Monday, October 5, 2009

The Many Tales of “Ardi”

The Inspirational Tale

Common metaphors in paleoanthropology are those of winning the lottery and being struck by lightning. But although fossils are rarely preserved or exposed at just the right place to be spotted, the assumption that fossil discoverers are supremely lucky is an unfair one.

Eugene Dubois decided he’d find a missing link on Java, moved his family there, and found the first Homo erectus fossils. Louis and Mary Leakey dedicated their lives to traversing East Africa for fossils and they found beautiful ones. It took most paleoanthropologists years and even decades of hunting, sometimes walking over the same doggone dust every season, just to spot a tiny speck of something spectacular.

This is the kind of tenacity it took for Tim White and his team to amass a fossil record of as many as 36 Ardipithecus individuals and to collect over 150,000 plant and animal specimens from the Middle Awash study area in the Afar region of Ethiopia. Not only that but “Ardi’s” fragile skeleton had to be prepared sliver by tedious sliver in the laboratory before the analysis could be accomplished. Last week’s special issue of Science came 15 years after the first announcement of the project because of the massive amount of collection and analysis that followed.

It's no surprise that paleoanthropologists were glued to their computer monitors last Thursday and Friday. Many of us are still carrying around our stack of 11 papers, picking our way through them every spare moment we get.

The Paradigm Shifting Tale

This is not just a new skeleton. Depending on your take, this is a renovation or an entire upheaval of the way we interpret the last 23 million years of ape evolution. Here is a list of just some of these new perspectives that Team Ardipithecus has presented to us:

1. Ardipithecus is a hominid yet it is much more primitive than chimpanzees and gorillas which are both highly derived. (This one is exciting because it links Proconsul (the 18 million-year-old ape that I work on) to hominids.)

2. Anatomical and behavioral similarities between ape species that have to do with suspension and locomotion-in-general are independently evolved. That is, they’re convergences or homoplasies. Even these traits in the fossil apes Dryopithecus and Ouranopithecus do not link them phylogenetically to living apes because they too evolved independently as homoplasies. Conclusion: Ardipithecus shows that phenotypic and maybe genotypic homoplasy is rampant in the hominoid (ape) fossil record.

3. Chimpanzees and humans share a deeper last common ancestor than the one estimated to have lived 6 million years ago (mya). Although the 6 mya hypothesis is currently “in vogue,” the last common ancestor actually lived more like 7-10 mya.

4. Even though the split between the chimp and human lineages occurred 7-10 million years ago, Ardipithecus from between 4-5 mya is our best model for the last common ancestor, better than any other fossil apes from Africa from between 6-11 mya such as Orrorin, Sahelanthropus, Chororapithecus, and Samburupithecus.

If paleoanthropologists were stifled, rather than stimulated, every time something new overturned old ideas, they would quit the field, or flunk out.

However, the suite of game-changers proposed by Team Ardipithecus are bound to raise not just eyebrows but hackles as well.

For example, they explain the ape evolutionary tree with multiple homoplasies (rather than with parsimony or the simplest explanation: that the traits evolved once in a common ancestor), but then they use parsimony analysis (which excludes homoplasies whenever possible) to claim that Ardipithecus is a hominid. Furthermore, the claims about ape homoplasies are grounded in the assumption that Ardipithecus is a hominid in the first place. Could they have done it any other way? I don't know the answer yet. But as it stands, it makes an already large pill even harder to swallow. The whole field is now charged to test these new hypotheses and we certainly have our work cut out for us.

The Snarky Tale

Tim White’s papers are always thrilling to read. Usually the science is tight and methodical, which is highly satisfying, but you can also count on enjoying a bit of snark, and when I write “enjoy” I guess I’m being snarky.

Here’s an example: As the authors explain how un-chimp-like Ardi is, and how derived chimps and gorillas are compared to hominids, they write,

“Some hold that our last common ancestors with African apes were anatomically and behaviorally chimpanzee-like, that extant chimpanzees can be used as “time machines” and/or that unique features of Gorilla are merely allometric modifications to accommodate its great body mass.”(White et al., Science 326(64): 75)

These hypotheses of "Some" paraphrased and quoted by Team Ardipithecus may not sound so strange to any of you. They don't to me. But we're being admonished. Sure, the authors can forgive Darwin for thinking this way because he was forced to work without a fossil record. But the rest of us seem to have no excuse even though up until now we’ve all had to do paleoanthropology without Ardi.

A decent hypothesis for what our last common ancestor with chimps looked like is that it looked a lot like chimps. It’s pretty simple. And it’s only a hypothesis, not a personality trait. And it’s going to remain a decent hypothesis until scientists beyond these authors also support Ardipithecus as a hominid.

If they must lash out at colleagues, the authors should wait until after we've ignored their findings or rejected their arguments. Bashing us over the head the second we get a chance to read about Ardi is something Ardi might do.


Ken Weiss said...

Another great post, Holly!
My reactions (as a non-paleontologist) are several.

1. Your account of the travails of getting all this material assembled and analyzed somewhat (but only somewhat!) absolves Team Ardipethicus from being so secretive and non-communicative about their finds for 15 years. They could have long ago released some of the major aspects of their finds, and I think that when federally funded that should be a formal requirement (quick and full release of data).

2. I can't comment on the phylogenetic arguments, but I do believe that all the splash that is always associated with these finds is wrong. When a field can be entirely turned upside down by a single find, and it's happened again and again and again, there is a need for much more humility on the part of the protagonists about what we know and don't know.

In that sense, to me personally, these things should appear, muted and more soberly, in journals like the Am J Physical Anthropology, than made into media events. A science with sound grounding should not be repeatably vulnerable to single new finds. That should happen, as we learn things, but not so predictably or regularly. I doubt many will agree with me, on this, however, as the media circus and food-fights are much of what makes paleontology fun for many people, public and scientist alike!

3. I think the homoplasy issues are interesting and can be accounted for in genetic terms without any particular stretching. I have written a recent installment of my Crotchets & Quiddities columns in Evolutionary Anthropology, on this subject (entitled "How the Persians Were Saved by Lightning", and available from my web page or the journal's).

The gist of the argument is that complex traits, like locomotion or skull shape or brain size involve many genes. Each will vary and there will be many different genotypes that can generate similar phenotypes. Genes typically affect many traits, so we see 'homoplasy' regularly even within species (two people with genotypes but the same height, for example).

Genetic drift, selection, mutation, and isolation can lead populations to have different assemblages of traits (and their underlying genotypes). It is a too-strong view of selection, in my opinion, that leads to expectations of coordinated suites of trait-change as if they are in lock-step genetically.

At least, the mixed homplasies are not in themselves very surprising, I think.

Anyway, it was a fine, very informative post!

Holly Dunsworth said...

Thanks Ken! And the authors do discuss the genetic implications for their homoplastic view of ape locomotor anatomy... I think you might really like what they have to say.

I think the biggest danger in taking so long to publish your discoveries is the ever deepening gulf between your new knowledge and the existing field's knowledge. It seems, these authors didn't handle it very well. They come out swinging with 15 years of built up secretive angst. It's not helpful.

Jason Hodgson said...

Great post Holly.

My favorite aspect of Ardi is that it keeps alive the near perfect streak of African hominoid fossils that are originally interpreted to be direct human ancestors (...perhaps I am even snarkier than White et al.).

It think it will be most interesting if Ardi turns out to be something other than a hominin.

After all, there must have been some hominoids running around that were not hominins. Yet we never seem to find them. ....incidentally, you note that paleontology is a lot of painstaking work and not just luck. But how about the luck of paleoanthropologists only ever finding hominins and never chimps or gorillas or extinct stem lineages. That has to be luck!

We will not really know the polarity of these locomotor traits until we know what ancestral chimps, gorillas and stems looked like. So I think it would be a lot funner if Ardi is not a hominin.'d think after 15 years they could have come up with a better name than "Ardi". When you find your complete Proconsul skeleton, please don't call it "Proco".

Holly Dunsworth said...

Here's my ideal version of a 15 year paleoanthropological study:

The discoverers form their team and do their analyses, meanwhile, the casts and samples they've sent off to morphologists and chemists are also being studied, but blindly. These other labs have ZERO information on age, geographic context, associated fauna, geologic context, and ZERO idea about what the discovery team is finding (i.e. the Party Line).

Then all of it is published side-by-side.

This kind of thing could only happen if set up from the outset.

And it's not even something we could do with "Proco" (haha) because we already know a lot about Proconsul.

Maybe we can pull it off with the next brand new thing we find....

Holly Dunsworth said...

Let the TV shows begin:

Holly Dunsworth said...

Regarding the length of time it took between announcing the skeleton and then revealing it to the world...

I want to point out to anyone who may not realize that paleoanthropologists aren't permitted to bring fossils out of their countries of origin.

This means that after they get the money to find the fossils in the first place, they have to get more money to travel back and work on them. Aside from funding issues, teaching, campus, family and community obligations also mean that fossils may sit for quite a while between study sessions, which are usually briefer than the scientist would like.

Paleoanthropologists should publish quickly, but others should be aware of the obstacles.