Until this week, there were two species of orangutan: the Bornean (Pongo pygmaeus) and the Sumatran (Pongo abelii), living on different islands. These were considered subspecies until 20 years ago, when the measured divergence of their mitochondrial DNA sequences led to them being separated as distinct species.
But now a new paper in Current Biology by Alexander Nater et al. (the “al.s” are numerous: see below for free reference, with the pdf here) adds a third species, P. tapanuliensis, also from Sumatra. Since new great ape species aren’t often described—the last was the Bornean/Sumatran orangs, and before that it was the bonobo, recognized as distinct from the chimp in 1933)—this has gotten a lot of attention, including in the BBC, in Science, and the Guardian.
But this biologist isn’t going along. Not only do I see this new “species” as merely an isolated and genetically differentiated population (as are many human populations regarded as H. sapiens), but I’d also contend that there is only one species of orangutan overall, with these three groups all being subspecies. Sadly, a lot of systematists don’t see it that way, as they seem to think that any isolated population, if it can be told apart morphologically or genetically from others, warrants being named as a new species. Yet to evolutionists, a “species” is not an arbitrary segment of nature’s continuum, but real entities that maintain their “realness” because they don’t exchange any (or many) genes with other such entities where they cohabit in nature.
On to the paper:
The new species, called the Tapnuli orangutan, as seen in the Guardian’s figure below, occupies only a small patch of forest south of lake Toba, about 1000 km², and numbers only about 800 individuals. They weren’t found until 1997 as the population (endangered due to habitat loss) lies south of the range of P. abelii and was assumed not to harbor orangs. The population is threatened because it’s small and much of the prime habitat is subject to logging.

Is this a separate species? While most of the media accept this uncritically, I’m dubious. That’s because the authors base the description on morphological differences and genetic divergence alone (with morphology of the skull based on a single individual, though one that lies outside the ranges of the other species). They give little information on reproductive isolation, and what information they give suggests that the two “species” of Sumatran orang did interbreed. (Data show genetic interchange between the two Sumatran “species” until about 20,000 years ago, when the two groups became completely isolated by the loss of suitable habitat between them.)
What we have, then, is a geographically isolated population of orang that has some morphological difference from the others, some genetic divergence from the others, but has no present opportunity to see if it can interbreed with others when they co-occur in nature: the criterion used in the “biological species concept”. Thus calling P. tapanuliensis a new species is a pure judgment call. Here are the results of the authors’ analysis:
- The three orangs are related like this: the Sumatran orang split off from the other two about 3.4 million years ago, so the Tapnuli and Bornean orangs are the most closely related—somewhat surprising as they live on different islands. But the area has been repeatedly sundered by volcanism and sea-level changes over the past several million years, so we don’t know the historical sequence of geographic changes in the ancestor.
- DNA analysis shows that the Tapnuli and Bornean orangs diverged about 700,000 years ago: about the same time as the divergence between the modern human and (extinct) Neanderthal lineage. The two populations on Sumatra are, as I said, about 3.4 million years old.
- The Tapnuli and (northern) Sumatran orangs continued to exchange genes (particularly Y chromosomes, since males are the most mobile sex) up to about 20,000 years ago; that stopped because the habitat between these two more distantly-related species prevented gene exchange. But clearly these anciently-diverged orangs were capable of hybridizing when they encountered each other, and the hybrids must have been at least partly fertile since Y chromosomes moved between them. It’s likely, then, that the more closely related Tapnuli and Bornean orangs would also be able to successfully hybridize were they to meet. We already know that the Bornean and Sumatran species hybridize well in zoos and produce fertile hybrids (called “cocktail orangutans”).
- One specimen of Tapnuli orang shot by a local was measured; its cranial dimensions put it outside the range of the other two “species”, so there are likely to be diagnostic differences in the skeletons of this and the other two species, though we need more speciments of P. tapanuliensis.
- The Tapnuli orangs have a different color and different calls from the other two “species”
So what we have is an isolated population of orangutans that split off about 700,000 years ago from a lineage now inhabiting Borneo. Somewhat isolated from the populations on both Sumatra and Borneo, it accumulated morphological and genetic differences, and possibly difference in vocalization, though these could be cultural?
But is it a new species?
I say no: it’s just an isolated population that’s somewhat different, with individuals being diagnosable. If you use the Biological Species Concept of evolutionists, which deems populations to be different species if they could not produce fertile hybrids when encountering each other in the wild, I’d say that the evidence of interbreeding until physical separation was complete only 20,000 years ago suggests that the Tapnuli and Sumatran orangs are a single species, which also means that the Sumatran and Bornean orangs are a single species as well. My guess would be that the new species would produce fertile hybrids with both of the other “species” in captivity.
If you consider the Tapnuli orang a new species, then you’ll have to consider Neandthals a species different from H. sapiens, for they diverged about 700,000 years ago and also hybridized successfully when they met: that’s why most humans carry some Neanderthal genes—the remnants of that ancient hybridization. Yet Neanderthals are considered by many biologists and anthropologists to be a subspecies of H. sapiens (H. sapiens neanderthalensis vs. H. sapiens sapiens).
If we call every genetically differentiated and diagnosable population a new species, then there’s no end to the mishigass. We surely don’t call genetically diagnosable populations of humans different species, and yes, if you have enough genes you can diagnose human populations pretty accurately.
Sadly, nearly everyone in the media has uncritically bought into this designation of a new species of great ape—save one scientist quoted in the Guardian:
Dr Andrew Marshall of University of York, said that the study highlighted the importance of conservation, and added that there might even be further species of great ape to be discovered.
But Professor Volker Sommer from University College London was less bowled over, pointing out that there is no clear criteria for what constitutes a new species. “Any bunch of expertised biologists can invent a new species, if they get their arguments together,” he said.
One more point: all the media, as well as the paper itself, emphasize that the new “species” is small in numbers and lives in endangered habitat, and so must in the future be the subject of intensive conservation efforts. Perhaps naming the population as a new species excites people because of the conservation implications, though in the U.S. the Endangered Species Act mandates the conservation of populations, not just species, so if these apes were in America, naming them as a new species would have no implications for conservation.
In the end, we have no strong evidence that this new species—nor the already-named Bornean and Sumatran orangs—would show any reproductive isolation from the other “species” were they to come into contact in nature. All we have is a geographically isolated population that has some genetic and morphological differentiation. To call it a new “species” turns taxonomy into a completely arbitrary exercise, which, as Dr. Sommer notes, becomes purely subjective. Until we have evidence of reproductive isolation or potential isolation between this population and the others, I’d say “forget about calling it a new species.” And on top of that I’d say that “calling the Bornean and Sumatran orangs different species is also a purely subjective decision”. If the latter two produced sterile offspring in zoos, I’d be down with calling those species different. But they don’t: they produce “cocktail orangs.” LUMP ALL THE ORANGS!
It would have been nice had the media asked a few evolutionary biologists whether this population qualifies as a new species. And I’d ask them this: if we really have an astounding new species of great ape here, why wasn’t it published in Nature or Science? My guess is because those journals would bridle at making that call on the basis of the data presented.
h/t: Matthew Cobb
UPDATE: Just as I finished my bit above, I got an email from Greg Mayer, who said he’d contribute his take on the orangs, too. I’m just pasting it in here without reading it so you can get his opinion. (The title of this post is for my bit alone.)
[Note from GCM: The title is Jerry’s, but I concur, so I’m adding it to mine as a second contribution. We spoke briefly by phone this morning, so knew each other’s reaction to the paper was dubious, but wrote separately.]
A new species of orangutan? I doubt it. II.
By Greg Mayer:
The paper’s chief conclusion—that there is a second species of orangutan on Sumatra—is not well-supported by the data presented.
The morphological analysis is risible. They have a single (1!!!) specimen, and compare it to only 33 others. The analysis boils down to “Can we tell this one from the other ones?” They say yes, but that’s not clear from the evidence presented. The most obvious method of doing so—presenting an ordination (e.g. principal components) of the morphological measurements, and showing that the single specimen lies well outside the cloud of points formed by the other 33—is not used. They do show several one-dimensional PC plots, and the single specimen is outside the range of variation on a single axis. (Curiously, their first few PC axes account for very little variation, which is highly unusual. This could be explained in the methods, but, as is usual these days, the methods are not in the paper, but in a supplement). They note, as though it supports their claims, that the single specimen is not in the interquartile range of 24 of 39 skull measurements. “Outside of the interquartile range” ?!?—that didn’t qualify as a subspecies in the heyday of oversplitting!
Of course, even if they could recognize this single individual, that wouldn’t mean anything other than individuals are different from one another. If you gave me 34 people, I’d be able to tell them all apart with 39 measurements– that wouldn’t, of course, come close to showing they belonged to different species. (It would be especially easy to tell them apart if 1 of the people was from Sicily, and the other 33 from Germany.) A minimal morphological analysis would require showing that the means and ranges of population multivariate phenotypes are such that individuals in the population are either diagnostically or statistically distinct.
The call differences are presented perfunctorily and incompletely, so no inferences can be drawn. Differences in calls, which are of potential interest as they may entail reproductive isolation (e.g., as in frogs), possibly have large cultural components in orangs (as they do in humans), and thus require careful study to determine their characteristics and effects.
The genetic analysis, taken at face value, indicates gene flow quite recently, until the habitats became geographically separated, directly contradicting their claim of specific status. That the single Y chromosome observed nests neatly within other Sumatran orangs again subverts their claim. I have been unable to determine how many individuals of the “new” species were genotyped in some way– Table S4 in the supplement should have that information, but that table is not in what downloads, and clicking on what seems to be a link to it goes nowhere.
I have additional issues: Does demographic reconstruction software really work? If species are anything I can can distinguish, doesn’t that make everything a different species? Have you tried applying that concept to extant Homo?
The plaintive defense of “operational species” concepts is a bit sad, and deeply confused conceptually. Citing Gorilla and Neaderthals as supporting their case again does the opposite, because there is only one species of Gorilla and Neanderthals are conspecific with sapiens, as the very genetic studies they cite prove. Their own studies prove the conspecificity of their “new ” species.
The above is obviously a quick reaction, not a complete analysis, but fell free to use any or all of it in your post.
I used it all.
___________
Nater, A., M. P. Mattle-Greminger, A. Nurcahyo, M. G. Nowak, M. de Manuel, T. Desai, C. Groves, M. Pybus, T. B. Sonay, C. Roos, A. R. Lameira, S. A. Wich, J. Askew, M. Davila-Ross, G. Fredriksson, G. de Valles, F. Casals, J. Prado-Martinez, B. Goossens, E. J. Verschoor, K. S. Warren, I. Singleton, D. A. Marques, J. Pamungkas, D. Perwitasari-Farajallah, P. Rianti, A. Tuuga, I. G. Gut, M. Gut, P. Orozco-terWengel, C. P. van Schaik, J. Bertranpetit, M. Anisimova, A. Scally, T. Marques-Bonet, E. Meijaard, and M. Krützen. 2017. Morphometric, behavioral, and genomic evidence for a new orangutan species. Current Biology, in press. DOI: http://dx.doi.org/10.1016/j.cub.2017.09.047