Prehistoric Overkill: Four Decades of Discovery and Debate

Paul S. Martin, University of Arizona

Real Audio Recording   

I've looked forward to this for weeks and months. This is exciting. I'm truly thrilled to be here. The reason I'm so excited about it is because I've spent much of my life devoted to some of the problems and issues that I think we'll touch on today, that certainly look that way from the description of the program. I've met many of the people who are on the program, and I haven't met them all. So it's thrilling to have an opportunity to cross paths with people devoted to an issue that I think is so important -- not just for the narrow disciplines involved, but for the history of our country, and the understanding of what America is truly about. If we know our country, we must know things that go beyond Columbus, that go back to the origins of human presence in the New World, and in other parts of the world.

But before I go into the pictures I've got to share with you, I have to acknowledge one other group of people who influenced me when I was a student at Cornell in the 1950s, who were spear carriers for Ernst Mayr and his wonderful new book, Systematics and the Origin of Species. Max Hecht, of Queens College, retired; Bessie Hecht; Danny Marien, geneticist at Queens College; and certainly the late Karl Koopman, here at the American Museum, sticking to his last . . . studying the mammals that he devoted his life to. Those were the people who helped set me on the path.

The path has led to discoveries that David Burney has summarized neatly in the American Scientist in this illustration. . . . David Burney's color illustration shows that the world has been divided into the haves and the have-nots as a result of extinctions in prehistoric time. We have Eurasia and Africa, the megacontinent, with megafauna remaining in it. And we have the rest of the world, whose continents are stripped of large animals as a result of ice age extinctions, or prehistoric extinctions after the end of the ice age. We have oceanic islands which don't show as well in Dave's illustration, though some of them certainly do, that lost heavily, with the exception of a few islands that serve as controls that didn't lose their tame native species of birds to any great degree until historic times -- specifically the Galapagos, the Seychelles, Reunion, Lord Howe and a few others that never were found until the historic period.

But most of the Pacific, the islands of the Pacific and many in the Atlantic, lost within the last few thousand years -- David Steadman estimates, in the case of the Pacific islands, some 2,000 species of flightless rails and Gallinules in the family Rallidae disappeared -- presumably coincidental with the first arrival of prehistoric people in those areas.

If we look back in time in North America, in the fossil record, back about 3 million years, the dotted line on the screen to your right is total number of genera, or total number of species, found in that time slice. The disappearance of genera is shown by the continuous line, and over the last 3 million years there's a bit of a bump, an extinction event, around 1.5 to 2 million years ago. But the number of genera going up over 30, or the number of species going up over 50, doesn't happen until the end, practically the very end, of the last 3 million years. And, in fact, if you sum all the known generic extinctions of large mammals in North America previous to 12,000 years ago, it's less than the number that disappeared 10 to 12,000 years ago. There's nothing comparable in the fossil record back that far; and I think John Alroy may summarize a fossil record that goes much deeper in time, and doesn't reveal such heavy loss of families or genera. We lost even a couple orders from North America and South America at the end of the Pleistocene.

These are flagship species. They aren't termites or daphnia, or races of drosophila. These are animals whose presence would make a huge difference. If Teddy Roosevelt had gone into the Amazon and come back with a giant ground sloth as a result of his hunting expedition there, this would resonate in ways that we have no idea about at present. The conservation movement lost symbols when it lost the megafauna in the Late Pleistocene. We cannot ignore this. If we're looking for restoration of diversity, and we don't consider large animals -- but we accept the bison, the elk and moose, and deer and antelope as sufficient in North America -- we're selling the place short. But I'm getting political already, and I want to go back to a little bit more of the data before I return to that theme.

Tony Stuart has brilliantly compared Europe and North America -- I hope I don't steal too much of his fire by showing some of his best data. The illustrations to your right are the kinds of large animals that we have to deal with. These are some of those 30 genera that we've already thought about occurring in their extinction only 12 to 10,000 years ago. And closer toward me is this wonderful record of how gradually we lose large animals from Europe and northern part of the Eurasian continent, starting sometime beyond the reach of radiocarbon. All of this argument turns on radiocarbon dating -- on good techniques, good measurements with radiocarbon -- now being done with such brilliance by Tom Stafford. We're losing temperate climate elephants and rhinoceros over 40,000 years ago. We're losing the hippopotamus out of Europe -- of course, it persists in North Africa and further down in Africa.

But the kinds of large animals that disappear end up in the northern coast of Siberia, with the last woolly mammoths 10,000 years ago. And it looked like that was the end of the line for this wonderful animal of the ice ages, until Wrangel Island was discovered with woolly mammoths on it 6,000 years later in time than had been previously known -- into the time of the pharaohs. The woollies almost made it -- and you think if only someone would look in Banks Island carefully enough, maybe they would find some still alive there, somewhere in the Canadian arctic archipelago. It's a fantasy, of course. But the musk ox survived until only 3,000 years ago in parts of the Old World, and it's fascinating. The musk ox had been allowed to be reintroduced into Siberia, out of stock from Canada that magically survived in the New World. One of the few cases in which a New World animal managed to pull through, as opposed to survival of large animals in the Old World, which happened more typically.

Now, by comparison -- the extinction of these creatures over here, when we can get good radiocarbon dates on them, co-occur right around 11,000 years ago. Mammoths, mastodons, camel -- mammoth of a different species; not the woolly but the Columbian mammoth -- horse, ground sloth, and sometimes there's several genera in these groups; sometimes several species. But this is where you have to consider some 30 genera of large animals. Now, maybe some of them disappeared earlier than 12,000 years ago, and maybe further radiocarbon dating will show that. But the dates that keep coming in, that cross my desk, are inevitably right in the ballpark that I've been talking about. Maybe some survive longer, and negative evidence is elusive. You cannot prove that anything is extinct just because you've got a young date on it. And you say: well, we've got ground sloths until 11,000 years ago in the Grand Canyon. They died then, they disappeared then. We don't know that for certain -- all we know is that that's the youngest dates that have been reliably obtained on such an animal.

So let's go look at the Grand Canyon briefly and see what we find. . . . In the Grand Canyon, at the mouth of Rampart Cave -- maybe somebody in this audience would recognize the guy in the white shirt, although I couldn't from right here. Twenty years ago he was a younger man, one of our most famous authors, most popular writers ... James Michener. And why did James Michener come to this particular cave in the Grand Canyon? Well, he came to see sloth shit. The deposit in the cave with dung balls on the surface, ideal for radiocarbon measurement. No problem of contamination in such a sample, and you can compare it with rhinoceros, which it resembles in texture, and cow shit, which it doesn't resemble because cow manure is much finer ground up. Radiocarbon dates on such deposits -- not only in Rampart Cave, but in other caves between Nevada and west Texas, some as different in their local environments -- Rampart is in the Mojave Desert; these high-elevation caves that also have sloth records in them are up in Pinion Juniper. The environment when these animals disappeared was different than it is now. It was cooler, and there were plants in the area that don't occur around the caves today. But the environment at high elevation was quite similar to the environment around Rampart Cave in the Late Pleistocene. So the environment itself -- the woodlands that the sloths lived in -- didn't disappear from the Southwest. And you can tell what their diet is from such a deposit. Much of it is globe mallow, and some is ephedra, desert plants that occur in the area now. These animals did not graze -- they fed on shrubs and forbs.

The stratigraphy in Rampart Cave, from bottom to top, with dates that go back on sloth dung over 30,000 years, and end at the top of the pile around 11,000 years ago. Now, this is old hat. There are people in this room who have seen these slides many times before. I give pretty much the same talk whenever I get a chance. The data are not radically different. But the trio of extinct animals that we deal with in the Grand Canyon goes beyond the ground sloths. It includes an extinct mountain goat that Jim Mead worked on for his dissertation years ago -- and he, too, had ideal material for radiocarbon dating ... namely, the horn sheath of an extinct mountain goat.

Coming out of another cave up in redwall limestone -- and this is what appears from one of the caves that Jim Mead investigated -- a wonderful skull of this extinct animal. This is not a mountain sheep, which occurs in the Grand Canyon at present. This is a close relative of Oreamnos americanus, the living mountain goat; and this extinct species is called harringtoni. And this is the deposit that specimen comes out of. This shows the quality of preservation. This material is dry as what you would find in a tomb of the pharaohs. The horn sheath -- this keratinous horn sheath, buried into the breccia-like deposit -- is part of the specimen we're looking at over here. The palate, this tooth row, is shown in the specimen that was recovered from the deposit. There's a lower jaw, a mandible of this same animal.

Now, you look at that and think: How could this be 12,000 or 11,000 years old? It's so fresh. But that's the testimony that the radiocarbon dating provides. Why shouldn't these animals last longer? Well, who knows. They don't, as far as our measurements are determined -- here's Tony Stuart's summary diagram of Harrington mountain goat radiocarbon dates. There are quite a few of them, back into as far as we can reach with radiocarbon. There's no clustering of dates near the top of the record -- the animals didn't leave us an opportunity to bias our sampling for youthfulness. The dates that Jim Mead got were scattered through the 20,000-year time slice that radiocarbon provides evidence of presence of these creatures. But when we plot their disappearance on a time line from 25,000 to 10,000 years ago, they disappear together.

Anyone identify the bird whose skull is on the right? Want to venture a guess what it is? Size of a dollar bill? It's a condor, a California condor. And if you were apprehended with this in your backpack in the Grand Canyon by somebody from Arizona Game and Fish, you'd be in for a felony, for having killed a condor. I mean, it looks like a freshly prepared -- macerated, maybe -- but pretty fresh cranium, with a beak attached. Perfect for radiocarbon dating. But you might not expect it was necessary to radiocarbon-date something like this -- it looks so fresh. Twelve thousand five hundred is the radiocarbon date that was recovered from this particular bird.

Steve Emsley studied not only caves in the Grand Canyon, but he got condor remains from other caves in New Mexico and west Texas. And the assemblage of radiocarbon dates from those localities deserves to be added to our time line and also radiocarbon dates not from the Grand Canyon, but from a source of material that's excellent for such measurement -- the saber-toothed cat from Rancho La Brea, Smilodon. And all of these creatures disappear around the same magic moment, some 11,000 years ago, by these dozens of measurements that have been assembled.

We're looking at the disappearance of two herbivores, one scavenger and one carnivore scavenger. Three are mammals; one's a bird. Of the two herbivores, one's arboreal in its origin and distribution -- the mountain goat. And one is neotropical -- the extinct ground sloth. If the climate were driving the cold-adapted species into extinction -- if things were getting too warm for Harrington's extinct mountain goat -- we might expect the ground sloth to do really well, to increase in numbers. Or if both herbivores were declining in numbers -- if they were both collapsing as the food supply declined -- we might expect to see their scavengers and carnivore predators disappear first in the fossil record, perhaps by several thousand years. If that's occurring, the time scale is too tight for us to resolve by radiocarbon so far. So I get at an ecological argument here that says that whatever is hitting this system is affecting different trophic levels, and different types of life forms.

Now, with French horns, the discovery of America has to be reckoned with. And it turns out that the radiocarbon dates on archeological sites in the New World linking the presence of the first people, link also the last occurrence of mammoths. And this starts in the state of Washington, on the Columbia River, at the Richie Roberts site, and continues down to the Mexican border, at Naco, with some 15 associations in between, many of which have been carefully dated by radiocarbon -- and, lo and behold, the radiocarbon dates are right there around 11,000 years ago for these sites. Now, there are two ways to look at this. One is: Hey, there's only 15 co-occurrences of mammoths, and none of ground sloths and none of extinct mountain goats -- never mind horses and camels, and a whole lot of other extinct animals -- with the presence of fluted points of these early people into the New World. Surely, that's not enough to explain the extinction of the megafauna by the arrival of people.

The other way to look at it, as Gary Haynes has pointed out, is that if you look for sites where we know elephants were killed in Africa 100 years ago in campaigns to remove them from lands that would become agricultural, the bones of those animals in known historic kill sites are not to be found. Fossilization is a rare event. Preservation of kill sites is unusual. In the whole of Eurasia, the co-occurrence of people and mammoths, has not yielded, according to another member of the Haynes team, who's unrelated to Gary, my colleague Vance Haynes at the University of Arizona. Yes, there are a lot of associations of archeological material with mammoths. A great deal -- some very spectacular ones -- but there are no clear-cut kill sites that have been described archaeologically, comparable to the Clovis sites, which we'll look at very briefly now. And I'm getting close to the end of the line with this story.

In Arizona, Clovis material is shown in this recent plot by Vance Haynes. Some of the Clovis points which can be identified easily may be picked up and moved into places by later people, or may even be in the possession of Native Americans now, who've recycled the stuff -- continued to keep it or use it. But the distribution of Clovis sites throughout the state is much greater than the known associations with mammoths, that we only have in southeastern Arizona. These sites should be international parks. There should be an arch over one of these places to match the St. Louis arch. We haven't begun to celebrate our country. I'm back political again, so watch out.

The sites themselves -- the four sites I'm referring to -- of these four, probably the best one for public display would be Murray Springs. Arizona is a transitory state. People come there and they don't decide -- they don't know whether they want to go to California or Texas. And, while they're making up their mind, they stay in Arizona. We don't have loyalists around, like the Mormons of Utah, that would really go out for an exhibit of something extraordinary in their fossil, record, I think. Or Californians, who have more loyalty to their state. Or Texans certainly have a great deal to theirs. Arizona is still trying to find its way. It can't even appreciate the fact that it was discovered, historically, in the 1540s. . . .

Back to the extinction of megafauna and the opportunity, if we could reconstruct this -- and it could be done; the map of this site is available. The deposit at Murray Springs of mammoths associated with tools and Clovis points, and even a shaft straightener unique to the New World found at that locality. I'm talking about this artifact up here, made of mammoth bone, and presumably used to straighten spears. The tools are in not good condition. If there were perfect Clovis points associated with those mammoths and bison at the site, at Murray Springs, they would have been saved by the people who were processing the bones of the extinct animals. Radiocarbon dates by Haynes in here -- right around 11,000 years for Clovis, a little bit younger for Folsom -- no extinct animals in Folsom time or subsequently, a test of this business of extinction by 1,000 years ago. And a summary diagram on one side, and new radiocarbon dates coming in, thanks to Russ Graham and Tom Stafford, and others who are generating new information of this sort. And I hope to hear more of this as we go into the day.

Every once in a while there's a date over here that's younger than 11,000 for all these different creatures, and those particular measurements, those dates, happen to be on genera or species of animals that are still around, like the raccoon -- or us. The extinct creatures over here include Cervalces. And, most wonderfully of all, at the bottom, Toxodon, the notoungulate from South America, member of an extinct order. There's much more radiocarbon dating to do to support this analysis, to test this argument.

The extinction of mammoths is easily driven by the arrival of a small number of potent predators, mammoths being relatives of elephants, and elephants being slow-reproducing species vulnerable to any extraordinary predation. Now, maybe that is not enough to explain the other extinctions, and the arguments of various folks who are here will illuminate ways in which the disappearance of mammoths, driven by people, might be amplified by phenomena that have nothing to do with human predation. Although I'd still argue that human predation should be important in the disappearance of ground sloths, for example, that ought to be extremely vulnerable to it.

Climate can't be involved. The climate has been changing so drastically, for such a long period of time -- measured by proxy data -- that it seems impossible that, by the end of a million-year period of swings from cold to warm, and back again, we could run into just one more that would make all the difference in the world.

To finish with one more thought, I just want to mention the disappearance of burros in historic time -- from the Grand Canyon -- as a result of policy and management aimed at a proxy species, or a surrogate for the extinct horses, that wasn't suitable in the opinion of the public and the agency involved in management of the Grand Canyon. Some of us opposed this. We thought that the proxy wild burros in the canyon should be given a chance to continue, but management ruled otherwise and the burros were removed. Not long after that, a condor-reintroduction program took shape, and we're now getting condors into Arizona. And guess what? The need for a food supply for those animals is going to be a major problem. So if we're going to restore anything of the sort that was lost, we need to think in new ways, along new lines. Not even the reintroduction of proxies for mammoths should be excluded from our hopes.

So if anyone out here has a big wallet, please talk to me before the day is over about introducing elephants into your ranch.

Q and A:

M: Do we have any questions for Dr. Martin? We have microphones, we have one right here.

Q: What is the earliest radiocarbon dates for humans in the New World, presumably in the northern part, in the Bering region, where they first came in?

PM: In Alaska, the earliest would be a little under 12,000. But it depends on who you talk to -- because there's a huge interest in searching for pre-Clovis presence. And there may be people out here, who know better than I, that there were, in fact, people in the New World before Clovis, and I don't want to get into that.

Q: [inaudible]

PM: Right. Our species reproduces fast enough, and is adaptable enough, to sweep from Alaska to Tierra del Fuego in less than a thousand years, if you accept models that have been proposed along those lines. It's amazing but doable. But 12,000 years, or 13,000 wouldn't fit the model, and those dates may occur -- I don't know.

interview | bio