View Electronic Edition

Bring Back the Elephants

This land is the mastodon's land; while "Home on the Range" commemorates buffalo, deer, and pronghorn it misses the mammoths, glyptodonts, and camels. There was a wild America considerably wilder than any brought to us on TV. Our late Pleistocene legacy means we can imagine more, not fewer, kinds of large animals on public lands.
- Paul Martin, 1992

A decade ago, biologist Michael Soul predicted that "the reintroduction of these large animals will be controversial, but I would not be surprised to read someday that cheetahs are helping to control deer and that mesquite is being 'overbrowsed' by rhinoceroses." Biologists Michael Soul and Reed Noss have proposed "rewilding" as the foundation of a continental conservation strategy. Central to this proposition is the recovery of existing top predators such as grizzlies, cougars, and wolves.

We would like to consider the ultimate in rewilding. America's charismatic megafauna was severely impoverished in the late Pleistocene, but we can turn to Africa and India for surrogates for restoration in the Americas. We suggest that the project begin by restarting the evolution of the most influential of the missing species, the extinct animals most likely to have exerted the greatest influence on their natural environment. Based on what is known of living megaherbivores in Africa and Asia, and based on the fossil record of the New World, there is one clear choice, animals as potent as fire in their dynamic influence on ecosystems. If we want the "super-keystone species," second only to our own in their capability for altering habitats and faunas, we should start with the restoration of living proboscideans - with African and Asian elephants.

We fully expect that the proposal for free-ranging elephants in the Americas will shock and confound many conservationists and naturalists. What could be more foreign in the New World than free-ranging elephants? Isn't this a heretical idea for those of us inclined toward deep reverence for the wild?

It all turns on what one regards as wild. For example, the gomphotheres, a family of neotropical elephants, prospered in the Americas for well over ten million years, but vanished at the end of the Pleistocene around 13,000 years ago, along with mammoths and mastodons. With such a rich fossil record and such a recent American extinction, it is natural to consider restarting New World evolution of the Proboscidea with whatever taxa of elephants are left.

We are keenly aware that living African ( Loxodonta africana ) and Asian (Elephas maximus ) elephants are not conspecific with fossil mammoths ( Mammuthus) or other native New World Proboscidea. But all are in the same family, and some taxonomists have considered Elephas and Mammuthus to be quite close, even congeneric; thus, an Asian elephant living today in Thailand is more closely related to the extinct mammoths of North America than to its African cousin. African and Asian elephants are the only members of the Order Proboscidea that were not lost in the megafaunal crisis of the late Pleistocene.

Unlike explosively reproducing aliens of the New World such as kudzu, Africanized bees, or zebra mussels, animals reproducing as slowly as elephants, with an intrinsic rate of increase of about five percent per year, should be controllable. To avoid unacceptable methods of regulation (for twenty years park rangers shot 300 to 800 elephants annually in the Kruger National Park, South Africa), Jay Kirkpatrick of ZooMontana and his collaborators have perfected a technique for limiting elephant populations by darting females with a long-lasting birth-control compound. Elephant forays beyond the perimeter of a reserve can be deterred, as in Amboseli National Park in Kenya, by an electrified wire. For a New World elephant park suitable for wide-ranging family units, we suggest a part of the lower Colorado River or the Rio Grande. Like most of North America, both regions were once ranged by mammoths. Both river systems are heavily invaded by alien Tamarix, riparian trees widely regarded as undesirable and a potential target for removal by elephants. The river banks support alien Bermuda grass ( Cynodon dactylon ), an African species eaten by elephants. Other potential sites for elephant introductions would be anthropogenic savannas in Central or South America - once home to gomphotheres - now pasturing livestock.

Raising the Columbian Curtain

In evolutionary times the flood plains, grasslands, and savannas of North America harbored a stunning variety of large animals - some forty-one species in western North America alone, more than three times as many as were left by the time Lewis and Clark detected bison, elk, pronghorn, and brown (grizzly) bears.

In planning New World restorations, conservationists have endowed only the large mammals of historic time with the exclusive status of hallmarks, or flagships, and overlooked the missing large mammals of the late Pleistocene. The animals that the first European explorers and settlers saw and wrote about became incorporated in ideas of what constituted American wildness. The viewpoint imposed by this "Columbian Curtain" is unrealistic in evolutionary time. Among the more common large herbivores of the late Pleistocene, only bison remain. The fauna of historic times lacks the largest and most representative animals of the continent, ghosts whose prehistoric presence is hinted at by sweet-tasting bean pods of mesquite, honey locusts, and monkey ear. Such fruits are the bait evolved to attract native large animals that served as seed dispersers.

Given their evident success for over 15 million years and the late hour of their New World extinction, we suggest that bringing back the Proboscidea is by no means as witless as it might seem at first. It is not the same as introducing goats or pigs onto an oceanic island, whose native plants long ago lost whatever defenses they once had to protect themselves against onslaught by the tongues and teeth of large herbivores.

A Deadly Syncopation

When megafaunal extinction struck North America in the late Pleistocene, at least seven species of proboscideans vanished. Unlike erratic background extinctions that sputter along randomly through the eons, often in step with evolutionary replacements, the late Pleistocene extinctions were catastrophic, and there were no replacements. The losses included native mammals in size classes to match the largest found in Africa and Asia.

Since unrelated groups of organisms, including marine invertebrates, did not vanish (as at the end of the Cretaceous, 65 million years ago), the end of the Pleistocene was not a time of mass extinction. Instead, what happened in America was an extinction of the massive (plus their parasites and commensals).

What caused such a loss, so late in the Pleistocene? Could it have been an asteroid hit, a circumstance many believe accounts for heavy extinction, including the loss of dinosaurs, at the end of the Cretaceous? Evidently not. There is no trace of an asteroid impact large enough to generate global repercussions that late in the fossil record.

Moreover, throughout the islands and continents of the planet, late Pleistocene extinctions were not synchronous, as would be expected in the case of a cosmic or climatic accident. Radiocarbon dates show that they were globally sequential, or what geologists call "time transgressive." While large animal extinctions impoverished Australia perhaps 50,000 years ago, they seem to have struck North and probably South America around 13,000 years ago. And the last population of woolly mammoths - including some dwarfs just two meters tall - vanished from Wrangel Island in the Arctic Ocean off Siberia only 4,000 years ago.

These prehistoric extinctions, significantly, follow the ancient footsteps of Homo sapiens , out of Afro-Asia and onto other continents and eventually even to remote oceanic islands, in what Ross MacPhee of the American Museum of Natural History calls a "deadly syncopation" of human arrival and faunal loss. It's impossible to fit this pattern to any known climatic or cosmic event. It does suggest what we call the "overkill hypothesis": that the extinction of Pleistocene mammals resulted from overhunting by humans.

In the long pull all species are doomed to extinction, just as death is the inevitable fate of all individuals. Most species that ever lived on Earth are no more. But this is a poor excuse for turning our backs on the extraordinary loss of flagship species on our watch. By "loss on our watch" we mean not just the extinctions of this or the last five centuries of European conquest in the New World; we mean the timescale of our species on this continent - the last 13,000 years at least. We have the opportunity to restart the evolution of proboscideans, along with horses, camels, and other extinct groups native in the Americas for millions or tens of millions of years. What can be done?

Mammoth Requiem

For starters, it is time to mourn our dead, especially the total loss of the mammalian Order Proboscidea - the American mastodons, the Columbian mammoths, the imperial mammoths, the woolly mammoths of the boreal and subarctic regions, the dwarf mammoths of Santa Rosa and other islands off the California coast, and the gomphotheres of the tropics.

In North America we need a "Mammoth Extinction Day" and in South and Central America a "Gomphothere Extinction Day." This might take place sometime around the summer solstice. Any of the numerous fossil localities known to yield bones of Proboscidea, such as Rancho La Brea with its magnificent Page Museum in Hancock Park, Los Angeles, would be suitable. An especially appropriate place for a wake would be at the Mammoth Site in Hot Springs, South Dakota, a paleoecological cathedral where 100,000 visitors a year pay modest admission to marvel at a unique in situ exhibit of splendidly preserved mammoth bones in the process of being excavated from the most concentrated natural deposit of mammoths known on the continent.

From the Hot Springs Mammoth Site tourists can drive to Wind Cave National Park to see a free-ranging bison herd. There ecologists study the interrelationship among short grasses, grazing, and fire. Bison are increasingly popular as a meat animal. But bison are a small part of the pre-extinction Wild West. And, according to the fossil record, bison entered North America only a quarter of a million years ago, well after the arrival and evolution of New World Proboscidea. It would be truly interesting to determine the adaptability of elephant family units mixed in with the bison.

We are not totally ignorant of the paleoecology of extinct American Proboscidea. Dung balls nine inches in diameter, discovered in the 1980s by a National Park Service team in southern Utah, proved too big, and the texture of the plant remains in the boluses too coarse, to match those of the Shasta ground sloths. The radiocarbon dates on mammoth dung balls were 14,500 calendar years ago, and the plant material indicated a cooler climate than occurs today. The extinct mammoths ate mainly grasses, sedges, and other riparian plants, salt bush, prickly pear, and even some needles of blue spruce.

African Visions

But much more about elephant ecology can only be learned from live animals. When elephants dig for water in the dry season, the water holes they leave behind attract other species. They thin out dense stands of low trees and shrubs. In the process, elephants improve forage production for other grazers. Undoubtedly the extinct mammoths, mastodons, and gomphotheres did the same. How might pachyderms affect the North American environment today?

The most interesting prospect for restarting the Proboscidea in America comes from what managers have discovered in Kenya's Amboseli Park, just north of Kilimanjaro. According to David Western, director of wildlife for Kenya:

If elephants and cattle had their way, they would trade places. In see herds of cattle filing into the park to graze, passing elephants headed out to browse. With elephants and cattle transforming the habitat in ways inimical to their own survival but beneficial to each other, they create an unstable interplay, advancing and retreating around each other like phantom dancers in a languid ecological minuet playing continuously over decades and centuries. Habitats oscillate in space like a humming top, driving and being driven by climate, animals, and people.

Our proposal to establish free-ranging elephant herds in the New World is to conduct not an agricultural but an ecological experiment. We have an extraordinary opportunity to learn more about how Nature works. How would fruits be dispersed if elephants were in the picture? What is the relationship among elephants, vegetation, and wildfire? Long smitten with the beguiling concept of a "forest primeval," Nature advancing toward a single natural equilibrium, North American conservation biologists have had to shift gears, adopting a more flexible concept: Nature jumping between a series of equally probable stable states. Over twenty years ago conservation ecologist Graeme Caughley found no single attainable natural equilibrium between elephants and forests in eastern and southern Africa. More recently, A.R.E. Sinclair reported that African elephants and fire reach multiple stable states. It appears that introduced elephants might have a great deal to teach us about the dynamic nature of wildness in America in evolutionary time. In the absence of elephants, inferences made on the dynamics of American vegetation types could be as one-sided as those made in the absence of fire.


The demise of Proboscidea in North America represents not only the loss of ecological relationships and evolutionary possibilities, but a foreclosure on entire realms of scientific inquiry. Clearly American ecologists suffer blind spots if the largest and most potent megaherbivores native to the continent are missing. David Western's vision of a timeless minuet between grazing cattle and browsing elephants in Amboseli fuels thoughts of how to attempt an American experiment. In the New World, we can substitute bison for cattle to see if bison, too, will dance the languid ecological minuet with African elephants, to the benefit of the American range! People, bison, and mammoths once coexisted in America. We see this in the Clovis sites excavated by the Arizona State Museum along the San Pedro River in southeastern Arizona. The early Americans speared and processed Proboscidea. We suspect they spent many days watching them very closely, as closely as David Western or the Masai watch African cattle and elephants today in Amboseli National Park.

From mammoths and mastodons the Clovis foragers would have learned much about edible wild plants. In the New World we suspect it was the extinct megafauna that introduced the first Americans to the sweet bean pods (or p'chita, an Indian name becoming part of borderland Spanish) of the mesquites, a valuable food plant for people living off the land. From the large mammals of the New World the newcomers learned the right season to rip apart dagger-leafed agaves for their sugary hearts, a rich source of calories. Surely the early Americans followed the game trails of the last New World elephants through the tropics, in the process learning about palm fruits and other fruits as attractive to people as to Proboscidea.

Today, African Loxodonta, or Asian Elephas, or both, could show us some of the coevolutionary secrets of America when it was truly wild. Beyond Pleistocene parks we need Pleistocene proving grounds, places to fathom as well as to celebrate our lost wildness. Above all, the time has come to consider restarting elephant evolution by enabling elephants to reinvent their ecology on the continent that once constituted an important part of their global range. What is at stake is complexity, joy, and the whole way of life of elephants.

Reprinted with permission from Spring 1999 Wild /Earth (PO Box 455, Richmond VT 05477. 802/434-0962, This shortened version does not contain the original's references to the scientific literature.