If you follow the popular science press, or read the evolution blogs, you’ve probably heard that, according a new paper in PLos Biology by Rohland et al., scientists have added another species of elephant to the two we have already. Up to now we all knew about the African elephant (Loxodonta africana) and the Asian elephant (Elephas maximus). Besides being located on different continents, these two species differ in size (the Asian is smaller) and other morphological traits (Asians have much smaller ears and, unlike Africans, only the males are tusked).
Previously, though, the Africans were divided into two subspecies, L. africana africana, called the “savanna” or “bush” elephant, and L. africana cyclotis, the “forest elephant.” These subspecies differ in morphology (though not as strongly as the African vs. Asian species), ecology—well, at least location; they are, as their names indicate, found in different habitats, although there is some co-occurrence (and putative hybridization)—and social behavior. (There’s no strong evidence that the morphological differences between the subspecies have a genetic basis rather than being induced by different ecologies (I suspect they are genetic, especially if the differences persist when the subspecies are raised in the constant environment of a zoo). I’m a bit warier about social differences, which can originate and be passed on culturally rather than genetically.
A paper by Roca et al., published in Science in 2001, examined the mitochondrial DNA of these two subspecies and found that they were quite divergent—far more diverged genetically (and presumably in time, since DNA differences largely reflect the time since populations split from a common ancestor) than was previously suspected.
The new paper by Roland et al. extends the earlier study to both mitochondrial and nuclear DNA, and confirms the earlier finding that the savanna and forest elephants are anciently diverged—perhaps by as much as 2 to 7 million years. They are still each other’s closest relatives, though, and are both more distantly related to the Asian elephant, which diverged from its own closet relative, the extinct wooly mammoth (Mammuthus primigenius), at about the same time that the two African groups diverged from each other.
Based on the long divergence between the two African populations, and their morphological differences, Roland et al. suggest that the savanna and forest elephants should be recognized as different species, bringing the total living species to three.
The paper is online for free, and if you want a shorter summary of the results, Greg Laden’s blog has a good discussion. I would like, though, to make two points about these results.
First, although I don’t have a huge quarrel with designating these forms as new species, I’m not particularly keen on designating populations that are geographically isolated, and have some morphological differences, as new species based largely on divergence time. That’s because I adhere (for reasons I outline in the book Speciation that I co-wrote with Allen Orr) to the “biological species concept” (BSC), which recognizes populations as different species if they exhibit reproductive barriers between them that prevent hybridization when the two populations co-occur. I outline my reasons for adhering to the BSC in chapter 1 and the Appendix, based on its connection to what most evolutionists see as “the species problem”—the existence of discrete, objectively recognizable groups in areas where they co-occur.
Now the savanna and forest elephants are mostly isolated in space, but there is some evidence that they hybridize when they do co-occur, though I can’t find information about whether those hybrids are fertile. (If they aren’t, then the two groups have no possibility of exchanging genes and are definitely different biological species). But their hybridization where they co-occur makes their status as biological species questionable.
What about the ancient genetic divergence and morphological differences? Well, unless the morphological differences are correlated with reproductive barriers, they’re not good indicators of reproductive incompatibility and status as species under the BSC.
We all know that two populations instantly recognizable by different appearances aren’t necessarily isolated reproductively.
Take the different human populations (“races” or “ecotypes,” if you will) that were pretty isolated before modern transportation began moving their members about. You would have had no trouble telling apart populations from eastern Asia from those of the Yucatan from those of sub-Saharan Africa. Yet we don’t think of those human populations as members of different species. Why not? Because we know now that they’re all reproductively compatible with each other. Any human male can in principle produce a fertile child when mating with any human female, regardless of where they come from. Are the morphological differences between savanna and forest elephants greater than those between human “races”? Who knows?
What about divergence time, as indicated by genetic difference? That’s a bit dangerous, too, because while genetic difference between populations can indicate the time since they last shared a common ancestor, it doesn’t necessarily say anything about reproductive compatibility or the possibility of gene exchange. Reproductive incompatibility is probably a byproduct of divergent selection, and although two populations are anciently diverged, they didn’t necessarily experience divergent pressures of natural or sexual selection that would cause reproductive incompatibility as a byproduct. If their environments are largely identical, for instance, there may be few types of selection that would make the populations diverge in a way that could produce reproductive barriers.
As an example, the African helmeted guinea fowl can produce viable hybrids (I don’t know if they’re fertile) with the domestic chicken, yet those species are separated by at least 55 million years. That’s a divergence equivalent to that separating humans and lemurs, taxa which of course can’t produce hybrids! The point is that different groups of animals and plants diverge evolutionarily at different rates, depending on the environmental differences they experience, as well as on other factors. And this means that it’s dangerous to infer anything about reproductive compatibility from divergence time alone—at least when that diverence time is a few million years or less.
Now Roland et al. could have at least alluded to this problem, since most evolutionists do employ the BSC, but nowhere in their paper will you find a mention of reproductive barriers. Their view that these populations should be different species comes solely from divergence time (and perhaps consideration of their morphological differences). In fact, the title of their paper is “Genomic DNA sequences from mastodon and woolly mammoth reveal deep speciation of forest and savannah elephants.” Deep speciation? What they really mean is “deep genetic and temporal divergence” between the elephant groups. And had I written the paper, those are the words I would have used.
Modern systematics, should, I think, be informed by evolution, at least to the extent that if you classify geographically isolated populations as different species, you should have a reasonable inference that they wouldn’t be able to exchange genes if they lived in the same place. (This is not impossible in some groups. In flies, for example, if you cross two allopatric groups and find that their hybrids are inviable or sterile, you are nearly 100% certain that they are biological species). It’s my opinion—and I know that others differ—that modern systematics needs to be informed by information about reproductive compatibility. And indeed, some modern alpha taxonomists (that is, those who name and classify different species) already do this. Roland et al. not only didn’t do this, but don’t even mention the issue. (This is, by the way, an ancillary point and not a severe criticism of their very nice paper.)
I’m not going to quarrel with the idea that there are now two species of African elephants. It would have been nice, though, had there been some discussion of their reproductive compatibility, and of the status of the hybrids when those two “species” co-occur.
My second point is that in many cases, and maybe even this one, species are named not only because of their biological properties, but for political reasons: it’s easier to get government protection for entities that are considered members of different species (or, in the US, of different subspecies) than simply different populations of the same species. This issue makes biologists more willing to name new species than they otherwise might have been.
As New York Times reported when discussing the elephant work:
The evidence means that conservation efforts may need to be re-evaluated, said Alfred Roca, an animal scientist at the University of Illinois and another author of the study.
“Conservation efforts tend to focus on savanna elephants,” he said. “But the forest elephant also really needs to be a priority for conservation.”
And, from National Geographic’s piece on the elephants:
Raphael Ben-Shahar, an elephant expert at Oxford University, says, “Up until DNA fingerprinting tests, species were defined on the basis of morphological and anatomical differences.” Using the old classification yardsticks, the forest elephant was merely a subspecies of the savanna elephant. However, there was widespread disagreement among taxonomists as to whether the differences between the two elephant types were significant enough to denote separate species, he says.
The DNA evidence should put the controversy to rest. [JAC note: DNA differences are not sufficient to put the controversy to rest, and a good reporter should have known that.]
“The impact on management strategies if there are two elephant species in Africa is huge,” says Ben-Shahar. “Now we will have two species that are less numerous than was thought before.”
Remember: we have known for decades that the forest and savanna elephants are morphologically different, and have known since 2001 that they are anciently diverged as well. These conclusions have not changed at all with this new paper.
Let’s face it: we biologists want to conserve everything, and will glom onto any strategy that lets us save as many populations as possible. That’s not a bad thing, for as a biologist I feel that the diversity of life has inherent value, both in terms of us not having the moral right to screw around with other equally-evolved forms, and in terms of biologists’ selfish desire to keep things around to study and learn more about nature.
But biologists often have to hide this motivation: we must pretend that we’re saving populations because we need to retain genetic diversity, or prevent inbreeding, or save rare alleles that could bring back a larger species. We can never divulge the real reason why many of us want to save things like the elephants—because they have inherent value as organisms, and because they’re fascinating. That’s why many conservation biologists are busy worrying about the species and subspecies status of plants and animals: they secretly treasure them for their own qualities, but have to make a different case to the government and public about why they need to be saved.
This isn’t just a theory of mine: I’ve talked to many conservation biologists who admit that the real reason they want to save species differs from the rationale they must offer the public and the government. (Not all conservation biologists are like this, of course: some really are motivated by the reasons they give the public and official agencies.)
All conservation biologists are doing a wonderful thing: they’re keeping our own oafish, selfish, and greedy species from driving every other species to extinction. Plants and animals have no defense against the depredation of humans; they have only the law—and the morality of right-thinking folks—to protect them. In some ways conservation biologists are the most important of all biologists, for without their efforts what would we have to study? Without them, ecologists and evolutionists would eventually be reduced to studying their own intestinal flora.
What a shame, though, that we have to manipulate biological nomenclature, and dissimulate about our motivations, to keep other species alive! Shouldn’t we name species based on biology rather than politics?
h/t: Geoff North
______
Roland, N., D. Reich, S. Mallick, M. Meyer, R. E. Green, N. J. Georgiadis, A. L. Roca, and M. Hofreider. 2010. Genomic DNA sequences from mastodon and woolly mammoth reveal deep speciation of forest and savannah elephants. PLoS Biology 8(12): e1000564. doi:10.1371/journal.pbio.1000564.
Roca, A. L. N. Georgiadis, J. Pecon-Slattery, and S. J. O’Brien. 2001. Genetic evidence for two species of elephant in Africa. Science 293:1473-1477.
I think it might be more interesting to determine how the proboscid elephant allegedly “evolved” from hyrax-like mammals in the first place. The massive diversity within eutherians just doesn’t make any evolutionary sense since there thee is considerable overlap within the ecological niches occupied.
Please, go haunt some other website.
you know what really doesn’t make any sense?
you.
I have to look up proboscid, hyrax, and eutherian.
But anyone who puts evolved in scare quotes is a dope.
… and has been a dope for, what, 150 years.
Sheesh! It is as if biologists didn’t know this by now, and have massive amounts of evidence. Next he/she will bring up “intelligent falling”.
Jerry-
confirms the earlier finding that the savanna and forest elephants are anciently diverged—perhaps by as much as 2 to 7 million years.
Has there been a consensus formed as to just how much time divergence is required via molecular data to decide when two populations have become different species, by that metric?
I’m a bit rusty on the topic, but last I recall, there wasn’t even close to being any consensus.
Is it really the case that there are hard and fast rules now?
No, there are NO hard and fast rules; it’s more or less completely subjective. (I grant that there’s some subjectivity to the BSC, too, if populations experience nonzero levels of gene exchange). But the use of time as a metric for speciation is COMPLETELY subjective. How much time makes a species? There isn’t, and cannot be, a fixed answer, even in one group.
But the use of time as a metric for speciation is COMPLETELY subjective
yes, this is what I recall, arguing taxonomies and cladistics as a grad student at Berkeley.
However, I can only assume after 20 years, there must be at least some lab trying to push for some objective measure?
Do you know who’s leading the charge there?
also:
I suspect they are genetic, especially if the differences persist when the subspecies are raised in the constant environment of a zoo
quibbles with the idea that a zoo environment is a constant one aside, shouldn’t information on behavioral differences in various environments be readily available in the lit?
Speciation might happen in a matter of minutes, via polyploidy. It can certainly happen very fast, under several kinds of circumstances.
If two groups can be shown to have been evolving as distinct groups for a long time (a few thousand years would do it for me), that counts as one line of evidence (one! Not conclusive) that they’re distinct species. But the fact that they’ve only been separate for a few years is not necessarily evidence that they’re the same species because strong reproductive barriers may have developed. Strong reproductive isolation is always conclusive, IMO, but unfortunately what we find in nature is everything from strong divergence (morphological, ecological, etc.) with complete interfertility through varying degrees of divergence and isolation to complete reproductive isolation with no observed divergence in morphology or ecology.
Evolutionary patterns are much messier and more tangled we’d probably prefer. It’s Darwin’s tangled bank on steroids. I often say that taxonomy would have been so much easier if only Darwin and been wrong — but systematics would then be impossible.
True, but then you can say that for everything, simple processes (laws) building complexity in specific cases. Electronics and computing would be so much easier if semiconductors didn’t exist, say.
But I grant that biology is messy (and yucky and slimy at times)!
Any human male can in principle produce a fertile child when mating with any human female…
Well, except for that “lack-of-a-uterus” thing.
Don’t be pedantic. Obviously, the female brings the uterus to the game as well as half the genes.
I’m not completely sold on the problems of revealing the “true” motives of biologists as there are many groups that seek to preserve the different species and they all presumably had slightly different motives.
What does concern me is when the scientific conclusions are driven by politics. Declaring new species not because there is a biological reason but for political ones will rightly bring many other scientific conclusions into doubt and confusion. Scientists may wear two hats, one where they gather and analyse the data for others to use; and one where they are active politically, possibly based on the conclusions they learned in their work. If those roles get confused and they start altering the data or conclusions to fit their political agenda, it can undermine the arguments that they – and every other group which relies upon their science – make.
It doesn’t take much to think back upon the many scandals where politicians tried (sometimes successfully) to alter scientific papers to suit their agenda. I can only imagine the fuel something similar would give to the anti-science groups or those that would remove all environmental protections if it turns out that biologists were engaging in similar shenanigans.
…in many cases, and maybe even this one, species are named not only because of their biological properties, but for political reasons: it’s easier to get government protection for entities that are considered members of different species (or, in the US, of different named subspecies) than simply different populations of the same species. This issue makes biologists more willing to name new species than they otherwise might have been.
For decades now, I have thought that it might be worth looking at the idea of preserving unique gene pools, which can be determined in an non-arbitrary fashion, instead of by species.
so, in this case, the two subpops of elephants DO have demonstrably divergent and unique gene pools.
why not push for preserving that, in and of itself?
it has become more and more apparent, as conservation biology has advanced, that maintaining diverse gene pools, is extremely important in preserving any given species. I’ve seen individual populations of Salmonids gain protected status, simply because of their importance in contributing diversity to the larger gene pool of a commercially important species.
I know this issue has been raised before, and perhaps the reason it hasn’t been pushed for is the same reason that the idea of preserving habitat over species hasn’t been; political expedience (hell, we seem to be losing ground on the idea of even preserving species of late).
Still, I wonder what people think conceptually of the idea of looking at preserving unique gene pools?
I think that politically it’s doomed. ‘Save the panda’ is one thing, ‘save the gene pool of Shushan Mountain pandas’ quite another. The level of biology education most people have makes it untenable. Sorry.
yeah, I’m aware of that. I probably should have been clearer with the “conceptual” part.
I’ve played this game on the hill, and one of the ways critics of the ESA attack it these days is to critique the “what even IS a species?” angle that is being pointed out here.
congresscritters maybe be ill-informed, but they DO sometimes hire informed staffers to present them with arguments.
Because we have the tech to delineate unique gene pools, AND make an argument about their economic and intrinsic value, and since the concept of species is being eroded continually anyway, why not formalize it into a conservation goal?
btw, again, this is not a new concept; it has been in practice in many realms of conservation biology for decades now, and even organized into various lab groups these days, RE:
http://www.vetmed.ucdavis.edu/vgl/wildlife/
…and even basic textbooks on conservation biology too:
http://books.google.co.nz/books?id=NLJ8IzE6_34C&pg=PA31&lpg=PA31&dq=population+genetics+and+wildlife+management&source=bl&ots=l-dy9Z0fM3&sig=tON-vW79XqjBqzz9ahYk4trbalI&hl=en&ei=ox4hTfz9BYaUvAPYwIz6DQ&sa=X&oi=book_result&ct=result&resnum=5&ved=0CDoQ6AEwBA#v=onepage&q=population%20genetics%20and%20wildlife%20management&f=false
“Modern systematics, should, I think, be informed by evolution, at least to the extent that if you classify geographically isolated populations as different species, you should have a reasonable inference that they wouldn’t be able to exchange genes if they lived in the same place.”
This was a very interesting post. Two brief points here regarding the quote above:
1. Modern evolution also needs to be more informed by systematics (e.g., see the Wake (2004) review of Speciation in American Scientist).
2. A systematist can most certainly be “informed by evolution” and not hold to the BSC.
“Any human male can in principle produce a fertile child when mating with any human female, regardless of where they come from.”
Can someone clarify this for me please?How do we know this is the case, without resorting to a circular argument? I am not meaning to nit-pick – and maybe I am being thick – but I just wondering why we are sure this is so.
Thanks!!
How about, “Any human couple can in principle produce a fertile child regardless of where they come from.”
We have ample historical examples of previously isolated groups coming into contact to produce fertile mixed-race offspring. (Our President is one such, if you hadn’t noticed.) Dutch in South Africa, Spaniards in Mexico, slaves in the American South, the list goes on and on. But we have no historical examples of mutually infertile human populations. Maybe you’re suggesting there are some combinations we haven’t tried yet (Icelanders and Tasmanians, say) and we can’t really be sure until we do, but it seems highly unlikely that we’ll find an exception now, in light of the historical record.
What I’m really seeing it’s that there is a common effort to protect endangered species in general at the expense of a complicated and problematic classification system. That is due to the little commitment that is assumed by the local authorities. I know that isn’t the correct form, but it should be hard to decide whether perform an objective classification or to aid in the conservation of the species (that likely are of my interest). Regarding the BSC, I think that demostrate the hybrids fertility won’t be an easy task!
Regarding the BSC, I think that demostrate the hybrids fertility won’t be an easy task!
actually, elephants are bred in captivity all the time.
my guess is that this experiment has already been done.
Are there zoological gardens where forest and bush elephants are held in a common display?
Well I wouldn’t say “all the time” as if it were easy. It’s very far from easy. Lots of zoos don’t even try, for that reason.
Among other things, few zoos seem to be interested in keeping male elephants due to their aggressive tendencies.
A quick google search indicates that both the National Zoo in Washington DC and the San Diego Zoo have African Elephant breeding programs.
Forest, bush, or both?
I agree with Ophelia, I don’t know about programs like this in zoos, but, if it has taken place, the record has the answer!
The way I’ve begun to think about species concepts is that really there is only one species concept. That is the idea of ancestral-descendent sequences of populations evolving separately from others.
All of the “species concepts”, I recently stopped counting at 26, really just seem to be operational definitions of what a species is. As you’ve said it’s subjective. The BSC doesn’t work for fossils and so we use the morphological species concept (MSC). The BSC is tough to apply strictly to elephants given the difficulties of ‘lab cross’ type experiments and so we can use the ecological species concept backed up with MSC and evidence of genetic divergence. Postzygotic barriers to reproduction lag behind ecological divergence and so I have no problem with them naming these two groups different species despite perhaps some level of gene flow.
Why hold all cases to the BSC when clearly there are instances it where does not apply and is very difficult to apply and there are other lines of evidence the bear on the species question?
Not my area of expertise (speciation) at all although I’ve thought a lot about these things recently, just genuinely interested.
If you can, I’d suggest you read Chapter 1 and the appendix of my book (sorry to be self-aggrandizing), because it takes up these very questions. There you’ll see that I don’t hold all cases to the BSC (asexuals, for example,can’t by definition conform to it). And you’ll see that your own evolutionary species concept has even more problems. For instance, put a population of lizards on an isolated oceanic island. From the moment they set foot (feet?) on the island they are evolving separately from their mainland ancestor. Do you then consider them a new species at the moment of colonization? If not, why not, since they are “evolving separately”?
I, too, have thought a lot about this stuff, and distilled what I’ve thought into those two chapters, so have a look if you can get a copy.
The problem with the ESC (yours) is that it can’t offer a solution to what the founders of the Modern Synthesis saw as “the species problem”, which really is a profound biological problem. It is this: why do organisms in one area fall into discrete, objectively delineated clusters? That is, why is nature discrete rather than continuous? The BSC helps answer that problem; the ESC can’t make a dent in it.
But, as I say in the book, what species concept you favor depends on what problem you’re addressing. I’m interested in the discreteness of nature and so have found the BSC most useful.
Thanks! I’ll get a hold of the book; again as it turns out. Unfortunately I only spent time on a few chapters last time.
I certainly see where you’re coming from with the lizard example. As I tried to answer your lizard questions I resorted to one of the other concepts depending on scenarios. I guess adding the evolutionary species concept as an umbrella for biological, morphpological, ecological etc. species definitions isn’t terribly helpful.
Thanks both, I was going to ask something similar: since genomic data is used both for extant species (BSC) and extinct species (MSC, apparently), it can’t really be tied to both; and I find the book tip helpful.
Meanwhile I’m still wondering though: I can see where the BSC would help with discreteness, presumably genetic observation will or have already helped clarifying the mechanisms. Similarly I expect it could help clarifying MSC trait (say, functionality) differences, even if only at a gene function (structural, enzymatic or control) level.
So I assume that, as Ichthyic speculates, the molecular data groups would come up with a divergence measure of populations for their own benefit, which would feed back into the above work? Maybe they shouldn’t call it “species” though.
http://www.youtube.com/watch?v=Lk34w6luoUg
Darwin’s Designer Dogs – Not
Very interesting essay! I’ll probably have more to say later.
JC: Shouldn’t we name species based on biology rather than politics?
Absolutely! Describing species for political reasons is a bit unethical, in my world. Of course, motivations are less important than the strength of the argument made. Taxonomists are often accused by economic interests of describing things for political reasons, but the thing that the developers ignore is that taxonomists also reduce populations/taxa to synonymy when the evidence for their distinctness erodes when more evidence arrives. I’d guess there are about as many names in synonymy as are accepted as apparently valid. That is strong evidence that we’re NOT just treating every possible population as a species to get it protected.
I’d also add that the specialization of the authors and time and funding constraints may also contribute to the shortcomings. However, it is usually possible to reword the paper to avoid making any claims which are not established by the work or previous work; working out estimates on the time of the divergence of the groups is certainly worthwhile. So now it is up to others to check up on hybridization and the fertility and viability of offspring – and then we’ll see if the claim of different species happened to be correct. Perhaps someone has already done the work and someone familiar with it will publish a letter.
More shocking than one additional elephant sp., at least to me, was the recent evidence that there are at least 6 distinct species of giraffe, even under BSC, plus several isolated sub-populations that would probably be counted as more species under one of the “terminal taxon in a cladogram” type species definitions.
“Extensive population genetic structure in the giraffe”
David M Brown, et al
BMC Biology 2007
Has anyone tested that?
Seriously though, and tying in to the discussion of speciation, I feel a bit odd seeing mammals and birds having large divergence times and still being cross-fertile.
It is nearly a given that a social and technological species would “experience divergent pressures of natural or sexual selection” than their closest kin. At the same time we don’t really know if, and when, we speciated.
Modulo the moral concerns, is it really impossible to look at artificial chimp-human hybrid fertilization? If it doesn’t take on cells, fine. If it does, it would ramp up the moral challenge, since I assume the next step must involve implantation of some form.
Speaking of moral challenge, wouldn’t a person that appeal to a philosophical “problem of is-ought” have a challenge to specifically derive rights from behavior?
Normally there is no such problem of course. Human rights derives from codifications of mostly legal rights AFAIU, in turn informed but not determined by ethical considerations, which at long last are related to, but not mapping to, moral behavior.
And as for humans and their closest social circles such as pets and husbandry there has started to form mutual moral behaviors. I don’t think there is much of “animal rights” as of yet though, nor that it would or could ever amount to much as long as we are natural omnivores and beleaguered by parasites et cetera.
But appealing to is-ought makes the usual (not so close) mapping between rights and morality fraught with problems.
[Incidentally, I can’t find the promised review of Harris’ book, which is probably my fault. If the philosophic specters (but as I argue above not actual social) of naturalistic and moralistic fallacies are dealt with there, I have missed that.]
Ultimately, for better or worse, all attempts at nailing down the definition of “species” (and, consequently, all definitions that depend on the term) are doomed to failure.
Why? Simply, you can’t quantize a fundamentally continuous analog concept.
When does “cool” become “lukewarm”? When does “yellowish-green” become “greenish-yellow”?
It’s easy to distinguish between extremes. As far as ambient temperatures are concerned, 0° F is cold, and 120° F is hot. Light with a wavelength of 460 nm is blue, and 640 nm is red. But what labels would you use to distinguish between 68° F and 69° F? Between 570 nm and 575nm?
That’s not to suggest that descriptive labels for temperatures, colors, or organisms are useless — quite the contrary. Color labels are still very, very, very useful, even if they’re arbitrary, subjective, and impossible to objectively define with precision.
Biology is no different.
The solution is to use the general, vague-and-fuzzy terms when they’re useful, and to use other, more precise language when not. If you really need to distinguish between 68° F and 69° F, specify the two exactly as I did. Pull out your spectrophotometer for precise color definitions, or at least use a suitable device-independed color space such as XYZ or Lab.
I won’t pretend to dictate to an evolutionary biologist what constitutes sufficiently precise language to use in the case of these elephants or any other grouping of organisms. Maybe the language has yet to be invented or reach widespread use.
But the problem here is one of linguistics, and attempting to use imprecise language when precision is called for. Stop worrying over whether the elephants are grayish-brown or reddish brown/gray.
If it’s a problem communicating with others in the field, that’s something that needs to be solved, stat. Perhaps an actual quantifiable measurement based on DNA variances or the odds of viable offspring? I don’t know; not my field.
If it’s a problem with communicating with the public…well, don’t obsess over whether or not the public understands all the subtleties and nuances. Yes, it’s vital to you for the work you do, but the public will never “get it.” Instead, explain why it’s important. That much enough laypeople will understand.
Cheers,
b&
Ben: Why? Simply, you can’t quantize a fundamentally continuous analog concept.
Living things are not like colors or temperatures, which are continuous. Living things are closer to being discontinuous, but with lots of linkages between groups because the groups evolve. One can take walk through the woods or desert or wherever and temperatures during the day will vary continuously, but the organisms seen will generally be organized into obvious groups based on morphological features. Looking closer at anatomy, genetics, etc. the same patterns will generally (but not always)hold. The problem populations are the minority but are the ones that require the most attention by biologists, so they’re the ones we talk about. Groups with extensive or strong linkages between species, those actively radiating into multiple forms right now, etc. are the most interesting and perplexing in all evolutionary respects. The elephant and giraffe problems appear to be less acute then they were, though questions remain. In neither of these cases do the organisms involved vary continuously. They are clustered, and it’s been a matter of resolving the number of clusters present. Maybe we’re not there yet, but our description of reality is now a bit better.
Actually, organisms are continuous all the way down to (but not past) the individual.
One of your great…great grandparents had a child whose great…great grandchildren are chimpanzees. But, along the way, for a long stretch of time, those whose children would mostly be human and those whose children would mostly be chimp were still interbreeding.
So which of your ancestors was the first human in your lineage?
The question is as meaningless as, “which wavelength of light is the least bluish?”
And what of ring species? Dogs? Ligers and tigons?
When it comes right down to it, we are all of us, eukaryotes, prokaryotes, and archaea, part of the same family. Until Dr. Venter’s latest magnum opus, the only question is how closely related our cousins are.
Who’s to say that one particular pair of very closely-related cousins, such as the elephants under question, are sufficiently distantly related to warrant a different family name? And which of their respective great…grandparents were the progenitors of the lines?
Cheers,
b&
And what of ring species?
we wouldn’t be able even to identify the pattern if we couldn’t delineate independent populations to begin with. It’s really not a continuous loop from start to “finish”.
Dogs?
you’re kidding, right?
dogs are all one species, especially when you look at the genetics. they’ve never even been considered otherwise.
Ligers and tigons?
IIRC, those hybrids cannot be crossbred successfully to themselves or back to the parent population.
again, I’m not sure why this is an example that doesn’t support the BSC?
I don’t think a liger or tiglon has ever been found in the wild. For one thing, tigers and lions now don’t occur in the same area. But even when they did, I think some form of reproductive isolation (probably a combination of ecological and sexual) kept them from mating. Tiglons and ligers are produced only in zoos.
For species in general, see Jerry and Allen Orr’s book. For lions and tigers as biological species, that’s been discussed here at WEIT before: http://whyevolutionistrue.wordpress.com/2009/03/14/caturday-felid-4/
The question is as meaningless as, “which wavelength of light is the least bluish?”
*blik blink*
that there ARE wavelengths of light suggests something, doesn’t it?
OK. I see your point.
Following one parent child line of individuals back does make separation of species in the line somewhat arbitrary. Though if Eldridge and Gould are right about punctuated equilibria this would be less so. The breaks could be made at the punctuations where the gene pool is changing rapidly.
But, what we’re dealing with most of the time are the terminal twigs in the hypothetical tree of living things, and they are usually fairly discrete units. We don’t have to worry about all those ancestors because we don’t have them to look at.
The 3 elephants are the last 3 lines persisting in the formerly much more diverse elephant group. Until recently we thought there were only 2 terminal twigs, one of them variable/diverging. Now it’s clear that three distinct terminal twigs (species) still exist, and the 2 African species have been evolving separately for quite a while. But, they separated from each other after the branching off of the line leading to the Indian E.
I agree that there are continuous links between all extant organisms and their ancestors. There are also groups of organisms (known as “species”) with no extant continuous links to other groups of organisms. Thus, although evolution is continuous, extant species can be objectively delineated.
You can do this for the same reason that allows qualitative characterization elsewhere, if there is structuring of the distribution. I.e. pdfs can be unimodal (say gaussians) or bimodal or more. Waves can be wavelets (photons). Water can be cold ice, lukewarm water or hot steam; here the essentially analogous distributions of bonds nevertheless allows emergent phase change of the larger system.
Recognizing speciation is basically no harder than recognizing tree branches.
Conversely the basic problem IMHO is if you want to study where the branch started or new branches forming; essentially the former is a problem of resolution (scale), the later that the qualitative measure breaks down (disappears).
Now admittedly I still have to read your book on speciation, but I am fairly sure that you would not be as comfortable with the BSC if you too were a botanist. And let us not get started with bacteria or clonally reproducing unicellular algae, eh? Nice to know that the BSC works for a few percent of the biological diversity on this planet, though. Where it does work, it is a very clear and testable concept, and thus scientifically most satisfying.
I hate to sound like Chris Mooney and keep saying, “Read my book,” but we do deal with the issue of plants explicitly (chapter 1). It turns out that plant species are far better demarcated and separated than most people (including botanists) think. This comes from the work of Loren Rieseberg, among others. And, as I said, I do deal with asexually reproducing species as well.
Really, there’s no need to be so snarky here, especially if you don’t know what I’ve already published on these topics.
Yeah, sorry, I have not read every book that exists on the planet. I may nonetheless form an opinion about the results of my own research.
Delimiting (plant) species with the BSC and then declaring that they are “are far better demarcated and separated than most people think” is question-begging (if that is what you are saying). Of course you will find that reproductively isolated populations are reproductively isolated. But if applying the BSC means that we should lump, e.g., the entire genus Polylepis into one species just because the 20-30 ecologically well differentiated and morphologically strikingly different currently recognized species are a wildly introgressing messy polyploid complex genetically, then hardly any plant taxonomist can follow you. (Plus an unpublished PhD in the USA has shown that even some related genera, one of them herbaceous, had samples nested within arborescent Polylepis. Fun!)
But if you deal with asexually reproducing species, how can you still apply the BSC anyway? At least by the definition it was presented to me?
My main problem with the BSC is that there is obviously a huge grey zone between “no gene flow” and “one interbreeding population”. I find the phylogenetic species concept (Nixon & Wheeler, Luckow) and the recent work on genic species (Wu, Lexer) very helpful as a way of thinking about species.
And I did not intend to come across as particularly snarky. The BSC is an ideal that would be nice if it could be made to work everywhere. But I maintain it cannot. Life is too messy for one simple template.
Yeah, sorry, I have not read every book that exists on the planet.
you only needed to read the one you attempted to comment on, ass.
I may nonetheless form an opinion about the results of my own research.
cite?
But if applying the BSC means that we should lump, e.g., the entire genus Polylepis into one species
it doesn’t.
but then, you’d know that if you actually knew how the BSC is actually applied…
OR READ JERRY’S BOOK.
The BSC is an ideal that would be nice if it could be made to work everywhere.
good thing nobody here has then, including Jerry.
And I did not intend to come across as particularly snarky.
fail.
utter fail.
This aggressiveness is really uncalled for. Coyne broke a lance for the BSC in his post, and I wrote that it does not apply very well to other groups, e.g. many plants.
Yes, I have not (yet) read the book, which should not be necessary to make a comparatively harmless comment on the BLOG POST (me can do uppercase too!) of a person I generally agree with in all other cases, but unless he has completely redefined the concept, it is “groups of actually or potentially interbreeding natural populations, which are reproductively isolated from other such groups”. By its very definition, this cannot apply to clonally reproducing groups, and that is the end of it. It also gets problematic in other cases, which I have alluded to above. If Coyne does not, as he wrote, “adhere to the BSC” everywhere it gets problematic, then we have no disagreement, but that did not come across in the blog post, which is what I replied to.
The species problem is difficult, with opinions among colleagues ranging from “species do not exist in nature” (seriously) over “all species should be monophyletic” (again, seriously) to an interpretation of the BSC that would indeed demand the lumping of the entire genus Polylepis (what do you know of its genetics, by the way?), of all of Prunella (the plants, obviously, not the birds) and many other clearly differentiated groups into one. And it is usually zoologists who suggest we botanists should use Mayr’s hammer to sort our china collection.
BTW, where do you draw the line if you say “reproductively isolated”? I have not heard of a fertile mule, but barriers in plants are not perfect. In Primula section Primula, for example, in some crosses nearly all hybrid seeds are not viable, but the odd one that is produces a very competitive plant that can potentially back-cross – so-called hybrid vigor. Is less than one introgressant in 100 individuals well enough “reproductively isolated”? Why not less than one in 10,000? This is not physics – it is fuzzy! Shouting me down does not change that.
This aggressiveness is really uncalled for.
I’m beginning to think it’s actually more and more appropriate!
Coyne broke a lance for the BSC in his post,
no, he didn’t.
it’s applied here because these were elephants, and he made it very clear where to look for his justifications in using it in this specific sense.
I wrote that it does not apply very well to other groups, e.g. many plants.
well, the problem is, see, even for MOST plants it works just fine and dandy, and this was NOT a thread about plants anyway, nor about the general applicability of the BSC.
Yes, I have not (yet) read the book, which should not be necessary to make a comparatively harmless comment on the BLOG POST
it is if you want to start off by critiquing someone’s position without even understanding what it is!
you are not intellectually honest, and not worth wasting time debating further.
If you want to debate the relative merits of any of the dozens of applied species concepts, suggest you do it where they are actually being discussed as to their application!
or…
READ JERRY’S BOOK, which does an excellent job of delineating which best applies where, and which exceptions are notable.
I’m not sure if I should call myself a conservation biologist – today, I mostly work with non-threatened populations of large mammals – but I have to object to your characterizations of conservation science. I can’t speak for those you’ve quoted in your piece, but for the most part, I’ve seen conservation biologists play by the rules both in public and in private. Sure, we’d all love infinite conservation resources, but most folks seem to be more realistic as to the money and time we have to commit to conservation. Given we only have so many conservation resources, it behooves us to get taxonomy right. If you “cheat” to conserve one more population over what’s really needed, or for false pretences, you’re in practice diverting money from taxons that really are deserving of protection.
I’m actually somewhat worried that you would say you’ve talked to people who do otherwise in private. Such people could bleed the system dry, when we have so little to work with already.
@2yaks
The world of primate conservation biology has elevated virtually every known subspecies (and almost every new population) to specific status more for the strategic, political reasons mentioned above than for science’s sake. Madagascar, for example, is now said to be home to ~100 living “species” of primates according to the new guide by Mittermeier and Co. Type sequences are replacing type specimens…
This sounds more like a case of one of the narrow “phylogenetic” species concepts being adopted by a recent worker in the group. Conservation issues may be secondary.
>Type sequences are replacing type specimens…
I don’t know exactly what this means in this context. Hasn’t a type series always been an option in zoology? I think entomologists commonly use type series, for example.
In botany we have just one holotype, but can have piles of isotypes and paratypes.
type mtDNA sequences!
give me a skull and skin too …please
Ah, of course!
They can’t get away with just a blob of tissue, can they?
It’s worse than type sequences, which if different enough might support that there’s some cryptic speciation going on. Some of the Lemur “species” come from horribly abused STRUCTURE results, from input data that badly violates some of STRUCTURE’s assumptions…
I suppose I’m familiar with some of the cryptic “species” of Lemur, and I suppose you’re quite correct about it being badly abused. Luckily, the talk I attended where one of the presenters (Yoder, from Duke) was heckled by the audience to the point where one person remarked that at the rate of her research, they would be finding one species per-tree.
I’m horrified to hear that some of that garbage has received traction.
I’m not a biologist, so you should take my comments with a grain of salt.
I see epistemic reasons for treating them separately, in that there are significant differences in the facts about the two groups. Whether having them separate subspecies is sufficient for those purposes is not up to a non-biologist to judge.
As for the interbreeding barrier issue, I tend to see that more as a guideline than as a rigid rule.
If it were up to me, I would go along with the separate species idea.
The wording is a little confused from a biological point of view, but the Endangered Species Act has always allowed subspecies of any taxonomic group and “distinct population segment[s]” of vertebrates (only) to be protected:
So, species status has never been necessary (under US law) to offer protection. The forest subspecies (I’m a bit more firm on this point than Jerry is) has long been known, and so there isn’t a new species, just a proposed change in rank. Cryptozoologists have long been interested in the third type of African elephant, the pygmy elephant (Loxodonta pumilio), which has never been clearly distinguished from the forest elephant (Loxodonta africana cyclotis); some cryptozoologists think there might even be a fourth type, the water elephant.
And this means that it’s dangerous to infer anything about reproductive compatibility from divergence time alone—at least when that diverence time is a few million years or less.
(…)
Modern systematics, should, I think, be informed by evolution, at least to the extent that if you classify geographically isolated populations as different species, you should have a reasonable inference that they wouldn’t be able to exchange genes if they lived in the same place.
Jerry, I think your determined adherence to the BSC is what is dangerous here. Using BSC is your choice, and you can argue till the day you die that it’s the only relevant one, but that is only philosophy. Please do tell why it is you think that applying a different definition is “dangerous”.
One could as well argue that the factor(s) that caused to lineages to diverge are what should be used to evaluate whether they are different species or not. Allen Orr recently gave a talk here at MSU, showing one case in Drosophila where reproductive isolation caused speciation. But suppose the two species of elephants diverged because of ecological factors, then one could argue that it is right to apply the ecological species concept. Besides, your tenacity gives off an air of certainty that BSC is flawless, and we know that it isn’t (e.g. ring species, degrees of hybrid fertility).
Let’s face it: we biologists want to conserve everything, and will glom onto any strategy that lets us save as many populations as possible.
No we don’t! There are species that we’d like to get rid of in certain places, and we are quite okay with one population of squirrels not existing on one particular mountain, if there are other benefits (see Gould on that one). But sure we do want to conserve both African species of elephant, and so saying they are different species is then the wrong thing to do because you like a different definition better?
I was going to ask about that, specifically in reference to this. I admit to being a little bit shocked when I read that the plan was to eradicate the species — my understanding (and maybe some expert in Guinea worms can correct me) is that this isn’t like smallpox where you can keep some in deep freeze for posterity, and in any case, it just sort of “feels” different to talk about making a virus extinct vs. intentionally making a species extinct — even a parasitic one. OTOH, since the only host is humans, I cannot fathom anything remotely like an ethical justification for keeping it alive. “Sorry, but a few of you are going to have to suffer a painful debilitating disease, in order that we can save this ‘endangered species.'” Yeah, um, no.
So clearly eradication is the right thing to do. It still feels weird.
Also, I can’t immediately tell if this is a joke or not…
Well, either it’s definitely a joke, or else they are really serious about it.
Next thing you know, you’re going to tell me that Pluto isn’t even a planet!
In any case, funny that this is getting such press in the popular media. I just got done reading The Ancestor’s Tale, in which Dawkins refers to this kind of thinking as “the tyranny of the discontinuous mind.” Not that the paper isn’t interesting, and there’s probably worthwhile discussion to be had here… but really, this is an obscure taxonomic point, that even most biologists probably ought not to care all that much about. Why does it capture the public imagination so much?
Simply: Most people are incapable of separating taxonomy from reality. To many, the map is the territory, and if some politically insensitive cartographer goes tweaking the notation she uses on the map, well, my house just might fall off a cliff!
Thanks for this post. I am only an interested layperson (my training is in mathematics) but I’ve always wondered “what constitutes a species”. After all, I see LOTS of geese during my training walks and runs and I have no trouble telling a snow goose and a Canadian goose apart; yet these sometimes mate and produce some cool hybrids. So are these really different species?
I suppose it really isn’t that clear, is it, and the “genetic difference” metric really isn’t that clear either.
May I recommend John Wilkins’ recent post on the topic of species definitions?
How many species concepts are there?
Thanks for that!!
>> “As an example, the African helmeted guinea fowl can produce viable hybrids (I don’t know if they’re fertile) with the domestic chicken, yet those species are separated by at least 55 million years.”
So, if the African guinea fowl and domestic chicken WERE able to produce a fertile hybrid, would they still be considered separate species by current standards? Dr. Coyne’s argument that the elephants perhaps shouldn’t be considered distinct species if they can produce viable offspring seems like it should also apply in this hypothetical case.
I should really read Dr. Coyne’s ‘Species’ book. But in my naivete, I suspect that there are other compelling reasons why the two bird types would still be considered separate species. For instance, one reason could be that the birds don’t mingle in the “wild” or at least never attempt to mate when they do; or perhaps their ancestors were not able to produce viable hybrids (but by some genetic quirk that ability is now possible); or maybe the hybrid itself can’t mate with either of the parent species.
Any thoughts on this? I’m sure my hypothetical case here – which appears from one naive perspective to circumvent the usual species “rules” – can be torn down with a very simple argument.
Perhaps reading the next post would have lead me to this comment (relating to elk and deer):
“You can force-mate them in zoos, of course, but that’s a one-way test. If they do produce fertile hybrids in a zoo, that doesn’t tell you that they’d do so in nature, for the enforced confinement of a zoo may break down reproductive barriers that would operate in the wild.”
That basically answers my question above, with respect to how Coyne and others would define a species. I’m going to go get that “Speciation” book, straight away.
Well, if you had actually looked at the cited Science paper (the media wrongly reported, as is reported here, that it was based on mtDNA), you would find a discussion of the biological species concept as it relates to the elephants. The two species do hybridize, but there is no gene flow (even common alleles in one species are not passed to the other species). So the hybrids are not reproductively successful, hence species consistent with the BSC.
Sorry, but I certainly did read the Science paper and did see their discussion of the BSC and their report of hybridization. If you had read MY post, you’d see that I said the PLos PAPER didn’t mention the BSC. As for the viability/fertility of hybrids, the data are simply too scanty to say anything about that at present, so I didn’t draw any conclusion.
Stop with the snark, okay?
Sorry for the snark.
The Science paper shows that alleles/haplotypes that are common or even fixed in savanna elephants north, east and south of the forest habitats are not found in forest elephants. Also, common indels present in forest elephant populations were detected at different ends of the forest block (indicating long-distance gene flow for elephants across forest habitats, using a marker not typically subject to homoplasy). These indels are also common in the hybrids, but are never present in savanna elephant populations. This indicates that genetic variants (SNPs and indels) have undergone gene flow across long geographic distances among the forest elephant populations, but never enter the savanna populations, despite the presence of hybrids at the forest edge.
While I agree with your overall view that there is too much splitting of species with little evidence, in the case of the elephants, the evidence supports long distance gene flow across the forest, but genetic variants present or even fixed in the forest elephants never cross into savanna populations (at least at detectable levels), nor is there evidence for savanna-to-forest gene flow, suggesting that they are two species under the BSC using he criteria of Mayr. This lack of nuclear gene flow is presumably due to lack of hybrid reproductive success.
The PLoS paper adds to this evidence by showing deep divergence, but it is the original Science paper that suggested they were species, based on the BSC. The PLoS paper does not discuss the BSC, but does refer to a series of references that do discuss it extensively.
Regarding the politics, there are conservationists who want to avoid splitting the elephant species in two since they fear that this will make it easier to justify resuming the ivory trade for the more common savanna elephants.
I’m a retired fish alpha taxonomist. I like the BSC. It works best where species are syntopic, of course. Unfortunately, there is a huge literature on hybridization between perfectly good fish species. So the fact that two species hybridize, even produce fertile hybrids, makes one wonder if, but does not prove that, only one species is involved. I’ve studied species which produce fertile hybrids and species which don’t even attempt to hybridize. The first is more interesting, but the latter is more satisfying. The real question is, in nature, how much introgression is there?
Basically, with allopatric populations, one can do crosses, as I have done. It is more common, however, to make an evaluation that these two allopatric populations look (more similar, as different) as two closely related syntopic species, therefore, we have (one, or two)species.
My only working encounter with subspecies was to raise two subspecies to full species status. Under the right circumstances, “I ain’t never seen one of them before!”, would lead to description of a new species. On the other hand, description of subspecies requires good knowledge of range, gene flow, distribution of variation, and statistical evaluation of variation. One objection to subspecies is that naming subspecies obscures patterns of variation in characters which do not follow the same distribution as those studied.
Elephant biology is changing all of the time. The traditional two species, has been changed with the forest elephant of West Africa, being reclassified into a distinct species. Also, with the elephants of Borneo, which were always though to be feral elephants brought there in historic times, have been reclassified into a distinct subspecies of the Asian elephant. DNA sampling of the Bornean elephants, shows that they have been distinct for up to 300,000 years.