Gigantopithecus has often been described as a bamboo eater, based on analogy with another kind of large herbivore in China, the giant panda. Giant pandas have several specialized feeding adaptations to support their bamboo diet. The most famous of these is the expansion of what we would call a wrist bone in most mammals, a sesamoid bone associated with the distal radius. In giant pandas, this bone projects from the arm in a way that makes it function similar to an additional digit for the hand, a solution described as "the panda's thumb". This "thumb" is used to grip the bamboo stems so that the teeth can work through the indigestible fiber and woody portions of the bamboo stems into the softer shoots. As a famous example in the history of evolutionary biology, the panda's thumb was celebrated by the evolutionary biologist Stephen Jay Gould as a unique evolutionary solution.
Some information on the dietary proportions of giant pandas is available from BBC News. The following is quoted from that site:
Ninety nine per cent of a panda's diet is made up of 30 species of bamboo. The remaining one per cent is made up of other plants and meat. Their digestion of bamboo is very inefficient; pandas only digest about 20 per cent of the dry matter of bamboo, whereas most herbivores assimilate about 80 per cent. This means that they must eat large amounts to obtain their energy requirements. They can eat between 12 and 38kg of bamboo shoots, leaves and stems per 24 hour period.
Giant pandas can maintain this dietary solution only by sustaining a high feeding rate. The digestibility of bamboo varies markedly across the year (Wei et al. 1999), and in the winter when new growth is rare or absent, there are very few nutrients available. Pandas do not have any significant digestion of the structural elements of cell walls or other fibers. They therefore must extract the proteins and simple carbohydrates from bamboo and pass the bulk as quickly as practicable. To this end, they have wide and flat molars and premolars compared to other bears. These are not teeth with high crowns and shearing surfaces. This makes them different from primates, like gorillas and colobus monkeys, that eat a high proportion of leaves and other vegetation. It seems that pandas are not really in the business of cutting fibrous bamboo into a pulp; but instead they crush the bamboo to extract as much of the cell contents as possible.
Gigantopithecus also had broad, flat molars and premolars. These teeth had relatively thick enamel. Enamel thickness is a tricky indicator of diet, because there are actually advantages to having enamel that wears through completely during life. If the goal is to maintain an effective shearing surface on the tooth for cutting fibrous plant material, then thin enamel exposes the softer dentin, which wears faster. The wear gradient between the two maintains a topography to the tooth surface that is a better shearing implement than a flat, thick-enameled tooth. So the thick molar enamel in Gigantopithecus would not be very useful for shearing bamboo leaves into an undifferentiated mush. But those teeth might have been used to crush bamboo to extract the cell contents while leaving the mass mostly intact.
The evidence suggests that Gigantopithecus differed from giant pandas in having a more varied diet. One of the world's experts on Gigantopithecus is the paleoanthropologist Russ Ciochon. He has
a very nice article about the species which appeared in Natural History magazine in 1991. This nice review features the history of Gigantopithecus discoveries, our current understanding of their anatomy, diet, and history, and Ciochon's own attempts to find fossil Gigantopithecus in Vietnam.
Ciochon describes looking for phytoliths on the teeth as evidence of diet. When the fossil teeth of Gigantopithecus were examined with scanning electron microscopy, dozens of phytoliths were found:
More than half of the phytoliths we observed were long and needlelike and could be attributed to the vegetative part of grasses, possibly bamboo. The rest were conical or hat shaped, attributable to the fruits and seeds of dicotyledons. Piperno tentatively identified them as fruits from a tree of the family Moraceae, quite possibly durian or jackfruit, both of which are common throughout tropical Southeast Asia. This proved that Gigantopithecus had a varied diet, although we still suspect that bamboo was its staple food.
This work is described in
Ciochon et al. (1990) in PNAS, which includes scanning electron micrographs of the phytoliths.
Of course the relative quantities of phytoliths do not directly address dietary composition, since different plants have different phytolith abundances. Likewise, one might speculate that the phytoliths on fossil teeth represent foods eaten near the time of death -- a "last meal" effect. This might explain the apparent evidence for one kind of fruit in the Gigantopithecus data: the individual died at the time that fruit was in season. In any event, Ciochon and colleagues (1990) conclude it likely that Gigantopithecus had a very broad diet, that nonetheless included bamboo as a staple. In support of this, they cite an examination of tooth wear by Daegling and Grine (1989 in abstract; later published in 1994 in SAJS) that found Gigantopithecus microwear to be similar to chimpanzees. Chimpanzees themselves eat a majority of fruit, with smaller proportions of leaves, insects, and meat.
Wei F, Feng Z, Wang Z, Zhou A and Hu J. 1999. Use of the nutrients in bamboo by the red panda (Ailurus fulgens). J Zool Lond 248:535-541.
Ciochon RL, Piperno DR and Thompson RG. 1990. Opal phytoliths found on the teeth of the extinct ape Gigantopithecus blacki: implications for paleodietary studies. Proc Natl Acad Sci U S A 87:8120-8124.
Dean MC and Schrenk F. 2003. Enamel thickness and development in a third permanent molar of Gigantopithecus blacki. J Hum Evol 45:381-387.
Daegling DJ and Grine FE. 1994. Bamboo feeding, dental microwear, and diet of the Pleistocene ape Gigantopithecus blacki. S Afr J Sci 90:527-532.
Ungar P. 1998. Dental allometry, morphology and wear as evidence for diet in fossil primates. Evol Anthropol 6:205-217.