Header image

john hawks weblog

paleoanthropology, genetics and evolution

Photo Credit: SK 46, CC-BY-NC-ND John Hawks, courtesy of Ditsong Museum of Natural History

Delayed maturation and Australopithecus robustus

I got a Twitter question today about whether any fossil hominins may have had delayed secondary development. The question arises in the context of development in male orangutans, which is the most extreme case of bimaturism (two distinct developmental trajectories) in living great apes. I was surprised to find I hadn’t written about that topic before. There has been at least one serious suggestion that robust australopithecines had a long period of arrested development in males, similar to silverback gorillas.

Gorillas and orangutans both exhibit substantial body size dimorphism, but males of these species develop in very different ways. Silverback male gorillas develop their large body size over a period of many years. Mountain gorilla males reach the adult female body size by around age 8-11, and lowland gorilla males are later at around age 11-14 (Breuer et al. 2009). At that stage, they are called “blackbacks”, and they transition into being silverbacks by a fairly long period of growth, up to 7 years or so. In western lowland gorillas, full adult body size is reached sometime around age 18 for males, in mountain gorillas around age 14-15.

Male orangutans can exhibit a delayed course of development of body size and secondary sexual characteristics, including their famous facial flanges. Male orangutans reach the size of adult females, and then what happens varies. Some go on to develop full adult body size and secondary sex characteristics rapidly, even within a single year. Others go into a prolonged period of stasis with little growth, which can last for many years.

Orangutan male with facial flanges
Male orangutan with facial flanges. Photo Credit: ucumari via Compfight cc

Not all orangutan males have this developmental hiatus. Some of them go on to develop large body size and flanges immediately upon reaching sexual maturation, others reach sexual maturation and are able to reproduce, but remain smaller and flangeless for many years. The development of flanges seems to be mediated by testosterone, with low-testosterone males more likely to suspend their flange development (Emery Thompson et al. 2012).

The flanged males have greater access to females in their wild habitat, even when there are unflanged males in the same home range (Utami et al. 2002). That observation suggests that the delayed development is a varying evolutionary strategy, with some males opting to take the stress and cost of development quickly, and others taking on these costs later. Development is costly, and a large male that exhibits sexual signals like facial flanges will be thrust into a stressful and potentially costly competition with any dominant male that already exists in the area. Anne Maggioncalda and colleagues (2002) showed that development of the secondary sexual characteristics was associated with very high stress levels in male orangutans, much more so than in adolescent males who delayed development. It may be natural to assume that the delayed development may be socially mediated. That is, males who grow up with a dominant male already resident might receive some signal that delays their development. But as discussed by Melissa Emery Thompson and colleagues (2012), the story is not so simple, as experiments in captivity are not consistent.

Now, that’s the story with sexually dimorphic apes and development. What about hominins?

TM 1517
TM 1517 Au. robustus mandible, skull and distal humerus, from Kromdraai, South Africa. CC-BY-NC-ND John Hawks, courtesy of Ditsong Museum of Natural History.

The impetus of this post is the study from 2006 by Charles Lockwood and colleagues. They considered the sample of crania and mandibles assigned to Australopithecus (Paranthropus) robustus from Swartkrans, Kromdraai and Drimolen, South Africa. They produced two rankings of these Au. robustus specimens, one in order of size, and another in order of age.

They could not take the same set of measurements on all specimens, so their ranking is just a general size ranking – a “big” mandible might be like SK 12 where nearly the whole thing is there, or it might be just a chunk of a mandible that preserves the corpus breadth. The age ranking is based on tooth wear, and ranges from third molars just erupted up to profound wear.

What they found is that the two rankings are significantly associated. Across their sample, bigger mandibles tend to be more worn than smaller ones. With the maxillae, they attempted to differentiate male from female samples, and then showed that the larger mandibles tend to be more worn among the males. The small group of four female maxillary specimens has no association of size and wear.

The crowns of the teeth do not grow during adolescence or after sexual maturity, but the maxilla and mandible may grow substantially – just as the faces and jaws of gorillas and orangutans become somewhat larger during the development of adult males. Lockwood and colleagues inferred that the greater degree of wear on larger mandibles and maxillae is a sign of prolonged growth among the males in the sample:

From this analysis, we infer that maximum (presumably male) size was greater among old adults than young adults (Fig. 1). Old adults in the higher size ranks also had the most well-developed morphology with respect to features diagnostic of the P. robustus face, such as the anterior pillars and maxillary trigon (16) (Fig. 2). Based on the ontogeny of sexual dimorphism in modern primates (17), we interpret this pattern as continued growth in males between early skeletal adulthood and full maturity. Minimum size, on the other hand, occurs throughout the age range. For example, SK 21 is the oldest specimen in the sample of maxillas, but also the smallest. Because small, relatively gracile individuals occur at every age, we conclude that females have reached full skeletal size by the time M3 has erupted or soon thereafter.

To arrive at this conclusion, they needed to assume that a very high fraction of the specimens in the sample – especially at Swartkrans – are males. Lockwood and colleagues reflected on the importance of the overrepresentation of males in the Swartkrans deposit:

An abundance of males is perhaps not surprising in a fossil sample that resulted largely from predation. Direct evidence of carnivore activity is present on several hominin specimens at Swartkrans, and member 1 of this site is among the most definitive examples of a predator-accumulated assemblage of hominins (27, 29). In dimorphic primates, nondominant males spend more time alone, on the periphery of a social group, or in small all-male bands (30). Solitary behavior places males at risk (31, 32). For example, when male baboons disperse, they suffer a mortality rate at least three times as great as that of group-living males or females (32). This degree of difference in mortality matches the male bias at Swartkrans. Females were apparently more shielded from predation. Putting this observation together with the conclusions about bimaturism and sexual dimorphism, we infer that the distribution of food sources allowed stable groups of P. robustus females to form in response to predation pressure, and males in turn sought to monopolize reproductive access to these groups (33).

This hypothesis hinges on the assumption that the Drimolen skull, DNH 7, was the first relatively complete specimen to be female. All the others, including the smallish Kromdraii TM 1517 and SK 48 from Swartkrans, must then be small males. The authors support their assumption in the online supplement with reference to the more complete craniofacial specimens, including SK 48, SK 46, TM 1517, SK 12, SK 83 and DNH 7. This is too small a set to make any statistical conclusions from – the authors rely on a larger set of more fragmentary remains to do that. But it is the set where they could consider multiple characters that might be related to sex. And among this set, the smaller male specimens also have smaller teeth. That’s not a likely consequence of delayed maturation in males. It might reflect body size variation in fully mature males, it might mean that these “smaller males” are actually females, or it might just be a statistical fluke. Some more examination might be useful.

Personally, I think it’s a good hypothesis that sexually dimorphic australopiths had delayed male maturation, if not exactly like the developmental trajectory of gorillas, at least broadly similar to it. This is the scenario that Lockwood and colleagues favored.

At the same time, it seems really unlikely that early hominins had the kind of strategy variation found in orangutan males. The arrested development morph in orangutans seems to be driven by two factors: the very high stress and cost associated with development of the secondary sex characters in males, and the greater opportunity that young males have to sneak around within the highly dispersed orangutan group structure. Early hominins lived in a very different ecology than orangutans, giving dominant males much more ability to monitor the social interactions of young males. Young male australopiths may have faced a similar situation as young male baboons or mandrills – forced by dominant males out of the center of groups and toward the margins where predation is higher. That part of the scenario sketched by Lockwood and colleagues seems very credible, and it does help to make some sense of Swartkrans.

That’s not the last word on the subject as much has been written about early hominin group structure since 2006. I’ll leave it for now and return to it somewhat later.


Breuer, T., Hockemba, M. B.-N., Olejniczak, C., Parnell, R. J. and Stokes, E. J. (2009), Physical maturation, life-history classes and age estimates of free-ranging western gorillas—insights from Mbeli Bai, Republic of Congo. Am. J. Primatol., 71: 106–119. doi: 10.1002/ajp.20628

Emery Thompson M, Zhou A, Knott CD (2012) Low Testosterone Correlates with Delayed Development in Male Orangutans. PLoS ONE 7(10): e47282. doi:10.1371/journal.pone.0047282

Lockwood CA, Menter CG, Moggi-Cecchi J, Keyser AW (2007) Extended male growth in a fossil hominin species. Science, 318(5855), 1443-1446. doi:10.1126/science.1149211

Maggioncalda A, Czekala NM, Sapolsky RM (2002) Male orangutan subadulthood: a new twist on the relationships between chronic stress and developmental arrest. American Journal of Physical Anthropology 118: 25–32. doi: 10.1002/ajpa.10074

Utami SS, Goossens B, Bruford MW, de RuiterJR, van Hooff JARAM (2002) Male bimaturism and reproductive success in Sumatran orangutans. Behavioral Ecology 13: 643–652. doi: 10.1093/beheco/13.5.643