Wevets, thanks for your post. There’s an important lesson in this one for you.
Before you bust somebody for doing something scientifically stupid, make sure that most scientists aren’t currently doing it that same way …
This method of comparing extinction rates is the standard way to do things. I was not explaining my own eccentric method. I was laying out how it is usually done, for people like yourself who might not be aware of some of the hidden pitfalls of comparing extinction rates.
I gave you the Regan citation. Take a look at that. They use EXACTLY THE SAME METHOD. Why? Well, I tried to explain that as well. Since we’re fighting ignorance here, I guess I lost that fight, but being a battler, I’ll go another round, here goes …
Comparing ancient extinction rates with modern ones cannot be done directly. This is because they look at different things. One looks at how many modern species have been reported by man as going extinct. The other one looks at the fossil record, and determines how many of the species that made it into the fossil record went extinct.
These, of course, are very different things being measured, apples and oranges. We would expect direct observation to find modern extinctions which will most probably never make it into the fossil record.
In order to compare them, we need to consider how many of the modern species will make it into the fossil record, and how many of those that do make it will be recognized from their fossils as being separate species. In other words, we need to estimate what the fossil record of our era will look like, and only then can we compare it with the fossil record of previous eras.
I may not have learned much on the cattle ranch, but I know prime beef from fertilizer. Or as we used to say on the ranch, “You can piss on my boots, Wevets … but you can’t convince me it’s raining …”
I won’t claim to know what backs this up, but the author of the most authoritative textbook on evolution, Douglas Futuyma, makes the following statement (Evolution, Box 7A, page 158):
“None of the mass extinctions of the past was caused by the actions of a single species. But that is what is happening today. Within the next few centuries, the diversity of life will almost certainly plummet at a greater pace than ever before [emphasis mine ].”
spazurek, thank you for the quotation and the citation.
Unfortunately, all we have there is another prediction of doom. While predictions of doom are an interesting field in themselves, without data to back them up they are just some guy’s best guess. Did Douglas Futuyma say why he believes that the extinction rate is likely to increase? Because if he believes it on the basis of the extinctions-area “relationship” postulated by E. O. Wilson, then we’re right back where we started.
Also, I get very nervous whenever a scientist starts saying things like “a greater pace than ever before” and “almost certainly”. Neither of those are scientific statements. I’m not quite sure how he might even know such a thing. By all accounts the asteroid impact was pretty quick, as was the PETM. However, the resolution of the fossil record doesn’t allow us to examine time periods as short as the 500 years that we are discussing here. And how is he so certain about the future trajectory of extinction?
Next, the peak extinction rates for birds and mammals were reached during the great period of European exploration. These were from the species which were introduced into Australia and the islands, which until that time had had only infrequent or no contact with outside species. This wave of extinctions peaked in about 1890, with about 1 species per year of birds and 0.6 species per year of mammals going extinct globally at that time. Ever since then, extinction rates have been decreasing. So if we’re going to extinguish species at “a greater pace than ever before”, that may well be the benchmark we have to beat. However, if that’s the case we’re falling down on the job, there has been nowhere near that rate of extinctions since the 1890s.
Why? Well, because by that point both Australia and most of the once-isolated island had seen the advent of the majority of the more dangerous modern introduced species. There are very, very few places left that have not suffered the depredations of imported dogs, foxes, mongoose, rats, human hunters, goats, pigs, cats, sparrows, starlings, and plants, plus the usual assortment of human, animal, and plant diseases.
And since extinctions from imported species peaked in 1890, and there’s very few untouched lands left out there, we will never see again the wave of extinctions that washed over the islands and Australia in the last few centuries. It is not widely recognized that for both birds and mammals, these extinctions in the islands and Australia, with only ~10% of the world’s habitable land mass, represent ~ 95% of all recorded extinctions, with only ~ 5% on the rest of the continents. Without those, we’ll have a very hard time even beating the record from 125 years ago, much less be the best in all of geological time.
Finally, my guess is that in the modern era (last 500 years), more birds and mammals have been driven extinct by rats than by man, with humans serving merely as the rat’s unwitting and unwilling vector … but like I say, that’s just a guess.
If I may impose on your running of the numbers, what are the figures like without deliberate conservation attempts, such as white lions and tigers- ie, assume those species are extinct (since human intervention is what preserves them).
No, I absolutely I do not believe that science=certaintly.
Science to me is notable precisely because it does include a numerical estimate of uncertainty. When you see numbers like “3 ± 0.18 metres/year”, you are in the region of science.
When you see claims like “a greater pace than ever before” or “almost certainly”, presented without numerical estimates of the uncertainty of those claims, you’ve left science and are wandering around in some other region.
So no, I’d agree with you that it is “quite the opposite”, that science = uncertainty … plus or minus a little bit, of course.
BrainFireBob, thanks for an interesting question. I had not looked at that.
The Red List has four mammals on the continents (excluding Australia and the islands) which are categorized as “extinct in the wild”, which I believe is what you are thinking of.
However, the CREO lists all four of them (Black Footed Ferret, Wild Horse, Saudi Gazelle, Scimitar-Horned Oryx) as “Currently extant under valid species name”, so they’re not even separate species. So there are no continental (excluding Australia and the islands) mammals which have gone extinct in the wild. OK, now I’ll check the birds.
… back again to report that the Red List shows one bird species on the continents (excluding Australia and the islands) which is extinct in the wild. This is the Alagoas Currasow, Mitu mitu. The Red List says “The extinction of this species was forecast almost as long ago as its discovery.” Why? Because it’s big (up to 90 cm), doesn’t fly that well, and is quite tasty … a very bad combination which was recognized even three hundred years ago.
Now, why are your white tigers and lions not on the list? Well, it is because they are not separate species, which is the subject under discussion, the extinction of species. In fact, they are not even sub-species, they are just colour variations … Wikipedia says:
and
Since white tigers and lions are neither species or sub-species but merely an unusual coloration which occasionally occurs naturally, they will be around as long as there are tigers and lions.
As I have already noted, they use EXACTLY THE SAME METHOD, and REACH A VERY DIFFERENT CONCLUSION than you have. You are comparing presumed continental extinctions (and I have already noted my disagreement with your rationale for choosing such a low number for observed extinctions) with the background rate for all mammals. That is not, as you claim, comparing “apples to apples”.
Island species, and other species with narrow geographical ranges, are much more susceptible to extinction than more cosmipolitan species. Thus, it is expected to be very likely that among continental mammals, the background extinction rate is lower than the background rate seen for all mammals taken together. Thus, even comparing your presumed lower number of observed extinctions against a corrected background rate will likely result in a calculated higher-than-background rate.
Further, even if the rate of extinctions among continental mammals lies within the “no reason for concern” range, it cannot be reasonably denied that all the other mammals are going extinct at higher-than-background rates. Species may not be dropping at Wilsonian levels, but they are most certainly dropping at greater than background rates. Regan, et al, agree with this assessment, which makes your claim that the so-called “6th wave of extinction” is illusory rather puzzling. It may not be happening to the extent that Wilson argues, but to deny that it is happening at all is to ignore the evidence.
Further yet, preservation bias may well not factor into the equation at all. Fossil lineages are seen to have “lifespans”; mammal species in particular seldom persist more than a few million years in the fossil record, with variation of the mean being seen anywhere between 1.4 milion years and 7.5 milion years, depending on the mammal group and the epoch. These data can be converted to probability functions based solely on the members of a given fossil assemblage and the likelihood they will persist into the future (i.e, given this 1-million year interval, what percentage of the current assemblage will likely persist into the next million-year interval?). The mean “lifespan” for mammal species, then, is an average of an average, taken across all mammal groups and assemblages. This, then, is how the “background rate” is generally calculated. It is unnecessary to calculate the probability of fossilization for extant species in order to compare current rates with background rates. At most, one might need to account for the possibility that extant mammals exhibit an average longevity which differs from fossil species. The background rate is a statement that answers the question, “Based on the average longevity of fossil species, what percentage (or other metric) of extant species might we reasonably expect to go extinct within a given interval?” And there are more species going extinct today than would be expected based on species longevities (that is, they are “dying young”). It is thus natural to ask “why is this happening”, and “what, if anything, can we do about it?”
Please do elaborate on any methodology commonly used in the literature to estimate the perceived rate of extinction 1,000,000 years in the future. :dubious:
I haven’t read it yet, but Darwin’s Finch has observed that Regan came to very different conclusions than you came to - specifically, that current extinction rates are 36 to 78 times the background rate.
I also note that there have been studies of forest birds in North America that came to very different conclusions than you have come to:
From: Pimm and Askins. 1995. Forest losses predict bird extinctions in eastern North America. Proceedings of the National Academy of Sciences of the United States of America, Vol. 92, No. 20. (Emphasis mine)
Interestingly, I think the authors have engaged in a little bit of the same special pleading that you have, but in the opposite direction.
However, you have made some errors they did not, including attempting to estimate estimates of extinction rates 1,000,000 years in the future, denying that the Carolina parakeet was a forest bird, ignoring examples that run contrary to your idea.
For example, you cite the CREO database related to bird and mammal extinctions, yet:
(Bolding mine.) Do you see the difficulty in citing them related to bird extinctions?
Nope, for that you have to be checking your head, not your boots.
Darwin’s Finch, I appreciate your response. You seem to believe that the record of fossil mammal extinctions is taken “across all mammal groups and assemblages”, and that it represents the “background rate for all mammals”. It is not, and it does not. It is only and solely taken across those mammal groups and assemblages that happened to make it into the fossil record.
On the order of 250,000 fossil species have been named and described, out of an estimate of 5 to 50 billion species that have existed. This means that only about one in 20,000 species has made it into the fossil record.
It follows that the species which made it into the record were those that were:
• abundant
• geographically widespread
• located in areas conducive to fossilization (shallow inland seas, areas of silt and sedimentation, places with slow-flowing water, and the like)
In other words, abundant continental species will be much, much more common in the fossil record than tiny localized island species. Or to quote from Regan et al. (previously cited),
The result is that the background extinction rate is, to at least a first approximation, the continental extinction rate. It is not, as you claim, a record “across all mammal groups and assemblages”.
Yes, as Regan found, including all of the island extinctions, our current extinction rate over the last 500 years is significantly higher than the background (continental) fossil extinction rate. But the current (continental) extinction rate is not statistically different from the background (continental) fossil extinction rate.
Does this mean there is nothing to worry about? Absolutely not. As I said before, extinction is always a concern, even local (as opposed to species) extinction.
My point, however, is simple. There is nothing in the record of extinctions that supports the idea that species are going extinct at the rate of 27,000 per year as E. O. Wilson has claimed, or at even a tenth of that rate. Extinction rates have dropped since about 1900, not risen as they would have if Wilson’s claims were true.
No thanks, do your own homework. I have detailed the methodology, given you my exact numbers, and pointed you to the the scientific reference. Your reply was:
You haven’t read the reference I cited, but you want me to provide more references? No thanks …
Pimm has not examined the theory posed by E. O. Wilson, who claimed that species/area relationships of extinctions affected all species. Pimm says that is not true, it only affects some much, much smaller number of species. While that is possible, it strongly supports my point, which is that there is no evidence that E. O. Wilson’s theory is correct.
I don’t deny that the Carolina parakeet was a forest bird. The Red List does. You don’t like that? Sorry, take it up with the Red List folks, tell them that wevets says they’re wrong, I’m sure they’ll fix their error right away …
And until you read the Regan paper, you won’t be able to appreciate the true stupidity of your statement that I have made an error by estimating what the fossil record of our time will look like in the future. Do your homework. For example, Regan says:
Using “paleontologic metrics” means we have to estimate what the fossil extinction rate of the current era will look like in a million years or so, so that we are measuring apples vs. apples.
I did not use them for bird extinctions, only for mammal extinctions … point me to one bird that I said was mentioned in CREO.
You don’t need to prove it. The proof is required “on the other foot” to misuse a phrase.
Extintions occured before humans due to natural processes. Extintions still take place. Claiming that humans have superceded all other processes is an extraordinary claim that requires extraordinary evidence, to quote Carl Sagan.
All major mammal groups (by which I mean Linnaean families, orders, etc.) extant today are represented in the fossil record. All extant species, however, are not necessarily thus represented, and I never claimed otherwise. The background rate is an average, it is not a representation of anything.
For someone who criticized Futuyma’s use of “almost certainly”, you have a curious habit of making “estimates” based on no real scientific data. Nevertheless, for the purposes of this discussion, it matters not one whit how many potential species there may have been. The background extinction rate is based on species longevities for known fossil lineages. The extant mammals today are members of those same lineages. If the fossil record shows that fossil members of Carnivora have an average longevity of 740,000 years during the Pleistocene, then we can reasonably expect that extant members of Carnivora will exhibit similar longevities. This is an average among members of Carniovora, so we should necessarily expect some divergence form that mean value. The rate for all mammals is an average across the various Orders.
More “almost certainly” from you. The background rate is exactly as I have stated: an average “across all mammal groups and assemblages”. Nowhere did I claim that it was a representation of the rate for all mammals from all biogeographical provinces, which is what you are mysteriously arguing against here.
However, even if we accept that the overall mammalian background rate is effectively the “continental” rate, there are still numerous biogeographical zones to be found on continents. Yet you only include recently-extinct forest-dwellers in your calculations, which you then compare against the background rate to claim that current rates are comparable to the background rate. You basically have two choices when making your calculations: include all documented continental extinctions (of which the IUCN lists 27 functionally extinct, 23 of which are documented to be extinct and another 4 of which are extinct in the wild) in your calculations, which you can then compare to the overall mammalian background rate; OR correct the background rate for forest regimes only, and compare the extinction rate for modern forest dwellers to the fossil rates. You have opted for neither of these, and instead have chosen to compare apples to oranges while trying to pass off those oranges as thick-skinned apples. Even if you wish to argue that fully half of those 23 extinct-in-the-wild species are invalid because of taxonomic vagueries, that still makes your calculations of current rates off by a full order of magnitude when compared against the overall mammalian background rate.
See above. You are not comparing continental extinctions against the background rate, you are comparing continental forest extinctions against the background rate. As I have already pointed out, rates will vary within and between biogeographical zones. The general claims for purposes of this thread is that species are disappearing at higher rates now than in the past (“in the past” meaning relative to fossil rates).
And, again, you are arguing against yourself here, as no-one in this thread had brought up Wilson’s claim before you did. This thread was not “Was Wilson right?”, but “Are species disappearing at ‘natural’ rates, or are they being ‘helped along’ by Man?”
I disagree, for reasons already mentioned. It is not necessary to predict the state of the Fossil Record…of the Future! in order to evaluate current rates. Indeed, as any good scientist would - or should - tell you, invoking speculative variables can only increase the margin of error in any calculations which include such.
Once again: the background rate is not based on the number of species which have ever existed, nor is it influenced by the whys and wherefores of how fossils were formed; it is based on the longevities of lineages found in the record. Unless you wish to argue that lineage longevities for extant (or recently extinct) species are markedly different from those in the past (in which case, one might just as well eschew any comparison to background rates at all), then all that’s necessary is a “simple” actuarial study, not a baseless prediction of how things might look in the fossil record millions of years from now. It was unnecessary when Regan did it, and it’s unnecessary when you did it.
Darwin’s Finch, thank you for an interesting post. I see that I still have not explained myself clearly.
As I mentioned before, it is necessary to compare apples with apples. We cannot directly compare the number of extinctions found in the fossil record with the number of extinctions observed by modern scientists, because they are looking at very different things. We would expect modern scientists to find many more extinctions in a given period than will ever make it into the fossil record of that period.
We do not, as you claim, need to “predict the state of the Fossil Record…of the Future!”. Neither I nor Regan et al. make that claim. What we need is to estimate how many of the currently observed extinctions will eventually make it into the fossil record. Then, and only then, can we compare that number (estimated present fossil record of extinctions) to the fossil record of previous extinctions. Then, and only then, will we be comparing apples with apples, fossil record with fossil record.
You say that the fossil extinction rate “is not based on the number of species which have ever existed, nor is it influenced by the whys and wherefores of how fossils were formed; it is based on the longevities of lineages found in the record.”
However, we need to note that, as you state, this is not all lineages. The estimate is based only on lineages found in the fossil record. These are, by and large, successful lineages that had large populations and widespread ranges. Lineages with tiny populations and small ranges are generally not included in the fossil record. Thus, the fossil extinction record will be biased towards the extinction rates of large population, successful species, which by definition will go extinct more slowly than small, limited range, generally unsuccessful species.
You, however, want to apply a fossil record that you admit does not contain all lineages, to all lineages, large or small populations, widespread or extremely local, not just those that are (or will be) found in the fossil record. That’s the problem.
You say: “Indeed, as any good scientist would - or should - tell you, invoking speculative variables can only increase the margin of error in any calculations which include such.” This is true. However, leaving out real variables does an even worse job of it, and that is what you are advising.
Why should a fossil extinction record comprised in the main of successful, widespread species give us an accurate estimate of the extinction record of tiny, local species? These species will have a much higher extinction rate, but they won’t be found in the fossil record. This virtually guarantees that fossil extinction rates will underestimate the true extinction rates, since the fossils being measured are mostly the very successful species.
In other words, the idea that we can just do a “‘simple’ actuarial study”, while tempting, is simplistic. While you are free to claim that both Regan and I are wrong, the number of times her article (“The currency and tempo of extinction”) has been cited (more than thirty citations in Google Scholar) indicates that yours is distinctly a minority view.
Hi intention – sorry for the extended delay to my response, but I just got married, so I’ve had a lot on my plate recently.
Getting right along to your point – I think that the Regan paper you cite does not actually provide evidence for your assertions (indeed, I agree with Darwin’s Finch that the Regan paper contradicts your conclusion.)
I think you’re misreading me here – not only am I saying that most scientists don’t do it that way, I’m saying that even Regan (who you cited) is not doing it that same way you do it.
Let’s examine your method side-by-side with Regan’s:
Full citation = Regan et al. 2001. The Currency and Tempo of Extinction. The American Naturalist, Vol. 157 No. 1 pp. 1-10. If you would like a PDF, I would be happy to send one.
Looking at just the equations, it looks to me like a unit change – Regan is adjusting a rate over a 400 year observation into a rate over a 1,000,000 year observation.
I disagree – Regan is not forecasting into the future here, she is putting a per 400 year rate into a per 1,000,000 year rate – merely a shift in units, not an actual prediction.
I can see how you would come to that conclusion, because she does say:
The problem here is that you are taking her statements to mean “what would be measured one million years from now” which is specifically not what she says. Her key words to remember here are “rate”, “per million years”, and “assuming the current rate continues” – the real giveaway occurs when she discusses how the rate has changed:
Intention, you may want to contemplate thinking critically about what you read: If Regan is arguing that extinction rates have changed rapidly, how can she also be arguing that you can extrapolate a constant rate from a 400 year sample period to predict a 1,000,000 year rate? Those two positions are inconsistent with each other.
There are three or four more important ways Regan’s method differs from yours, so let’s look at those too.
First, there’s no need to shout. Second, she does not use the method you use.
Here are the major differences:
Regan compares morphological species to morphological species (apples to apples), you compare morphological species to genetic species (apples to mangoes).
You use an unreliable method to try to tease out “forest species” (without defining “forest”) while Regan compares all species without trying to distinguish between habitats.
You exclude islands and Australia while Regan does not. Your island count is likely as unreliable as your forest count.
You assume that specifically selected, non-random subsamples represent the whole, Regan does not.
Number 1:
There’s your problem – you apply a definition to current extinction records that cannot be applied to paleontological records. You’re no longer comparing apples to apples, you’re comparing apples to mangoes.
Some background for you: Biologists use different species concepts to define the word “species.” In the case of CREO, as your quote above states, they are defining species based on genetic characteristics (a genetic species concept.) Paleontologists don’t define species that way because, despite what Jurassic Park tells us, DNA from fossils is rare and far between. Instead, paleontologists use a morphological species concept or “morphospecies” to define species.
Regan knew this and was then able to compare modern morphologically defined species with fossil morphologically defined species. You have compared modern genetically defined species with fossil morphologically defined species; so you have not used the same method she did.
What are the effects? Well, Blake mentioned the extinct Auroch (Bos primigenius) which was given a different species name than the cow (Bos taurus) because they have different features (they’re different species under the morphological species concept.) DNA testing results place them together as the same species under the genetic species concept. You go from 2 modern extinct morphological species to 1. However, if you have 2 extinct paleontological species that would be the same genetic species, you cannot apply the same test, so you still have 2 extinct species in the fossil record.
This technique of comparing apples to mangoes allows you to reduce the number of modern extinctions without reducing the number of ancient extinctions, biasing the results in favor of a lower modern extinction rate vs. fossil extinctions, but it’s just a numbers game – consistency demands that morphological species be compared to morphological species, not genetic species.
Number 2:
You use the IUCN Red List to try to tease out “Forest” mammals and birds only.
I think the mistake here is not reading the introduction for the Red List:
A system designed for determination of relative risk of extinction may not include comprehensive or consistent information on species that have already become extinct, and a search bears this out. Take the famous case of the Carolina Parakeet, one of the better-documented extinctions.
A search for extinct forest parakeets turns up zilch.
However, the Red List itself says the Carolina Parakeet lived in cypress and sycamore trees, and only records its habitat as “Terrestrial”. (see http://www.iucnredlist.org/search/details.php/5268/sum)
So it’s easy to see how your search method has missed extinct birds and mammals – you’re using the database to search for something it was not designed to do and does not contain enough information to do. If you’re interested in extant, but threatened, animals, perhaps it’s a better resource. But no database can do everything, and treating one as definitive just because you’ve used it for a search that it wasn’t designed to accommodate doesn’t produce reliable results.
Later on you modify your results:
But as far as I can tell, you’re still using the same method, searching the Red List (and CREO) database for extinct animals limited by factors the Red List doesn’t have for all its extinct animals. If the method didn’t work, maybe it’s time to select a new method?
The technique of selecting just one habitat type to compare fossil extinctions to modern extinctions allows considerable freedom to play with the numbers – you can just classify an animal that doesn’t show up in the list as “not a forest animal” and – poof – your number of modern forest extinctions decreases, but your number of fossil extinctions doesn’t.
Number 3:
You are excluding islands and Australia vs. including continental mammals and birds.
The technique of saying “I’m not going to count islands” means you can eliminate 34 modern mammalian extinctions while not eliminating any fossil extinctions, allowing systematic bias in favor of decreasing the modern extinction rate relative to fossil extinction rates.
The technique of saying “I’m not going to count Australia” means you can eliminate about 18 modern mammalian extinctions, while it’s not clear from your method that you eliminated any Australian fossil extinction. Since you cite rates from Foote (1997), and he did not eliminate fossil Australian mammals from his calculations, this leads to systematic bias in favor of fossil extinction rates being estimated higher than modern extinction rates.
It also just doesn’t make a lot of sense – islands are disproportionately small in land area and disproportionately large in terms of species numbers.
You also use the same method to find island mammals that you did for forests – see above.
Number 4:
You engage in selective subsampling, and treat it as if it were random subsampling. Regan engages in selective subsampling, and treats it like selective subsampling.
Your math here depends on the assumption that extinction rates are consistent across widely different taxa, which your own source (Regan) says is not true. By contrast, note how Regan in her discussion is careful to state that her calculations based on mammal numbers are applicable to “mammalian extinctions” (page 9) and her separate discussion of Australian angiosperm extinction rates (pages 7 and 8.)
It’s a basic rule of statistics that in order to obtain a representative sample, your sample must be random. Let’s take Regan’s citations on total species:
If there are as many as 4,760 (a high estimate) mammal species, that would be 0.16% of the total number of species if there are 3 million, and 0.016% of the total number of species if there are 30 million. It would be inappropriate to take a systematic sample so small and treat it as if it were representative of the whole.
To sum up: You’ve introduced three sources of systematic error that are not present in Regan’s work.
You compared apples to mangoes (morphospecies to genetic species.)
You use an incomplete method to tease out forest species.
You selectively exclude areas with significant extinctions (islands and Australia).
You have also engaged in some misunderstandings of Regan’s work:
You think she is forecasting when she is adjusting units or measurement.
You treat nonrepresentative samples as representative of the whole while Regan treats her nonrepresentative samples as nonrepresentative samples.
It doesn’t look to me like you’re using Regan’s method here at all, and you may not have understood her paper as well as you believe.
We’re overdue an asteroid strike,I suspect that will outdo us on the extinction front.
Perhaps if we actually averted this disaster by our own actions this might in some way atone for the extinctions we HAVE caused but we 'll have to start working on the solution in a practical way as opposed to theorising about it right now as when we spot the asteroid with our name on it it will be far too late to put the technology in place.
