Thomas "Cody" Prang

A morphometric assessment of fossil hominin feet in the Turkana Basin

This talk will review the functional and evolutionary implications of fossil hominin foot morphology in the Turkana Basin with special emphasis on a new partial foot (KNM-ER 64062).

FULL TRANSCRIPT

Okay, good afternoon. I just wanted to echo the sentiments of several other people in acknowledging and thanking Stony Brook and TBI and the conference organizers, Lawrence, Fred, and also Alicia, for bringing us all together to celebrate the life and scientific legacy of Richard Leakey. Unfortunately, I have no isotope data or rocks, and I am discussing a talus but not the geological kind. So, when Bernard and others reminded us of the falsification of the single species hypothesis in the 1970s, it reminded me that we're sort of experiencing something similar in the sort of locomotor functional morphology side of things, where in recent years it's become obvious that fossil hominid and post-crania are sort of far more diverse than previously thought, and that there's maybe more locomotor diversity among good, well-adapted bipedal hominins. 

And so, as bipeds, humans have numerous postcranial morphologies that improve locomotor economy and endurance compared to chimpanzees and almost certainly compared to the very earliest fossil hominins such as Ardipithecus and maybe also compared to Australopithecus as well. And these things include increased hind limb length, spring-like soft tissues in the lower limb, large muscle volume, and a large proportion of slow twitch muscle fibers and on and on and on. The human foot is one of those regions that contributes to locomotive economy and endurance in humans. And we're well aware of all the really big changes between the feet of apes and other primates and the human foot, including the loss of a grasping hallux. Note that this is a nice smaller pointer I appreciate that. This is a dorsal view, a top view of a right foot, and this is a bottom view or a planter view of the opposite side. 

But we know that humans evolved a more adducted big toe or hallux with the digits more in line with the other toes from some kind of ancestor that was adapted to climbing an arboreality. There are numerous other features of the foot in humans that differ between chimpanzees and other primates, including more obvious things like having reduced toe length compared to that of a chimpanzee. But there are more subtle features of the foot, especially of the midtarsal joint and on some of the other intrinsic joints of the foot in humans that likely reflect the evolution of longitudinal and transverse arches that also likely contribute to economy of locomotion. 

And so this is an inside view where a medial view of the foot skeleton in a human compared to a chimp and the human foot is clearly sort of arched from front to back in a longitudinal manner simply due to the geometry of the tarsal and metatarsal elements. So the human foot has a structural arch compared to the foot of a chimpanzee. The transverse arch of the human foot, which runs side to side or mediolaterally, has recently been shown to contribute substantially to the stiffness of the foot during the stance phase of bipedalism. And so morphological sort of predictors of longitudinal and transverse arch either presence or functional mechanism are less well understood compared to things like the presence or absence of a grasping hallux, but yet they are sort of key to understanding locomotor morphology in fossil hominids. 

Now, in the last, I don't know, 10 or 15 years or so, our knowledge of foot morphology and fossil hominins has expanded somewhat, which is surprising to say given that we only have a handful of partial feet in the early part of the human fossil record. Of course, there are many other specimens that are not identifiable to particular individuals that represent these attacks through time and space. But the Ardipithecus foot first published in 2009 was a huge surprise because it displayed a divergent grasping hallux morphology or big toe morphology in a skeleton that showed some hints for early bipedalism.  Yohannes earlier this week reminded us about the Burtele foot and its more overall, more primitive morphology associated with enhanced grasping compared to the foot of Australopithecus afarensis. Most of these feet sort of post afarensis and post OH 8, no offense to anyone studying foot morphology much later in time. Most of those foot elements are essentially modern human-like in their functional morphology with small differences between us and those earlier populations. And this is also a little bit of a cheat here because the Dmanisi foot is actually a composite foot from multiple individuals, so there's even fewer partial hominin feet than this. 

And so most of those big changes, those big phenotypic differences between humans and apes are well-documented even in a small sample of partial feet. We can add another partial foot to this list. This is KNM-ER 64062, this is a partial foot that was discovered by Meave and Louise Leakey's team some time ago. And as you can see, it's a fairly complete foot compared to the other known partial feat in the early part of the fossil record. And so given its completeness, given its age and its location, there's a lot to say about this important specimen. Most of that will be saved for a forthcoming description and analysis in collaboration with Louise and Carrie and several other people. But today I'm just going to talk about the talus of this particular specimen in comparison to the other tali that are sort of classically known from the Turkana Basin. 

So this is a side view, again, a medial view of a chimpanzee foot and the talus is the element, the bone that connects the leg to the rest of the foot. And from a morphological or functional morphological perspective, tali morphology is significant because it preserves joint surface associated with ankle joint dorsiflexion and plantar flexion, right, bringing the foot upward or downward, which are important components of both chimpanzee and human positional repertoires. But the talus also preserves joint morphology beneath the bone related to the use of inversion and eversion. And so, the talus historically and in the literature has been a focal element in the study of the functional morphology of the human foot. 

So, this is essentially the highlights of fossil hominin foot material from the Turkana Basin. There are other elements including some fragmentary metatarsals. There's a more recently published complete metatarsal from Omo. But most of the time when we think about fossil hominin foot elements from this region, we're usually thinking of these tali, which have been written about over and over and over again in various papers. Notably, there are two tali, KNM-ER 1464 and 1476, which in the literature are sort of presumed Paranthropus boisei. But unfortunately, none of these individuals have craniodental material directly associated with them, which makes their taxonomic assignment quite challenging. 5428 is a very interesting specimen because it's slightly later in time compared to 1476 and 1464. And in the literature, it's presumed to represent Homo erectus. I think probably because it's big outside of that there aren't many sort of morphological criteria that we could use to reasonably assess the taxonomic affinity of those specimens. 

The Omo talus in 2006 was noted by Jibo and Schwartz to share similarities with more derived hominin morphologies and dissimilarities from TM 1517 a Paranthropus robustus or partial talus from Kromdraai and also compared to OH 8. And so, with the addition of 64062 and the talus in particular to this sort of collection of elements, it becomes clear that there seems to be two groups of tali morphology represented here. And one of them is associated with or is most prominently represented by the morphology of 1464. And so, you can see that the tailored trochlear surface itself in the tailored specimens above is very wide compared to the narrow surfaces of the tali below. And so, this is a behind view, a posterior view of the talus, the calcaneus, the heel bone would be down here, and the big toe would be somewhere over here. In association with that width there's also an asymmetry of the trochlea surface in 1464, 1476, and to a lesser extent, 5428, where the outside part of the trochlea, the lateral side on the right is much wider than the medial side.

The trochlea is very deeply grooved, and I significantly think this articular surface for the fibula is distinct in those specimens where the sort of outside wall has a more vertical orientation with a larger angle between that wall and the rest of the articular surface, whereas these guys tend to have much more gently sloping articular morphology with a less vertically oriented wall. Lastly, 1464 on one end compared to 64062 on the other differ greatly in the size and morphology of a groove on the back surface of the talus, which is this groove here, the flexor hallucis long groove. And the flexor hallucis longest groove houses the tendon for a muscle outside of the foot that causes flexion of the hallux. So the features observed among these tali variably tend to be associated more with ape-like morphologies with a more sort of arboreal locomotive repertoire. So, the question is, might there be two sort of distinct morphs in tali morphology in this region around this time? And so, I'm going to present two metrics, two univariate metrics that are thought to reflect functiona things of relevance as well as the overall sort of 3D shape of the tali specimens. 

So this is one metric which captures the grooving of the trochlea, and it's a metric that's quantified with something called a quadric surface. And a quadric surface is essentially just like a mathematical surface that's generated by a fit to an articular surface that's separated or segmented from a 3D model of a bone. And so quadric surface fitting was popularized by Matt Tocheri and Mary Marzke a number of years ago for quantifying the saddle shape curvatures of the trapezial metacarpal joint. And so here we're just looking at one dimension, the mediolateral or side to side dimension of that curvature. And so humans have dramatically reduced curvature compared to most African apes. And the traditional functional explanation for this is that African apes tend to use more varied foot positions during vertical climbing, including dorsiflexion and inversion compared to humans. And that increased trochlear grooving may help to stabilize the talus within the ankle joint. Notably, mountain gorillas overlap substantially with humans, which Matt Tocheri also showed several years ago reflecting their higher frequency of terrestriality in their locomotive repertoire. And so this is a variable that plausibly reflects frequency of arboreality and terrestriality reality rather than sort of bipedalism versus quadrupedal.  

So, we can add the fossil hominins in, we see that fossil hominins are quite variable in their trochlear grooving. If you've read any of the tali papers, you'll know that each of those papers always includes a posterior view of the talus to document variation in trochlear grooving. But it's been quite challenging to provide a quantitative assessment of that. And so, we can see that 1464 has the greatest trochlear grooving of any of the fossil hominins followed by OH 8, followed by 5428, followed by 1476. So that's sort of more sort of primitive or ape-like tali morph seems to be associated with increased trochlear grooving, whereas the 64062 specimen has a very flat, very human-like trochlear morphology, which overlaps with Homo naledi. It falls just below LB1 Homo floresiensis and just above the Dmanisi specimen D4110.

I'm moving on to the second metric. This is an angular metric that represents the orientation of the talus head, the front part of the talus relative to the ankle joint. And I mentioned the structural arch of the human foot. And one of the reasons why humans have that structural arch is because it enables through the right amount of stability and mobility, it enables the stretch of the plantar soft tissue on the bottom of the foot, the plantar fascia, or the plantar aponeurosis. And if those plantar tissues become overstretched, it can be painful, which is known as plantar fasciitis. But it turns out that the stretch of those plantar soft tissues temporarily stores elastic potential energy, which is mostly returned to the system. And that's sort of the energy saving or the mechanical energy saving mechanism of the human longitudinal arch. And so due to that difference between humans with an arched foot and apes with a non-arched foot, we would expect humans to have more plantarly oriented or downwardly oriented articular surfaces of the talus and also of the calcaneus. So this is the angle between the ankle joint and the navicular facet. The front of the talus called the tali navicular joint angle, and humans have a much more plantarly oriented facet compared to apes, which is associated with the raising of the midfoot and a structurally arched foot. 

Interestingly, most of the fossil hominins actually don't have particularly human-like tali sort of morphological geometry. The most humanlike of them includes 64062, the Omo talus, and interestingly enough, Australopithecus afarensis. I think there's a really good reason for that pattern in afarensis. I'm happy to talk about it at the bar later. Unfortunately, we don't have that much time today to dive into the complexities of Australopith morphology, but like before 5428 and 1464 have sort of more ape-like angular orientations implying that there's variation in these geometries among fossil hominins that reflects differences in longitudinal arch development. 

And finally, I'll show some data comparing the 3D shape, the overall sort of 3D shape of the talus in humans, chimpanzees, and gorillas to this sample of fossil hominins. And I'm lucky to have a talk earlier today explaining geometric morphometrics. But essentially this is an approach where homologous landmarks or hopefully homologous landmarks are collected across a range of individuals, including so-called fixed landmarks in red landmarks that slide along curves in yellow and landmarks that slide across surfaces in blue. And so after a generalized procrustes analysis or an alignment of the coordinate data, typically those shape coordinates are subjected to principle components analysis, which is a multivariate technique that enables us to take a lot of variables and condense them down into fewer new variables called principle components that are each sort of combinations of the original dataset, right? So, we can go from many landmark variables to many fewer variables using PCA. 

And so here we've got humans on the right side of PC1 and African apes on the left side of PC1 and variation along PC1 is associated in humans with an overall sort of stouter more robust talus with a less medially oriented or inside oriented tali head, a flatter trochlear surface that's less angled, a robust tali navicular joint or tali head here that's downwardly oriented and some subtle modifications to the angular relationships of the joints beneath the talus, the subtalar facets. And like most analyses of fossil hominin morphology, the fossils fall intermediately between humans and apes, which is no surprise. But one interesting thing that's happening here is that 1476, 1464, and 5428 all sort of fall along the same area along the higher end of PC2, whereas the 64062 foot and the Omo specimen fall further down more toward the sort of negative side of PC2, we can add known homo specimens like Homo floresiensis. 

There's two specimens of Homo naledi and the Dmanisi specimen, which also tend to fall sort of on that more negative side of PC2. One thing to note is that despite the importance of the OH 8 specimen historically and in the literature and all of that, the OH 8 talus is incomplete. And so, some degree of virtual reconstruction is necessary to include a fossil like OH 8 in an analysis such as this. Variation along PC2 shows that higher values like gorillas versus chimpanzees or 1464 versus 64062 is associated with a more robust tailored trochlea with increased side to side sort of width of the anterior part of the ankle joint, which is plausibly associated in apes with vertical climbing, a less curved trochlear surface, an overall increased robusticity.

So, it seems that 1464 and 1476 do seem to share morphologies in common with one another that differ from 64062, the Omo specimen and the known sort of early Homo or more basal Homo specimens including Dmanisi and Homo naledi. And so, there's a really interesting quote from the 1975 paper in which Leakey and Walker say alternative concepts relating to single species versus multiple species concepts, especially those concerning niche divergence in sympatry should now be formulated. And one possibility is that locomotive behavior is one part of that sort of divergence as represented in something such as tailor morphology. And so just before wrapping up, I just want to take a brief moment to acknowledge and thank Bill Jungers for his contributions to the field and also the impact that he had on me personally and also professionally. So thank you so much.

View all Human Evolution videos