It has always been suggested, by a wide range of evidence, that australopiths in general, and robust australopiths in particular, have a higer degree of sexual dimorphism than chimps, and possibly, dimorphism in body size as high as one sees in any ape. Resent research on growth patterns, just published in Science, examines this.
In dimorphic primates, the larger size of the male is usually obtained by an extended period of growth of the males, so they essentially "grow past" the females. In addition, there are usually additional traits that the larger, adult males have that re not seen in the young males. IN other words, younger and older females in most primates (but not humans) look fairly similar to each other, while younger and older males look more different from each other. You can estimate the age of a primate on the basis of a number of factors such as tooth wear, or at least, you can rank specimens on the basis of tooth wear and figure that you have a series of differing age individuals, in order.
This brief study simply ranks individual remains of South African Paranthropus robustus by size and apparent age. The larger individuals are older, but the smaller individuals are both younger and older. The researchers then conclude that there is dimorphism and that this involves extended growth periods in males.
They also discovered that there are far more males in the sample than females. This is reminiscent of a study conducted by Diana Webber (a former student of mine) for her honors thesis, in which she looked at mangabeys, in comparison with a study done by Wrangham and others of the remains under a monkey-eating eagle's nest in an area where these eagles tended to eat a lot of mangabeys. I have observed male mangabeys turning on attacking eagles and intervening between them and the troop, and the only monkey's I've ever seen killed by eagles were big males. In addition, the Efe (Pygmy) hunters tend to end up with adult male mangabeys when they go for these monkeys. Dian'as study of skeletons of mangabies collected by a researcher in the 1950s showed that while females and young males were shot in random places and in random directions, the adult males were all shot from the front. The researchers working with the South African material hypothesis that there is a link between predation and high rates of male deaths, in relation to this particular set of fossil depoists.
The P. robustus pattern contrasts with that of some other australopithecine collections. For example, the sample of Australopithecus africanus is either biased toward females or shows no bias. The A. africanus accumulation from Sterkfontein member 4 may also be the product of predator behavior, but this conclusion is less certain than for the Swartkrans sample. If both Swartkrans member 1 and Sterkfontein member 4 hominin collections are largely attributable to carnivore activity, the difference in sex bias raises the possibility that P. robustus and A. africanus differed in social structure.
What the researchers fail to consider is the possibility of trimorphy of the kind found in orangs. It is possible that among the males there are both small and large morphs.
LOCKWOOD, C. A., MENTER, C. G., MOGGI-CECCHI, J. & KEYSER, A.W. (2007): Extended male growth in a fossil hominin species.. Science, 318, 1443-6.