16

I have heard that Earth is experiencing another mass extinction event, with many species going extinct at a rate comparable to Earth's many other past extinction events.

This has been reported in the news. The BBC reports the claim like this:

The Earth has entered a new period of extinction, a study by three US universities has concluded, and humans could be among the first casualties.

The report, led by the universities of Stanford, Princeton and Berkeley, said vertebrates were disappearing at a rate 114 times faster than normal.

Is the Earth experiencing a mass extinction as reported?

  • I've limited the question to the original claim and reopened. – Sklivvz Jun 22 '15 at 17:41
  • 8
    What would convince you one way or the other? A three-university study is generally pretty convincing, and it agrees with a past study, according to the article you cite. – HDE 226868 Jun 22 '15 at 18:38
  • @Sklivvz It might be right to keep the question short, but I think you should have kept the reasons why the claim is not just obvious. There are some reasons to be skeptical. – matt_black Jun 22 '15 at 19:15
  • 1
    @matt_black that would only add bias to the question. Let the facts speak for themselves. – Sklivvz Jun 22 '15 at 19:17
  • 5
    @matt_black: it demonstrates that we made this question better by making it answerable. I have no opinion about the answer, as a moderator I notice that it's referenced and adds value. You seem to disagree with the conclusions of it. This is fine, just downvote it. If you feel even more strongly, provide an alternative answer. That's how stack exchange works, we privilege answers. We discourage debate. Disagreeing in a comment (or worse, in an edit on a question) is not the correct way to use the site. – Sklivvz Jun 22 '15 at 23:09
23

tl;dr: Yes, the paper provides strong evidence that we're undergoing a mass extinction event. The media reporting is basically right (for once).

What you are asking

I'm going to assume that what you're asking amounts to: "Is there any obvious flaw in the methodology of the recent study on this topic, that the (notoriously poor) mainstream science journalists have overlooked?". The reason I'm making this assumption is because I think that, if the claims of the study are as reported, the claim mentioned in the question must follow. If for some reason you're not convinced by the scientific method in general, then I doubt we can help you here.

The actual findings of the study

Okay, so first let's look at the study, not the bbc news article. It's authored by reasonable looking academics, at well known and important institutions. It's published in a good place. Because of these two things, I'm inclined to take whatever they actually say in the study at face value, so let's look at what it says.

If you read the study, you find that the authors define a mass extinction to be a time period when the rate at which species go extinct is much higher than the "background rate". The rate that species go extinct is expressed in somewhat funny sounding units, but basically it's the fraction of species that go extinct each year. That is, the number of species that went extinct, divided by the total number of living species at the start of that year. This lets us sensibly compare periods of high and low biodiversity.

The authors use what is described as a "very conservative" estimate for how often species go extinct during normal life (i.e. not during a mass extinction event). This estimate is arrived at by rounding up the latest best estimate of how often mammals have gone extinct (derived from the fossil record), and then assuming that all types of organisms go extinct that fast, even though we have good reasons to think that this is untrue (apparently invertebrate species live longer?). The authors overall find that for every 1,000,000 species you have alive at the start of a year, about 2 will be extinct by the end of the year. This means that a typical (mammalian) species will live for about 500,000 years.

The authors then look at the records of species we know about, and count how many of those have become extinct in the last century. They find that far more than expected have become extinct. About 8-20 times as many using the "very conservative estimate", and more like 50-100 times as many using a less conservative estimate. This handy graph gives you a pretty good image of what they find. The dashed "Background" line at the bottom shows how many species should have gone extinct, while the other lines show how many actually did.

Criticisms

Okay, so what can we conclude from the paper? Well, taken at face value, it looks like there's a mass extinction in progress. Some arguments against this conclusion might be:

  • The paper doesn't actually offer a statistical analysis of this phenomenon. I think this is a little sloppy, even though the effect is so large that it almost doesn't matter. Basically they need to look at the variance in the baseline extinction rate, and give us a range of normal values. However, I rather suspect that the range won't be 10-20 times larger than the mean value.

  • The paper relies on observed extinctions. It might be the case that lots of species go extinct all the time, and we just haven't gotten very good at keeping track of things like that until now. I think this is a valid concern. It could be that there are massive and systematic biases in the fossil record, and that species are dying off in large numbers all the time. However, the shape of the curve the authors present suggests otherwise. If species were dying all the time, then certainly after about the 1850s, when citizen science got big and people were actually doing a good job of keeping track of numbers of different species, we should expect the extinction rate to be basically flat. It shouldn't be increasing if this is the usual rate stuff dies at.

EDIT:

  • In the comments, @matt_black and others raise a criticism of the paper on the basis of its authorship. I didn't realize this was a controversial issue when I started writing the answer, so I didn't investigate it then, but have now.

In particular, the second author, Paul Ehrlich, has made many widely publicized and incorrect predictions. Most notable was the prediction that hundreds of millions of Indians would die of starvation during the 1970's and 1980s (Counterpoint: Ehrlich is quoted in the given Wikipedia link in the 1990's claiming that the statement in the book was an example, not a prediction, and that the core message of the book was still basically correct). As such, some people would be inclined to discount the paper's findings because its author has a history of publishing somewhat sensationalist and incorrect predictions. As a counterpoint, we can note that this is a logical fallacy: if the paper's methodology is sound, then it doesn't matter who is making the claims. However, going a step further, the paper's opening sentences are:

The loss of biodiversity is one of the most critical current environmental problems, threatening valuable ecosystem services and human well-being (1–7). A growing body of evidence indicates that current species extinction rates are higher than the pre-human background rate (8–15)

Examining the cited references, we find that they include a number of papers that I would characterize as having a high, apparently positive, impact on the scientific community. For the first claim (That the loss of biodiversity is a big problem), we see this paper, which boasts nearly 600 citations over 12 years. I did not look at every instance of it being cited, but a sample of 10 citing papers I selected at random contained no citations I would characterize as negative, being roughly evenly split being those I would characterize as endorsing the findings and those I would characterize as using the manuscript to support specific factual claims in their own paper.

For the second claim, the most cited of the 7 listed papers is this one, which I would again characterize as having an extremely high impact on the scientific community, which appears to be largely positive. The paper's opening lines make the claim that the extinction rate is 100 to 1,000 times larger than the estimated baseline rate, which is actually a much more outlandish claim than the one made in the paper we're discussing in this question (about an order of magnitude larger). The paper has more than 1800 citations over the last 20 years. Because the paper is older, and scientific consensus changes over time, I looked at the first two pages of google scholar's list of papers that cited this one in the last 2 years, and checked all the papers that were not paywalled. Included among them is a second study by some of the same authors from the last two years, which again claims (without citation) an extinction rate 1,000 times the baseline. So at least we know the original authors still endorse this view. Of the remainder, I find 4 that restate the key claim (100-1,000 times the baseline rate) with citation, 5 that cite it to support a somewhat watered down version of the claim (diversity is declining "fast"; Humans are "dramatically altering biodiversity around the globe"; Human actions have "generally negative" impact on biodiversity; Ample evidence that extinction rates are increasing), 3 that cite the paper in support of another, more minor claim (diversity is decreasing in specific regions; certain factors making extinction more likely; reasons for certain fauna being rare in Britain; there are lots of species of fungus we don't know about yet), 1 that does not cite the paper at all, 1 that is the paper we are discussing, and 2 self citations.

To summarize, I found 9 papers citing the core claim as fact (with more or less hedging of the wording); 3 citing it for other reasons, and 3 miscounts by Google Scholar. Overall, while my sample of 15 papers is small, I'd characterize this as a claim that is probably broadly accepted. If there is controversy, it seems to be well hidden.

  • 2
    "It's authored by reasonable looking academics, at well known and important institutions." One the authors is Paul Ehrlich. – Andrew Grimm Jun 22 '15 at 22:03
  • 4
    @John be careful about the use of metrics - problematic scientists often have better metrics than good ones. – Andrew Grimm Jun 22 '15 at 22:22
  • 4
    To be clear: I'm using reputation as a proxy for "didn't just make these results up", "didn't go fishing for p-values" and the like, which is a depressingly common practice when looking at scientific papers on controversial topics. It does seem possible that the interpretation of the results might be exaggerated by the authors, but even this doesn't seem very obvious to me. – John Doucette Jun 22 '15 at 22:26
  • 2
    The paper isn't that long or that technical. People should be making methodological criticisms rather than essentially saying stuff along the lines of "one of the authors published some screwy pop culture books back in the day". That doesn't invalidate the paper. – KAI Jun 22 '15 at 23:12
  • 2
    Even better, we should be reporting what the scientific community thinks of the paper. This one is a bit too young to have citations, but I understand there are other, older, papers making similar claims. Have they been approvingly or disapprovingly cited? – Oddthinking Jun 23 '15 at 0:49

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .