Today is Ardipithecus day. Eleven papers in tomorrow’s issue of Science describe the research on one exceptional skeleton (numbered ARA-VP-6/500, nicknamed “Ardi”) as well as more than thirty other individuals, mostly represented by isolated teeth with a few partial sets of teeth.
I have a lot of material to share about these papers and how they change things in paleoanthropology – so much that I can’t possibly fit it all into one post.
So I’m starting out with a basic overview of the main points, organized as an FAQ. Over the next few days I’ll be exploring some of the most central issues in closer detail: in particular,
How we now interpret the earliest hominins in light of Ardipithecus.
What the skeleton means for our understanding of the human-chimpanzee common ancestor.
How Ardipithecus relates to the first australopithecine, Australopithecus anamensis.
How the crushed pelvis became a 3-d model, and whether we should believe it.
Can Ardipithecus be consistent with genetic estimates of human-chimpanzee divergence time?
What was the locomotor adaptation of Ardipithecus really like?
How was the "Discovering Ardi" documentary feature?
I expect I’ll be posting a new story every day for the next week or so. This initial post will be the central location for the series, and here I’ll try to give the most general-interest information.
I will also have a short article coming out in Seed sometime this week, I will post a link when that is up.
I will be editing this post recurrently – I’ve been speed-writing for the last couple of days and so I have some work yet to do on adding references, fixing typos, rephrasing, etc. This will be a stable document after the first week.
UPDATE (2009-10-03): OK, it's been a couple of days, so I'm closing out the post and adding a jump. I'll continue to update links inside as I round out my reactions to the papers.
What’s the big deal?
If you want a basic description of the facts, here they are. Today’s series of papers is basically unprecedented in paleoanthropology. There are eleven papers in total, giving comprehensive coverage of the anatomy, paleoenvironment, and evolutionary interpretation of a new skeleton of Ardipithecus ramidus and dental remains representing more than 30 additional individuals. They have been published simultaneously in a coordinated effort including excavation, faunal correlation, microscopy, palynology, CT-scanning, three-dimensional reconstruction, isotopic analysis, and lord knows what else.
It’s the closest thing we’ll ever see to a big science effort in the little field of human evolution – like Tim White was building a supercollider under everybody’s noses.
The skeleton has been nicknamed, “Ardi” and it is 4.4 million years old. The site is Aramis, Ethiopia, in the Middle Awash field research area. The skeleton includes most of both arms, except the humeri, both hands, both feet, the right leg, the left ox coxa and part of the right ilium, a bit of sacrum, a couple of vertebrae, and a near-complete skull and dentition. It’s a bit more complete than Lucy, although preserving different parts.
The skull and pelvis were badly crushed, both of these were given CT-assisted reconstructions which are presented in separate papers (Suwa et al., 2009; Lovejoy et al., 2009c). Additionally, the series includes three papers on the paleoenvironment, a complete description of the dentition, and separate papers on the forelimbs and feet. The central paper in the series, by White and colleagues 2009b, is accompanied by a summary paper by Lovejoy examining the human-chimpanzee common ancestor in light of Ardipithecus.
As one of the papers puts it (White et al., 2009b), it represents a previously-unknown “adaptive plateau” for the hominins. Considering that really only three such “adaptive plateaux” were known before this – roughly, australopithecines, robust australopithecines, and humans – that gives some impression of the amount of difference evident in these remains from later hominins.
As I’ll describe, some substantial ambiguities and questions remain, which will no doubt shape the progress of paleoanthropology for many years to come.
Why did it take so long?
White and colleagues 2009b give a detailed overview of the state of preservation of the skeleton
The bony remains of this individual (ARA-VP-6/500) (Fig. 3) (37) are off-white in color and very poorly fossilized. Smaller elements (hand and foot bones and teeth) are mostly undistorted, but all larger limb bones are variably crushed. In the field, the fossils were so soft that they would crumble when touched. They were rescued as follows: Exposure by dental pick, bamboo,and porcupine quill probe was followed by in situ consolidation. We dampened the encasing sediment to prevent desiccation and further disintegration of the fossils during excavation. Each of the subspecimens required multiple coats of consolidant, followed by extraction in plaster and aluminum foil jackets, then additional consolidant before transport to Addis Ababa.
Still, it’s possible to overstate this explanation. Bad preservation of remains is not uncommon in archaeological contexts. In this case extreme care was obviously warranted. But just as important may have been the opportunity to interpret and guide the reconstruction of the fossil using CT scanning and other technological enhancements.
To me, that is the central message of today’s announcement and papers. The big science version of paleoanthropology is one that brings an interdisciplinary and technological approach to fossil remains right from the very start. Coordinating such an extensive interpretive project requires time – in this case, fifteen years.
I can see a shiny nugget of goodness in that depressing span of time. The initial publication of the distorted Sahelanthropus skull led to substantial disagreement about the anatomy of the skeleton. Later CT reconstruction appeared to clarify some aspects of the anatomy. Arguably, it would have been better to delay publication until the CT reconstruction could be done. Obviously White and his team wanted to minimize the opportunity for error in their interpretations. They’ve covered their bases.
But that example also shows the danger of the wait-and-see approach, in that it tends to silence skeptical inquiry. Are there morphological details of the Ardipithecus skeleton that are obscured rather than clarified by reconstruction? At the moment, we don’t know.
What was the story before today?
Back in the Cold War, CIA analysts and other folks would read carefully through Pravda and other Soviet publications, parsing every word to look for the barest hint of the Politburo’s intentions. There was a word for those people: “Kremlinologists.” It seems to me that somebody quoted in Ann Gibbons’ book, The First Human said that paleoanthropologists basically do the same thing with Ardipithecus — poring through every publication or interview, looking for hints about the fossils hidden from the field for fifteen years. I don’t remember right now who, and I don’t have the book in front of me (I reviewed the book here in 2006).
On the other hand, there are people who follow every Twitter about their favorite celebrity, recording the GPS coordinates of sightings, and “running into” them at the openings of exclusive clubs. There’s a word for these people, too: “stalkers.”
Paleoanthropologists for the last fifteen years have been a little bit of both. It’s hard to help it – Tim White let slip ten years ago that the skeleton’s locomotor style was like something out of the Star Wars cantina and, well, let’s just say that some people hear his voice and think of Weird Al Yankovic.
If you weren’t following paleoanthropology in 1994, you may not remember Ardipithecus at all. For a brief, shining moment, it was the earliest hominin. Well, except for the Lothagam mandible, but nobody ever seems to remember Lothagam. It doesn’t even show up in the current series of papers.
Then, the species fell under a veil of secrecy. The initial find was from a locality called Aramis, within the Middle Awash field concession worked by Berkeley paleoanthropologist Tim White and colleagues. The news escaped that further fossils from Aramis had been found, including a partial skeleton. After initial examination of the skeleton, White and colleagues (1995) submitted a brief comment to Nature in which they changed the genus name of the first specimens. Instead of Australopithecus ramidus, they would henceforth be Ardipithecus ramidus. After that, silence.
Research at other localities in the Middle Awash uncovered earlier remains, which Haile-Selassie and colleagues 2001; 2004 attributed to a new species of Ardipithecus, Ar. kadabba. These were never the earliest hominins (predated at their initial discovery by Orrorin and later by Sahelanthropus, but at 5.5 million years old they were not far off. In their 2004 paper, Haile-Selassie and colleagues even suggested that all of these terminal Miocene hominins actually represent variations of a single species. An unstated implication is that the species would then be called Ardipithecus tugenensis.
Sileshi Semaw and the Gona Research Project 2005 found Ardipithecus downriver from Aramis, at a locality called As Duma. This represented approximately the same age as the Aramis horizons, and showed that Ardipithecus ramidus was not just a one-off. But the remains were only a few fragments. Based on the paleoecology of the immediate find, they suggested that the species had lived in a ”mosaic” of environments, bringing together elements of the fauna from both woodland, wetland and grassy woodland facies. That interpretation becomes a point of contention in the current series of research articles.
Other hints about Ardipithecus morphology have been dropped over the years. In a key 1999 paper, Owen Lovejoy along with Martin Cohn and Tim White described the Ardipithecus pelvis. They didn’t show it or say it was Ardipithecus, but there it was nonetheless. The interpretations of tooth size in the other, more fragmentary Ardipithecus remains (referred to as “relatively small”) made the body size of the skeleton fairly clear, which enabled interpretation of an radius earlier found at Aramis as a relatively long forelimb. And so on, the main conclusions have been foreshadowed elsewhere.
One thing stood out as a surprise. Ardipithecus had a grasping foot.
Did Ardipithecus really have a grasping foot?
Short answer: Yes.
The paper about the foot remains, by Lovejoy and colleagues 2009a, is full of just the kind of impenetrable prose you’d expect for a paper about foot remains. I have a lot of affection for people who know feet, but all the “fulcrumating” has me fulminating.
If we hack through the verbiage, the feet send a clear message:
Several elements of the Ardipithecus ramidus foot are preserved, primarily in the ARA-VP-6/500 partial skeleton. The foot has a widely abducent hallux, which was not propulsive during terrestrial bipedality. However, it lacks the highly derived tarsometatarsal laxity and inversion in extant African apes that provide maximum conformity to substrates during vertical climbing. Instead, it exhibits primitive characters that maintain plantar rigidity from foot-flat through toe-off, reminiscent of some Miocene apes and Old World monkeys. Moreover, the action of the fibularis longus muscle was more like its homolog in Old World monkeys than in African apes. Phalangeal lengths were most similar to those of Gorilla. The Ardipithecus gait pattern would thus have been unique among known primates. The last common ancestor of hominids and chimpanzees was therefore a careful climber that retained adaptations to above-branch plantigrady.
“Unique among primates.” I hate it when they say that.
From the point of view of a foot specialist, this foot has many interesting aspect that can illuminate the evolution of stance and locomotor behavior in Miocene apes and the ancestors of the hominins.
From an Anthro 101 point of view, it’s an ape foot.
Still, Lovejoy and colleagues 2009a; 2009b describe the anatomy of the Ardipithecus foot as clearly different from Australopithecus, but different from chimpanzees also. The confusing thing is that it isn’t intermediate between those two forms. In their account, chimpanzee feet are specialized for more grasping, while the Ardipithecus foot retained a more generalized form. The confusion comes from parallelism in apes after Proconsul, which left Ardipithecus resembling monkeys more than apes in certain aspects of its foot anatomy, but more recent apes more than early apes in others.
The metatarsus of Ar. ramidus, chimpanzees, and gorillas presents a striking contrast to their metacarpus. Like the foot phalanges, the metatarsals also appear to have been universally shortened in all hominoids subsequent to Proconsul. The basis of this universal shortening, however, is somewhat unclear, because tarsal evolution contrasts dramatically in hominids and African apes. The modern ape foot has obviously experienced functional reorganization into a more hand-like grasping organ. The Ar. ramidus foot did not. This suggests that substantial elements of a more lever-based, propulsive structure seen in taxa such as Proconsul and Old World Monkeys [robust plantar aponeurosis; retained quadratus plantae; robust peroneal complex] were preserved in the GLCA/CLCA. These structures were sacrificed in both African ape clades to enhance pedal grasping for vertical climbing (Lovejoy et al., 2009b, 102)
That may be all I want to say about the foot for now. You can see that this is one of the most important anatomical aspects of the specimen in terms of understanding the origins of bipedality. Ardipithecus was not an obligate biped in any sense applied to Australopithecus.
OK, it wasn’t a biped, then. So how do you explain the pelvis?
The pelvis of Ardipithecus, as reconstructed by Lovejoy and colleagues 2009c, is intermediate between the chimpanzee and australopithecine morphology. In particular, the blade of the ilium is short and relatively curved compared to the long, flat chimpanzee ilium. But it does not approach the pelvis of Lucy or Sts 14 in those aspects, and the ischium is very chimpanzee-like in shape. The pubic symphysis was shorter than the long chimpanzee morphology, and the auricular surfaces appear consistent with a relatively shorter sacrum than in chimpanzees.
In reconstruction, it looks like a blend of hominin-like and chimpanzee-like anatomies. Lovejoy and colleagues further argue that the proximal femur indicates that a somewhat humanlike gluteus maximus insertion was in fact primitive for apes, with chimpanzees and gorillas having a derived non-humanlike form.
So what does this mean for locomotion? In their description (Lovejoy et al., 2009b), the African ape pelvic morphology is derived as a way of stiffening the lower back, in coordination with shortening the lumbar spine. If the African ape gluteal morphology is also derived (are you counting parallelisms yet?), then neither the ilium nor the proximal femur (excepting the possibility of bone distribution data not observable in Ardi) are useful markers of bipedality.
In other words, even though the Ardipithecus pelvis may look intermediate between chimpanzee and australopithecine morphologies, it’s not indicative of bipedality. Ardipithecus might have the locomotor morphology of the human-chimpanzee common ancestor.
To me, that seems shocking. More on this later.
What was Ardipithecus’ environment like?
The Middle Awash field team was able to do a very interesting thing in its paleoenvironmental reconstructions. The layer at Aramis containing the Ardipithecus skeleton and other remains is essentially a 3 to 5 meter thick series of paleosol, alluvial silt and fossilized bone and wood of various kinds. It is underlain by a glassy tuff and above by a basaltic tuff, which presumably represents some kind of pyroclastic event that swept through the area. The two tuffs are statistically indistinguishable in age, and the team guesses that the time between them represents something like a thousand years, maybe an order of magnitude more or less. So what they have is a thin sandwich of paleoenvironments, spread over the extent that the twin tuffs cover.
Now, this sandwich outcrops across roughly 9 km of linear distance (White et al., 2009b). So the team could sample distinct localities across this entire transect. What they found was that the line represented a range of habitats from open and grassy at one end to closed and wooded for (most) of the rest. They found Ardipithecus exclusively in associated with the wooded environment – complete with fossil wood, lots of monkeys and tragelaphines. They found no Ardipithecus at all in the localities representing more open environments. White and colleagues 2009a argue that this is a very strong test of habitat preference for Ardipithecus — it liked the trees.
Several aspects of Lower Aramis Member larger mammal assemblage abundance data constitute strong indicators of ancient biofacies and biotope. The locality-specific subassemblages are remarkably consistent in their taphonomy and taxonomy across the 7 km distance from the easternmost (SAG-VP-7) to westernmost (KUS-VP-2) Ar. ramidus localities.
Contemporaneous localities between the two tuffs farther south of the modern Sagantole drainage (SAG-VP-1 and -3, at the southeastern paleotransect pole) are relatively impoverished. They lack this diverse and abundant mammal assemblage and contain no tragelaphines, no monkeys, no fossil wood or seeds, no birds, no micromammals, and no Ardipithecus (table S1). Complementary structural, taphonomic, and isotopic data from localities on this pole of the paleotransect suggest a more open landscape that supported more crocodilians, turtles, and hippopotamids, presumably associated with water-marginal settings more axial in the drainage basin (White et al., 2009a).
That reconstruction makes sense with the locomotor anatomy. It also makes sense with the isotopic data on diet. After sampling carbon and oxygen stable isotopes in five Ardipithecus individuals, they conclude that it had a C4 plant consumption much less than later australopithecines, while higher than the very minimal value in chimpanzees, and that it habitually lived in mesic (not too wet, not too dry) habitat.
What does Ardipithecus tell us about hominin origins?
The paper by Owen Lovejoy, “Reexamining human origins in light of Ardipithecus ramidus” is possibly the most interesting in the collection. It will take me some more reflection to figure out what I think about the whole paper, but here I can abstract out the main ideas.
Much of the paper is speculative, concerning the “reproductive biology” of the human-chimpanzee common ancestor. In this paper, Lovejoy uses the acronym CLCA for ”chimpanzee last common ancestor,” which I find totally confusing. Since this is so close in time to the human-gorilla common ancestor, I’ll just take advantage of the new taxonomic scheme and call these ancestors the “stem hominines.” Lovejoy’s interest in reproductive biology is longstanding, as it formed the centerpiece of his 1981 article on human origins.
In many ways, this current article is an update of that one, because they arrive at the same singular focus: the association of canine reduction with increasing bipedality. Canines, in Lovejoy’s description, are principally a function of mating biology, and so any indirect evidence we have about the evolution of mating systems in humans or chimpanzees becomes very relevant to the factors that caused hominin origins.
Ardipithecus clearly shows that the canine reduction came first, bipedality later. Lovejoy integrates this fact into his earlier model, that the change in mating biology caused the change in locomotor strategy, as males substituted provisioning and food sharing as modes of mating competition in the place of aggression.
However, I think this is short-sighted. We already know that some degree of canine reduction occurs in several Miocene ape lineages, and that mating competition is highly variable among living apes and primates generally. What Ardipithecus shows, if we assume a connection between it and earlier candidate hominins like Sahelanthropus and Orrorin, is that the reduction of the canines preceded the evolution of effective bipedality by more than three million years. It is very difficult to conceive of mating biology as a cause of the locomotor evolution, when it is so removed from the change in time. It’s as if we stubbornly insisted that bipedality was the cause of stone tool transport.
The most interesting part of this paper is what Lovejoy says about the relevance of chimpanzees. This was also anticipated in an earlier paper, this one by Sayers and Lovejoy 2008, which argued that a chimpanzee model is too restrictive as a way of understanding the initial biology of hominin ancestors. Here, Lovejoy makes that view explicit in terms of the arboreality of Ardipithecus:
The primitive nature of the craniodental and postcranial anatomies of Ar. ramidus suggest that the CLCA, unlike African apes, was predominantly arboreal. However, all of its descendants have since developed relatively sophisticated adaptations to terrestrial locomotion. What was the CLCA’s socio-reproductive structure before these events? Whereas African apes ahve, in the past, almost invariably been selected as CLCA vicars [stand-ins], Ar. ramidus now allows us to infer that they have undergone far too many pronounced and divergent specializations to occupy such a role (Lovejoy, 2009, 74e4).
Lovejoy and colleagues discuss this concept in more detail in the paper outlining the Ardipithecus postcrania (Lovejoy et al., 2009b). I will be spending much more time on this paper, which makes several provocative assertions about developmental biology. But the conclusion of the paper
Ar. ramidus implies that African apes are adaptive cul-de-sacs rather than stages in human emergence. It also reveals an unanticipated and distinct locomotor bauplan for our last common ancestors with African apes, one based on careful climbing unpreserved in any extant form....
Ardipithecus has thus illuminated not only our own ancestry, but also that of our closest living relatives. It therefore serves as further confirmation of Darwin’s prescience: that we are only one terminal twig in the tree of life, and that our own fossil record will provide revealing and unexpected insights into the evolutionary emergence not only of ourselves, but also of our closest neighbors in its crown (Lovejoy et al., 2009b, 105–106).
OK, I don’t remember Darwin saying anything about the neighbors in our crown. But you get the point – the stem hominines weren’t like chimpanzees or gorillas.
What about the skull?
Gen Suwa headed the cranial reconstruction (Suwa et al., 2009). Most of Ardi’s skull is represented on one side or the other, except for the basicranium. The team did have the temporal bones from another specimen, ARA-VP-1/500 (previously described by White and colleagues 1994). These temporal bones were too big to fit together with Ardi’s skull, so they digitally shrank them – sort of like reducing on a photocopier.
First of all, they did a really cool thing – they reconstructed the spatial relation of the two temporal bones by aligning the semicircular canals. Those tiny structures of the inner ear are like a miniature three-dimensional coordinate frame – part of the vestibular system that senses the position of the head. I’m sure they’re the first to do that, but it’s pretty neat to be able to align two temporal bones with no contact points between them.
I mention that because their reconstruction of the temporals determines the position of the line between the carotid canals on the base of the skull – the bicarotid line. This element of anatomy was very important in our consideration of Sahelanthropus (Wolpoff et al. 2006), as the measure between the bicarotid line and basion was a key indicator of the position of the foramen magnum. Jim Ahern found that this distance actually overlaps substantially between humans and chimpanzees, and most australopithecine crania actually fall into the chimpanzee range. That makes the trait questionable as an indicator of habitual head posture. Here, Suwa and colleagues found that the basion-bicarotid distance in Ardipithecus is as low as seen in the lowest known australopithecine cranium.
Suwa and colleagues advance an old hypothesis for this basicranial form. It’s not about upright posture, it’s about the brain.
The Sahelanthropus and Ardipithecus crania securely associate a relatively short basicranium with small cranial capacity. The hominid basicranial pattern and associated morphologies [such as foramen magnum orientation] are widely held to be related to bipedality and upright posture, despite a lack of empirical evidence to clearly support a functionally based correlation. The Ar. ramidus cranium raises the alternative possibility that early hominid creanial base flexion was associated with neural reorganization that was already present in Sahelanthropus/Ardipithecus, as suggested for Pliocene Australopithecus. Such a hypothetical supposition is in part testable by both future fossil finds and by anticipated advances in our understanding of genomic expression patterns pertaining to brain function, structure, and morphogenesis (Suwa et al., 2009, 68e6).
I say, “old” because that was Raymond Dart’s interpretation of the Taung endocast — it was humanlike because of a neural reorganization.
Are they right? I’ll say this: if the basicranium does not reflect posture in these fossils, then there is no compelling evidence for posture at all.
If brain reorganization was underway in these ancient species, there’s no indication of it in the size of the brain. Ardi, like Toumaï, had a small brain — they estimate only 300–350 ml for its endocranial volume.
The rear of the skull – the nuchal plane of the occipital bone – was not preserved, so the most important remaining comparison with Sahelanthropus is the supraorbital region. This is small in Ardi, and extremely large and thick in Toumaï. Suwa and colleagues propose that this morphology matches the assessment of female sex for Ardi, which seems entirely reasonable. In the context of later Australopithecus, Ardi’s supraorbital torus might even be large for a female of its size. The difference between Ardi and Toumaï in browridge size would be surprisingly large considering the relatively slight difference in size between the two skulls. Still, with only two specimens to compare, a species with very large browridges in males might show this kind of difference on occasion.
Why don’t any of the papers have a cladogram?
This is an interesting omission, no? There’s no cladogram. What we get is this weird phylogenetic diagram that looks like a sectioned spinal cord:
If we could find a way to repair it, Homo would regain feeling.
Setting aside the aesthetics — which I’m sure were a lot of work — this set of scenarios is very unsatisfying. In all three, some version of Ardipithecus is the stem for later hominins. They haven’t shown that at all. None of the scenarios include chimpanzees or gorillas — yet no matter what you think about genetic estimates of divergence, the stem hominines were large and diverse populations with long-term interactions. Maybe these stopped before 6 million years ago, but none of the genetic data suggest that now at all.
Cladograms would oversimplify some aspects that should be considered complex, but maybe we could have one?
White and colleagues 2009b give a long table of “derived” characters in Ardipithecus and Australopithecus, but they are “derived” only with reference to their inferred state in the human-chimpanzee LCA. But elsewhere in these papers, they argue that some of these “derived” characters are actually primitive morphologies for apes, for which chimpanzees are independently derived. For many of the dental features, if we supposed a Miocene ape ancestor, the broadened mandibular body, thicker enamel and so on would look primitive, not derived. In the table, they list upper and lower canine traits separately, and break them up into six or more for each. That’s a quick way of making one morphological change look like twelve or more instances of independent evolution. Talk about atomizing traits!
So I wonder if a real cladistic analysis might not place Ardipithecus with the australopithecines. Especially if it included a proper sampling of Miocene ape taxa.
Maybe worse, a real cladistic analysis that did place Ardi with Australopithecus would probably put the earlier Ardipithecus kadabba as an outgroup to both. That would make Ardipithecus paraphyletic.
I wouldn’t typically care, because I don’t think taxonomic rules should direct the science. But it does seem like a delicious taxonomic dilemma. The likely solution would be to lump Ar. kadabba with either Orrorin or Sahelanthropus, or both. Orrorin kadabba would have priority.
But if one were feeling saucy, she could publish the cladistic analysis on the teeth, point out the dilemma, and then offer a novel genus name for the Kadabba sample. Maybe somebody’s already thought of that — there are a lot of journals out there.
Haile-Selassie Y. 2001. Late Miocene hominids from the Middle Awash, Ethiopia. Nature 412:178–181.
Haile-Selassie Y, Suwa G, White TD. 2004. Late Miocene teeth from Middle Awash, Ethiopia, and early hominid dental evolution. Science 303:1503–1505.
Lovejoy CO. 2009. Reexamining human origins in light of Ardipithecus ramidus. Science 326:74e1–74e8.
Lovejoy CO, Latimer B, Suwa G, Asfaw B, White TD. 2009a. Combining prehension and propulsion: The foot of Ardipithecus ramidus. Science 326:72e1–72e8.
Lovejoy CO, Suwa G, Simpson SW, Matternes JH, White TD. 2009b. The great divides: Ardipithecus ramidus reveals the postcrania of our last common ancestors with African apes. Science 326:100–106.
Lovejoy CO, Suwa G, Spurlock L, Asfaw B, White TD. 2009c. The pelvis and femur of Ardipithecus ramidus: The emergence of upright walking. Science 326.
Sayers K, Lovejoy CO. 2008. The chimpanzee has no clothes: A critical examination of Pan troglodytes in models of human evolution. Curr Anthropol 49:87–114.
Semaw S, Simpson SW, Quade J, Renne PR, Butler RF, McIntosh WC, Levin N, Dominguez-Rodrigo M, Rogers MJ. 2005. Early Pliocene hominids from Gona, Ethiopia. Nature 433:301–305.
Suwa G, Asfaw B, Kono RT, Kubo D, Lovejoy CO, White TD. 2009. The Ardipithecus ramidus skull and its implications for hominid origins. Science 326:68e1–68e7.
White TD, Ambrose SH, Suwa G, Su DF, DeGusta D, Bernor RL, Boisserie JR, Brunet M, Delson E, Frost S, Garcia N, Giaourtsakis IX, Haile-Selassie Y, Howell FC, Lehmann T, Likius A, Pehlevan C, Saegusa H, Semprebon G, Teaford M, Vrba E. 2009a. Macrovertebrate paleontology and the Pliocene habitat of Ardipithecus ramidus. Science 326:87–93.
White TD, Asfaw B, Beyene Y, Haile-Selassie Y, Lovejoy CO, Suwa G, WoldeGabriel G. 2009b. Ardipithecus ramidus and the paleobiology of early hominids. Science 326:75–86.