Arguing over Extinction

A paper published in Nature two weeks ago sent scientists and press outlets reeling. Have scientists overestimated recent extinction rates due to habitat loss? According to the authors of the study, in a way, they have.

The way they have is a mathematical equation to calculate species loss called the species area relationship, or SAR.

Peter Roopnarine, the Academy’s curator of paleontology and geology defined SAR for me via email:

A species area relationship describes, for a given area or region, the number of species that will be encountered if you sample a part of the region or area. The number of species levels off as you increase the sampled portion, implying that there is a reasonably fixed, or average number of species to be found in a given area. If you increase the area, you will discover new species at a rate described by the relationship.

But Stephen Hubbell of UCLA and Fangliang He of Sun Yat-Sen University in China report in their paper that this overestimates extinction rates. In fact, Hubbell and his colleagues have known about this error for a while. Here he is quoted in the New York Times Dot Earth blog:

We realized the problem with the method 8 years ago, but we only figured out how to prove it mathematically last year, and we wanted a proof to make it completely clear why the problem was real.

Peter clarifies the reason the SAR is flawed:

The problem with this, as reported in the paper, is that this inversion assumes that species are distributed randomly in a given area. This is almost never true, and you will in fact tend to discover most individuals of a population very quickly. But discovering the last individual takes significantly more effort. What this all means then, is that is you estimate extinction rates by reversing the species area relationship, you ignore this last great effort required to lose the last individuals of the species, and you therefore overestimate the rate of extinction.

Their conclusion is that many published estimates of rates of ongoing extinction, and rates used for conservation and management purposes, are probably overestimates. They are very careful to point out that this does not mean that there isn't an extinction crisis happening now, because there certainly is. But perhaps we have more time to turn things around for many more species that we previously thought.

As Peter says, Hubbell is clear that even though the equation is wrong, this doesn’t mean extinction isn’t happening. He simply wants more thorough data, as quoted in Nature News:

I hope our paper is an argument for jump-starting efforts to conduct 'bio-blitzes' to get a detailed understanding of the distribution of life on Earth.

Bio-blitzes like the current Philippine Biodiversity Expedition the Academy is undertaking with its Filipino colleagues.  (You can also read more about the importance of counting species this way here.)

As you can imagine, press coverage of this study went many ways—slanting right, left and sideways. Sheril Kirshenbaum’s great blog in Wired can take you to each side and back. This also includes many scientists’ attacks on the study.

Peter even raised an issue in our correspondence:

One concern that I do have with the paper is that while extinction indeed does not occur until the last individual of a species has died, a species is generally doomed well before that point. First, there is functional extinction, when a species becomes rare enough that it ceases to have any meaningful interactions in its biological community or ecosystem. It simply no longer matters to predators, prey, parasites, etc. That has an impact on the ecosystem, even though the species is not extinct. Second, when a population declines to a certain level, the probability of extinction by random causes increases significantly. For example, an abundant species might not have to worry about seasonal storms, because mortality is never high enough to cause extinction. When rare, however, the probability is much greater. Another problem would be the increasing difficulty of reproduction, for example finding a mate. All these factors contribute to extinction in a way not really captured by their analysis.

There are many more arguments online, too. A commentary on KCET gives a good, long-term picture titled, “Endangered Species Habitat: Still A Big Deal”. Perhaps most of us can at least agree on that.

Do you? What are your thoughts? Please feel free to comment below.