The best known are probably the mammoths and mastodons, which ranged in size from Columbian mammoths two storeys tall at the withers to pygmy varieties that stood only head high to a human being. On the grasslands, they may have roved in densities similar to those of elephants in African parks today, more than three animals per square kilometre. But these trunked and tusked beasts were nothing more than the Pleistocene’s comforting opening act. North America was home to pampatheres, which resembled armadillos the size of overturned dories, and another armoured family, the glyptodonts, at their largest the size of a subcompact car. There were ground sloths — amiable-looking herbivores that, standing on their hind legs to browse the woodlands, would have been an awesome presence, with the largest weighing in at nearly three tonnes. There were herds of wild horses, some as heavy as today’s draft horses; and tapirs rooting through the wetland; and an antelope called the saiga, with a pouchy snout that acted as an air filter (it still exists today, as a critically endangered species in Central and Northeast Asia). Wild oxen drank at watering holes alongside camels that would tower over today’s dromedaries. One such humped species, the camelops, could be found in herds from the US Southwest to the Subarctic, where its bones sometimes washed out in Klondike gold digs. There were giant moose, giant llamas, giant elk, giant boars. There was a vampire bat twice the size of any known today. There was a beaver the size of a black bear.

Consider the carnivores. Packs of dire wolves were widespread, the animals heavier by ten kilograms than modern wolves but still far from the most fearsome predators in the Pleistocene wilderness. That title might go to the giant shortfaced bear, a dedicated flesh eater large enough to look into your eyes while still on all fours. The greatest feline, still haunting the pop-up book nightmares of children, was the sabretoothed cat, the ultimate ambush predator, with serrated canine teeth as long as chef’s knives, and a body smaller than that of today’s lions but nearly twice as heavy. And there were prides of American lions, akin to those in Africa and India in every way but one: they were larger. Many of these species still survived as recently as 10,000 years ago, in climes similar to ours. Put another way, lions and sabre tooths lived within the myth time of North America’s First Nations. They lived at a time only as distant from the founding of farming in Europe as that founding is from us today.

The list goes on, from scavenging birds with wingspans of nearly five metres to beetles adapted to rolling the dung of giants, but lest we forget: there were tortoises. Six big species roamed widely below the glaciers, enduring the coming and going of ice ages. Among them was the bolson tortoise, which might once have grown to 160 kilograms, ten times the size of the largest individuals today.^{3Anecdotal evidence suggests that very large bolson tortoises — perhaps up to a metre long — survived into the twentieth century. For example, in 1946 a Mexican mule herder told a story of waking up in the Chihuahuan Desert to find his saddle missing. He followed a set of tracks from the place where it had been, and eventually found it on the back of a tortoise. The animal had apparently taken shelter beneath it, then walked away with it stuck on its shell.} Picture the tortoise basking. Picture it in its burrow, the heavy tromping of megafauna overhead. Picture it at its peak as a species, roaming the breadth of today’s Chihuahuan Desert from what is now New Mexico, Arizona, and western Texas to almost the entire Mexican Plateau, an area larger than the Yukon Territory.

The men and women who came together for the Ladder Ranch workshop concerned themselves with a simple question: do modern human beings have ecological, aesthetic, and even moral reasons to attempt to recreate the Pleistocene? The answer they arrived at was “maybe”. The group began to moot what they would later describe as “a series of carefully managed ecosystem manipulations using closely related species as proxies for extinct large vertebrates.” This “Pleistocene rewilding,” the group agreed, had the potential to revitalize conservation biology (and, by association, the environmental movement) by steering it away from “managing extinction” in last-ditch protected areas. The discussion was, despite the rather cautious language of science, the makings of a manifesto that would declare an end to the conservation century and the dawn of an age of restoration.

The proposed manipulations — introducing experimental wild populations of horses, camels, and elephants, for example — had as a natural end point “the ultimate in Pleistocene rewilding for North America”: a free-living population of lions, limited only by the kind of perimeter fencing that encloses African parks, somewhere among the southern Great Plains states of Colorado, Kansas, Oklahoma, New Mexico, and Texas. But the program needed an initial stepping stone, a species that could test the rewilding hypothesis on a short timeline, with few ecological risks, and a comparatively low level of controversy. What was needed, as one tortoise researcher would later put it to me, was “the medical marijuana of Pleistocene rewilding.”

It was pure serendipity that two grasslands ecologists, Carl and Jane Bock, now professors emeriti at the University of Colorado, Boulder, had arrived at the Ladder Ranch with tortoises on their minds. During a break in the meeting, the Bocks approached Joe Truett and his colleague Steve Dobrott. Might the Turner Fund be interested in the little-known bolson tortoise? The Bocks had worked for thirty years near Elgin, Arizona, on a research preserve founded by a naturalist, Ariel Appleton, who in 1973 had taken a colony of Mexican bolson tortoises into captivity as a backstop against the species’ possible extinction. Appleton had died earlier in 2004, and the animals needed a new home. It was Harry Greene, a Cornell reptile expert, who overheard the conversation and made a synaptic leap: the bolson tortoise met all the group’s criteria for the rewilding experiment. “The bolson tortoise would be perfect,” he said.

In August 2005, the rewilding proposal appeared as a commentary in the journal Nature. The response was explosive. Some opposing scientists pointed out that the article ran in the same issue as a study showing that lion attacks on humans in Tanzania had risen 300 percent in the past fifteen years. Good Morning America broadcast a brief interview with rewilding proponents, followed by a clip from the film Jumanji of elephants crushing cars. It was the first time many people had heard that North America was once home to giants and monsters, and reactions tended toward the visceral. Several threatening letters sent to Cornell were judiciously turned over to the authorities. Surprise offers of rewilding habitat rolled in from ranchers in Texas, Arizona, Kansas.

Then, on August 23, Tropical Depression Twelve, south of the Bahamas, began to gyre toward the US shore, and within a matter of days the Pleistocene was forgotten once again amid the horrors of Hurricane Katrina’s landfall in New Orleans. The excitement over rewilding had lasted less than a month, but it had been, in Josh Donlan’s words, “the largest ecological history lesson in American history.”

Tracing the history of the human relationship to our ecology is a new academic endeavour, just a few decades old, and its first principle could be said to be this: to know what is, you must know what was. Consider the Caribbean population of the green sea turtle, a distant relative of the bolson tortoise, which now sits at 300,000. Measured against the green turtle’s near-extinction in the early twentieth century, the figure is a triumph, to the extent that some biologists question whether the International Union for Conservation of Nature should continue to list the green turtle as an endangered species. Look at that datum from a more distant historical baseline, however, and it rapidly loses its cheery sheen.

In 2006, the Scripps Institution of Oceanography led an attempt to determine historical peak numbers of Caribbean sea turtles based on 163 sources, including Charles de Rochefort’s 1666 The History of the Caribby-islands, and “the first American novel,” William Williams’ semi-autobiographical Mr. Penrose: The Journal of Penrose, Seaman. These yielded an estimated original Caribbean green turtle population of 91 million adults. Try to imagine the bounty: sea floor grass beds grazed by turtle hordes; beaches roiled by the flipper churn of nesting females; turtles peering from reefs, lolling in surf; an endless shoal of turtles, a flock of turtles, a drift of turtles, a tedium of turtles —more than 300 times as many as today. Can we really say that the green turtle is no longer endangered?

In 1995, Daniel Pauly, a professor with the Fisheries Centre at the University of British Columbia, coined the term “shifting baselines” to describe the tendency of his peers to measure the health of fish stocks against the length of their own careers. Each generation bemoaned the losses that occurred on their watch, but failed to acknowledge the accumulation of extirpations across centuries, if not millennia — “a gradual accommodation,” Pauly wrote, of “creeping disappearance.” The revelation came to him in part from a study he published as editor of an academic journal concerned with aquatic resource management. The article amounted to a scientific review of Sea of Slaughter, Farley Mowat’s 1984 chronicle of Atlantic overfishing, which he based on five years spent reading anecdotal historical accounts. The authors of the study concluded that biomass — literally the weight of life itself — in the North Atlantic has likely declined by 97 percent since written records began. Mowat, exhausting language in his effort to express this reality, refers to the drawdown as “a massive diminution of the entire body corporate of animate creation.” The report’s authors applied the term “genocidal.”

Scientists do not typically embrace anecdotal evidence, however — the authors of the Sea of Slaughter study refer to Mowat’s data as “data,” in quotation marks. In 2005, Andrea Sáenz- Arroyo, science director of the Mexican non-profit organization Comunidad y Biodiversidad, led an attempt to quantitatively test the shifting baselines syndrome. She and her team interviewed fishers in eleven communities along the central Gulf of California, which separates the Baja Peninsula from mainland Mexico. The world remembered by fishers who have passed their fifty-fifth birthdays proves far richer than the real world of the present, with five times as many commercial fish species and four times as many productive fishing grounds.^{4The older fishers’ memories are supported by historical documents and can’t be dismissed as “fish stories.” A naturalist visiting the area in 1932 described Gulf grouper “in unimaginable numbers,” at sizes that had given rise to the local fishers’ slogan “a ton an hour.”} It is a world in which a fisherman could hope to boast of twentyfive Gulf grouper as a day’s catch, enough to feed 700 people or more, each fish with an outraged, grey-lipped mouth that could take in a man’s head and shoulders. Most older fishers consider the Gulf grouper stock depleted; the new generation, who on a good day might catch one or two of the great predators, each smaller by twenty kilograms than the fish their grandfathers caught, does not. A study of the upper Gulf of California published in March 2008 confirmed Sáenz- Arroyo’s work and made this critical connection: “It is crucial for the restoration of this ecosystem that young fishers and the Mexican public are able to visualize previous states of their local ecosystems.”

The same could be said of almost any people in almost any place — probably every place — on earth. The effect has been called a “double disappearance”: first, we lose a forest, or a species of tortoise, or the abundance of a fish; then we lose the memory that it ever existed. Standing on the Armendaris Ranch, the Chihuahuan Desert looks to me something close to untouched. The land itself seems to whisper that history here is as rare as rain and as thin as leaf litter. I find it hard to believe that the bolson tortoise could have been eradicated from the whole of this single, sprawling basin, let alone nearly every square inch of North America’s largest desert. What could work this dark miracle?

To answer that question, we might follow the thread from the present day backward through time. We would start in the Bolsón de Mapimí in Mexico, the last redoubt of the wild bolson tortoise, and reel in first the recent decades, when anyone who cared to look would find excavated burrows, animals living as ranch house pets, carapaces used as water bowls or abandoned alongside cowboys’ campfire ashes. Reel in all those years when truckers would stop along the highways to pick up campesinos’ oxcart loads of tortoises, bound south for Mexico City or north to the border (turtle soup was popular in California, they say). Take away the new toll highway, and old Highway 49, which made it possible to actually live in the desert, in the godforsaken bolsons, if you happened to have nowhere else on earth to live. Take away the pemex oil survey roads, and the hungry road crews, the cruelty of young men gathered too long away from women and elders and children.

The 1920s, when the bolsons were razed for farming, the mule-driving teams remembered for culling ten or more big tortoises for Saturday-night fiestas, the chief driver saving the largest for his breakfast. ^{5Only a few years earlier, in 1918, an American naturalist stuffed two specimens that sat unnoticed in storage at the US National Museum for forty years before they were recognized by an intern as a new species. They were collected in a part of Mexico from which the tortoise has long since vanished.} The Great Wax Rush of the early 1900s, the candelilleros skiping every limestone slope and coulee for the candelilla plant, or gathering the creosote wood to fuel the factory boilers that cooked off the waxy scurf, the men and their burros prospecting as far as packed water would allow and living off the land, eating jackrabbits, eating tortoises. In 1882, the Mexican Central Railway, opening “the last frontier,” and the crew that built the rail line, 300 badly paid men roasting tortoises at their fires, or shipping them south to Mexico City, already home to a quarter million residents.