go and the oil exploration canals can be filled (if possible), or dammed with earth barriers (in the second case, the work paid for by oil companies). The value of the Louisiana wetlands in the Mississippi Delta have been calculated to be worth $846 per acre per year in fishing benefits, $401 in furs, $181 in recreation, and $7,500 in storm protection benefits. These are coastal wetlands so their value in hurricane protection and fisheries is more than that of upstream floodplain lands, though other calculations put the value of any natural wetland at $6,000 to $8000 per acre.

Most modern rivers are controlled by dams. Dams create discontinuities in riverine ecosystems. They prevent fish from using all of a river. Below a dam, the warm silty water of a plains stream emerges clear and cold, good habitat for the introduced trout, but not for native fish, which are adapted to the warm silty summer flow. The water is cold because it is released from the bottom of the reservoir. By storing much of the silt and organic material moving through river systems, dams impede the movement of material and energy through a drainage basin. Thus dams inevitably cause the erosion of riverbanks, riverbeds, deltas, and sea beaches at river mouths, and change the habitat of estuaries where marine fish spawn. They alter the thermal regime to which a river’s fish and invertebrates are adapted. Temperature is often used as a cue for egg hatching or larval development; the wrong temperature makes eggs hatch or larvae develop at the wrong time. Cold flows from dams may be too cold for the larvae of native fish. By changing a river’s thermal regime dams change its faunal structure. Dams also alter habitat terrestrial habitat. Bennett Dam in British Columbia stopped the flooding of the Peace/Athabasca Delta in Lake Athabasca and thus eliminated the muskrat habitat (and with it the muskrats) on which the native peoples depended. Dams for the La Grande hydroelectric complex on James Bay more or less eliminated the char populations of the rivers and also reduced the numbers of lake trout. Whitefish, pike and chub increased. None of the fish will be edible for the next 20 to 100 years because of methyl mercury released by bacterial decomposition of rotting vegetation in the lakes. Because of the fluctuating water levels, the margins of the reservoirs, unlike those of the rivers that preceded them, are rarely used by wildlife like ptarmigan, beaver or moose. About 10,000 pairs of waterfowl lost nesting sites in the flooded riverine wetlands.

The reservoirs behind dams are often centers for transmission of fish diseases; and are inhabited by unstable assemblages of fish. (Typically the fish population booms in a new reservoir with the release of nutrients from newly flooded soils, and then falls.) But some fish introductions have been successful, such as the trout fisheries below the high dams on the Colorado; or (for a while) the striped bass fishery in Lake Mead. (That fishery was also based on an introduced shad as a forage fish and on algae fertilized by the river water. Building Glen Canyon Dam on the Colorado upstream of Lake Mead shut off much of the supply of phosphorus the algae needed and the bass fishery precipitously declined.) Introduced fisheries are often successful at the expense of the native species in the river. Trout, for instance, eat the young of native Colorado fish such as razorback suckers and pikeminnows and make recovery of their populations, already hindered by modification of the riverine habitat (especially the colder water and the loss of floodplain backwaters), more difficult.
Dams slow the flow of water to the sea. Five times the water volume in streams is now held behind dams worldwide, lowering sealevel by a little over an inch and slightly slowing the rotation of the globe. It takes several times as long for water to move through a dammed river’s basin (up to a year, in some cases). Reservoirs emit methane, a greenhouse gas, from the decomposition of organic material washed into them (or left in them when they are filled). Methane emissions from reservoirs are largely a tropical problem. Half the reservoirs in Brazil emit more greenhouse gases than would a coal-fired power station of the same output. Flooding low gradients like those of the Mississippi Valley to generate electricity uses 200 times more land per kilowatt-hour than photo-voltaic collectors, but hydro-electric power stations with reservoirs produce 200 times the energy that goes into them (the energy used to build, maintain, and fuel the generating equipment), compared to 40 times for wind turbines, 20 times for heavy oil, 10 times for coal, probably less than 10 times for photo-voltaic collectors. The reservoirs behind dams are kept full with spring runoff so as to generate electricity, supply irrigation water, and for water supply (all demands that rise in summer), so they are not much use in summer floods. During the 1993 floods on the Mississippi 12 times the storage capacity of the reservoirs moved through them. The reservoirs reduced flood peaks by a few percent (a few feet) of the total rise. There has been a statistically significant increase in warm season floods in the Mississippi Valley from 1940 to 1990, thanks largely to an increase in the frequency and magnitude of heavy rainstorms. More frequent and heavier precipitation is predicted for a warming world. Insurance companies may favor realistic methods of flood control, such as letting rivers occupy more of their former floodplains. Controlling floods this way doesn’t take up a large amount of land (perhaps 3% of urban and agricultural land in the lower 48 states). Large dams are engineered to last 250 years but have an economic life of 50 years. Then, as their concrete starts to deteriorate from contact with the water and their reservoirs fill with silt (1% per year on average, but a half life of 20 years on China’s Yellow River), they reach the age of repair. In the United States dams are licensed for 50 years. Many licenses are coming up for renewal.

Some biological systems are difficult or impossible to manage, especially if they exist in alternative stable states. We see managed biological systems as stable, but natural systems experience, and may require, fluctuations, some seasonal or episodic, some chaotic. Stabilizing water levels in wetlands reduces the diversity of plant species and results in a long-term decline in productivity (as with reservoir fish); stabilizing water levels in prairie potholes reduces the wet meadow and marsh zone, important in spring for some species of ducks, reduces the overall diversity of plants and lets cattails take over the wetland. Stabilizing water levels in the Everglades also lets cattails take over. Important seasonal habitats may not appear attractive: dabbling ducks use flooded sections of plowed fields (vernal pools: full of newly hatched invertebrates) in early spring, move to more permanent water in summer. In many ecosystems drought, flood, and fire are necessary agents of biological renewal. Such catastrophic events interfere with human economic development.

For the most part we have not tried to see rivers as biogeophysical systems connecting the hills to the sea, with all their interconnected productivities, but as untamed beings to be turned into stable and benign industrial entities: straighter, with a constant water level, useful for navigation, water supply, as a source of power. River control schemes were part of the Progressive Movement (1890-1920), that sought to transform the natural world into an efficient, productive and controlled system for maximum human benefit. Whether in the end controlled rivers are worth more than natural ones is a matter of debate, and depends on how the ecosystem services of a natural river are priced (these include lower floods; the turning of riverine silts and nutrients into trees, waterfowl and fish; cleaner water; abundant riverine and ocean fisheries; expanding deltas and offshore islands; healthy coral reefs; growing beaches; protected coasts), compared with the river’s value for navigation and electric power, and the use of its floodplains for farms, powerplants and houses. Real estate values rise near swimmable rivers with fish and eagles; and one can keep many of the services of a wild river and still use some of its water and some of its power. The cost of maintaining a controlled river, including the cost of rebuilding after floods (currently something like $2 to $4 billion a year in the United States) also has to be taken into account.

The United States has 620,000 miles of controlled rivers. Some of the dams that control them (those that are old, abandoned, or of less economic value than the fisheries they displace) may be removed as their licenses expire (as was Edwards Dam on the Kennebec in Maine), but most will not be. The Napa River in California has flooded periodically for decades. After a flood in 1986 caused $100 million in property damage and the evacuation of 5000 people in Napa City, along with 3 deaths, the residents of the City of Napa decided, rather than further reinforce the levees, to let the river reclaim part of its historic flood plain. The Corps of Engineers removed dams and levees along 7 miles of river. Approximately 650 acres of wetlands were restored. The expenses not covered by the federal government, which included buying up some of the floodplain and moving a set of railroad tracks and some buildings, were paid for by a 0.5% rise in the sales tax in the City of Napa. In this case, the rise in real estate values in the city from ending the risk of flooding, turned out to be considerably greater than the cost of the additional taxes. Napa residents also avoided the costs of flood insurance, the inconvenience and danger of flooding, and the occasional death.

Once a river is dammed, engineers favor damming the whole drainage so as to control flow all along the river. The best way to restore the rivers of the upper Middle West would be to remove all the dams (and thus barge transportation) on the Missouri from its junction with the Mississippi and all those on the Mississippi and its tributaries above St. Louis. Freight such as grain and coal would be shipped by rail on new lines, their construction subsidized, to natural deep water ports further down the river or on the Great Lakes. Levees would be moved back and a sufficient amount of the original flood plain bought up to absorb floods. Watersheds would be managed to reduce the drainage rate into streams. Less land would be irrigated from the rivers that cross the plains. This isn’t going to happen, but controlled rivers can be managed so as to take better advantage of their services. Riverside fields that require pumps to keep them dry can be returned to the floodplain. Lowering levees to let 10-year floods take over farm fields (which could be farmed 9 years out of 10 and the other year compensated for) would let the river produce more fish and the fields and floodplain remove more nutrients. Run-of-the-river flow schemes time water releases from dams to mimic natural cycles (spring floods, summer low flow), and try to maintain a certain level of flow, against the demand for industrial water (for power generation, moving barges through locks, irrigation), and so restore aquatic habitat. Under them, populations of insects, mollusks, invertebrates and fish all rise and their growth rates increase. Letting rivers like the Illinois flood during their normal flood period rather than storing all the water behind dams, and letting them reclaim more of their floodplains, would improve the functioning of the riverine habitat. The floodplains would absorb the fall and winter flood and the warm season floods from thunderstorms. Scheduling lock and dam repair for low water, in low water years, would make such work interfere less with the natural flow of the river. Sequestering dredge spoil (full of toxics and nutrients) would keep it out of the river. Ideally, dredge spoil would be processed to remove the contaminents and the remaining soil spread on farmland. Ending industrial releases of toxics and increasing fines for spills would make future dredge spoil less toxic. Changing agricultural practices, such as rotating row crops with legume hays, strip cropping, the use of winter cover crops, and vegetating the banks of channelized streams would reduce the flow of silt and nutrients into the river. The ecological services of a healthy river are not owned by anyone but are part of the general good. However many uses of private property connected with the river will reduce them; and thus reduce the general good.

Restoring shallow floodplain lakes, such as those along the Illinois, is a more difficult matter. Lakes tend to store nutrients, silt and pollutants. The muds of the lakes are now saturated with phosphorus. Release of stored phosphorus from these sediments is likely to maintain algal blooms in the lakes for a long time after agricultural input of the nutrient falls. In summertime, the lakes rapidly stratify, with warm, oxygenated water on top and cool, oxygen-poor water below. Stratification is accompanied by the rapid development of anoxia in the sediments and the release of more phosphorus into the water column. Every 7 to 10 days a windy period mixes the migrating phosphorus into the water column and breaks up the anoxia, setting the stage for another algal bloom (which returns much of the phosphorus to the sediment in the bodies of the algae). Bottom-feeding carp and brown bullhead in the lakes transform the particulate matter of the sediments into soluble nutrients through their digestive systems, also at a rate sufficient to maintain the algal blooms. The blooms help prevent the re-establishment of rooted underwater plants, and thus help eliminate competition for the carp, which would be partly controlled by a healthy population of pike. Floodplain lakes can be dredged and their nutrient-rich sediment (if not too toxic) pumped onto nearby farmland. Together with agricultural practices that reduced nutrient losses and erosion this would solve much of the nutrient problem. Permanent improvement means re-establishing aquatic vegetation. Since the turbid lakes are now in a relatively steady biological state, and are kept turbid by periodic prairie winds, this is difficult, but probably not impossible.

Control of a river allows its overflow land to be farmed or otherwise developed (perhaps 120 million acres in the Mississippi drainage). The economic return from agriculture on drained land in the greater Mississippi Valley (about $36 billion at $300 per acre gross return) is balanced by the costs of maintaining the controlled river (dredging, dam and levee construction and maintenance; a few billion a year) as well as the other costs associated with a controlled river, such as flood insurance and the yearly $2 to $4 billion in clean-up costs, an 80-90% loss in riverine fisheries and an unknown one in ocean fisheries (another few billion), more expensive water purification systems (this cost is unknown, but in general, $1 spent on protecting water resources saves $7 to $200 in the cost of water treatment and filtration facilities). The profit involved in developing the most vulnerable overflow land (that nearest the river or with the highest water tables) is less than zero. Commodity agriculture is not very profitable (returns on capital are usually under 3%). Government-supported loans, grants for farm improvements and price supports for crops make it work. Subsidies to farmers in the industrialized world amount to a billion dollars a day. Farming is so little profitable because so much of its income goes to the larger economy: to steel makers, oil drillers, chemical manufacturers, farm equipment makers, car dealers, grain brokers, railroad and barge transportation systems, electric power generators, insurance salesmen, banks, seed companies, bio-tech laboratories, and so on. So agriculture contributes to the growth of the larger economy. Its total contribution to the U.S. Gross Domestic Product is estimated at 12.5%. Some of this is lost if overflow land is not drained. One could also argue that the free economic goods provided by a less disturbed landscape, such as flood control, clean water and abundant and edible wild fish, are less economically desireable than the active management required by the current riverine landscape (levees, dams, a managed river, yearly flood clean-up costs, more expensive water purification systems, fish farms). The more manipulated a landscape is, the more work it provides. (Of course, such complete control often has unpleasant consequences. The elaboration of work required to maintain a landscape under increasing human control resembles in many ways the elaboration of financial investments, issued on increasingly slim margins of credit, during a period of financial boom—just before the collapse.) While the riverside communities and farmers cannot afford the cost of maintaining the river, and that money must come from taxes on the economy as a whole, the economy as a whole (those steel-makers, car manufacturers, and so on) benefits, and the tax monies are re-distributed to the economy through the work required to maintain the man-made river. A similar logic would congratulate us for cutting down all the old-growth forests in the United States, since global warming is going to destroy them anyway (and at least the money went into other development); for continuing to log the Amazon (the opinion has been voiced, that if the forest is going to collapse, we might as well make money off it while that’s possible); or for fishing out wild ocean fisheries (an immensely lucrative pastime). The question is whether ecosystems and ecosystem services (which, like wild animals, are not owned by anybody) have sufficient value to justify regulating behavior on private lands; or in the public lands of the oceans.

Economic arguments favoring development work because ecological goods such as flood control, wild fish and clean water are not given sufficient economic value. Ecological services acquire economic value when their loss causes economic harm. While the economic value of a landscape varies over time, its ecological value doesn’t. A newly cutover piece of timberland is worth little as timberland (who wants to pay taxes on it while the trees grow?) while 200 years later its timber may be worth a fortune; but all along the streams that flowed through it depended on the state of that landscape for their health. The economic costs of ecological conservation measures are in most cases trivial compared with their economic benefits, but in the end the economic argument misses the point. We could not live without clean air, clean water and fertile farmland. Many ecological benefits (swimmable rivers, birds that return in the spring, the roar of the spring flood) are difficult to value (perhaps they are incalculable). Others are too remote from human notions of value to be easily assessed in monetary terms. To replace the cycling of nitrogen by bacteria, for instance, we would have to manufacture the nitrate, spread it, and then control its leaching over the whole surface of the planet. To replace the work done by cellulose-decomposing bacteria we would have to deal with the 100 billion tons of plant material that die every year, 10 times the mass of the fossil fuels extracted yearly. Luckily (or as far as we know) we have not yet much interfered much with the control of the planet by microbes. The rising seas associated with global warming will wipe out much of the real estate value of the east coast ($250 billion in Miami alone, much of it no longer insurable), as well as the remaining ecosystems the land once held (the estuaries, sand dunes, mud flats, tidal marshes). A 2-foot rise in sealevel will change half the freshwater marshes of the Everglades to salt marshes. But waves and rivers will create new and similar ecosystems inland, over the foundations of former settlements. Did we do the right thing in extinguishing the value of coastal lands as ecosystems in order to create all that human value (since it’s now going to be lost anyway)? It would seem some coasts, barrier islands, and river floodplains should never have been settled with permanent structures. Hurricane Katrina sent a storm surge 35 feet high over the barrier islands and 18 feet high into the coast of Mississippi, wiping out 90% of the structures within a half mile of the coastline. The same buildings had been wiped out 37 years before by Hurricane Camille. Such coasts should never have been settled, but left as habitat for fish, shellfish, deer, eagles, cougars, as wooded refuges for exhausted, northward migrating birds. Since the quest for economic benefit is the only thing that causes significant change in a capitalist society (and does so efficiently), many people look for a way to give ecological goods (like empty land) a reasonable dollar value, in order to prevent a market economy from further degrading the natural world. The hope is to guide the market economy toward improving the natural world. Currently, this hope seems problematic, with the average value of terrestrial ecosystems put at $466 per acre per year, probably a trivial number to a suburban developer. (At 5% interest, it would require a payment of $9320. per acre into a fund to ameliorate the loss of ecosystems services by the development.)

Numbers put on the value of ecological services are still somewhat arbitrary. (I am thinking of the value of flood control, clean water, of carbon capture by grasslands or forests, of the protection of coasts by mangroves and coral reefs.) Some calculations put the value of global ecosystem services equal to that of global annual human production, some put it at 10% of that. Land in a modern economy is still cheap, compared to human services, like war, pensions, or medical care. About 12% of the earth’s surface is currently protected, much of it uninhabitable mountains and deserts. A writer estimates that raising the area of protected landscapes to 10% of every major region would cost $23 billion annually for protection and policing; buying the land, another 2 million square kilometers, would cost $4 billion. Such costs seem very low; a plan to put 12-25% of the United States into permanent conservation status is estimated at $360-$980 billion for purchasing the land or easements on it. The cost of such schemes, especially in developed countries, and their necessary size (20-40% of each landscape is better than 10%), makes one realize that—even if such schemes come to fruition—a gentler way of using the human landscape is probably the only way a functioning natural landscape (overall) will survive. An advantage of natural systems is that their return, while small compared to man-made systems (probably 2-3%), is very long-term. If lands are selected carefully, their ecological value in human terms is endless, and if they are sufficiently connected, their value continues through periods of changing climate. While investment in human systems must be renewed at intervals of 30 years (roofs) or 250 years (large dams), investments in natural systems renew themselves. Overflow lands constitute about 7% of the lower 48 states, and probably half of that is necessary for healthy riverine ecosystems. As the world develops, the value of functioning rivers, woodlands, and coasts rises; this is scarcity value, now expressed in real estate value. In an edible world, such landscapes provided food, clothes, drinking water, building material. These values are lost in a modern marketable landscape, but the effect of living biological landscapes on real estate values still grows. Since the degradation of ecological function always affects someone somewhere (the value of his lands, the condition of his water and air, the state of his health), more and more schemes to protect the natural functioning of landscapes are probably in our future.

Ecosystem services are part of the general good of nature, and of the human world, not owned by anyone, but heavily influenced by human economic behavior. So-called stakeholder capitalism attempts to take account of the needs of a company’s stockholders, its workers, and the communities in which it operates. We should add to this the welfare of the larger natural environment on which any civilisation floats, the welfare of our agricultural soils, and of our farm animals. Expanding the capitalist view is not always a simple matter: Henry Ford was sued by his stockholders when he raised his worker’s wages by $1 a day, even as the time taken to manufacture a Model T fell from 12 hours to 2.5 hours. But views evolve. Protective ecological schemes are a likely a matter of survival. Biologically functioning landscapes also give one hope, as well as a connection with the non-human world in which we lived up several millennia ago, and to which, if our civilisation collapses, we will return.