Hominins living on the sedge

The evolution of C₄ photosynthesis was a signal shift in the history of life on Earth, including at least 50 independent origins in 19 families of higher plants (1). Today, C₄ plants account for a quarter of the primary productivity on the planet despite representing a small fraction (∼3%) of the estimated 250,000 land plant species (2). Most C₄ plants are grasses (family Poaceae, 4,500 species), followed by sedges (family Cyperaceae, 1,500 species) and dicots (1,200 species). A few C₄ plant species contribute significantly to the human diet, either directly as domesticated cereals (e.g., maize, sorghum, millet) and additives (e.g., cane sugar, corn syrup) or indirectly via animals raised on C₄ crops and pastures. As a result, human tissues can have high δ¹³C values because C₄ plants and their consumers are enriched in ¹³C (3). In recent years, this telltale signature of a C₄-based diet has also been found in a wide range of hominin species (4), a finding that has both informed and fueled debate. Now, paleoanthropologists must contend with a unique and surprising finding.

In PNAS, Lee-Thorp et al. (5) report carbon isotope data from the tooth enamel of Australopithecus bahrelghazali, a hominin species that lived in the Chad Basin ca. 3.5 million years ago. Their findings, which indicate an early dependence on C₄ biomass (∼55–80% of the diet by linear extrapolation), are fascinating because they raise challenging questions concerning the diet of this species. The magnitude of ¹³C enrichment, which, among australopithecines, is eclipsed only by Paranthropus boisei (6), suggests that the carbon in their diet was derived mainly from C₄ plants rather than the tissues of C₄ grazing animals (5). This inference led the authors to focus on sedges, a graminoid plant that is perhaps more promising than grass as a food source for hominins. Indeed, the thickly enameled, low-cusped (bunodont) teeth of A. bahrelghazali and P. boisei would appear to be functionally incompatible with a diet of grass blades (7). Could sedges, then, bring consilience to the C₄ conundrum?

C₄ Conundrum

Grine et al. (8) put it well when they wrote, “Dietary signals derived from microwear and isotope chemistry are sometimes at odds with inferences from biomechanical approaches, a potentially disquieting conundrum that is particularly evident for several [hominin] species.” In an attempt to reconcile these discrepancies, termed the C₄ conundrum (9), Lee-Thorp et al. (5) call attention to Cyperus papyrus, a familiar C₄ sedge with “chewy and pleasant tasting” pith (10). A diet premised on pith is expected to require laborious loading of the jaws and teeth, a behavior that could have favored the evolution of massive chewing muscles, such as those of P. boisei (6). Pith is also soft and unabrasive, traits that correspond to the molar microwear of P. boisei (11). However, chimpanzees turn to papyrus pith only in extremis (12), probably because it is exceedingly fibrous and almost devoid of protein (13). In theory, few primates could survive on a diet of 55–80% papyrus pith. A puzzling exception is the Alaotran gentle lemur (Hapalemur griseus alaotrensis), a rare primate from Madagascar that specializes on hydrophilic species in the Poaceae and Cyperaceae families (>95% of the diet); in fact, the pith of the papyrus-like Cyperus madagascariensis alone accounts for about 65% of its diet (14).

Another type of ¹³C-enriched plant tissue lies underground (15), and Lee-Thorp et al. (5) call attention to the potential importance of corms in the genus Cyperus (Fig. 1A). Corms are starch storage organs that feature in the diets of many baboons (16–21) and some human populations (22) (Fig. 1B). In fact, the corms of Cyperus esculentus, a C₄ species (23), were once widely cultivated as a food crop in ancient Egypt (Fig. 1E), perhaps because only 150–200 g of corm tissue per day can satisfy human lipid requirements (24). The high quality of Cyperus corms and their role in the diets of humans and savanna-dwelling baboons suggest that C₄ corms could have been a significant contributor to the elevated δ¹³C values of A. bahrelghazali and P. boisei.

Fig. 1.

The appearance of C₄ sedges is quite variable. (A) Many savanna species, such as Cyperus cristatus, resemble grass. The corms of C. cristatus and Cyperus blysmoides account for about 15% of the diet of baboons in Laikipia, Kenya [e.g., Barton et al. (17)]. (B) The corm of Cyperus fulgens is consumed widely across southern Africa. Because the outer tunic (C) is often gritty with quartz particles (Inset; SEM), it is usually removed before ingestion (D) (photographs courtesy of Ben-Erik Van Wyk, University of Johannesburg, Auckland Park, South Africa). (Scale bar: Inset in C, 200 μm.) (E) Painting in the Tomb of Rekhmire (Theban Tomb 100) illustrating the volumetric measurement of C. esculentus corms, ca. 1504 to 1425 B.C. (photograph ©2001 Francis Dzikowski, Brooklyn, NY).

A problem with this hypothesis is that corms are gritty (21), and microwear data argue against such a diet. It is plausible that the exogenous grit on corms exerted a strong selective pressure on hominin manual dexterity, which, in turn, was preadaptive for tool use. Corms have an outer peel, or tunic (Fig. 1C), that humans can remove to avoid grit and minimize tooth wear (Fig. 1D). Thick mats of discarded corm tunics at some Middle and Later Stone Age sites in South Africa (25) are a testament to our aversion to grit and the ecological value of corms as a food source. For P. boisei, the regular but imperfect peeling of C₄ corms might explain the paradox of how it can have deeply chipped teeth (26) and yet, at the same time, a microwear signature that stands at odds with its robust craniodental morphology (7, 8).

Regardless of whether A. bahrelghazali was consuming papyrus pith or Cyperus corms, or both, Lee-Thorp et al. (5) have

Lee-Thorp et al. have thrown a spotlight on the importance of sedges in hominin evolution.

thrown a spotlight on the importance of sedges in hominin evolution. A collaborative focus on the nutritional composition, isotopic ecology, and mechanical properties of sedges could prove very instructive, and researchers in the disciplines of primatology, human ecology, and paleoanthropology are well poised to collect these data. For example, an improved knowledge of sedge phytoliths, such as the one recovered from the dental calculus of Australopithecus sediba (27), could help to distinguish between sedge species or possibly even the type of tissue that was consumed. Congratulations to Lee-Thorp et al. (5) on a fine paper.