Never Mind That Boiling Kettle…

Last week I briefly mentioned some stark estimates about the potential extinctions that could be triggered by global warming. Since then, some global warming skeptics have tried to pour cold water on these results by making some dubious claims about natural selection and extinctions. While I have reported about global warming from time to time, I leave blogging on the subject to others (particularly David Appell over at Quark Soup). But in this case, evolution is drawn into the mix.

Here, in a nutshell, is what the scientists wrote last week in their Nature paper (which the editors have made available for free). They studied over 1,000 species in all sorts of terrestrial habitats, from Mexican deserts to Australian rainforests. Using information on their ranges, the scientists estimated the range of temperature, moisture, and other climate conditions in which the species currently persist (what they call a "climate envelope"). As the climate changes in the coming century, these climate envelopes will change shape. Some may expand towards the pole, while others will slide, and others will shrink. To estimate the shape and position of these new climate envelopes, the scientists looked at the shift in these ranges in three different scenarios for life in 2050–a "small" change of .8-1.7 degrees C, a medium change of 1.8-2.0 degrees, and a big change of over 2 degrees. (These ranges come from IPCC projections, which in turn are based on a range of potential future emissions of carbon dioxide. For comparison, the planet has warmed an estimated .6 degrees C in the past century.)

Some species that can spread quickly may be able to move into their new climate envelope. Others that disperse slowly–animals that only live in isolated patches of heath, for example–may only be able to survive in the overlap between today’s climate envelope and the envelope of 2050. Others still may simply have nowhere to go–Australia’s rainforest, for example, are on the northern coast of the country. Global warming is predicted to chew away at their habitat at the south, but they can hardly find new territory in the ocean to the north.

The researchers estimated how much range all of these species will lose. Ecologists have long known that the number of species that a territory can support is proportional to its size. That means that if a forest gets cut into a fragment, it will lose some species. If a species is found nowhere else, it becomes extinct. Based on this relationship between area and biodiversity, the scientists concluded that global warming will knock out a significant percentage of species. At the low end of global warming scenarios, they estimate 18% becoming extinct, and at the high end the number is 35%.

Potent stuff. In terms of speed and magnitude, these results suggest that global warming alone could trigger mass extinctions on par with some of the all-time great catastrophes. In the news coverage of the research and in later comments, skeptics tried to mock it. One line of mockery took on the science of predicting extinctions. The best example of this comes from Gregg Easterbrook, who called the research "nonsense." His blog was promptly picked up approvingly on the blog of the conservative magazine Reason, and will no doubt continue to circulate in such quarters.

(A note on Easterbrook, which you are welcome to skip: In the past, I have pointed out how his understanding of evolution can get downright foolish. This point doesn’t bear more repeating. I’m revisiting him now only because he offers a common sort of poorly reasoned "skepticism" directed at the science of extinction. Granted, Easterbrook accepts the reality of global warming and claims to be concerned about the potential for extinctions in the future. But his criticisms of this particular research represents a pretty widespread argument that needs to be challenged.)

COMPUTER MODELS: First, Easterbrook complains that the study is, in his words, "entirely a computer simulation." He claims that "as anyone familiar with this art knows, computer models can be trained to produce any desired result….Computer models are also notorious for becoming more unreliable the farther out they project, as estimates get multiplied by estimates, and then the result is treated as specific. This is a 50-year projection, and everything beyond the first few years should be treated as meaningless statistically, given that tiny alterations in initial assumptions can lead to huge swings at the end of a 50-year simulation. Nature is a refereed journal, but it appears that all the peer-reviewers did was check to make sure the results presented corresponded to what happened when the computer models were run. There does not appear to have been any peer-review of whether the underlying assumptions make sense."

This sort of complaint only makes sense if you haven’t bothered to read the paper and become familiar with some of the referenced papers on which it is built. The species-area relationship is an iron-clad rule in ecology that’s been tested and retested many times. The concept of climate envelopes has been tested as well; it has proven its mettle by allowing scientists to accurately predict how species shifted their ranges as Ice Ages altered past climates. Moreover, the researchers–well aware of the uncertainty that can plague these sorts of studies–came up with extinction estimates with three different methods for analyzing the loss of habitat. They got the same results with all three tests, which is evidence that while the estimates are not precise (and aren’t claimed to be) they are robust. So Easterbrook’s claims about the study being sensitive to tiny changes in the underlying assumptions doesn’t hold up.

COUNTING SPECIES: Easterbrook and others often try to raise doubts about extinctions by confusing the numbers. He points out that the lead author, Chris Thomas, was quoted in the Washington Post as saying as many as 1.25 million species could go extinct. Amazingly, it seems, Thomas is unaware that the International Union for the Conservation of Nature, which says that 12,259 speces are threatened. "The jump from 12,259 imperiled species to 1.25 million extinctions is a hundredfold increase!" Easterbrook claims. "The rate of species imperilment will rise in a short period to 100 times the current rate, and based solely on climate change? That sounds extremely implausible. In fact it sounds like cockamamie galimatias."

It sounds this way if you ignore the difference between the recorded numbers of total and threatened species and the estimate numbers. Scientists know that there are far more total species on Earth than have been recorded. All you need to do is go to a tree in some remote rainforest and pick off the insects, nematodes, and other critters living on it. Many will be new to science. Based on the proportion of new species scientists find each time they take a survey, they can estimate how many species await discovery. There’s a lot of debate over just how many species there are out there, but all good estimates are in the millions–in some cases in the tens of millions. Thomas’s study looked at the percentage of extinctions recorded in each habitat. Because the extinctions all follow a general pattern, he then calculated what 37%–his worst case estimate–means in terms of the actual number of species on Earth. If you assume 3.3 million species,, you get 1.25 million extinctions. But the fact is that Thomas was actually using one of the lowest estimates of the total number of species. If there are 15 million species, the figure would be 6 or 7 million.

As for the good folks at the IUCN, they have an even harder time than the scientists who look for new species. They need data on a species’s historical and current range, its historical and current population, the threats to its habitat, its life history, and many other pieces of information before they can decide that a species is threatened. This takes many years of field work, and so only a small fraction of species have gotten that sort of attention. It is no surprise that they only register 12,259 species as threatened. Certain groups of species are much better studied than others–birds, for example, have been the focus of naturalists for centuries, and most species have probably been found. If you look at these well-studied groups–among which most threatened and extinct species have been identified–you find a pretty consistent rate of extinctions. Stuart Pimm calculates it to be 100-1000 times higher than the typical "background" rate of extinctions in the fossil record.

RECENT EXTINCTIONS: Another common tactic for questioning extinction science is to claim that we should already have a long record of human-caused extinctions and don’t. Easterbrook asks why, for example, we haven’t seen a lot of extinctions in the past 20 years in the Pacific Northwest, a place that has experienced warming as well as habitat destruction and whose wildlife is very well studied by scientists. "For anything even remotely close to Thomas’s 1.25 million extinctions to be a hard number, we should already be seeing the bow wave in the form of dozens if not hundreds of extinctions in well-studied areas like the Pacific Northwest. Instead we see, um, zero."

I don’t see how Easterbrook gets to decide exactly what evidence in recent history is enough to falsify this study. Did he run the computer model with data from the Pacific Northwest over the past twenty years and come up with 20 extinctions? (I’m reminded of creationists who used to point out that there were no intermediate fossils between land mammals and whales. That proved evolution was false. Then, when a species of whales with feet was found, they claimed that the lack of a species in between this intermediate and true whales was proof that evolution was false. "Keep moving the goal posts" is the strategy.)

In setting up his goal posts, Easterbrook reveals many of the ways in which he misunderstands the study of extinctions. Different habitats experience different rates of extinctions. For one thing, different regions undergo different rises in temperature and shifts in rainfaill. For another, the most sensitive species are the ones with the smallest ranges. That’s why the deforestation of the northeastern US wiped out relatively few birds–because most have big ranges and could therefore survive. The Pacific Northwest is not a particularly major hotspot of these small-range species.

The biggest misunderstanding that Easterbrook and his fellow "skeptics" have is that extinctions happen overnight. In fact, extinctions take time. Thomas and his coauthors talk about 18-37% of species being "committed to extinction" by 2050–not actually extinct. Studies in African and Asian forests have shown that species don’t disappear immediately from fragments of forests. Instead, they may need as long as 50 years or more to finally give up the ghost. Any extinctions driven by global warming in the past couple decades might not come to pass for decades. That doesn’t mean, however, that animals and plants aren’t already responding to climate change with longer growing seasons, shifting ranges, earlier arrivals at breeding ground and so on. The wheels are in motion.

Easterbrook was hardly the only voice raised against the Nature study. In the Washington Post article, for example, we learn that "one skeptic, William O’Keefe, president of the George C. Marshall Institute, a conservative science policy organization, criticized the Nature study, saying that the research ‘ignored species’ ability to adapt to higher temperatures’ and assumed that technologies will not arise to reduce emissions."

United Press International gave more attention to O’Keefe’s claims–

"As with everything climate-related, however, the verdict is not unanimous. In a review of the literature on the subject published in July 2003, the George C. Marshall Institute in Washington, D.C., concluded: ‘The facts do not support claims of mass extinctions arising out of climate change. Whether through adaptation, acclimation or migration, available research suggests that the threats may be overstated.’

"The Marshall report, by Sherwood Idso, Craig Idso and Keith Idso, said to expect ‘a biosphere of increased species richness almost everywhere on Earth in response to the global warming and increase in atmospheric (carbon dioxide) concentrations of the past century and a half that have promoted a great expansion of species range throughout the entire world.’"

The Idso report does not appear in a scientific journal, but simply lingers on the George C. Marshall Institute web site. You can search for a reference to it in the major scientific journals on global change, but you will search in vain. Along the way you won’t find papers offering support for the report’s wide-eyed conclusion. How then does it–and the Marshall Institute–earn a place in reports on this new research? As Chris Mooney suggests, in the case of the Washington Post, a naive notion of how to get both sides of the story may be to blame. I suspect something else is going on in the UPI article. UPI is owned by the conservative Rev. Sun Young Moon, who also owns the Washington Times, which ran UPI’s story–along with many previous anti-global warming stories. The George C. Marshall Institute meanwhile has helped promote other sketchy research that claims to overturn predictions about global warming. (See David Appell’s article, for example.)

Adapations to global warming are certainly an important factor in how the natural world will respond to all our greenhouse gases. Animals and plants don’t simply keel over if things get a little warm. They have strategies encoded in their genomes to adapt in their lives to changing environments. They are plastic. And on top of that, over time they can evolve to become better adapted to a changing environment. Over the past 20 years, evolutionary biologists have documented rapid natural selection in the wild. In the face of intense fishing, for example, some salmon have evolved to be 25% smaller in just a few decades. If global warming emerges as predicted, it will be a particularly powerful selective force, raising temperatures and altering climate at an astonishing clip compared to past climate change. Life is already adapting to the warming world. In the Proceedings of the Royal Society of London, Canadian researchers report that red squirrels in the Yukon have responded to the warming climate by breeding 18 days earlier than they did in 1989. Part of this shift was the result of changing genes.

It’s possible that rapid evolution will let some species avoid extinction by adapting to the new climate. But some species may find their evolutionary path blocked. Julie Etterson and Ruth Shaw of the University of Minnesota studied the potential evolution of partridge peas in Minnesota. Adaptations to warmer, drier climates, they found, won’t come for free. Those changes will interfere with how the pea plants grow, which will lower their overall fitness. As a result, the slow evolution of the plants will lag behind the climate change, and may leave them badly suited for survival. We therefore can’t assume that natural selection will save us from the risks of climate change.

Climate change, of course, is nothing new on Earth, and so the fossil record can offer some clues to the balance of adaptation and extinction in the coming century. The Idsos imagine that a few decades of high levels of carbon dioxide and elevated temperatures can whisk us back millions of years, merging forests into "super forest ecosystems" far more diverse than anything that exists today. It’s actually possible to do something beyond blithe hand waving of this sort with the fossil record, though. In a recent issue of The Journal of Mammalogy, for example, paleontologists looked at how the diversity of North American mammals responded to different rates of climate change. So far, they conclude, the changes that have happened so far are within the normal variability of mammal history. But within a few decades, global warming will transform the community of mammals beyond anything seen in the past 60 million years, causing widespread extinctions.

But the fossil record is missing some crucial elements of today’s world. Even if an animal or plant might be able to spread into a new climate envelope, we humans may block their path. It’s hard to imagine how the trees, fungi, insects, reptiles, and other species could move out of a rain forest preserve and into a farming region or a swath of industrial properties.

Despite the speed with which global warming appears to be occurring, we still have a lot to learn about how many extinctions it will cause. There is indeed a lot of uncertainty and plenty of room for debate. But we don’t have time to waste with the distraction from the likes of Easterbrook and the Marshall Institute.