The discussion about dimorphism between the sexes in the genus Australopithecus has been an ongoing debate in the world of physical anthropology and paleoanthropology. For many people interested in human evolution, this is an interesting question because sexual dimorphism can explain early mating systems.
I won’t get into that just yet, but I will overview some of the key points raised in the last 25 years with a specific emphasis on recent back and forth brouhah from two different camps. To start, in all honesty, I don’t know who really sparked the debate about investigating sexual dimorphism in early hominids, but I think it is safe to guess that it all happened after Lucy was discovered in the mid ’70’s.
One of the first traceable accounts of the debate started with one of my undergraduate professors. She introduced us to the questions she raised in the ’80’s about whether the different species of australopithecines are truly separate species or representatives of extremes in sexual dimorphism. At the time, and even now, a probability of this magnitude did not come without controversy. It got a lot of people talking about robust australopithecines being representatives of the male sex, whereas the more gracile ones are females. This consideration has been called going ‘beyond the orthodoxy’ by some.
Since then, a very thorough compilation of the debate was synthesized by Frederick Grine in the late ’80’s, who by the way, just recently put out a new revision of the text, “Evolutionary History of the “Robust” Australopithecines.” The debate has been going on still, but the general consensus fell in line with those made by Henry McHenry in 1992: Australopithecus afarensis was more dimorophic than modern day humans.
Much of the problem with the debate was that there just are not that many samples of australopithecines to reconstruct a good, statistically valid, understanding about the amount of variation in a population. That’s one of the biggest shortcomings in anthropology… the fact that we’re trying to extract as much information as possible from limited number samples. It is no fault of anthropology, that’s the nature of the research… When one finds a fossil, that’s basically all they find, a single fossil. We’d all like to find complete skeletons, but that’s rarely the case.
Furthermore, Australopithecus fossils do not come out from the same stratigraphic layer nor the same location. The fossils that do come out from the same layer and location, are most often not attributable to the same individuals (i.e. not from one definite skeleton). Nor are the fossils consistent in quality and composition of the physical fossil. In other words, the fossil record of Australopithecus, one of the more exhaustive collections in hominin species, is still not complete enough.
Enter the discovery of the A.L. 333 locality in Hadar, Ethiopia. What was unearthed at A.L. 333, is what people call the “First Family” of Australopithecus, an abundant but fragmentary collection of fossils from the same stratigraphic layer. The discovery was made from the same area where Lucy was found, and in the mid ’70’s, but it wasn’t published until the early ’80’s. The assemblage contained over 200 fragments of fossil bone with a minimum number of individuals at 9 and a maximum number at 22. Since they were all concentrated in the same stratigraphic layer and in close proximity, many consider the site a simultaneous death assemblage — basically a catastrophic event killed at least 9 early hominins in one fell swoop. This discovery seemed fulfill of the biggest shortcomings in analyzing Australopithecus, having a comparatively large collection of fossils from the same time, come from the same locality.
Fast forward to August of 2003, when PNAS published a paper that analyzed the dimorphism present in A.L. 333 collection. The title basically tells it all, “Sexual dimorphism in Australopithecus afarensis was similar to that of modern humans.” The authors made about 30 different measurements of bones from A.L. 333 and compared them to Lucy as well as the Makapansgat humeri. Unfortunately, because 9 complete skeletons weren’t found, the fossils used in the analysis are not all one anatomical element… rather, some are humeri, some are radii, others are femora, etc. The used the measurements from these different anatomical elements to make an estimation on the femoral head diameter, which is based off the best preserved Australopithecus afarensis femoral head, Lucy’s. This is called the template method, when a certain element is not represented in all the samples, so another element’s measurements that is present in the sample, is used to make a relative estimation about the missing element. To all you who live and breath statistics, this maybe a big bad faux pas — but really what else are you to do, make no analysis at all?
The authors of this PNAS paper extend their conclusion from the lack of dimorphism present in Australopithecus afarensis collection in A.L.333, to a discussion about a potential monogamous behavior in Australopithecus afarensis. In humans, the lack of a large dimorphism between males and females, is linked to monogamy. Why? Since human babies are born at a less developed stage and require more help and care to raise the baby, the males have lost their dimorphic stature while focusing on raising the baby. Without a large stature they aren’t able to harem up females that more dimorphic species like silverback gorillas have done. I’m not to sure I understand this linkage because it seems like one of those, “we don’t have empirical evidence but this story makes sense,” sorta situations.
These two conclusions irked the camp that want to think Australopithecus afarensis was largely dimporhic. And the most outspoken voiced their opinion in a 2005 issues of the Journal of Human Evolution. The paper, “Sexual dimorphism in Australopithecus afarensis revisited: How strong is the case for a human-like pattern of dimorphism?“John Hawks covered this two years ago, so I won’t rehash the points too much. In a nut shell, the authors of this slam the A. L. 333 study on several points, the primary one being that the A. L. 333 specimens represent males. They also raise questions about whether or not skeletal measurements, the only thing that can be analyzed from fossil remains, correlate with sexual dimorphism as well as if skeletal dimorphism can be used to predict social or behavioral features of primates. The authors were especially irked with prediction of monogamy due to the lack of dimorphism in the A.L. 333 samples.
The A.L 333 study offered up a quick and terse response that same year. Here’s the paper, “The case is unchanged and remains robust: Australopithecus afarensis exhibits only moderate skeletal dimorphism. A reply to Plavcan et al. (2005).” The authors address how skeletal dimorphism is only one aspect of size dimorphism, and how it can never tell us the total body size dimorphism by itself. About the monogamy claim, they clarify that no living primate model provides a basis to construct for A. afarensis sexual behavior, and therefore it is necessary to construct one. That’s why they merely raise the possibility that A. afarensis wasn’t exclusively, definitively polygamous, it coulda been monogamous as well as polygamous.
We’ve heard little about this debate since this 2005 cat fight. That all changed when this paper came out late last year in the American Journal of Physical Anthropology, “Strong postcranial size dimorphism in Australopithecus afarensis: Results from two new resampling methods for multivariate data sets with missing data.” The authors of this new paper offer up a really promising new methods to take out inaccuracies of estimating a missing measurement from another element. Furthermore, they take a multivariate analysis, which means more than one measurement which brings in more data that can be compared and scrutinized. They ultimately figure out that A. afarensis was as dimorphic as gorillas and orangutans.
I read this paper and man, I must admit, I have no idea what their two methods they are. I got so frustrated in making sense of out it. They claim their methods can compensate for inconsistencies using template method, but it just sings to me as an abstracted, obfuscated template method. Furthermore, if you wanna make sense of the method, maybe perhaps crunch out the numbers yourself, the authors provide you with this mind-numbing array of equations that I’ve uploaded to your right. I’m sorry if I can’t provide a more intelligible critique but really what are they trying to do — make their data so unapproachable that it can never be validated nor refuted? I’ve been told by many different professionals that if the data and the methods seems too complex, if they bring derivatives and calculus into to the fold, then something is fishy.
So the latest word is that A. afarensis dimorphism is similar to gorillas and orangutans — who’s male and female skeletons look like completely different species to the untrained eye! The jury is still out, there definitely seems to be two camps that are continually hashing this out, and until we find more complete comparable specimens, we’ll be reading this sorta stuff for a long time.
I do want to make a comment about where I stand on the issue. Like many others, I want to believe that A. afarensis was dimorphic… at least comparable to the levels seen in chimpanzee and bonobos. Why? Because, they are one of the earliest hominins we know of and have a lot of specimens of. Furthermore, from what we know of A. afarensis birth canal size and brain size, they weren’t as undeveloped as humans. Rather, they were born at similar developmental stages seen in chimpanzees. The argument that males lost dimorphism to help raise the kids doesn’t apply when we consider this fact. Now, the evidence from A.L. 333 tells us otherwise, and I trust that evidence because it comes from one stratigraphic layer and time. Furthermore, I trust the simple straight forward analysis on the A.L. 333. Not only does it come from a reputable source, Owen Lovejoy, perhaps one of the more preeminent biological anthropologists out there, but it also is followable — the statistical analysis, the data, the methods are clear and thorough.
- Gordon, A.D., Green, D.J., Richmond, B.G. (2008). Strong postcranial size dimorphism in Australopithecus afarensis: Results from two new resampling methods for multivariate data sets with missing data. American Journal of Physical Anthropology, 135(3), 311-328. DOI: 10.1002/ajpa.20745