Requiem for a River

"More than 80 years ago, seven western states hammered out a pact dividing up the water in the Colorado River..."

More than 80 years ago, seven western states hammered out a pact dividing up the water in the Colorado River. Agriculture was king and Las Vegas just a railroad watering stop in the middle of nowhere. Today, after an eight-year drought, the river is in crisis. Tim Folger traveled from its snow-fed headwaters to the feeble trickle that enters the Gulf of California, asking everyone he met: What comes next?

Snake Valley, Nevada
[Elevation 5,300 feet]

Somewhere on the road between the lonely McMansion where the Mormon polygamist's senior wife lives and the dried-up spring where the wild horses died of thirst, I put my foot in my mouth. "How big is your ranch?" I ask Dean Baker, the lean and weathered owner of much of the land around us.

My question seems innocuous enough, but an embarrassed silence envelops the packed Chevy Suburban in which I'm riding with eight Nevada ranchers.

Before Baker can answer, Hank Vogler, a hefty man with a long, droopy gray mustache, interrupts. "Well, that's a bad question," he says. "That's like me asking you what does your wife look like naked. Your reply should be, 'That's none of your damn business.'"

I realize that I've inadvertently put Baker in the uncomfortable position of being asked to reveal his net worth to a stranger in front of his friends. Later I learn that he recently refused a $20 million offer from a real estate speculator for his 12,000-acre ranch, land that he and his family have worked for more than 50 years. The speculator wasn't interested in Baker's modest home, which stands behind a gas station, or even his land. He wanted the water rights. In the nation's driest state, water is a precious commodity.

On this late April afternoon, Baker has organized what he's calling a water tour to show me and ranchers from some of the neighboring valleys what eight years of drought have done to the local springs and water table. Baker raises cattle, alfalfa, barley, corn, and other crops here in the Snake Valley, about 300 miles northeast of Las Vegas. The valley is one of many in the Great Basin, the vast, arid, sparsely populated 200,000-square-mile plateau that sprawls across nearly all of Nevada and half of Utah. West of us, beyond a plain matted with sagebrush, juniper, and greasewood, the Snake range walls off the horizon, with the snowy peak of 13,000-foot Mount Wheeler lambent against the azure high-desert sky. Due east, in Utah, looms the Confusion range, named for its chaotic geologic strata. North, beyond sight over the curved rim of the planet, barren salt flats spread like a bridal train to the Great Salt Lake.

Our first stop is at Needle Point Spring, across the state line in Utah, the site of what was once a small pond. Six years ago, 12 wild horses were found dead there. The pond, on which they depended, vanished nearly overnight after a nearby rancher tapped it to irrigate some new fields. The only sign that water ever flowed there is some dead, brittle gray grass and an empty cattle trough.

The demand for water here, exacerbated by the growth of Las Vegas, has never been greater. Las Vegas, built in the middle of the Mojave Desert, gets 60,000 new residents-and four inches of rain -- each year. To secure the water it needs to maintain that growth, the city plans to build a $2 billion pipeline to pump groundwater from the valleys of northern Nevada. Baker and his fellow ranchers believe the pipeline will be a disaster, not just for them but for the Great Basin ecosystem, which is one reason we've driven to Needle Point Spring. If a single farmer can suck a spring dry, what will happen when a city of nearly two million starts pumping groundwater here?

The remote Snake Valley is but one of the many fronts in a battle for water rights that will play out in the decades ahead across the entire Southwest. The Brobdingnagian twentieth-century system of dams, reservoirs, tunnels, and canals that made possible the explosive growth of desert metropolises like Las Vegas is overtaxed. An unprecedented drought has depleted Lake Mead and Lake Powell, the huge reservoirs on the Colorado River that supply water to some 25 million people and support a $1.2 trillion regional economy. With the onset of global warming, an already bad situation is likely to get much worse. Some climate scientists suspect that the current drought is not an aberration but the start of a transition to permanently drier conditions in the fastest-growing -- and most arid -- region in North America.

Disputes over water are nothing new here. Without the Colorado River, Las Vegas, Phoenix, and Tucson would still be small towns, and the Sonoran Desert would reclaim the implausibly green Imperial Valley in California, one of the world's richest agricultural regions. Even San Diego would be running out of water.

Late last year, the seven states that share the Colorado River's water -- Arizona, California, Colorado, Nevada, New Mexico, Utah, and Wyoming -- agreed to new federal guidelines for managing the river that should prevent the drought from morphing into a full-blown catastrophe. But that agreement won't end the region's water wars. During four months of traveling from the river's wild, clear headwaters in the Colorado Rockies to its trickling, ignominious end in Mexico, I frequently heard two aphorisms, the first attributed to Mark Twain, the second anonymous: "Whiskey's for drinking, water's for fighting over," and "Water always runs uphill to money."

Glacier Gorge
Rocky Mountain National Park, Colorado
[Elevation 10,600 feet]

The river that sustains seven states begins in the sky, when moisture borne from the Pacific on westerly winds settles as snow on the peaks of the Rockies. This high, white reservoir is the source for roughly 70 percent to 80 percent of the water that flows in the Colorado. The potential energy stored in the falling snow and transferred to rushing mountain freshets sculpts the Grand Canyon and spins the 35-ton turbines at Hoover Dam.

Right now, though, the snow is mostly just making me cold. Our small party -- two scientists, a park biologist, another journalist, and I -- has paused on the northeast edge of a frozen lake, known simply as the Loch, high in Rocky Mountain National Park. We've spent this early-spring morning passing single file through a soundless white forest, buoyed atop five feet of snow by our snowshoes. We've followed the path of a stream, which courses invisibly beneath snow and ice, up the slope of a steep, U-shaped valley gouged by glaciers more than 12,000 years ago. We're nearly past the tree line now; our destination is a weather station on a ridge above the lakeshore, where we'll begin to collect snowfall and streamflow data.

"If you have an extra layer, now would be a good time to put it on," says Jill Baron, an ecologist with the U.S. Geological Survey (USGS). Baron, a crisp, energetic woman with short brown hair, dons a yellow and gray Gore-Tex parka. A large backpack covers nearly half of her five-foot frame. She has made this climb hundreds of times over the past 24 years. "Good data, hard won" is how she describes her work.

We tuck our heads deep into our hoods as protection against frostbite and set off across the lake. With our heads bowed and cowled, our small column must look like a strange sect of mountain penitents. A half hour after crossing the lake we reach CO98, the prosaically named weather station. Baron immediately puts me to work shoveling snow away from some solar panels that power a variety of instruments, and I'm grateful for the warmth of activity.

CO98 is one of more than 250 such outposts in a continent-spanning network called the National Atmospheric Deposition Program, a project jointly operated by the USGS, the Department of Agriculture, and assorted state and local groups. Researchers like Baron visit all the stations weekly, recording the amount of rainfall and snowfall, checking streamflow gauges, and testing water, rain, and snow for acidity or pollutants like mercury.

While I help Baron remove some paper charts from a drum attached to a rain and snow gauge, Karl Cordova, a biologist with the National Park Service, and Donald Campbell, a USGS hydrologist, retrieve a couple of five-gallon plastic buckets of snow from a simple but ingenious collector. The device consists of a lidded box that holds the buckets and a small motor that opens and closes the lid. When snow or rain falls on two small metal plates that almost touch, the moisture conducts electricity, creating a circuit between the plates. The current triggers the motor to open the lid and expose the buckets. When the precipitation stops, so does the current, and the lid slides back into place.

Cordova loads the two snow-filled buckets into his cavernous backpack, adding maybe 20 pounds to his load, and we start back down the trail, where one more chore awaits. At a stream flowing out of the lake, we stop to collect a water sample. To reach the stream, Cordova digs a hole through several feet of snow. Baron removes her snowshoes, drops into the hole, and dips a flask into the frigid brook.

Arduous research like this over the past decade has revealed two unsettling facts about mountains throughout the West: from the Cascades to the Rockies, less snow is falling each year and the dwindling alpine snowpack is melting earlier.

"Peak runoff when we first started doing our studies was from the middle to the end of June," Baron tells me as we start back. The wind has subsided; we actually manage to enjoy the Loch's icy beauty now and even talk comfortably as we clomp side by side over the snow. "So snowmelt would start somewhere from the end of April to early May. In the past several years we've been seeing it occur three to four weeks earlier."

Baron is cautious about making the connection, but we're probably witnessing the effects of global warming on one of the highest, coldest parts of the country. Climate scientists predict that for every 1.8 degree Fahrenheit increase in temperature, mountain snow cover will retreat upward by 500 feet. The West's total snowpack could be reduced by as much as 40 percent in the next half century.

As the region heats up, and as more precipitation falls as rain and less as snow, it might seem that rainfall would simply make up the deficit. Even leaving aside the likelihood that precipitation of all kinds will decrease, studies of groundwater chemistry show that in many parts of the West, up to 90 percent of groundwater comes from melting snow. During the spring thaw, a few months' worth of water flows down the mountains within weeks, and the concentrated rush penetrates dry terrain more deeply than intermittent rains can.

"It's really interesting to ask what it means for the biology up here," Baron says. "Are we likely to see a change in when plants flower or put on leaves? If you get a big pulse of snowmelt earlier than before, does that mean these ecosystems dry out earlier in the year? And what would that mean for tundra, for meadows, for forest? Do they go into drought stress earlier, and is it more severe, and what will that do?"

The day after my hike with Baron, I drive north along the western slopes of the Rockies through a blizzard that clears as I descend into Kawuneeche Valley, the location of Grand Lake, the largest natural body of water in Colorado and part of the river's headwaters. The stream feeding into the lake here is not yet out of the mountains, but already it is being diverted. Grand Lake is the first of many locations where water that would otherwise run unimpeded to the Gulf of California is sluiced away toward cities and farms.

The diversion project, though enormous, is mostly invisible. On the southeast shore of Grand Lake, concealed beneath its surface, is the entrance to a 13-mile-long tunnel bored in the early 1940s clear through the Rocky Mountains. The nine-foot-wide Alva B. Adams Tunnel can bring up to 550 cubic feet of water every second across the Continental Divide to the cities and farms on Colorado's dry eastern plains. Perversely, while 80 percent of Colorado's precipitation falls on the west slope of the Rockies, 80 percent of the state's population lives east of the mountains. The tunnel spreads the river's bounty all the way to the northeasternmost corner of the state.

Just north of Grand Lake, with a few feet of snow on the ground and winter still stubbornly holding out in the mountains, one could almost forget the drought and doubt the existence of global warming, were it not for the dead trees. Tens of thousands of lodgepole pines cover these slopes. Many of them have been standing for well over a century; now a large number are dying, victims of a decade-long mountain pine beetle infestation. The beetles killed close to five million lodgepole pines in Colorado in 2006, nearly a fivefold increase from the previous year. About 43 percent of the state's lodgepoles have been infected. Trees at this elevation had been largely immune from the assault, but warmer temperatures have increased the beetles' range. Rust-colored and dead, the forests are a somber warning sign of climate change.

After parking my car along the roadside, I trudge through deep snow a few hundred yards into the forest. Beyond a stand of trees the Colorado glints into view, a shadowy umber band chiseled on a field of white. In the distance, clouds of blowing snow flare like shaggy manes from the peaks of the Never Summer range. The river in this snow-covered meadow is really still a mountain brook, barely six feet across, the water only ankle deep. Delicate sheets of ice jut from the banks, but the swift current prevents the river from ever completely freezing over.

The Colorado got its name from the color it acquires from sediments downstream, but here, some 70 miles northwest of Denver, its waters are clear. Although some other American rivers are longer, none is more crucial to its watershed than the Colorado. The lowlands' contributions to the river are relatively modest: about half of the water that reaches Hoover Dam comes from just 15 percent of the watershed's land area here in the mountains. But whatever the water's provenance, not a drop of it will ever reach the sea. Seven states, two nations, 14 dams, farms, cities, and evaporation will claim it all.

Page, Arizona
[Elevation 4,118 feet]

You know you're living in a nation that's congenitally warped by easy abundance when, on a 90-degree april afternoon, you see a large boat being hauled along a highway through the desert. On the road to Page, Arizona, assorted vessels on trailers are not an uncommon sight, and the welcome sign at the town limits sports a sailboat painted on a blue background.

Fifty-one years ago Page didn't exist. Nor did Lake Powell, the 186-mile-long reservoir that stores about two years' worth of the Colorado River's flow behind the 710-foot-tall concrete wall of Glen Canyon Dam. Page, which sits atop a mesa 600 feet above the lake, was originally a camp for the workers who began building the dam in 1957. Now it's a resort town with three golf courses and nearby marinas.

The dam was controversial from the start. The reservoir it created submerged some of the most spectacular red rock canyons in Utah, with formations rivaling those of the Grand Canyon. The rising waters drowned plants and wildlife, inundated towns and archaeological sites, and turned the canyon's solemn ancient buttes into archipelagoes.

On my second day in Page, I succumb to the lure of the immense flooded canyon and join 69 other tourists for a daylong excursion on Lake Powell. The drought's toll is visible everywhere. A fat white stripe of calcium carbonate deposits left by evaporation stretches 100 feet or more above the waterline, staining the red sandstone canyon walls. Once underway, our boat hits 20 knots and kicks up a good wake, passing dozens of multimillion-dollar houseboats. One has four Jet Skis lashed to its stern deck; another has a hot tub.

Lake Powell, with its many inlets, has a coastline longer than California's, and over the course of the day, Melvin Howard, our captain, pilots us into side canyons so narrow that passengers on either side of the boat can touch the smooth sandstone walls. These canyons once harbored cottonwoods, willows, and springs; the many verdant, secluded glens led John Wesley Powell, the nineteenth-century explorer, to give Glen Canyon its name.

At a public meeting at a Marriott hotel in Page, some 100 people are watching a PowerPoint presentation by Tom Ryan, a hydrologist with the Bureau of Reclamation's Salt Lake City office. The bureau manages Glen Canyon Dam, which creates Lake Powell, and most of the other big engineering projects on the river, including Hoover Dam. Ryan is talking about how the drought is depleting Lake Powell and Lake Mead, 250 miles downriver, and what that means for the millions of people who depend on the approximately 16 trillion gallons stored in the two lakes when they're full. Lake Powell hasn't been full for more than eight years, and no one knows if it will ever be full again.

Ryan's slide show is filled with grim statistics. "This is July 18, 1999," he says, pointing to a spot on a graph of water levels in the reservoir. "The lake was filled to 97 percent of its capacity. In the next five years the reservoir dropped almost 150 feet. Then 2002 -- that was a killer," he continues. "Lake Powell dropped 47 feet, and there was no runoff into the lake at all. It was the second-driest year in 100 years of record keeping on the Colorado River. We reached a low point on April 8, 2005: 33 percent capacity. After 2005 we rebounded a little bit.

"The good news now is that Lake Powell is half full," Ryan says. "But based on this year's snowpack, we won't get above 51 percent. Snowpack basin -- wide above Lake Powell is 50 percent of what it usually is this time of year. The bad news is that we don't know if we're in the eighth year of an eight-year drought or the eighth year of a 15-year drought." Or, I can't help thinking, the eighth year of an arid new era.

Conditions aren't any better at Lake Mead, the only reservoir larger than Lake Powell in the country. Shortly before the meeting begins, I buttonhole Terry Fulp, a colleague of Ryan's at the bureau. "At Lake Mead, we're about 100 feet down from a full pool," Fulp tells me. "We are about 72 feet above the minimum power pool. If we drop below that, we can't generate hydroelectric power."

While the loss of power generated by the Glen Canyon and Hoover dams wouldn't be catastrophic -- together the dams provide less than 10 percent of all power used in the Southwest-the reservoirs' water is crucial. "If Mead were to fall another 122 feet," Fulp tells the audience in Page, "Las Vegas could no longer draw water from the lake. That's about 1.9 million people whose taps wouldn't have water."

To prevent any such disasters, the bureau is modifying some of the most hidebound rules governing the river's use. These are codified in decades of labyrinthine decisions known collectively as the Law of the River, the keystone being the 1922 Colorado River Compact. Congress ratified the compact six years later after long, contentious negotiations among the southwestern states. It divides the states into two basins: an upper basin consisting of Colorado, New Mexico, Utah, and Wyoming; and a lower basin made up of Arizona, California, and Nevada. Each basin now receives half the Colorado's average annual flow, which is further subdivided among the individual states.

In 1922 Las Vegas was an obscure railroad watering stop in the middle of nowhere. The nation was largely rural; no one doubted that the vast share of the water allotted in the compact should go to farms. The compact was written for a world that no longer exists, for a river that seemed inexhaustible. Nevertheless, no one wants to scrap it and become mired in endless lawsuits while hammering out a new agreement.

Until the end of 2007, no incentive existed for any of the states to limit their annual withdrawals from the river. In fact, the system encouraged states to use their full allocation or risk losing the water to another state. The reservoirs could not carry over unused water from year to year, and if water remained unclaimed, other states could request to use it. Throughout the 1980s and 1990s, California lived on this unused water.

In December 2007, the seven states agreed to new conservation guidelines proposed by the Bureau of Reclamation. The new rules let states save their unused water. Water banking, as it's called, is intended to encourage conservation; states can put away water for future dry years or trade stored water among themselves. The rules also impose what amounts to a "water tax," which allows the bureau to retain 8 percent of the states' banked water in Lake Mead to offset withdrawals and evaporation. The new guidelines also specify how the states will share the burden of shortages if -- or when -- the reservoirs can't deliver in full.

The reforms are crucial, Fulp says. "We can store four times the river's average annual inflow in Powell and Mead, which is why we have been able to get through this drought without significant water shortages. When we walked into this drought in 1999, we were fortunate to have a relatively full system. That may not always happen."

In fact, the odds of that happening again are not good. Studies of the Southwest's past climate show that the twentieth century was anomalously wet. Ironically, the years leading up to the signing of the Colorado River Compact were among the wettest the region has ever known.

Tucson, Arizona
[Elevation 2,389 feet]

There's a treasure beneath the bleachers of the University of Arizona's football stadium. In a dusty warren of rooms that share corridors with concession stands and storage lockers is the university's Laboratory of Tree-Ring Research, home to the world's largest collection of dated wood samples.

"This is not simply a University of Arizona collection. This is a world heritage collection. It's unique," says Rex Adams, the public face and de facto curator of the collection. "It contains a record of environmental history for the last 8,800 years." Adams, a compact man with a trim gray beard and thinning gray hair, is leading me down aisles that wend through boxes of wood samples piled to the ceiling. Outside, the city shimmers in the heat of a spring morning; yesterday the temperature in Tucson hit a new all-time high for April.

"Just across the way is the international collection. You can see material here from South Africa, Australia, New Zealand, Argentina," Adams says, his voice fading momentarily as he disappears down a cool, dark cardboard-box canyon. "We've had researchers come from some of these places and ask, ‘Do you have a specimen from the mid-1970s, because we don't have that forest anymore.'"

Andrew Ellicott Douglass, an astronomer at Lowell Observatory in Flagstaff, started the collection in 1904, hoping to find evidence of sunspot cycles in tree rings. He never succeeded in that quest, but he founded the modern science of dendrochronology by precisely matching tree-ring samples to specific calendar years. Today the lab houses more than a million wood samples from more than 10,000 sites around the world.

"The single most important aspect of dendrochronology is the ability to provide absolute dates," Adams tells me. "If we say this year is 1748, it is 1748 -- not 1747 or 1749."

Researchers here have used tree-ring data to reconstruct the Colorado River's flow for much of the past several hundred years. And that written-in-wood record tells a story very different from that of the data collected with river-flow gauges, which cover only the past century. When the Colorado River Compact was signed, 20 years of measurements suggested that the river's average annual flow was 16.4 million acre-feet. (One acre-foot equals 325,851 gallons, the amount of water that would cover one flat acre to a depth of one foot.) A family of four typically uses about half an acre-foot of water in a year. Lakes Mead and Powell now hold about 25 million acre-feet between them.

The tree-ring record for the Southwest shows that the years from 1900 to 1922 were probably the wettest of the past 500 years; in the recent drought the river's annual flow has plummeted as low as 3.8 million acre-feet. In short, states have been allocated more water than the river actually contains.

David Meko, a hydrologist at the tree-ring lab, has looked as far back as anyone into the river's past. Water managers don't like to think about his research, which shows that droughts far more severe than today's have afflicted the Southwest, although those past droughts differed from the current one in one crucial respect: they ended.

"The Colorado basin had worse droughts during the medieval years," Meko says. On his computer he shows me some precipitation graphs derived from tree-ring studies. "This drought," he says, pointing to a span that began in 1121, "lasted around 60 years. It's really unusual for not having any very wet years. That would be important for the Colorado River today, because we rely on those wet years to fill the reservoirs. So now we're sitting around waiting for one of those to come along again, and it's not happening. Those very wet years may not be coming anymore; the climate of the Southwest may be moving into an almost perpetual period of drought."

A few days before my meeting with Meko, the journal Science published a startling paper by a team of climatologists from several universities. The researchers took 19 of the world's leading computer climate models and used each to project the Southwest's climate for the remainder of the century. I spoke with one of the paper's authors, Richard Seager, of the Lamont-Doherty Earth Observatory at Columbia University in New York City.

"We plotted the results for all the models one at a time and found that they were all doing the same thing," Seager told me during a telephone interview. "I was shocked when I saw the results." Almost every model predicted that a state of severe drought will become the norm for the entire Southwest. All of them show that the storm tracks that bring moisture to the region -- that is, the winds that drop the snow that feeds the Colorado River -- will be shifted northward by global warming. When I asked Seager about this permanent megadrought, he corrected me.

"You could call it permanent drought, but drought is a term that implies an anomalous dry period. It's really a shift to a drier mean climate. No one says the Sahara is having a drought."

Unfortunately, none of the climate models has been refined to the point where water managers in the Southwest could use it to plan precisely for shortages. For now, state and federal agencies are in the disconcerting position of relying on measurements of the Colorado River's flow for the past 100 years to gauge the future, even as it becomes increasingly clear that those past flows will never come again.

So how dire is the outlook for the Southwest? Could Lake Powell drop another 150 feet, as it did between 1999 and 2004, leaving Las Vegas high and dry? Tom Ryan, the Bureau of Reclamation hydrologist, told me in a telephone interview last December that the situation is serious but manageable. The drought has forced change on the region, he said. During its first few years, the states of the Southwest continued to operate as they had for most of the previous decade, when the system had a surfeit of water. The early failure to recognize the magnitude of the drought exacerbated the problem.

"There was still this mentality that the reservoirs were pretty full and we could just use whatever we needed to meet our demand," Ryan told me. "That mentality has obviously changed. And we've gone to normal deliveries, away from surpluses." The days of releasing 8 million acre-feet of water to the lower basin states are over, at least for now; the figure has been scaled back to 7.5 million acre-feet annually.

After eight years of the worst drought of the past 100 years, both Lake Mead and Lake Powell are still nearly half full. While the two reservoirs may never again be completely full, Ryan said it's unlikely they'll get so low that a cutoff of water to any of the region's cities would occur. Some straightforward conservation measures would probably prevent that calamity. Even now, in the midst of a drought, the cities of the Southwest are still wasting water. According to a study by the Pacific Institute, a California-based think tank, Las Vegas alone could save 86,000 acre-feet of water annually -- nearly 30 percent of Nevada's total allotment from the Colorado River-with the installation of water-efficient fixtures and appliances in homes and hotels. And of all the cities in the Southwest, none is more vulnerable to the vicissitudes of drought.

Las Vegas, Nevada
[Elevation 2,001 feet]

Sin City. Pathetically easy to vilify. Where else, outside of an oil emirate, would there exist anything like the illuminated fountains at the Bellagio Hotel, jetting melted mountain snow 240 feet into the hot Mojave air, the nightly eruptions pulsing in time to music? Or for more "Desert? What desert?" attitude, take a stretch Hummer down the Strip to the Mandalay Bay Resort and Casino, with its 11-acre beach, rolling artificial surf, and cabanas.

But in truth, when it comes to water, the city-and the entire state of Nevada-have been shortchanged. The Colorado River Compact gives Nevada just 300,000 acre-feet each year, the smallest allotment of any state. California receives more than 10 times Nevada's share, and Lake Mead loses twice as much each year to evaporation alone.

Las Vegas has no more ardent advocate than Pat Mulroy, the driven, get-to-the-point general manager of the Southern Nevada Water Authority. Mulroy is one of the most powerful people in the Southwest. No one here would be surprised to see her occupy the governor's mansion in Carson City someday.

"The Las Vegas Strip uses 3 percent of southern Nevada's water, but it is the biggest economic engine in the state," Mulroy tells me. We're in her spacious office not far from the Strip. She's sitting behind a polished desk so big it would be classed as a light truck if it had wheels. She's well tanned, has elegantly coiffed, thick silver hair, and wears a chic pantsuit with a high-necked white sweater. Every few minutes she cants forward in her seat, speaking forcefully for emphasis.

"We recycle everything that hits the sewer system." she says. "It either gets reused on a golf course or in the Strip's fountains, or we treat it and return it to the Colorado River. And for every gallon we put back in, we can take a gallon over our allocation out."

Mulroy describes the city's considerable conservation efforts. Even with those measures, outdoor watering accounts for 70 percent of the city's water use-a typical figure for most of the urban Southwest. So the Southern Nevada Water Authority offers a two-dollar bounty for every square foot of turf that homeowners rip up, which comes to nearly $3,000 for a typical yard. All homes built since 2003 must have water-conserving landscaping in the front yard, and lawns can cover no more than half the backyard. There's a moratorium on golf course construction. There are plans to construct an intake pipe 24 feet in diameter at the bottom of Lake Mead, a $400 million insurance policy in case the lake's level falls below the city's existing pipes.

When I ask Mulroy whether cities in the Southwest will ultimately have to enact some sort of growth limitations in the face of a drought exacerbated by global warming, she leans toward me across the vast plain of her desk; I reflexively lean away.

"Water has never constrained growth in the Southwest. What is critically important is not the rate of growth but how we grow. I came here in 1974, and for me to sit here today..." She gestures to the high-rise skyline outside her office window. "We're getting more compact, more of an urban-core community. To the extent that we live on less land and we get away from the quarter-acre with grass all around it, we've met a lot of the water resource objectives that we need to.

"Everyone who asks me that question, I ask them a question: Where do you want these people to live? People are going to go where the jobs are, and every community is going to create as many jobs for their children as they can. If they're not in Las Vegas, they'll be in Colorado. So tell me where you want them."

Perhaps in a part of the country that isn't a desert, I want to say, but prudence or gutlessness keeps me silent.

Las Vegas gets 90 percent of its water from the Colorado River, and by 2015 or so it will face shortages without another supply. That is why Mulroy is pushing a plan to build a $2 billion pipeline to pump more than 180,000 acre-feet of water to the city each year from the rural valleys of northern Nevada.

In April 2007 Nevada's state engineer partially approved the first stage of that plan, giving Las Vegas access to groundwater in Spring Valley, which lies just west of Snake Valley in the Great Basin. Mulroy was hoping to get 91,000 acre-feet a year, but the state engineer, citing concerns about the health of Spring Valley's aquifers, limited the initial amount to 40,000 acre-feet. However, the decision will be followed by others as the Southern Nevada Water Authority goes after water in the Great Basin's other valleys. "At full development I think we'll have 50 percent of our water coming from the Colorado and 40 percent from groundwater supplies," Mulroy says.

In any case, she says, the city doesn't have any other options.

She leans back in her chair. "At the end of the day this organization has to face almost two million people and explain why no water is coming out of their tap." She leans forward. "We have to do this." She's speaking very loudly now. "We don't have a choice."

Callao, Utah
[Elevation 4,348 feet]

"I wish she would stop and think about it," Cecil Garland tells me as we drive around the remote community of Callao, Utah, in his Dodge Ram V8 -- "she" being Pat Mulroy. "They're talking about an 84-inch pipeline up here to take water out. You could just about drive this pickup through that. It can't happen."

Garland is a fit and active 82-year-old rancher and former Vegas craps dealer. Callao, home to five ranching families, lies in the heart of the Great Basin at the north end of Snake Valley, about 70 miles from Nevada's Highway 50, once singled out by Life magazine as "the loneliest road in America." I met Garland at his adobe-walled home, which is shaded by two huge, 100-year-old cottonwoods. He greeted me barefoot -- "My wife made me take off my muddy shoes" -- wearing a denim work shirt and overalls. He's sturdily built, about six feet tall, bald, with a gray mustache.

"They look on the map and they don't see any roads or any big towns," he told me as we sat on his porch. "They just automatically assume that there's a vacuum here, a huge vacancy. And it isn't so. What you find here is people living tranquilly. Our biggest concern is, did we have a good calf crop, can we get our hay up without it getting wet-bright, clean hay. The reason you don't see any cities and towns here is there's damn little water. That should be elementary. If there was a vast amount of water here, we'd be farming and using it, because we've got all the land in the world."

To make his point, Garland spends a couple of hours driving me around the irrigated fields of Callao. From the air, this is one of those mysterious places where you see perfect circles of green surrounded by bare, brown desert. We've stopped by one of those circles now, and are looking at the big pivot irrigation system that created it. The system consists of a quarter-mile-long pipe fitted with sprinklers and wheels. Driven by electric motors, the pipe pivots around a hub, typically making a complete circuit every seven days.

Garland says that the 40,000 acre-feet of Spring Valley water awarded to Las Vegas would be equivalent to 150 pivot systems, 75 miles long if laid end to end, running day and night for six months. "Those of us who have practical experience, we know you can't do that."

He drives me to a field where he's preparing to plant alfalfa. First he had to clear the field of greasewood, a spiny-branched shrub with bright green leaves that stand out against the desert's gray-green sage.

"It takes five to six years to make the transition from greasewood to alfalfa," he says. "There's a brief period in there when you can't irrigate during the winter, and you've already ripped up the greasewood, so you've exposed the soil." He kneels and scoops a handful of powdery sand-a single winter's wind has removed all the topsoil. Garland fears that the entire valley could look like this if Las Vegas has its way.

The Southern Nevada Water Authority insists that if its pumping shows any sign of damaging any of the valleys' aquifers, the operation will stop. But no one in Snake Valley believes that. Once the water starts to flow, people here are convinced it will be impossible to turn off, even for someone with Pat Mulroy's political clout. "Giving Las Vegas water is like giving a drunk a drink," Garland tells me.

As we drive back to his home, I mention Mulroy's assertion that agriculture uses 90 percent of Nevada's water and that 75 percent of the state's population-concentrated in urban areas-uses just 10 percent while generating more than 60 percent of the state's economic activity. Farms contribute just over $300 million to Nevada's $100 billion economy. After my recitation of facts, Garland continues to look straight ahead as he drives, speaking with resignation and sadness.

"You know, to say that, to even think that, speaks of a callousness that I despair of. Even if they took all the humans out of this valley and said, we'll take the water-that would be the least of it," he says. "The true destruction would be to destroy the marshes and seeps and springs. And for what? For more sprawl? I cherish my ranch because it offers me a way of life. It's hard work a lot of the time. Many times I've been on my side trying to straighten out a calf inside a cow that couldn't be born the way it was, and I wonder why I quit dealing craps in Vegas and came up here. But I don't wonder very damn long."

Leaving Callao, I recall my conversation with rancher Dean Baker in Snake Valley. "We drilled and beat our heads against a wall with four or five wells," Baker told me. "What we learned, and you can imagine how much it cost us to learn, is that it's a closed dish of water. If you pump one of those and start sucking water, a half mile away the level of water keeps going down." The Southern Nevada Water Authority, he said, will probably learn the same hard lessons.

I asked Mulroy about some of the concerns voiced by Baker and Garland.

"First, there is no question that the initial 40,000 acre-feet permitted to the Southern Nevada Water Authority in Spring Valley is available annually," she wrote to me in an e-mail. "The Nevada state engineer, who is historically very conservative in granting water rights, verified that more than that amount is available for use on an annual basis."

She added that pumping from one well would not necessarily affect springs at another location. "It is also important to understand that there are subterranean barriers such as geologic faults that affect where and to what extent effects propagate. If we sank a well adjacent to a spring, as Mr. Baker and others have done, it would very likely affect its flow. However, our hydrologists will construct exploratory wells so that we can evaluate how the aquifer will respond to a given pumping rate and frequency. We will use that information to site the wells and manage withdrawals for minimal effect."

Jim Deacon, a biologist recently retired from the University of Nevada at Las Vegas who has studied Great Basin ecosystems for more than 40 years, says the best available evidence contradicts Mulroy's claims. A paper that he co-authored last fall in the journal BioScience found that pumping on the scale planned by the Southern Nevada Water Authority could lower groundwater levels in 50,000 square miles of the Great Basin. Depending on local geology, water tables could plunge by anywhere from 50 to 1,600 feet over the next several decades. Moreover, a study issued last fall by the U.S. Geological Survey showed that the aquifers in Spring Valley and Snake Valley are connected.

"The system is more thoroughly integrated than we thought," Deacon told me during a telephone interview. "So taking water from one valley is going to take water from another in that area."

Ironically, Nevada's water laws are relatively good compared with laws in other states and were written to promote the sustainable use of groundwater. In a sense Mulroy is right in saying that 40,000 acre-feet could be taken from Spring Valley. "The catch is, this definition of sustainability assumes that the only useful purpose of water is to help supply people," Deacon said. "There is no provision for maintaining viable ecosystems."

He acknowledged that Mulroy is under tremendous pressure. "Pat understands very well that her job is to supply water to Las Vegas, and that if she starts saying there isn't enough for sustainable use, she's going to get fired." That happened, he said, to two previous directors of the Las Vegas Valley Water District, as it was once called, who tried to make the case in the 1960s and 1970s.

Dean Baker agrees with Deacon. "This will be [Senator] Harry Reid's and Pat Mulroy's legacy of disaster 30 years from now, with billions of dollars to pay off and environmental problems and lawsuits that are insurmountable, with not enough water to put in the pipeline."

Imperial, California
[Elevation 59 feet below sea level]

Several hundred miles to the south of the great basin, in a desert several thousand feet lower and a couple dozen degrees hotter, the Colorado River flows into the Mexican state of Baja California. Between the booming border town of Mexicali and the Gulf of California, the riverbed will run almost dry.

Before the river crosses the border, the Imperial Dam near Yuma, Arizona, diverts about 75 percent of its water-more than three million acre-feet annually-into the 82-mile-long All-American Canal, the world's largest irrigation canal.

The canal creates an agricultural "green zone" called the Imperial Valley -- or, as local boosters like to call it, America's Winter Salad Bowl. (In wintertime, when production in California's huge Central Valley tapers off, about 90 percent of the nation's produce comes from here.) Most of the valley's 400 or so farming families have been here for three or four generations, working 1,500-acre farms in the middle of the Sonoran Desert and generating about $1 billion in revenue every year.

"We haven't had rain down here for about a year and a half. It has been dry," George Moses tells me as we cruise past fields of cotton and spring wheat in his pickup. Bordering the fields is a long, narrow canal that runs to the hazy horizon, part of a network of roughly 3,000 miles of canals and drainage ditches that water nearly a half-million acres. "What I'm showing you is the circulation system of the valley," he says. Moses, who has been working since 5:30 in the morning, is a zanjero, a job title derived from the Spanish word for ditch, zanja, and idiomatically translated as "ditch rider." He has thick gray hair pulled back in a tight ponytail and a craggy, seamed face, one cheek rounded by a chaw of tobacco.

Moses and the several dozen other zanjeros employed by the Imperial Valley Irrigation District (IID) work in shifts around the clock, opening and closing canal gates, taking orders on cell phones from the district's control center. Many of the canals are a century old and were dug with nothing more than horses, mules, and shovels. Even today there are few pumps; most of the system operates on gravity alone, making it the biggest gravity-fed irrigation project on the planet. The valley is also the single largest user of the Colorado River's water, receiving 3.1 million acre-feet each year under the Colorado River Compact-about 70 percent of California's allotment-thanks to water rights staked out by farmers in the nineteenth century.

It's hard to put a price on all that water. Just ask Maureen Stapleton, the general manager of the San Diego County Water Authority.

I meet Stapleton at the Water Authority's headquarters, on the outskirts of San Diego. A slender woman with straight blond hair, she is wearing a dark blue pinstriped suit. If she and Pat Mulroy ever teamed up as negotiators, Stapleton would have the good-cop role. She makes her points with quiet, unflappable reasonableness.

After eight years of disputes with the irrigation district, Stapleton managed to close a deal that she calls "the largest and longest water transfer in the history of the country." Over the next 75 years the deal will transfer a total of 30 million acre-feet-enough to fill Lake Mead-from the Imperial Valley to San Diego. By 2020 the valley will supply 22 percent of the city's drinking water. In return, San Diego will pay the valley roughly $50 million a year.

In addition to purchasing water from the Imperial Valley, Stapleton tells me that San Diego is pursuing a strategy of diversification, to avoid reliance on any one source of water.

"If we've learned anything, it's that you don't want any single supply to dominate your portfolio," she says. "Been there, done that. Didn't work. Conservation is not the silver bullet; desalination is not the silver bullet; water transfers are not the silver bullet. It has to be a combination of appropriate resource planning. Couple global warming into that, and that's where you have to look at the adaptability of your plans in order to be responsive."

Yet San Diego would be hard-pressed without the Imperial Valley agreement. The transfer represents a historic shift in the West, where agriculture has always dominated water use. As cities continue to grow, deals like this one will become increasingly common. Agriculture holds the West's last and largest remaining stores of water. New dams won't solve the region's looming water shortages, not while existing reservoirs can't be filled. If ambitious engineering projects like Hoover Dam, Lake Mead, and Lake Powell made possible the growth of the Southwest in the twentieth century, water that is bought, borrowed, traded, or otherwise appropriated from agriculture will allow the region's cities to survive the twenty-first.

Although the agreement has been in effect for four years, it remains controversial. Just a few days before I met with Stapleton, the San Diego County Water Authority, to avoid more legal battles, agreed to pay IID an additional $40 million over the next 10 years.

"Water in the Imperial Valley is their lifeblood," Stapleton says. "There was a serious concern about whether this would diminish the agricultural economy. But I think it has proven to be an opportunity, both for the agricultural community, which can benefit from an infusion of dollars, and for this urban agency, which needs the certainty of water through transfers. We see it truly as a win-win opportunity."

That's not a point of view widely shared in the Imperial Valley. At the irrigation district's headquarters in Imperial, a hot, dusty town about 120 miles east of San Diego and 20 miles north of the Mexican border, I meet with John Pierre Menvielle, a third-generation farmer in the valley. He's wearing a crisp white button-down shirt and blue jeans; his eyes are narrowed to a permanent squint, and his voice is raspy enough to file steel. When I mention Stapleton's "win-win" comment, he scoffs.

"For Maureen to tell you it's a win-win is crazy," he says. "No way. A win for them, because they get the water, which helps their economy. It kind of got shoved down the IID's throat. Nobody likes it. It's a natural resource that belongs in this valley. People don't like the outsiders coming in here; it leaves a bad taste in everybody's mouth. And the transfer is for too long-75 years. There was no up-front money to help things get going."

The problem with the transfer, Menvielle says, is that all the water the valley sends to San Diego is supposed to be surplus water created by improved conservation practices by the valley's farmers. But it costs money to install more efficient drip-irrigation systems, or to buy pumps that will return water to the head of a field to be reused, or to replace the zanjeros with computer-controlled canal gates. And the price paid by San Diego for the valley's water doesn't cover those costs.

Menvielle estimates that the total cost to modernize the valley's irrigation system could approach $300 million. In addition to the computer-controlled canal gates, IID plans to install pumps below the main canals to capture water seeping into the soil. Until those conservation measures are in place, the only way the valley can meet its obligations to San Diego is by fallowing fields. Last year farmers took 20,000 acres out of production, with consequences that have rippled through the valley's economy. Farmers here refer to fallowing as the F word.

"It hurts farm service providers. The harvesting people. Your fertilizer people, twine people, combine people-anybody associated with a fallowed field," Menvielle says. "We're kind of stuck with this transfer. Hopefully in the long run it will turn into a win-win, as Maureen Stapleton says it is. But it could be 10 years before farmers actually start seeing money in their pockets. And that's a long time to wait."

Menvielle also worries that the water transfer agreement with San Diego won't be the last. Of the 170,000 people who live in the valley, only about 400 families run farms.

"The farm community was king," Menvielle says. "But that's changing. When you have a five-person, publicly elected board, eventually you'll end up with non-pro-ag people on there, and they're just going to jam it to the farmers. We worry that future boards may transfer more water, and then there won't be enough water to farm your land with, and your land becomes worthless."

Vince Brooke, a powerfully built farmer and former president of the Imperial Valley Farm Bureau Board, says the water transfer is not just a local issue. It has national implications, and raises a fundamental question: Is the country willing to sacrifice agriculture-and perhaps jeopardize food safety as we rely more on imports-for the sake of urban growth?

"I think at some point the state and federal governments are going to have to address that," Brooke says. "Water is a critical issue, but so is a consistent and safe food supply. And I believe that they will collide."

The Southwest managed to pull off a small miracle throughout the twentieth century, sustaining both unbridled urban expansion and productive agriculture. It's a run that couldn't possibly have lasted, not in these deserts. Global warming and the drought are accelerating a process that has been under way for decades, forcing issues that the region would have inevitably faced even in the absence of climate change.

The crisis facing the Southwest isn't so much due to any lack of water-even in the driest years the Colorado River can satisfy the needs of millions. The real crisis is a demographic one. Is urban development a goal to be pursued at any cost? Or as Cecil Garland, the rancher in tiny Callao, Utah, put it, do we want lawns or lettuce? Craps or crops?