Escolar Documentos
Profissional Documentos
Cultura Documentos
Anthropology Department and Institute of Ecology, Baldwin Hall, University of Georgia, Athens, Georgia 30602-1619, USA
b
477 Kay Ave, Menlopark, Pretoria 0081, South Africa
Received 31 December 2003; accepted 16 November 2004
Abstract
A small minority of Africas wild plant foods are C4. These are primarily the seeds of some of the C4 grasses, the
rootstocks and stem/leaf bases of some of the C4 sedges (especially papyrus), and the leaves of some of the C4
herbaceous dicots (forbs). These wild food plants are commonly found in disturbed ground and wetlands (particularly
the grasses and sedges). Multiple lines of evidence indicate that C4 grasses were present in Africa by at least the late
Miocene. It is a reasonable hypothesis that the prehistory of the C4 sedges parallels that of the C4 grasses, but the C4
forbs may not have become common until the late Pleistocene. CAM plants may have a more ancient history, but oer
few opportunities for an additional C4-like dietary signal. The environmental reconstructions available for the early
South African hominid sites do not indicate the presence of large wetlands, and therefore probably the absence of
a strong potential for a C4 plant food diet. However, carbon isotope analyses of tooth enamel from three species of
early South African hominids have shown that there was a signicant but not dominant contribution of C4 biomass in
their diets. Since it appears unlikely that this C4 component could have come predominantly from C4 plant foods,
a broad range of potential animal contributors is briey considered, namely invertebrates, reptiles, birds, and small
mammals. It is concluded that the similar average C4 dietary intake seen in the three South African hominid species
could have been acquired by diering contributions from the various sources, without the need to assume scavenging or
hunting of medium to large grazing ungulates. Eectively similar dominantly dryland paleo-environments may also be
part of the explanation. Theoretically, elsewhere in southern and eastern Africa, large wetlands would have oered early
hominids greater opportunities for a C4 plant diet.
2005 Elsevier Ltd. All rights reserved.
Keywords: Carbon isotopes; Paleodiet; C4 forbs; C4 sedges; C4 grasses; CAM plants; Ecology; Papyrus; Paleoanthropology
) Corresponding author.
E-mail address: chaspete@uga.edu (C.R. Peters).
0047-2484/$ - see front matter 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhevol.2004.11.003
220
Introduction
The two Australopithecus species that were
recovered from cave breccias in the Transvaal,
South Africa, before and immediately following
World War II were thought to have had dierent
diets. Paranthropus robustus, with its massive jaws
and large molars, was considered to have been predominantly a vegetarian, while Australopithecus
africanus, with its more humanlike dentition, was
thought to have been omnivorous, and perhaps
included a fair amount of esh in its diet
(Robinson, 1954). The demonstration that carbon
isotope studies on the bones of prehistoric humans
indicates aspects of their diet (Vogel and van der
Merwe, 1977) opened up the possibility of
throwing some light on the diet of the early
hominids. The technique is based on the fact that
the tissue of plants that utilize the four carbon
mode of photosynthesis (C4 plants) has a relatively
high content of the rare stable carbon isotope 13C,
which uniquely distinguishes these plants from
those using the more common three carbon
photosynthetic pathway (C3 plants). Furthermore
the isotopic signal is passed on to the animals that
feed on these plants (Vogel, 1978).
The application of photosynthetic systems in
the African context results from the nding that,
although woody plants are typically C3, the vast
majority (O95%) of grass species in the savannas
and bushvelds of the warm summer rainfall
regions of southern Africa are C4 plants (Vogel
et al., 1978; Ellis et al., 1980). This also has been
found to be the case in eastern Africa (Cerling
et al., 1997). As a consequence, in these regions, the
bones of grazers, such as the zebra and wildebeest,
have a carbon isotope composition that clearly
distinguishes them from browsers, such as the
girae and kudu (Vogel, 1978; Ambrose, 1986), as
well as from predominantly vegetarian omnivores
like the chacma baboon (Lee-Thorp et al., 1989).
The isotopic values found in carnivores that prey
on ungulates show the relative amount of C4
grazers in their diet (Lee-Thorp et al., 1989).
The early hominids that have thus far been
investigated are from southern Africa. They show
13
C values intermediate between that of a C3 and C4
diet, which has been interpreted as indicating the
Methods
Evaluation of the plausibility of a C4 plant food
diet for early hominids requires a broad perspective on the taxonomy and ecology of C4 plants.
Our approach was initially the dicult one of
reviewing an extensive but often unsystematic
literature on the identication of C4 species. That
approach was problematic because of limitations
in available listings, screening methods, and
taxonomy. However, a comprehensive review of
the higher taxonomic distribution of C4 photosynthesis then became available (Sage et al., 1999).
This encouraged us to start again, with a more
general taxonomic perspective.
The rst step in the analysis presented here is
a brief overview of the worlds C4 families, along
with descriptive statistics on the number of C4
genera found in those families. This step is
important for clarifying the limitations of the
analysis as applied to African plants. Not all of the
species in the C4 genera identied by Sage et al.
(1999) have been screened for photosynthetic
pathway, and the unvouchered botanical status
of many screenings results in additional uncertainties, but the general patterns at the generic level are
strong enough to permit a preliminary application
of the analysis to African plant taxa. Where the
genera appear to be exclusively C4, the untested
African species are treated here as probably C4.
Results
The C4 syndrome is apparently restricted to the
owering plants, or angiosperms (Sage et al.,
1999), the dominant class among the seed plants.
221
Table 1
Worldwide occurrence of C4 photosynthesis in angiosperm families and C4 genera from those families also represented in the wild food
plants of sub-Saharan Africa and plant parts eaten for probable C4 species1
Angiosperm
subclass
Dicotyledones
Acanthaceae (1/250)
Aizoaceae (5/126)
Blepharis (80)
Gisekia (5)
Sesuvium (1*/12)
Trianthema (17)
Zaleya (6)
Achyranthes (1/7)
Aerva (1*/10)
Amaranthus (60)
Celosia (1*/45)
None
Heliotropium (6*/250)
Cleome (2*/150)
Seed (1)
Leaf (1)
Potherb (1; may be C3)
Leaf (1)
Vegetable (1)
--Potherb (1)
Leaf (4)
----Fruit:berry (2; may be C3)
Potherb (1)
None
None; note: Chenopodium is C3
Euphorbia (250/1500)
Amaranthaceae (11/74)
Cyperaceae (28/131)
Hydrocharitaceae (1/16)
Poaceae - Gramineae (372/ca. 800)
Mollugo (3*/35)
Boerhavia (2*/20)
None
Portulaca (40)
None
Tribulus (25)
Zygophyllum (1*/80)
Cyperus (ca. 500/650)
Mariscus (200)
None
Brachiaria (ca. 100)
Cenchrus (22)
Cynodon (10)
Dactyloctenium (13)
Digitaria (ca. 220)
Echinochloa (ca. 35)
Eragrostis (ca. 350)
Eriochloa (ca. 30)
Imperata (8)
Leptochloa (27)
Leptothrium (2)
Miscanthus (ca. 20)
Panicum (ca. 260/370)
Paspalum (ca. 320)
Pennisetum (80)
Rottboellia (4)
Setaria (ca. 110)
Sporobolus (ca. 160)
Stipagrostis (50)
Themeda (18)
Urochloa (ca. 120)
Collated from Sage et al., 1999 (and personal communications) and Peters et al., 1992, with some additions and corrections.
223
Table 2
Common ecological characteristics of Africas wild C4 plant food genera: dicot leaf taxa and grass seed taxa (cf. Table 1)1
Angiosperm
subclass
Geographic distribution
Elevation
General habitat
Leaf-food dicot
genera (10)
Seed-food grass
genera (18)
1
Synthesized from: dicots, Jerey, 1961; Verdcourt, 1961; Heine, 1963; Elers et al., 1964; Nabil El Hadidi, 1985; Townsend, 1985;
Agnew and Agnew, 1994; Whitehouse, 1996; Pooley, 1998; Maundu et al., 1999; grasses, Clayton, 1972; Clayton et al., 1974; Clayton
and Renvoize, 1982; Gibbs Russell et al., 1991; plus references cited in Peters et al., 1992.
224
usually weeds (sometimes perennials) and/or include seasonal wetlands in their habitat associations. Echinochloa is characterized by freshwater
hydrophytes and marsh grasses on clay soils.
Eragrostis can be locally common on moist sandy
soils. Panicum and Sporobolus are found on
a variety of soil types, but generally prefer damp
situations. Limited but similar information on
many of the genera suggest a reconstruction of
relatively high natural productivity on seasonally
(briey) inundated disturbed ground, e.g., well
grazed pasture at the edge of wetlands. The
seasonal pattern would probably be vigorous
vegetative growth early in the rains, followed by
setting of seed before loss of soil moisture. As the
spiklets begin to dry out they easily shatter and the
seed is dispersed. This may occur while the ground
is still wet. Intensive grazing by a variety of
herbivores may follow, rst on the sandy, then
clay, soils as the ground dries out. Dry season
disturbance may also occur, e.g., warthogs digging
for grass rhizomes.
C4 plant food values and deleterious side eects
The fresh young leaves of the edible forbs are
good sources of moisture, minerals, and vitamins
(sometimes including vitamin C), but because of
their high water content, they provide very limited
amounts of carbohydrate and protein, and almost
no fat (Wehmeyer, 1986; Maundu et al., 1999).
The fresh tubers, bulbs, and rhizomes of Cyperus
are primarily sources of moisture and carbohydrate (Wehmeyer, 1986; Peters, 1999). Fresh grass
seed is a source of minerals and some vitamins,
and is high in carbohydrate, but intermediate in
protein (which may not be balanced in amino
acids; Ward, 1971), and low in fat (Wehmeyer,
1986; Maundu et al., 1999).
Potential deleterious side eects from ingestion
include digestion inhibition (dicult to evaluate)
and toxicity (relatively easy to document). Very
few of these plants can be expected to be
poisonous, although many (if not most) have been
used as medicinals (Watt and Breyer-Brandwijk,
1962). Those that have been reported as poisonous
are forbs and grasses that develop toxic principles
(especially hydrocyanic acid) under conditions of
225
C4 plant evolution
The earliest unequivocal C4 grass is from the
late middle Miocene (ca. 12.5 Ma) of California
(reviewed by Cerling, 1999). For Africa, the
evidence for C4 grass in the Miocene is indirect.
Stable carbon isotope analyses of East African
paleosol carbonates provide indirect measures of
local C4 biomass (Cerling, 1992). Combined with
carbon isotope analysis of fossil tooth enamel
(Cerling et al., 1997), they have been used to draw
conclusions about changes in the evolution of C4
vegetation in Africa. Isotopic values from paleosol
carbonates sampling the late Miocene to the
terminal Pliocene (from ca. 10 to 1.8 Ma) are
226
CAM plants
A further category of plants, those that exhibit
Crassulacean acid metabolism (CAM), also may
have contributed to the C4 dietary intake of the
early hominids. Our review of these plants has
relied primarily on the material in volumes 30 and
114 of the Springer series Ecological Studies,
Analysis and Synthesis (Kluge and Ting, 1978;
Winter and Smith, 1996).
Although CAM is diversely represented among
the vascular plants, most CAM plants are succulent species inhabiting dry regions or tropical
forest epiphytic bromeliads and orchids. A few are
submerged aquatic macrophytes, e.g., Isoetes.
Smith and Winter (1996) provide a list of plant
genera containing species capable of CAM. In
most cases all of the member species have not been
screened, and the CAM status of African species
reported elsewhere as edible usually is not conrmed. African CAM genera with wild edible
species are best considered as possible candidates
for a CAM contribution to early hominid diets.
227
50
100%C4
-10
-8
-6
-4
-2
Fig. 1. The relative 13C content of tooth enamel of selected species from three early hominid sites in South Africa, expressed as the
parts per thousand (&) deviation from the PDB reference standard. The horizontal bars show the range for each group and the short
vertical lines give the average for each group. The approximate C4 carbon contribution to the diet is indicated on the top of the
diagram. From top to bottom: Makapansgat Member 1. Browsers: the averages and total range of eleven ungulate species, identied
as C3 plant consumers on the basis of the data (n Z 37). Grazers: the averages and range of ve ungulate species classed as C4 plant
consumers, as above (n Z 18). Austalopithecus africanus, four specimens. Sterkfontein Member 4. Parapapio spp., ten specimens
consisting of P. broomi, P. jonesi, P. whitei. Swartkrans Member 1. Browser: ve specimens of Tragelaphus cf. strepsiceros. Grazer:
three specimens of Connochaetes sp. Papio spp., ve specimens of P. robinsoni and three of P. ingens. Theropithecus sp., ve
specimens. Panthera pardus, seven specimens. Paranthopus robustus, six specimens. Homo ergaster, three specimens. Data compiled
from Lee-Thorp et al. (1994, 2000), Sponheimer and Lee-Thorp (1999), Luyt (2001).
228
229
Table 3
Possible C4 foods for early hominid diets1
Lifeform
Plants
C4 sedges
Examples
papyrus
C4 grasses
seed grasses
C4 forbs
amaranthus
Animals
C4 plant eating
invertebrates
tortoises
eat green grass seed plus insects; mass nesting in rainy season
small ungulates
various
medium to large
ungulates
wildebeest
1
Animal notes based on Ward, 1971; Boutton et al., 1983; Senzota, 1983; Alden et al., 1995; Capaldo and Peters, 1995, 1996;
Kingdon, 1997.
230
231
232
233
Acknowledgements
We thank J. M. Maguire for her help in the
initial stage of the project, and R. F. Sage for
assistance with the C4 plant species assignments
and general encouragement, which helped revitalize the project midway. Partial support was
provided by the Anthropology Department, University of Georgia, U.S.A., and the CSIR, South
Africa. M. K. Bamford and M. Murray-Hudson
provided opportunities for observations in the
Okavango Delta, Botswana.
References
Agnew, A.D.Q., Agnew, S., 1994. Upland Kenya Wild Flowers.
East African Natural History Society, Nairobi.
Alden, P.C., Estes, R.D., Schlitter, D., McBride, B., 1995.
National Aubudon Society Field Guide to African Wildlife.
Alfred A. Knopf, New York.
Ambrose, S., 1986. The isotope ecology of East African
mammals. Oecologia 69, 395e406.
Avery, D.M., 2001. The Plio-Pleistocene vegetation and climate
of Sterkfontein and Swartkrans, South Africa, based on
micromammals. J. Hum. Evol. 41, 113e132.
Backwell, L.R., dErrico, F., 2001. Evidence of termite foraging
by Swartkrans early hominids. Proc. Natl. Acad. Sci. 98,
1358e1363.
Bamford, M., 1999. Pliocene fossil wood from an early hominid
cave deposit, Sterkfontein, South Africa. S. Afr. J. Sci. 95,
231e237.
234
235
236
Vogel, J.C., van der Merwe, N., 1977. Isotopic evidence for
early maize cultivation in New York State. Am. Antiq. 42,
237e241.
Ward, P., 1971. The migration patterns of Quelea quelea in
Africa. Ibis 113, 275e297.
Watson, V., 1993. Composition of the Swartkrans
bone accumulations, in terms of skeletal parts and
animals represented. In: Brain, C.K. (Ed.), Swartkrans:
A Caves Chronicle of Early Man. Transvaal Museum
Monograph No. 8. Transvaal Museum, Pretoria, pp.
35e73.