Science Got Ardi Wrong or: The Enigma of Ardipithecus [UPDATED]

There may need to be a significant revision in the recent description of one of humanity's oldest ancestors. Ardipithecus ramidus (or "Ardi" for short), the 4.4 million year old hominid fossil discovery, has been a godsend to paleoanthropologists (pun intended). But one of the key researchers has made what could be a serious error in his interpretation.

Christopher Ryan, who writes for Psychology Today at his blog Sex at Dawn (also the title of his forthcoming book) has discovered evidence that could undermine Owen Lovejoy's argument about human sexual evolution ever since Ardi:

In a nutshell, Lovejoy argues that the evidence he and his colleagues presented indicates an absence of sperm competition in the human line and shows male provisioning of females that eventually led to the modern nuclear family. Simple enough. But to make this all-too-familiar argument, Lovejoy misleads, misunderstands, and mis-states his own findings to the point where, if this were a graduate-school paper, his professor would demand a re-write.

Ryan demonstrates multiple factual errors in Lovejoy's paper "Reexamining Human Origins in Light of Ardipithecus ramidus" (the 11th and last in the special series, free access here) and has submitted a letter (along with Todd Shackelford) to the editors of Science.

Two of the most damning points have to do with evidence pertaining to sperm competition. Lovejoy states that the ratio of human testes size to body size is similar to that of gibbons, a monogamous "lesser" ape. However, Ryan points out that the actual figures show something completely different. Citing the authoritative textbook Primate Sexuality by Alan Dixon (which I recently used as a teaching assistant for a class of the same name at Duke) he points out that gibbons have a testes to body ratio of 1 to 1000. Humans come in at 1 to 160. Dixon also shows that chimpanzees and bonobos lie at a ratio of 1 to 36, significantly closer to humans than gibbons are. [UPDATE: A reader has pointed out that these values are in error. See comments below.] This is a glaring oversight on the part of both a distinguished scholar as well as the journal itself.

Another very troubling error is in the case of human sperm production. Lovejoy states that human spermatogenesis efficiency (the daily sperm production per gram of testes) is only a measly 0.06 x 10^6 and he cites Peirce and Breed in their 2001 paper in the journal Reproduction as his source. Using this figure Lovejoy then concludes:

The estimated corresponding value in chimpanzees is greater than that of humans by two orders of magnitude.

That would be a significantly large difference, except that it appears Lovejoy made a mistake because the actual figure from that paper is 6 x 10^6, not 0.06 x 10^6. This, interestingly enough, is two orders of magnitude greater than the figure he cites, making chimpanzees and humans identical. This completely undermines his argument that humans have low sperm production.

Here's a screen capture of Peirce and Breed's paper for those who can't access it past the pay wall:

And here's the direct quote from Lovejoy's paper:

Another measure, spermatogenesis efficiency (daily sperm production per gram of testes), "varies from about 2.65 x 10^7 in rabbits to

Why does this matter? Testes size and sperm production are both key predictors of primate mating strategies. Chimpanzees and bonobos, for example, have a highly promiscuous society where females mate with multiple males during their period of oestrus. In the arms race that developed to outcompete other males, chimpanzees and bonobos have evolved extremely large testicles with correspondingly high sperm volumes to "wash out" an earlier male's contribution. In contrast, gibbons are socially monogamous (though not strictly sexually monogamous, both partners are known to have "extra-pair copulations" with neighbors) and have correspondingly small testes and reduced sperm production as a result.

By mischaracterizing these values Lovejoy has given a false impression about the evolution of human mating systems. We are much, much closer to chimpanzees and bonobos than we are to gibbons. It's actually extremely shocking that Lovejoy would have made such a major mistake on what is common knowledge among evolutionary anthropologists.

Ryan suggests the following possibility for why these blatant mistakes weren't identified in the rigorous peer review process:

Perhaps editors, fact-checkers, and general readers are over-eager to accept even the weakest arguments, as long as these arguments support the notion that sexual monogamy is characteristic of our species' evolutionary past. This is what they expect and hope to be told.

In Pablo Neruda's poem "The Enigma" the brilliant Chilean writer questions the ability of the human species (a handful of genes short of being a chimpanzee) to obtain unbiased information about themselves from the natural world.

I walked around as you do, investigating

the endless star,

and in my net, during the night, I woke up naked,

the only thing caught, a fish trapped inside the wind.

Neruda suggests that all of us, poets and paleontologists alike, cast our net out into the world to collect and assemble various facts, but all we end up finding is our own ideas reinforced. While I personally think the scientific method is the antidote to Neruda's enigma, his point is well taken that our own personal biases can frequently shape the way we interpret (or, rather, misinterpret) information to fit our own views. The value of science as a cultural tool is to root out these errors and to mercilessly critique one another's results to expose these personal biases. I, for one, will be eagerly awaiting the response from Lovejoy and the editors at Science about these mistakes in such a major research paper and look forward to untangling the enigma of Ardipithecus.

For more on this see my post: Reexamining Ardipithecus ramidus in Light of Human Origins.

References:

ResearchBlogging.orgLovejoy, C. (2009). Reexamining Human Origins in Light of Ardipithecus ramidus Science, 326 (5949), 74-74 DOI: 10.1126/science.1175834

Peirce, EJ and Breed, WG. (2001). A comparative study of sperm production in two species of Australian arid zone rodents (Pseudomys australis, Notomys alexis) with marked differences in testis size Reproduction, 121, 239-247. DOI: 10.1530/rep.0.1210239

Categories

More like this

"a testes to body ratio of 1000 to 1." Really? I think you mean body to testes, or else gibbons would have a hard time moving around, never mind escaping predators. Or heck, administering the sperm in all those massive testes. Nice image, though.

By fizzchick (not verified) on 15 Oct 2009 #permalink

Thanks for catching that. Fixed. See how important the peer review process is?

Where are you guys are getting the 1:160 testes to body size ratio for humans? By my calculations from the table on page 219 of Dixon, the ratio is:
1:373 in Chimps
1:1000 in Gibbons
1:1266 for Humans
1:2114 in Orangs
1:3280 in Gorillas.

The human and gibbon are both below the regression line in the table on the same page, while the chimpanzee is far above it. Perhaps I'm overlooking something in a different section of the book?

Eric,
Please note that I've edited my blog entry, which you so kindly linked to, to reflect my own miscalculation. My figure of 160/1 body mass to testicular mass in humans is in error. The number given by Zinjanthropus above, is closer to the mark, though Dixson gives no exact mean body weight for humans on the graph in question. Hoist on my own petard, for sure. My apologies.

Please, clarify all this thing about testes to body ratio. 1000 to 1 in gibbons may sound weird, but if you fix the figures by turning them around, then you get this 1 to 160 for humans that I don't find very comforting either. I mean, I am a male human of about 80 kg of body weight and I guess my testes fall very short of the half a kilo that should be expected with a 1:160 ratio. And what about that 1 to 36 for chimps! Are they really able to climb trees?
I guess that some units are missing...

Luis,
Yes, you are correct. In my own calculations, I left off a zero. Pretty embarrassing, when I was calling attention to Lovejoy's errors. I'm sorry for the confusion.

As you've already shown here, Christopher is something of an amateur scientist. By his own admission in his blog, he has no professional experience as an investigator.

I would recommend taking his findings with a grain of salt -and most certainly checking all the facts- before subscribing to his armchair musings.

I guess you can consider this a shill post...I try to make several arguments about how it is possible to have monomorphism in body mass (or any other intrasexually selected trait) in spite of overt male-male competition. This has implications for how we interpret ancestral mating systems from fragmentary fossils.

Here's the link

When it comes to Owen Lovejoy and his longstanding claims of early monogamy in hominids "I would recommend taking his findings with a grain of salt". He does quite nicely on his speciality of biomechanics; his hominid behavior stuff is very poor and always has been.

By anthrosciguy (not verified) on 17 Oct 2009 #permalink

I thought that the average weight differences between male and female humans indicated partial monagamy and partial straying.

Your two main condemning points against Lovejoy's paper are that he made mistakes regarding testes size & sperm production.

As Zinjanthropus has already pointed out, when it comes to testes size, humans & gibbons both fall below the regression line & the ratio for humans is much closer to gibbons than any other species.

As far as sperm production is concerned, yes there is a typo in Lovejoy's paper as you pointed out. The correct value for humans should have been 6 x10^6. While it is unfortunate that such a typo made it to the final print, it is ultimately irrelevant. As many others have pointed out chimpanzee sperm production is much higher & remains higher after subsequent ejaculations as compared to humans. E.g., mean chimpanzee sperm production is 1278 * 10^6 [1]; easily more than 2 orders of magnitude higher than in humans, which is precisely what Lovejoy stated. His description & analysis remains valid either way.

You should be less quick to criticize others for making mistakes, especially when you have made your own typos in your critique of them.

[1]J. Marson, D. Gervais, S. Meuris, R. W. Cooper and P. Jouannet (1989) Influence of ejaculation frequency on semen characteristics in chimpanzees (Pan troglodytes). J. Reprod. Fert.