THE history of life on Earth is a story of boom and bust. Along the way, five major extinction episodes have each seen the death of at least three quarters of all species, followed by life’s resurgence. Are we now living through the sixth extinction as our own activities destroy ecosystems and wipe out diversity? That’s the doomsday scenario painted by many ecologists, and they may well be right. The trouble is we don’t know for sure because we don’t have a clear picture of how life changes between extinction events or what has happened in previous episodes. We don’t even know how many species are alive today, let alone the rate at which they are becoming extinct.
A new project aims to fill some of the gaps. The Paleobiology Database aspires to be an online repository of information about every fossil ever dug up. It is a huge undertaking that has been described as biodiversity’s equivalent of the Human Genome Project. Its organisers hope that by recording the history of biodiversity they will gain an insight into how environmental changes have shaped life on Earth in the past and how they might do so in the future. The database may even indicate whether life can rebound no matter what we throw at it, or whether a human-induced extinction could be without parallel, changing the rules that have applied throughout the rest of the planet’s history.
But already the project is attracting harsh criticism. Some experts believe it to be seriously flawed. They point out that a database is only as good as the data fed into it, and that even if all the current fossil finds were catalogued, they would provide an incomplete inventory of life because we are far from discovering every fossilised species. They say that researchers should get up from their computers and get back into the dirt to dig up new fossils. Others are more sceptical still, arguing that we can never get the full picture because the fossil record is riddled with holes and biases.
Advertisement
The disagreement has opened up a debate about the value of trying to measure global diversity. Is it a meaningless waste of resources, or can history teach us lessons that will help in tackling the current biodiversity crisis?
Fans of the Paleobiology Database acknowledge that the fossil record will always be incomplete. But they see value in looking for global patterns that show relative changes in biodiversity. “The fossil record is the best tool we have for understanding how diversity and extinction work in normal times,” says John Alroy from the National Center for Ecological Analysis and Synthesis in Santa Barbara. “Having a background extinction estimate gives us a benchmark for understanding the mass extinction that’s currently under way. It allows us to say just how bad it is in relative terms.”
To this end, the Paleobiology Database aims to be the most thorough attempt yet to come up with good global diversity curves. Every day between 10 and 15 scientists around the world add information about fossil finds to the database. Since it got up and running in 1998, scientists have entered almost 340,000 specimens, ranging from plants to whales to insects to dinosaurs to sea urchins. Overall totals are updated hourly at . Anyone can download data from the public part of the site and play with the numbers to their heart’s content.
The database encompasses four working groups that include researchers from all over the world. Hallie Sims from the Smithsonian Institution in Washington DC heads the palaeobotany group, which aims to reconstruct the history of plant diversity. A “taphonomy” group studies how dead things turn into fossils. There is also a marine group. Alroy runs the vertebrate group.
Already, the database has thrown up some surprising results. Looking at the big picture, Alroy and his colleagues believe they have found evidence that biodiversity reached a plateau long ago, contrary to the received wisdom that species numbers have increased continuously between extinction events. “The traditional view is that diversity has gone up and up and up,” he says. “Our research is showing that diversity limits were approached many tens of millions of years before the dinosaurs evolved, much less suffered extinction.” This suggests that only a certain number of species can live on Earth at a time, filling a prescribed number of niches like spaces in a multi-storey car park. Once it’s full, no more new species can squeeze in, until extinctions free up new spaces or something rare and catastrophic adds a new floor to the car park.
So the current extinction episode might lead to a rapid burst of evolution over the next few million years, Alroy says, though he warns that the global influx of introduced species could complicate things. “We might end up with a world of vermin,” he says.
Alroy has also used the database to reassess the accuracy of species names. His findings suggest that irregularities in classification inflate the overall number of species in the fossil record by between 32 and 44 per cent. Single species often end up with several names, he says, due to misidentification or poor communication between taxonomists in different countries.
Repetition like this can distort diversity curves. “If you have really bad taxonomy in one short interval, it will look like a diversity spike – a big diversification followed by a big extinction – when all that has happened is a change in the quality of names,” says Alroy. For example, his statistical analysis indicates that of the 4861 North American fossil mammal species catalogued in the database, between 24 and 31 per cent will eventually prove to be duplicates.
Of course, the fossil record is undeniably patchy. Some places and times have left behind more fossil-filled rocks than others. Some have been sampled more thoroughly. And certain kinds of creatures – those with hard parts that lived in oceans, for example – are more likely to leave a record behind, while others, like jellyfish, will always remain a mystery. Alroy has also tried to account for this. He estimates, for example, that only 41 per cent of North American mammals that have ever lived are known from fossils, and he suspects that a similar proportion of fossils are missing from other groups, such as fungi and insects (Proceedings of the National Academy of Sciences, vol 9, p 3706).
Not everyone is impressed with such mathematical wizardry. Jonathan Adrain from the University of Iowa in Iowa City points out that statistical wrangling has been known to create mass extinctions where none occurred. It is easy to misinterpret data. For example, changes in sea level or inconsistent sampling methods can mimic major changes in biodiversity. Indeed, a recent and thorough examination of the literature on marine bivalve fossils has convinced David Jablonsky from the University of Chicago and his colleagues that their diversity has increased steadily over the past 5 million years (Science, vol 300, p 1133).
Adrain believes that fancy analytical techniques are no substitute for hard evidence, but he has also seen how inadequate historical collections can be. When he started his ongoing study of North American fossils from the Early Ordovician, about 500 million years ago, the literature described one genus and four species of trilobites. Just by going back to the fossil beds and sampling more thoroughly, Adrain found 11 genera and 39 species. “Looking inward has maybe taken us as far as it’s going to take us,” he says. “There’s an awful lot more out there than is in the historical record.” The only way to really get at the history of biodiversity, say Adrain and an increasingly vocal group of scientists, is to get back out in the field and collect new data.
This approach seems to be validated by research such as the Panama Paleontology Project, which is perhaps the most extensive field-based project ever. The PPP began in 1986 as an attempt to study evolutionary changes resulting from climate change and the rise of the isthmus of Panama, which sealed off the Caribbean Sea from the Pacific Ocean between 2.5 and 6 million years ago. Now, with more than 500,000 specimens collected from both sides of the land bridge, off the shores of Panama and Costa Rica, the PPP is revealing the urgent need for more thorough fossil collections elsewhere, says PPP-cofounder Jeremy Jackson of the Smithsonian Tropical Research Institute in Panama and the Scripps Institute of Oceanography in La Jolla, California.
The PPP collections have quadrupled the number of samples from some places. In one case, the PPP and a parallel project in the Dominican Republic dug up more species of a certain kind of bryozoan – a sedentary invertebrate – than had previously been documented in the whole of North America. PPP researchers have also revealed that there was a mass extinction of molluscs about 1.7 million years ago, and that new species took their place, leaving diversity at the same level. Results like these suggest that there are more holes in the data than there are data, Jackson says, especially in the tropics, where biodiversity is greatest but often sampled the least.
“There is no reason why palaeontologists could not launch a series of regional studies like PPP that could be used in combination to ask how global diversity has changed over some period of time,” notes Jackson. “We could answer a question like that using about 20 studies comparable to PPP and get a meaningful answer. But it is a waste of time to even ask the question now that we know just how woefully inadequate is the data in the global databases.”
However, one of the biggest supporters of such regional studies, Geerat Vermeij from the University of California, Davis, thinks any attempt to create global diversity curves is pointless, because lumping numbers together only obscures the details of local trends. “Global numbers mean absolutely nothing,” he says. “An organism in the tundra couldn’t care less about how many species there are in the rainforest.” Instead, Vermeij argues, extinction should be viewed purely as a regional phenomenon. As an example, he points to the Pliocene, between 5.2 and 1.6 million years ago, when few extinctions happened in the Indo-West Pacific, but up to 70 per cent of bivalves disappeared from the western Atlantic. In that case, a global analysis might turn up only a minor extinction event, even though the event was catastrophic on a regional scale among a select group of creatures.
Vermeij’s view is extreme – most experts support some combination of regional and global approaches to understanding the history of biodiversity’s booms and busts. Knowing that half of the globe’s species disappeared at a specific time would be less illuminating than knowing that half of the world lost all of its species while the other half stayed the same, says Arnold Miller of the University of Cincinnati in Ohio. To understand what’s really going on, he argues, we need to know how global and local factors work together. The current crisis among corals worldwide is a good example. It is often blamed on global warming, but if a series of regional studies showed that corals are suffering in some places but not others, that would suggest that local phenomena, including such things as overfishing and pollution, could be equally important factors.
The Paleobiology Database could eventually provide the information required to detect shifting patterns of extinctions and speciation at both local and global levels. But some people still argue that, whatever the scale, simply counting and documenting past species will give little insight into the current diversity crisis. According to Jablonski, this approach fails to account for how species relate to each other and to the world around them. “I’m not claiming that every single beetle in the Amazon rainforest is essential,” he says. But some might be crucial as pollinators or links in the food chain. “It’s all tied together. You don’t know which ones you can spare.” This sort of thinking makes it clear that even if species are not yet disappearing at the same rate as they did during past mass extinctions, by disrupting the balance of nature we may be storing up trouble for entire ecosystems.
No matter how you cut it, Jablonski says, major extinctions change the way the world works. And it is becoming clear that even the most significant ecological changes charted in the fossil record are often a matter of chance. If dinosaurs hadn’t disappeared, for example, humans might not be here today. “People used to think mammals won because they are smart and cute, while dinosaurs were dumb and ugly,” he says. “We don’t believe that anymore. The dinosaurs died because a rock fell out of the sky. Mammals were handed the world on a silver platter.”
Then there’s the phenomenon that Jablonski calls “dead clade walking”, in which certain lineages or “clades” of organisms may emerge from an extinction event intact, but subsequently fail to diversify and only lead to evolutionary dead ends (Proceedings of the National Academy of Sciences, vol 99, p 8139). These kinds of trends only become apparent long after an extinction event has ended. “Contrary to popular TV shows,” he says, “not every survivor is a winner.” So, surviving a mass extinction does not in itself guarantee that a species will persist. For conservationists with an eye on the future, the fossil record might help focus their attention on groups that need it most.
There may be ways of working some of this uncertainty into models that try to predict what impact human interference is having on biodiversity. One approach has been pioneered by Sean Nee from the University of Edinburgh. In what he calls the Saving Private Ryan effect, Nee thinks of species as soldiers and asks how shooting them at random would lead to the wiping out of entire families. The answer depends on the size of each soldier’s family and how many soldiers get shot. Nee uses mathematical models to see how many branches of the tree of life would be slashed if various numbers of species were to disappear. “If you have an unusual species with no close relatives,” he says, “maybe it should have higher priority for conservation than one with many relatives.”
A complete Palaeobiology Database would allow Nee and others to compare today’s extinctions to patterns in the past. But predictions about what will happen next can only be as accurate as our knowledge of the number of species alive today. And that is far from certain. Fewer than 2 million species have been identified and, while most experts suspect the total falls between 10 and 20 million, the number could be as high as 100 million. With an inventory of all living species, ecologists could start to put the current biodiversity crisis in historical perspective.
Although creating such a list would be a task to rival even the Palaeobiology Database, it is exactly what the San Francisco-based ALL Species Foundation hopes to achieve in the next 25 years. The effort is essential, says Harvard biologist Edward O. Wilson, who is alarmed by current rates of extinction. “There is a crisis. We’ve begun to measure it, and it’s very high,” Wilson says. “We need this kind of information in much more detail to protect all of biodiversity, not just the ones we know well.”
Let the counting continue.