How Many Species Will We Lose?
A conversation with Stuart Pimm
Contrarians argue that environmental problems such as global climate change, stratospheric ozone
loss, the loss of species, and human population growth are not serious problems. One of the glibbest of the
contrarians is Julian Simon, a business professor who picks and chooses anecdotes to paint a rosy picture of
environmental issues. In a July 19, 1996 letter to Science magazine, he said there is no evidence that the rate of
species extinction has increased in the last century, and he questioned whether human population growth has
caused an increase in the rate of extinctions.
In response to this and similar pronouncements on the subject by non-experts, a group of scientists has examined the historical evidence of the loss of species and genera from different groups of animals and plants in many different parts of the world. Not only is Simon's argument easily rebutted, but scientists now are beginning to understand the causes of extinctions well enough to predict where and how many will occur in the future.
We spoke with Stuart Pimm about why and where species extinctions are occurring and where they can be expected to occur. Professor Pimm received the Ph.D. in ecology in 1974 from New Mexico State University. He is presently a professor in the department of ecology and evolutionary biology at the University of Tennessee, Knoxville.
ER: Professor Pimm, how did you get involved with debating contrarians?
SP: A few years ago Steven Budiansky wrote an article in US News and World Report. He titled it, "The Doomsday Myths". He tackled the lot: global warming, the ozone layer, and the issue of species extinction. He argued that massive forest losses did not cause species extinction. He was clearly wrong, he didn't understand the science and I decided he needed to be corrected.
The particular counter example he used was the fate of species in Eastern North America. But we have an extraordinarily good history of what we have done to the continent since Europeans first settled it in 1620 or thereabouts. Almost as soon as the settlers began to move away from the Northeast and coastal Virginia, naturalists were out there with their guns collecting specimens. We know John James Audubon was particularly thorough - he didn't miss much - and he found a number of species that we don't have today. The Labrador duck and the great auk were hunted out quite early on, but they were marine species. The eskimo curlew, which is gone or nearly so, nested in the remote arctic but stopped on the Great Plains on migration. It suffered a double whammy of losing its migration routes and losing its wintering grounds in South America. In the eastern forests of North America there were only four extinctions: the passenger pigeon, the Carolina parakeet, the ivory billed woodpecker, and Bachmann's warbler. They are all more or less southeastern birds, the most northerly would have been the pigeon.
Here we have the apparent anomaly that having destroyed so much of Eastern North America, we have lost so few species. In Budiansky's article he has a map of the forests of Eastern North America in 1620, it is mostly forested and then he has a map showing what old-growth remains today - the only old-growth forests remaining. It looks as if we have destroyed almost all our forests and what has happened? Only four bird species out of over 150 have gone extinct. The implication he draws is that we ecologists don't know what we are talking about, we are scare mongering, we are part of some global conspiracy to alarm the public and to keep ourselves employed.
Bob Askins and I decided to take a closer look at that argument. There are two parts to the forest story; a local one and a regional one. The local story is, the colonists cleared the forests, farmed for a decade or so and then let the forests grow back again. The example that comes to mind is George Washington's estate at Mt. Vernon. If I remember correctly, that estate was cleared of its forests at least three times. They needed an enormous amount of wood; the houses were poorly insulated, the stoves were inefficient, they were appallingly wasteful in their use of wood. As forests claimed old fields, they cleared new areas. Colonial agricultural yields were only about a quarter of what they were in the Old World. Here in the Colonies land was cheap, labor was expensive, and there was limited use of fertilizers. At this local level we cleared fields as other ones were growing back into forests: there were always some forests.
At the regional level the story is much more familiar. We settled in New England and on the Virginia coast, moved into the Ohio valley, and kept moving progressively westward. As we moved west, the eastern forests recovered. Detailed records show the New England forests being cleared in the early 1700s, then the Ohio valley cleared by the late 1700s early 1800s, Tennessee was being cleared by the mid-1800s and so on. Settlers arrived here in Tennessee about 200 years ago. I live just on the south side of the river that runs through Knoxville. There is a drawing in Harper's Weekly from the 19th century of where I now live. It was completely bare of trees, it was just grazing land at that time. Today it is forest again. As late as 1890 Tennessee was only five to ten percent forested; Tennessee now is a beautiful, largely forested state.
This movement of people westward, clearing as they went, is so much part of our American culture and history that it is hard to stand back from it and look at the ecological story that it tells. What is happening today in your part of the world, in the Pacific Northwest, happened here in Tennessee eighty or ninety years ago. The fights over setting aside the Great Smoky Mountains National Park are eerily like the battles that are being played out in the old-growth forests of the Pacific Northwest today.
Getting back to the extinction story: patterns of deforestation tell us that at no time was America more than about fifty percent deforested. 1870 was the low point and we have more now.
The second part of the extinction puzzle is a relationship we ecologists call the species-area curve. It is a calibration of species against area. We ecologists love to go to islands. Off we go to the Caribbean with our butterfly nets or our plant presses or our binoculars. We visit many different islands of different sizes and count the number of species there. If you plot the number of species against the size of the island, you find a relationship that is not linear but curved: if you double an area you do not double the number of species. A simple way to express it is this: An area that is half the size of a larger area does not lose half its species, it has about eighty-five percent of the original species in it.
We can use this area-species relationship to make predictions about the consequences of clearing forests. Suppose we cut half of an area's forest, then a fifty percent loss of habitat will lead to a fifteen percent loss of species. The prediction says we have lost fifty percent of Eastern North American forests so we should have lost twenty-four species; that is, fifteen percent of the 160 bird species in Eastern North America.
One of the things that has always annoyed and exercised my friend Paul Erhlich is that there is abundant evidence that species go extinct; by that he means there is much evidence that we lose species locally. Almost all the species that once lived in Manhattan are now extinct, for their homes are now under concrete. Unfortunately, this idea of counting local extinction confuses people because they say they are not really extinct.
So let's play the hard nosed rule that says, if a species is extinct it means it is gone for good, a global extinction. How many species would we lose if we chopped every last tree down from Maine to Florida, from Louisiana to Minnesota? The answer is, a surprisingly small number of species: there are only twenty-eight species of birds that are found only or mostly within that range. Many of the birds we think of as familiar birds in Eastern North America are also the birds you have in your garden in Washington state. These widespread birds would not have gone globally extinct even if we had cut every tree east of the Mississippi. Now, let's restrict the analysis to the ones that are found only in Eastern North America. We call these endemics - they are species with small geographical ranges. Fifteen percent of twenty-eight species is almost exactly four; in close agreement with the four extinctions that actually happened.
ER: Is the Carolina parakeet an example of having a restricted range?
SP: They all were. The only birds that could go globally extinct would be the birds that were restricted to the
East; all four of those species were. Three of them: the parakeet, the woodpecker, and the Bachmann's warbler were all restricted to the Southeast. It was the southeastern forests that got hammered hardest, and there were no more than about a dozen species that were found uniquely in the Southeastern US. Of those dozen species, we lost three, a twenty-five percent extinction rate!
Put simply, we lose species that have small ranges. So we can use our history as a key to the future. In contrast, in the Amazon by and large, most of the rainforest is still there, and we have to make projections about what we think will happen in the future. Particularly in the tropics there are small areas where, in the size of a few states, there could be a hundred or two hundred species found only there.
ER: There can be hundreds more species per acre in the Tropics than in higher latitudes.
SP: Yes, but they also pack more endemic species in. I talk about the cookie cutter model: the idea that deforestation stamps out little areas. If that cookie cutter lands on an area with many endemics, you lose many species. In fact, if you had a malevolent cookie cutter and you selected the eighteen areas worldwide that Norman Meyers calls hotspots, you could wipe out twenty-five percent of all species in an area that is no bigger than Eastern North America.
It is alarming that these places are all vulnerable. For example, the Atlantic forests of Brazil have broadly the same number of species of bird as Eastern North America but about 150 of those species occur only in that area. It is an area ten or fifteen times more vulnerable than the Eastern USA. Brazil's population is growing quickly and people are moving to these coastal forest in large numbers.
ER: What are hotspots?
SP: They are areas where there is a large number of species that have small ranges.
ER: You estimated the overall extinction rate before humans to be about one extinction per million years per species. Is that also true of hotspots?
SP: That pre-human extinction rate means that only one in one million species should go extinct each year. It is based on the most abundant fossils - mostly marine invertebrates. We don't have fossils for hotspots, so we don't know what their extinction rates are. Some have argued that extinction rates on islands, or island-like hotspots in places like the Andes, may be naturally very high.
I think there is a quite powerful but indirect way of getting at that question: It is that natural extinction rates cannot be too different from natural speciation rates. If natural extinction rates exceeded speciation rates, there would be no species left. And we do know something about the rates of speciation and evolution.
The field of molecular genetics is beginning to give us a valuable insight into the dynamics of speciation, and it is not that fast. And as molecular clock data become available, we are often able to estimate the time it takes to generate new species for islands like Hawaii, where many species have gone extinct. We are finding that species are going extinct hundreds or thousands of times faster than evolution can produce new ones.
ER: What are the primary mechanisms of extinction?
SP: Most people think the main cause was and will be habitat destruction. I am not so sure about the past causes. There is a quite disturbing difference between the mechanisms of past extinctions and what we think will be the mechanisms of the future extinctions. We wrote a paper a couple years ago, where we compiled all the individual studies of extinction hotspots: of birds on islands, of mammals in central Australia, plants in the Fynbos of South Africa, fish in Africa and North America, mussels in North America. The extraordinary thing is the people who did those studies all considered introduced species to be the principle cause of extinction. Goats on oceanic islands, the nile perch in African lakes, eucalyptus trees in the South African Fynbos, and foxes and rabbits in Australia - they have all caused massive habitat damage.
If you look at what ecologists project is going to be the principle cause of extinction in the future, it is habitat destruction. My experience working in Hawaii is this: When you look at the tiny little fragments of forest that remain there, you see an enormous amount of damage done by introduced species. When you fragment a habitat; when you leave little postage stamp-size bits of habitat behind, the species within them are vulnerable.
They are surrounded by areas, all of which are full of weedy, successful species which not only thrive where they are, but penetrate those habitat fragments and do damage there as well. Put simply, habitat losses cause extinction, then the species in the surviving habitat fragments suffer extinctions from introduced species. Fragmentation makes it easier for the introduced species to come on in, and when they get there they are able to do a lot more harm.
Our projections of what is going to happen into the future are probably underestimates. We have said, if you chop ninety percent of the forest down, you will only lose fifty percent of the species. The problem is that those fifty percent of the species remaining in their little postage stamp nature reserves are going to be under an enormous amount of threat from introduced species. That area-species formula we talked about only takes habitat fragmentation into account. So we have got bad news and we have got worse news: We are going to lose some species, and the species that remain behind are going to be under threat from the organisms that have replaced them.
The reason small patches of habitat lose species is a whole litany of problems: The obvious one is that when populations become very small, individuals often can't find mates or don't have enough young. The example I always give as a long term, detailed study of a small population and all the disasters that befall it, is British royalty. For 900 years they have suffered one disaster after another. They are repeatedly unable to find the right mate, and to produce an heir and a spare. There are also genetic problems, inbreeding, that arise when all the potential mates are close relatives. Royalty's problem is that they want to be a small population; they don't want to marry commoners. And for all small populations, life is hard. Unlike royalty, many species have small populations, not by choice but from loss of habitat.
The other part of the problem for small populations is the intrinsic variability of nature. Even when the population is large enough not to worry about the problems of finding a mate, there are still concerns about the variability that nature throws at you. Working with endangered species I am constantly aware of the considerable variability that nature imposes on rare species.
The numbers of a small bird can vary ten-fold over a twenty year period. These are not endangered species, just ordinary, common, well-established species. So we set aside a small reserve, a national park - the one I am working in at the moment is Everglades National Park - then you find you have a population of maybe 5,000 individuals of an endangered species there. The normal range of variability means you have a fifty percent chance of seeing numbers as low as 500 within twenty years or so. With the best will in the world we have set aside preserves, but not really thought about the natural variability in numbers that their species encounter. We wake up and find the reserve no longer has 5000 but 500 individuals, a number well below the threshold for being critically endangered. Even big parks, even quite large fragments of area are sometimes not big enough to hold sustainable populations.
We talked earlier about the Atlantic Coast forests of Brazil where only maybe one or two bird species have gone extinct. A large number of species are critically endangered. They are playing a game of demographic roulette; they might survive one disaster and maybe they will come back, but in all probability another disaster will wipe them out.
ER: It is like a gambler playing with a small stake.
SP: It is exactly that process: a couple bad rolls of the dice and you are out of the game. There may be some gamblers that linger longer, but some will go fast. Incidentally, this creates a devil of a problem for managers because they are almost fighting chance. You don't know what will hit you next. When I first became involved in Everglades National Park about eight years ago, the problem was too little water, and the last few years it has been too much, at least too much in one particular place. And in between we have seen a hurricane, we have seen fires. It is almost Biblical; fire, flood, there is one catastrophe after another. Species survive those catastrophes by spreading the risk, they occur in many different places. A catastrophe might wipe out a population here, but it will survive there. Next time it will be the other way around. Increasingly we are making all species play with small stakes.
ER: Can we predict extinction rates in the future?
SP: I think there are many fewer problems doing this than the critics would allow. Critics would say, No you don't know what is going on. This is Julian Simon's point: you only know that a few species are going extinct.
Well, increasingly the groups of species that we know are going extinct are growing as our knowledge increases. I mentioned in my rebuttal to Simon in Science recently, the Nature Conservancy has just published a small but important report that looks at eighteen different groups of species in North America. These are not just birds; there are mammals, amphibians, lizards, flowering plants, ferns, dragon flies, beetles, all species that for a variety of reasons, people like and so we know them quite well. They are very different kinds of species in very different kinds of habitats. Across that range you find that birds are proportionately the least threatened group. It is not that I picked birds as examples and birds are wimpy and go extinct easily. If anything it is the other way around, birds are actually quite robust.
Simon's argument is easily rebutted; there is nothing peculiar about the species we pick, other than that we know them well. They are completely different kinds of species; birds and ferns, tiger beetles and mussels. We know that an increasing proportion of species are being threatened with extinction, in many cases they have already gone extinct.
Moreover, we understand the mechanism of extinction in a quantitative, predictable way. If we know the patterns of endemism, we can predict the proportion of species that are going to go extinct. That is a no-nonsense, quantitative assessment of humanity's impact. My group has used this recipe to provide quantitative answers to such questions as: How much forest have we lost and how many species have we lost as a consequence?
At the moment, my research group is mapping out the patterns of endemism, and then combining that information with information about forest losses. The information comes from satellite imagery taken over the last twenty years, and even older information from aerial photography. We put those images together so we can come up with detailed predictions of species extinctions in different places. My student Tom Brooks is currently in Kenya; I will be joining him soon. We chose Kenya because we have a fifty-year database of deforestation in three different areas of Kenya. This is a joint project with a team from the National Museum of Kenya supported by the National Geographic Society. Our argument is: Here is a forest fragment that was isolated twenty years ago, and it has a certain number of species in it. Here is a similar fragment that was isolated fifty years ago. How many more species has the older fragment lost? We are beginning to get at the very details of how quickly species are going; and from what size of fragment we will lose how many species.
ER: This is fine tuning or calibrating your extinction curve.
SP: Exactly. It is a matter of calibration. It is also a matter of giving advice to those who establish and manage national parks. They want to know how much time they have, how many species are they going to keep? How many of those species are going to be so rare that they are going to have to spend a lot of resources to protect them.
ER: Those who deny environmental problems strike a receptive chord in many people. What truth do they come close to?
SP: These nay-sayers are a very potent combination of a number of things. First, I don't think you can dismiss the fact that many of them get their support from powerful interests; from people who have a great deal at stake in there not being any legislation to limit what they do. Your Senator Gorton's version of the reauthorization of the Endangered Species Act on which I had to comment last year, looked to me as if it were written by the timber interests of your state. Many environmental critics have access to resources that are very impressive: the Heritage Foundation, the Cato Institute, and all those sorts of places.
These critics also touch a receptive chord. I think in part that the First Amendment to our Constitution, our commitment to free speech, the sense of fair play, the sense that journalists must present both sides of a story is a factor. This is a deeply ingrained tradition that we all admire and we all respect. Unfortunately it means that if you have one uninformed person against a complete scientific consensus, that the one uninformed person is going to have his views aired. There are all sorts of disastrous cases: a man who says that HIV does not cause AIDS for example. Not all of Julian Simon's stuff is benign. He writes, Don't worry about the ozone layer, the increased ultraviolet may "have beneficial effects" by reducing the incidence of rickets. Of course there are many medical people who are worried about the increase in skin cancer. I lost my first graduate student to skin cancer.
The public feels that we should always have both viewpoints. But does this mean that every time NASA shows a picture of the Earth from space, they have to have somebody from the Flat Earth Society present to say that the Earth is really flat? There is a right to free speech, but that doesn't mean that uninformed people have a right to be heard in every debate. We all realize that is a horribly difficult issue and I think most journalists who don't know the science say very sensibly, I am going to cover both viewpoints. At some point however, they have the responsibility to weigh the evidence and chose to ignore the flat earthers.
As society gets more complex, it is not easy to see what the issues are: It is much more difficult to decide whether there is an ozone hole than whether the Earth is flat or not. We have to struggle with the issues. That is why we value education. That is why we are not still living in caves.
The final criticism is the view that says scientists are just a special interest. It argues that we are just like every other group of people who are trying to make a living. When you are in a group of people who blatantly flaunt the truth for political or economic gain, it is hard to believe there are other people who care passionately about it. There are scary implications in the view that the evidence for the ozone hole, global warming, and species extinction is politically debatable. The policies to deal with these issues involve politics, but the Earth is roughly spherical, however you decide to vote. As I said to an amused Senate committee last year, I hoped that the journalists then about to testify about how few extinctions there were would not then extend their amateur meddlings to my father-in-law's profession of brain surgery.
1. Pimm, S.L., Russell, G.J., Gittleman, J.L., Brooks, T.M. Science, 269, 347-350 (1995)
2. Pimm, S.L, Askins, R.A., Forest losses predict bird extinctions in eastern North America, PNAS (USA) 92:9343-9347 (1995)
Copyright 1996 Environmental Review