Where have all the rivers gone?

SOME OF THE EARTH'S MAJOR ARTERIES ARE NOW SO CONSTRICTED THAT THEIR LIFEBLOOD NO LONGER REACHES ITS NATURAL DESTINATION. THE RESULT IS A MASSIVE FAILURE OF ECOSYSTEMS AND ECONOMIES.

In 1922, American naturalist Aldo Leopold journeyed by canoe through the great delta of the Colorado River. What he reported seeing there was a verdant waterscape, where for millennia the river had been depositing its rich silt and building up a diverse ecosystem before entering the Sea of Cortez - known north of the border as the Gulf of California. He saw deer, quail, raccoon, bobcat, vast fleets of waterfowl, and even the great jaguar, el tigre - "the despot of the Delta." The meandering river, slowing as it spread out through countless green lagoons, led Leopold to muse, "for the last word in procrastination, go travel with a river reluctant to lose his freedom in the sea."

Leopold never returned to the Delta for fear of finding this "milk-and-honey wilderness" badly altered. His fears were justified; today, the Colorado's freedom has been lost to a degree even the prescient Leopold could scarcely have imagined. Except in years of unusually high precipitation, the Colorado River no longer reaches the sea at all - it literally disappears into the surrounding desert. Much of the abundant wildlife is gone. Off the coast to the south, the once-productive fisheries in the Sea of Cortez have declined dramatically. In striking contrast to Leopold's experience, author Philip Fradkin more recently characterized what is now a dessicated place of mud-cracked earth, salt flats, and murky pools as "the most inhospitable terrain on the North American continent."

What has happened to the Colorado is but an extreme example of a disturbing worldwide trend: more and more rivers are running dry as dams and diversions siphon water off for burgeoning cities and thirsty farms. In Arizona, the Salt and Gila rivers used to converge west of Phoenix; but now they dry up east of the city because of extensive diversions for irrigated farms in the region. In California, some 35 kilometers of the San Joaquin River have been so permanently dewatered that thickets of trees have sprung up in the dry riverbed, sand and gravel are mined from it, and developers have even proposed building houses in it. In China, about 50 kilometers south of Beijing, villagers say the Heaven River dried up 20 years ago. And in the water-deprived Middle East, where surface streams are extensively overtapped, the lower stretches of the Jordan River have dwindled to a salty trickle.

It is the arresting decline of the world's larger rivers, however, that most graphically conveys the magnitude of the problem. The Nile, the Ganges, the Amu Dar'ya and Syr Dar'ya, the Huang He (or Yellow River), and the Colorado are each now so dammed, diverted, or overtapped that for parts of the year, little or none of their fresh water reaches the sea. Their collective diminution portends not only worsening water shortages and potential conflicts over scarce supplies, but mounting ecological damage. That damage, in turn - from degraded river deltas and species on the brink of extinction to shrinking inland lakes and disappearing wetlands - places the economies and people who depend on them at growing risk.

Many regions have now fallen into a zero-sum game - in which increasing the water supply to one user means taking it away from another. More water devoted to human activities means serious and potentially irreversible harm to natural support systems. With population and consumption levels rising at record rates in many parts of the world, it is a dilemma with far-reaching consequences - one that calls for a wholly new approach to valuing and managing rivers.

RIVERS AT RISK

Human efforts to control rivers date back thousands of years. The Assyrian Queen Sammu-Ramat, who ruled during the late 9th century B.C. in what is now northern Iraq, is reputed to have had inscribed on her tomb: "I constrained the mighty river to flow according to my will and led its water to fertilize lands that had before been barren and without inhabitants." In Zhengzhou, China, about 700 kilometers southwest of Beijing, stands a statue of the emperor Yu, under whose reign the Chinese people were said to have prospered from the building of dikes and irrigation canals to control the mighty Yellow River.

But it was not until this century that engineering schemes began to alter natural water courses on a massive scale. The construction of dams to store water and diversion canals to transport it to cities and farms has been central to the economic growth of regions wet and dry alike. A controlled supply of water for irrigation became critical to boosting food production as population and consumption grew. Large-scale hydroelectric power fueled urban and industrial expansion. And the taming of flood waters allowed farms and towns to situate on fertile soils and near shipping channels.

Indeed, for a time, river basin "development" became the sine qua non of economic advancement. Winston Churchill, noting the Nile's importance to the entire northeast Africa region, after a military campaign on that river in 1908, prophesied that "One day, every last drop of water which drains into the whole valley of the Nile...shall be equally and amicably divided among the river people, and the Nile itself...shall perish gloriously and never reach the sea." He was not speaking sardonically. Four years later, a California engineer, Joseph Lippincott, foresaw a similar fate for the Colorado River. "We have in the Colorado an American Nile awaiting regulation, and it should be treated in an intelliggent and vigorous a manner as the British government has treated its great Egyptian prototype."

The construction of the great Hooever Dam on the lower Colorado in the 1930s broke all engineering records up to that time. Some 220 meters high and able to store 1.7 years' worth of the river's average flow, Hoover Dam presage, an engineering frenzy that was to tame many of the world's rivers over the ensuing decades. Around the world, the number of "large" dams (those more than 15 meters high, and backing up months' or years' worth of their rivers' flows) climbed from just over 5,000 in 1950 to about 38,000 today. More than 85 percent of large dams have been built during the last 35 years.

Hoover Dam was also the first of a generation of structures that became known as "superdims" - those more than 150 meters high. Today, there are more than 100 of these megaliths, most of them built during the past four decades. Besides Hoover, they include such giants as the Bhakra Dam in India and the Itaipu in Paraguay, as well as such familiar U.S. dams as Glen Canyon on the Colorado and Grand Coulee on the Columbia.

Worldwide, dams collectively store on the order of 6,000 cubic kilometers of water - equal to 15 percent of the earth's annual renewable water supply. Thousands of kilometers of diversion canals siphon water out of rivers and reservoirs and deliver it where and when needed to expanding cities and farming regions. Globally, Water demand has more than tripled since mid-century, and the rising demand has been met by building ever more and larger water supply projects. Many rivers now resemble elaborate plumbing works,w ith the timing and amount of flow completely controlled by planners and engineers so as to maximize the rivers' benefits to human activity.

But if the intended triumph of such plumbing was to bring rivers as neatly into the service of human convenience as a bathroom faucet, it hasn't worked out that way. Rivers are central to the planet's ecology; turning them on and off at will damages other parts of the system. And because aquatic organisms cannot live long without water, large reductions in streamflow - even for short periods of time - can be damaging or deadly to them. In addition, most water-based animals are well-adapted for living either in the flowing waters of rivers and streams or in the standing waters of lakes and ponds; relatively few thrive in both. When rivers are dammed, and flowing water is replaced by a permanent reservoir, many river species are placed at risk.

In short, the manipulation of river systemms is wreaking havoc on the aquatic environment and its biological diversity. Fresh waters contain extraordinary concentrations of animal life - including, for example, about 40 percent of the 20,000 recognized fish species. According to some estimates, the total diversity of animal life per unit area of rivers is 65 times greater than that of the seas.

In North America, the American Fisheries Society lists 364 species or subspecies of fish as threatened, endangered, or of special concern - the vast majority of them at risk because of habitat destruction. Many are found in arid regions where intensive water demands and rising salinity are destroying vital habitat. More than a third of the desert fish species of the American Southwest, for instance, are now listed as threatened or endangered. As Professor Alan Covich of Colorado State University puts it: "We have often ignored the high species richness associated with inland waters and have allowed many fresh water habitats to be dammed, channelized, drained, eroded, and polluted with nutrients, salts, silt, and chemicals. Biodiversity and ecosystem integrity are declining in a wide range of locations throughout the world...."

AND THEN THERE WAS NONE

The Colorado River ranks among the most heavily plumbed water courses in the world. Controlled by 10 major dams, it now irrigates some 800,000 hectares (about 2 million acres) of farmland, serves the household needs of more than 21 million people, and generates nearly 12 billion kilowatt hours of energy annually. Its waters fill swimming pools and sprinkle green lawns some 400 kilometers away in Los Angeles, power neon lights in Las Vegas, and irrigate thirsty crops in the deserts of California, southern Arizona, and northern Mexico.

Ironically, it was in 1992, the same year Leopold experienced the wet, wild abundancee of the Colorado River delta, that seven U.S. states started diving up the river. They signed the Colorado River Compact, which gave 7.5 million acre-feet (about 9.25 billion cubic meters) to the upper basin states per year and an equal amount to the lower basin. Mexico finally got an assured supply in 1944, when the two countries signed a treaty committing the United States to send an average of at least 1.5 million acre-feet per year across the border.

Unfortunately, there were two major problems with these allocations. First, a total of 16.5 million acre-feet had been committed to the seven states and Mexico, but the long-term average flow of the Colorado produced only about 90 percent of that; more water had been promised than the river could reliably deliver. Second, none of the compacts and treaties dividing up the Colorado's water designated any flow for the river environment itself, including the delta and its abundant wildlife. As long as human demands remained well below the river's flow, this was not a problem. But except for unusually high flood years, virtually the entire flow of the river is now captured and used - and has been for some time (see graph on page 13). Indeed, flow readings at El Meritimo, the southernmost measuring station on the Colorado, were discontinued in 1968 because there was nothing to measure.

Only recently has much attention been given to the effects of these diminished flows on the river delta and estuary. The delta and upper Gulf of California comprise the largest and most critical desert wetland in the American Southwest, as well as one of the world's most diverse and productive sea ecosystems. Besides drying up wetlands and causing a severe deterioration in water quality, the reduction in freshwater flow has also cut the flow of nutrients to the sea and reduced critical habitat for nursery grounds.

Catches from the upper Gulf shrimp fishery have dropped off steeply, and other fisheries are in decline as well. Unfortunately, there is little hard data to correlate these declines with the drop in river flows. But local fishers believe that, along with overfishing, it is a major cause. Indeed, the "golden days" to contemporary fishers in the upper Gulf apparently refer to the mid-1980s, after the high flood year of 1983, when huge snowmelts in the upper Colorado basin caused the river to flow at rates not seen for several decades. And according to Alejandro Robles, executive director of Conservation International's Mexico program, many upper Gulf fishers believe that the rare flood flows of January 1993 temporarily brought back substantial numbers of fish species that had become scarce in the region.

But such rare events canot reverse the long-term decline of the ecosystem or the economic, social, cultural, and ecological toll it is causing. A large number of species that depend on the lower Colorado-upper Gulf ecosystem are now threatened or endangered, including the green sea turtle, the Yuma Calpper Rail, the desert pupfish, the bonytail chub, and the razorback sucker. Much attention has focused on the vaquita, the world's smallest porpoise and most endangered sea mammal, whose population in the upper Gulf is believed to number just a few hundred. Also of special concern is the totoaba, a steel-blue fish that grows up to 2 meters in length and 135 kilograms in weight and that once supported a popular sports and commercial fishery. The totoaba used to breed in large numbers in the brackfish waters of the Colorado estuary, while spending most of its adult life in the deeper waters of the nearby upper Gulf. Between, habitat degradation and overfishing, the totoaba is now on the verge of extinction.

The Cucapa, or "people of the river," a 2,000 year old culture of fishers and recessional (flood-based) farmers, are fading with the ecosystem around them. Just 40 to 50 families remain south of the border. There is little work for the younger tribal members, and many have migrated to the cities. Traditionally, the Cucapa ate fish three times a day, but now they are lucky to have it once a week. Since the water is too salty to grow certain traditional crops, such as melons and squash, their diets have become less healthy. As Anita Alvarez de Williams, an expert on the Cucapa told a National Geographic writer several years ago, these people "have been around for a couple thousand years. But barring a miracle, you're seeing the last of them."

Other communities of the upper Gulf - including El Golfo de Santa Clara, San Felipe, and Puerto Penasco - were initially founded as fishing camps, and fishing remains the basis of their economic and cultural viability. According to researchers Marcela Vasquez Leon, Thomas McGuire, and Hernan Aubert, shipyards are closed, packing plants are operating well below capacity, local businesses are suffering, and households are struggling to survive. "The most direct way to revive the economies of the upper Gulf," they write, "is to revitalize the upper Gulf itself..."

EGYPT'S LIFELINE

All it takes is a single flight into Cairo in order to grasp what the ancient Greek historian Herodotus meant when he called Egypt "the gift of the Nile." A vast sea of desert sand surrounds a narrow green strip of life on either side of the river. This strip, along with fertile lands in the Nile delta near the Mediterranean Sea, have supported Egypt's civilization for millennia.

Like the Colroado, the nIle is a lfieline for this desert country that gets virtually no rain. It sustains 60 million people and irrigates some 3 million hectares of cropland. Moreover, with Egypt's population climbing by 1 million people every 9 months, and the need for drinking water and food rising proportinately, demands on the river are intensifying - even as upstream countries contemplate siphoning off more Nile water for themselves.

Egypt has practiced irrigation perhaps longer than any other society, having drawn upon Nile waters for agriculture continuously for at least 5,000 years. For much of Egyptian history, the basic pattern of water use was that of "basin irrigation," in which a series of canals formed 5,000-to-10,000-hectare basins that stair-stepped slightly downstream. Nile water was diverted into the higher basins, flooding them and depositing nutrient-rich silt. The water then drained successively into each lower basin, until at the end of the sequence it re-entered the Nile to flow out into the Mediterranean. Egyptian geographer Gamal Hamdan described the ancient basin system as an ecologically sustainable adaptation to the natural environment - one that had a large enough water-to-land ratio to prevent a buildup of salt, and that let the bulk of Nile water run to the sea.

This system, however, limited crop production to just a third of the year. It was during the nineteenth century that Egypt converted to perennial irrigation with its extensive system of barrages (small dams) and canals. The culminating structure in the network was, of course, the High Dam at Aswan, which was constructed during the 1960s to provide virtually complete control over the Nile's waters and a crucial hedge against drought. Lake Nassar is able to store fully two years of the Nile's average annual flow.

Prior to construction of the Aswan High Dam, some 32 billion cubic meters of Nile water reached the sea each year, equal to 38 percent of the river's average flow. After the dam was built, flow to the sea dropped dramatically - to some 6 billion cubic meters - as flood waters were halted at Lake Nassar and diverted for crop production. With greater control over the river and continued expansion of irrigated lands, outflow to the sea sank to about 3 billion cubic meters in the mid-eighties. Today, the amount of freshwater from the Nile reaching the Mediterranean totals just 1.8 billion cubic meters per year, and all of it is released to the sea during the few winter months when crops need less irrigation. Moreover, in order to expand irrigated land further, the government hopes to be storing 80 percent of this remaining outflow in one of the northern coastal lakes by the end of the decade, which would leave less than 400 million cubic meters of fresh water flowing to the sea - scarcely one-half of one percent of the Nile's total annual runoff.

At this point, it should be noted that the patterns of diminution vary greatly from one river to another; and while very little of the Nile's fresh water reaches the sea, a fair quantity still reaches the delta. In contrast to the Colorado, which is virtually gone by the time it reaches its delta, the Nile still has a substantial amount of its water flowing far enough to irrigate its delta's rice, cotton, and other crops. From there, what is released to the sea each year is some 12 billion cubic meters of salty and polluted farm drainage. With the nation's water demand already bumping up against supply limits, the government plans to reuse as much of this drainage as possible.

Not surprisingly, the High Dam has greatly altered the river system, although with the Nile, as with the Colorado, cause-and-effect linkages are not altogether clear. Out of 47 commercial fish species thriving in the Nile prior to the dam's construction only 17 were still being harvested a decade after the dam's completion. And in the eastern Mediterranean, the annual sardine harvest dropped by 83 percent - a likely side-effect of the reduction in nutrient-rich silt entering that part of the sea.

Perhaps the most threatening long-term consequence of the Nile's diminished flow is that the delta, so essential to the country's economy, is slowly falling away into the sea. Most river deltas naturally subside from the weight of their own sediment, but under natural conditions this is usually countered by deposition of silt brought in by the river. The Nile transports an average of 110 million tons of silt each year, much of it fertile soil from the Ethiopian highlands. For thousands of years, 90 percent of this silt reached the coast to replenish the delta, while the remaining 10 percent was deposited on the Nile floodplain. The delta stopped growing about a century ago, after the first barrages were built by the British. But since completion of the High Dam, and the trapping of virtually all the silt in Lake Nassar, the delta has actually been in retreat. Borg-el-Borellos, a former delta village, is now 2 kilometers out to sea.

Global warming and the anticipated rise in sea level that higher temperatures will bring greatly increases this threat of inundation. Much of the northern delta lies only 3 to 4 meters above sea level. Researchers at the Woods Hole Oceanographic Institution in Massachusetts calculate that Egypt could lose 15 to 19 percent of its habitable land within about 60 years, displacing a similar portion of its population - which by then would likely total well over 100 million - and wiping out some 15 percent of its economic activity.

In the meantime, the paltry amount of Nile water reaching the sea is a harbinger of difficult times ahead for Egypt. Prospects for increasing Nile supplies have diminished since a joint project with Sudan to channelize part of southern Sudan's Sudd wetlands has been put on hold indefinitely because of the Sudanese civil war. Pumping more water from wells and recycling more of the drainage from farms will help somewaht, but these options are limited. And Egypt, last in line of the 10 countries in the Nile basin, also faces the prospect that its upstream neighbor Ethiopia - source of 85 percent of the Nile's water and now enjoying greater political stability - will soon begin capturing more of the river's flow for its own economic development.

SHRINKING SEAS

Some of the most dramatic consequences of river depletion are found where rivers empty into inland lakes or seas. In the Sahel region of Africa, for example, the combined assaults of prolonged drought and diminished inflow - especially from the Logone and Chari rivers, which have been heavily diverted for irrigation - have shrunk Lake Chad by about three-fourths over the last 30 years. This lake and its tributaries harbor some of the richest fisheries in the world. The Grand Yaeres, an area of wetland twice the size of Luxembourg along the Logone River floodplain, is a critical haven for fisheries, livestock, and recessional agriculture in the central Sahel. Diversions from the Logone River at Maga Dam, much of it to irrigate rice, are drying out these important wetlands and causing water tables to fall beneath some 150,000 hectares of the Logone floodplains.

The most publicized example of a lake dying for lack of river water is the great Aral Sea in central Asia. Once the world's fourth largest freshwater lake, the Aral has steadily been shrinking since 1960 because of the diversion of its two major sources of inflow - the Amu Dar'ya and Syr Dar'ya - for irrigation. Prior to 1960, these two rivers poured 55 billion lion cubic meters of water per year into the Aral, a little over half their average combined flow. Between 1981-1990, their combined discharge to the Aral dropped to an average of 7 billion cubic meters - just 6 percent of their total annual flow (see graph on page 14). Much of the time, these rivers now run virtually dry in their lower reaches.

The Aral Sea's demise may have been foreordained after companies of the Karakum Canal in the late 1950s. Some 1,300 kilometers in length, the Karakum transports water from the Amu Dar'ya westward toward the Caspian Sea. Combined with a vast network of inefficient irrigation systems built to expand cotton production in the desert, this huge diversion of water began a process of slow desiccation that would eventually result in the bizarre sight of seagoing ships sitting on bone-dry land - the sea having literally evaporated from under them.

Under the direction of Moscow's central planners, irrigated agriculture in the basin expanded by half during the 1960s through the 1980s, reaching some 7.5 million hectares. As more and more river water has been siphoned off, the sea has continued to shrink steadily. In 1988, it actually split in two - into a larger lake in the south and a smaller one in the north. By the mid-1990s, the Aral's area had dropped by half and its volume by three-fourths. Salinity levels had tripled.

This still-unfolding chain of ecologicaal destruction and human suffering ranks the Aral Sea as one of the planet's greatest environmental tragedies. Some 20 of the 24 fish species in the Aral have disappeared, and the fish catch, which totaled 44,000 tons and supported some 60,000 jobs in the 1950s, has dropped to zero. Abandoned fishing villages dot the sea's former coastline. Each year, winds pick up anywhere from 40 to 150 million tons of a toxic duct-salt mixture from the dry sea bed and dump them on the surrounding farmland, harming or killing crops. The low river flows have concentrated salts and toxic chemicals, making water supplies hazardous to drink. Coupled with poor sanitary conditions and heavy pesticide use, contaminated drinking water has contributed to rampant disease. The incidence of typhoid fever has risen nearly thirtyfold, and that of hepatitis, sevenfold. The rate of esophageal cancer in Muynak, an old fishing port, is 15 times the Soviet average.

Both river deltas have become severely degraded by the reduction in river flows. According to Philip Micklin of Western Michigan University, a leading U.S. authority on the Aral Sea basin, the tugay forests - assemblages of willow, poplar, tamarisk, and other water-loving trees and shrubs that were common in the delta regions - have been decimated, in turn destroying vital habitat for the region's animal life. Wetlands have shrunk by 85 percent. That, combined with high levels of agricultural chemical pollution, has greatly reduced waterfowl populations. In the Syr Dar'ya delta, for example, the number of nesting bird species has fallen from an estimated 173 to 38.

Growing interest, support, and study have been directed to the region by groups ranging from the United Nations Environment Programme, the World Bank, European Union, Russian and Central Asian scientists, and a variety of local and international NGO's. Since the dissolution of the Soviet Union in 1991, responsibility for the Aral basin tragedy has shifted from Moscow to the former Central Asian republics (Uzbekistan, Kyrgyzstan, Tajikistan, and Turkmenistan) and Kazakhstan, all now independent countries. They have formed the Interstate Council for the Aral Sea and several supporting institutions - key steps toward corporation. Ameliorating the human suffering and stabilizing the Aral Sea environment is going to take an extraordinary level of commitment and funding by these young nations and international donors. And here, as in Egypt, high rates of population growth compound their water problems: the combined population of the five basin countries is projected to climb by 14 million people during the next 16 years.

DRY SEASON BLUES

Throughout Asia, where monsoon climates prevail, the management of rivers poses special challenges. Much of the region has substantial precipitation overall, but gets the bulk of it during a relatively short rainy season. India, for example, gets 80 percent of its rainfall during three to four months, with much of it coming in just a few monsoon storms.

This leaves many regions prone bo both serious flooding and drought. Dams that capture and store runoff during the wet season have been a principal means of evening out the water flows - reducing flooding during the rains and assuring irrigation during drought. Even as population and food needs continue to expand, many Asian rivers are completely tapped out during the drier part of the year, when water supplies are so critical to irrigation. According to a 1993 World Bank study, "Many examples of basins exist throughout the Asia region where essentially no water is lost to the sea during much of the dry seaosn."

Among the largest of these is the Ganges, which rises in the Himalaya mountains of Nepal and then flows through India and Bangladesh before emptying into the Bay of Bengal. In the early 1970s, India completed the Farakka barrage with the aim of diverting Ganges water into the Hooghly River so as to improve navigation and increase water supplies for the port city of Calcutta. Newly independent Bangladesh was concerned that not enough Ganges water would cross into its territory during the dry season, and that its crop production would be reduced.

The two countries agreed in 1977 to a short-term solution for sharing the Ganges' dry-season flow, and also guaranteed Bangladesh a minimum amount of water during extremely low-flow periods. That agreement expired in 1982, and was replaced with an informal accord that did not include the guarantee clause for Bangladesh. A follow-up agreement expired in 1988. Since then, the two countries have been deadlocked, leaving Bangladesh with no assurance of minimum flows for its dry-season irrigation needs. In 1993, the dry season flow into Bangladesh was the lowest ever recorded. As river beds dried up and crops withered, the northwestern region suffered severely. The Ganges Kobadak Project, one of this poor nation's larger agricultural schemes, reportedly suffered an estimate $25 million in losses.

India's unilateral diversion of the Ganges at Farakka also means that during the dry season the river no longer reaches its natural outlet through Bangladesh into the Bay of Bengal. One result has been the rapid advance of a saline front across the western portion of the river delta, causing serious and possibly irreversible damage. Rising salinity levels are damaging coastal mangrove forests, many species of which require fresh or brackish water. Mangroves are not only important sources of wood for cooking (rice can't be eaten without cooking), but also provide critical habitat and breeding grounds for many species of fish and shellfish. Unless more water is allowed to flow into the delta during the dry season, the damage to vegetation and loss of fisheries is likely to continue, along with a spreading disruption of the area's economy.

A similar problem is arising in the delta of Thailand's Chao Phraya basin. Water demands in the basin already exceed available supplies. Indeed, a team of World Bank water specialists has noted that "Inevitably, as nonagricultural demands increase, and if no additional water becomes available, supplies for dry season cropping will tend to decline." Flows for navigation are consistently below optimum, and water supplies to the Bangkok area are not sufficient to alleviate the severe overpumping of groundwater there, which is causing portions of the land to subside. Yet unless water is allowed to flow into the delta during the dry season, a saline front will advance through this delta, as well, threatening irreversible harm to the ecosystem.

In China, water projects that include major diversions for irrigation have dried up the lower reaches of many rivers during portions of the year. The Yellow River, known as "China's Sorrow" for its long history of flooding, is now frequently bone dry during the dry season. Indeed, competition for water is particularly keen in the north China plain, where surface waters are nearly fully exploited and groundwater tables are falling at alarming rates. The drying up of surface streams signals mounting water scarcities in this country where some 100 cities and towns already suffer shortages.

CHANGING COURSE

An old Inca proverb says, "The frog does not drink up the pond in which it lives." It presages one of the looming challenges of our time - reconciling the growing water demands of expanding populations and economies with the need to protect water's fundamental ecological support functions.

The age-old notion that any runoff to the sea is "wasted" reflects a narrow view of what a river's work - or evolutionary purpose - really is. That purpose includes delivering nutrients to the seas, with their complex food webs; sustaining economically and culturally important fisheries; protecting wetlands, with their capacity to filter out pollutants; providing habitat for a rich diversity of aquatic life; safeguarding fertile deltas; protecting water quality; maintaining salt and sediment balances; and offering some of the most inspirational natural beauty on the planet.

Restoring and maintaining the integrity of river systems is going to take the deployment of new technologies, policies, and management strategies. It will take unprecedented cooperation both within and between countries. And, most fundamentally, it will take a new ethic of sharing water - not only with each other, but with nature as well.

A critical first step is for societies to recognize that there are limits to the amount of water that can be diverted from rivers. Exactly how much water needs to be left in-stream will vary with the time of year, the habitat requirements of riverine life, the system's sediment and salt balances, and other factors specific to each river basin. But setting even preliminary "minimum flows" for both average and low-flow periods would provide a needed degree of insurance for the health of river systems - even as scientists progress toward a better understanding of their complex ecological workings.

In regions where rivers are already overtapped, as in much of the western United States, meeting such minimum requirements will involve shifting some water away from farms and cities over to the environment. Even a decade ago, this might have seemed a radical notion. But thanks to a flurry of court decisions, legislative actions, administrative rulings, and citizens campaigns, the process has begun.

In late 1992, for example, the U.S. Congress passed legislation dedicating 800,000 acre-feet (987 million cubic meters) of water annually from the Central Valley Project in California, one of the largest federal irrigation projects, to maintaining fish and wildlife habitat and other ecosystem needs. Among other things, it set a goal of restoring the natural production of salmon and other anadromous fish (those that migrate from salt to fresh water to spawn) to twice their average levels over the past 25 years. Two years later, in December 1994, California and federal officials reached an agreement to limit the amount of fresh water that can be diverted from the San Francisco Bay delta-estuary, a highly productive aquatic environment that is home to more than 120 species of fish. In years of normal rainfall, 400,000 acre-feet must be reserved for the ecosystem. Farmers stand to lose the most water from this reallocation, while cities will likely face cutbacks mainly in dry years. But all Californians will gain in the long run, as economic activity comes into better balance with the water environment that supports it.

In yet another landmark shift in California, the courts have greatly broadened use of a legal doctrine called the "public trust," which asserts that governments hold certain rights in trust for the public and can take action to protect those rights from private interests. The California Supreme Court used this principle to require the city of Los Angeles to stop siphoning water from tributaries to Mono Lake on the eastern side of the Sierra, which had shrunk in volume by half - and dropped by 45 feet - because of decades of the city's water diversions. Since the city had legal rights to this water, which constituted some 15 percent of its supply, it fought the ruling for many years. Finally, however, in September 1994, Los Angeles reached an agreement with state and environmental officials that will allow Mono Lake's level to rise sufficiently to halt its ecological collapse. Since even existing water rights can be revoked to prevent a violation of the public trust, this doctrine offers a potentially powerful tool for restoring the aquatic environment.

While such site-specific cases are promising steps forward, broader government leadership will be needed to fully incorporate protection of the aquatic environment into water policies and river management. In the United States, for example, some impetus has been provided by Daniel P. Beard, Commissioner of the Bureau of Reclamation, which has been one of the world's predominant dam-building agencies during this century. In May 1994, Beard told an international gathering of water specialists that "the dam building era in the United States is now over," and went on to clarify a wholly new mission for his agency: "Every problem we must address has a common theme. That is: there isn't enough water in the river.... To solve these problems, we cannot build new reservoirs. Instead, we will have to encourage the movement of water from one use to another. We believe conservation, demand management, efficiency improvements and reuse offer our best opportunities for doing this."

Consistent with this policy shift, in January 1995, the Bureau issued new draft guidelines and criteria for the development of water conservation plans by all the irrigators and cities it serves. These guidelines state that the Bureau will be "proactive" in identifying new opportunities for improved water efficiency and that the agency's funding assistance "will generally be directed toward those investments where a portion of saved water can be dedicated to environmental restoration or enhancement."

A good candidate for such proactive effort is the lower Colorado River and delta. The Mexican government took a major step toward protecting this declining ecosystem and its threatened species in June 1993, when Mexican President Carlos Salinas de Gortari declared some 9,300 square kilometers of it a biosphere reserve and banned all fishing within a core area, noting that here, "ecology has priority over politics," and that "today, we are just realizing what we are at the point of losing." Underpinning his political statements are the potential economic benefits of restored fisheries, and possibly eco-tourism. Action by Mexico alone, however, will not be sufficient. Unless, the United States joins its southern neighbor in restoring river flows to the delta ecosystem, the decline of this region is only likely to spread.

One possible source of "surplus" water could be the large quantities used for irrigation in the lower Colorado River basin. In the Welton-Mohawk Irrigation District in southern Arizona, for example, the federal government heavily subsidizes three water-related operations - supplying water to the District, removing salty drainage from the soils, and desalting drainage water so that it is fit to send across the border to Mexico. Were the government to buy out these irrigators, it would not only save taxpayers money, but would free up on the order 370,000 acre-feet (about 450 million cubic meters) of water annually for restoring the lower Colorado ecosystem.

Saving water for nature will be far more difficult in developing countries, where demands for food and drinking water are rising apace with population growth. But in those countries as well, ensuring minimum water flows to satisfy ecological needs is critical to protecting fisheries, delta economies, and - as the tragedy in the Aral Sea basin underscores - the health of local people.

In all water-scarce river basins, cooperation among countries is essential not only to optimizing economic benefits from the river but to safeguarding its ecological functions. Unfortunately, while most water-sharing agreements have specified how much water each river basin country is allotted, they have allocated nothing for the river system itself. But if all countries in a basin agreed on a portion to be left in-stream to satisfy ecological needs, they could then divide up the remaining river water equitably among them.

For example, if scientists determined that at least 10 percent of the Ganges flow must reach the delta during the dry season in order to prevent the advance of a damaging saline front, this 10 percent would have to be agreed to and guaranteed by all three riparians - Nepal, India, and Bangladesh. Bangladesh's fair share of the Ganges dry season flow would then be added to this 10 percent to determine how much of the river must flow across the border into Bangladesh during the dry season. To be realistic, in the negotiations of such agreements, downstream countries such as Bangladesh would likely need to yield a slightly larger portion of their allocation to maintain minimum flow because they accrue the economic benefits of the delta ecosystems.

In the case of the Ganges, the ongoing dispute among the adjoining countries suggests that prospects for such an agreement in the near future are dim. Along the Nile, however, prospects look brighter. At a February 1995 meeting in Tanzania, the water affairs ministers of a majority of the Nile basin countries - including Egypt and Ethiopia - agreed to form a panel of experts representing all 10 basin countries with the aim of developing a basin-wide framework for water-sharing and cooperation that would lead to "equitable allocation of the Nile waters." This is a striking development, one that may not only avert conflict over water in this basin, but set the stage for more sustainable water management and use.

In rich and poor countries alike, meeting irrigation, industrial, and household water demands while also protecting the aquatic environment requires much greater incentive to use and allocate water more efficiently. In most developing countries and dry regions of wealthier countries, agriculture accounts for 75 to 90 percent of water use. In such areas, reducing irrigation needs by 5 to 10 percent can free up substantial quantities of water. Switching from sugar cane or rice to less water-intensive crops, investing in drip irrigation lines or low-pressure sprinklers to reduce evaporation losses, and scheduling irrigations to more closely match a crop's water needs are just a few of the ways farmers can save.

If farmers were required to pay prices for water that reflect its true cost, many would make these efficiency improvements. By heavily subsidizing water, governments give out the false message that this resource is abundant and can be affordably wasted - even as rivers are drying up, fisheries are collapsing, and species are going extinct. Yet virtually all governments do subsidize water use - typically by building large water projects and then charging farmers only a fraction of the water's true cost.

In some regions, the ability to sell river water can create incentives to use it more efficiently. Such water marketing implies clear property rights to the water, which do not currently exist everywhere in the world, and cannot replace governmental responsibility to ensure that ecosystems are protected. But where laws and infrastructure make marketing possible, as in the western United States, this approach can benefit the aquatic environment in two ways. First, rather than damming and diverting more river water, cities can purchase water from farmers who have saved it by improving their irrigation efficiency or have decided to fallow a portion of their cropland.

Second, the existence of markets allows private organizations and government agencies to purchase water rights and dedicate them to restoring river flows. For example, the Nature Conservancy, based in Arlington, Virginia, has returned of private instream water rights as well as by working with state agencies to transfer existing water rights to instream uses. In Colorado, a coal-mining subsidiary of the Chevron Corporation donated $7.2 million worth of water rights on the Black Canyon of the Gunnison River to the Nature Conservancy, which then turned those rights over to the state Conservation Board for conversion to an instream water right. As a result, additional water will remain in this portion of the river to benefit the trout fishery, as well as three endangered fish species.

Such creative solutions mark an emerging era in water management. Pricing, marketing, regulations, and a brace of conservation technologies all have important roles to play in meeting human demands for water while at the same time restoring and protecting rivers. The tens of thousands of dams and vast lengths of diversion canals built over the last century stand as concrete testaments to impressive engineering skills and human control over nature. But in the years ahead, we may come to judge our success at water management quite differently - by our ability to share water equitably, to do more with less of it, and to restore life and integrity to the earth's rivers.

KEY SOURCES

FURTHER READING

KEY SOURCES FOR THIS ARTICLE:

ASIT K. BISWAS, MOHAMMED JELLALI, AND GLENN STOLIT, Water for Sustainable Development in the 21st Century (Oxford, UK: OXFORD UNIVERSITY PRESS, 1993), VARIOUS CHAPTERS.

ALAN P. COVICH, "WATER AND ECOSYSTEMS," IN PETER H. GLEICK, ED., Water in Crisis: Guide to the World's Fresh Water Resources (NEW YORK: OXFORD UNIVERSITY PRESS, 1993).

BRIAN GRAY, "THE MODERN ERA IN CALIFORNIA WATER LAW," Hastings Law Journal, VOL. 45, JANUARY 1994.

HARALD FREDERIKSEN, JEREMY BERKOFF, AND WILLIAM BARBIER, Water Resources Management in Asia (WASHINGTON, D.C.: THE WORLD BANK, 1993).

PHILIP MICKLIN, "THE ARAL CRISIS: INTRODUCTION TO THE SPECIAL ISSUE," Post-Soviet Geography, VOL. 33.5, MAY 1992.

FRED PEARCE, The Dammed (LONDON: THE BODLEY HEAD, 1992).

GILBERT WHITE, "THE ENVIRONMENTAL EFFECTS OF THE HIGH DAM AT ASWAN," Environment, VOL. 30.7, SEPT. 1988.

FRANK WILSON, "A FISH OUT OF WATER: A PROPOSAL FOR INTERNATIONAL INSTREAM FLOW RIGHTS IN THE LOWER COLORADO RIVER," Colorado Journal of Internation Environment Law and Policy, VOL. 5.249, 1994.

DONALD WORSTER, Rivers of Empire (NEW YORK: OXFORD UNIVERSITY PRESS, 1985).

FOR FURTHER READING:

JIM CARRIER, "THE COLORADO: A RIVER DRAINED DRY," National Geographic, JUNE 1991.

GARY PAUL NABHAN, Gathering The Desert (TUCSON, AZ: UNIV. OF ARIZONA PRESS, 1985), ESPECIALLY PAGES 151-164.

SANDRA POSTEL, Last Oasis (NEW YORK: W.W. NORTON, 1992).

MARC REISNER, Cadillac Desert (NEW YORK: PENGUIN BOOKS, 1986).

Sandra Postel is Director of the Global Water Policy Project in Cambridge, Massachuseets, and a senior fellow of Worldwatch Institute. She is author of Last Oasis: Facing Water Scarcity (W.W. Norton, New York, 1992).