It is generally accepted that modern humans evolved from some chimpanzee-like ancestor (Pilbeam, 1996). Consequently we use our data on the nutritional ecology of modern chimpanzees to draw conclusions about how nutrition, and in particular, how macronutrient chemistry, may have been involved in human evolution. We will be discussing only the plant component of the diet; although modern chimpanzees are omnivorous (Teleki, 1981), plant foods provided the great majority of the food seen eaten by these primates during our year-long study (ca. 99%).
The data presented in the first half of this report comes from a study we have recently completed comparing the diet of chimpanzees with that of three sympatric frugivorous monkey species in Kibale Forest, Uganda (Wrangham et al., in press; Conklin-Brittain et al., submitted). The study species were: chimpanzee (Pan troglodytes), blue monkey (Cercopithecus mitis), redtail monkey (C. ascanius), and gray-cheeked mangabey (Cercocebus (Lophocebus) albigena).
We included two troops of each monkey species, each with slightly overlapping ranges and complete range overlap with the chimpanzees (Figure 1). The forest compartments designated K14 and K30, for purposes of mapping the forest, were adjacent compartments. A border was also shared between K14 and K15. The "C" stands for chimpanzee, "B" for blue monkeys, "R" for redtail monkeys, and "M" for mangabeys.
The six principal plant food categories were: ripe fruit pulp, unripe fruit pulp, seeds, leaves, flowers and pith (Figure 2). There was relatively little diet overlap in species lists between chimpanzees and the monkeys as a group. More species of seeds and leaves were consumed by monkeys, while more species of pith were consumed by chimpanzees. Even within the fruit categories, there was little fruiting- tree species overlap (Wrangham et al., in press).
The proportion of each food type in each primate species' annual diet is shown in Figure 3. The chimpanzees were ripe fruit specialists while the monkeys split their time equally among ripe fruit, unripe fruit and seeds, and leaves (Wrangham et al., in press; Conklin-Brittain et al., submitted). The "C" stands for chimpanzee, B14, M14, and R14 were blue, mangabey, and redtail monkey troops respectively in the K14 compartment of Kibale Forest, while B30, M30, and R30 were blue, mangabey, and redtail monkey troops respectively in the K30 compartment. Pith = PI, leaves = LV, unripe pulp and seeds = FU+SD, and ripe pulp = FR.
Figure 4 shows that chimpanzees tracked the availability of ripe fruit quite closely, while the monkeys did not. The solid lines, and left-hand scale, show the percentage of feeding time on plant items spent eating ripe fruit (upper line: chimpanzees; lower line: mean and ranges for six monkey groups). The dashed line, and right-hand scale, shows the mean percentage of trees with ripe fruit in the range of the seven primate groups, also called the fruit availability index (Wrangham et al., in press).
For most of the year a mean of 25% or less of the monkeys' feeding time was spent eating chimpanzee foods. However, when fruit was abundant, the blues and redtails tracked the chimpanzee ripe fruits where as the mangabeys did not (figure 5).
When ripe fruit was not available, the chimpanzees significantly turned to piths as their primary fall-back food (Figure 6). Meanwhile, the redtails increased their unripe fruit and seed intake as ripe fruit availability decreased. The blue monkeys did not track ripe fruit, instead they tracked leaves, with seeds and unripe fruit being their fall-back food when leaves were scarce. Mangabeys did not track any single food category significantly (Conklin-Brittain et al., submitted).
Dietary diversity for all species increased with more feeding records (Figure 7). However, in relation to observation time, all monkeys had higher dietary diversity than chimpanzees.
In summary, the monkeys had much more similar diets to each other than they did to the chimpanzee, whether one considers the percentage of food items shared or the percentage feeding time per month on shared foods (Figure 8). There exists a striking diversity in the plant parts and plant species consumed by these frugivorous monkeys as compared to the chimpanzees in Kibale forest. Chimpanzees were much more frugivorous.
Given that chimpanzees and monkeys differed in their diets, we ask here whether these differences affected their intake levels of macronutrients. For Figures 9-13 (Conklin-Brittain et al., submitted) the values for each month were weighted averages, and the weighting coefficient was % of time spent feeding on the different plant food items that month.
We found some striking similarities, for example, the crude lipid or fat content of the diet (Figure 9). There were three important points regarding the crude lipid content of the diets: 1. there were no significant differences among the primate species in the fat content of their diets, 2. the seasonality in fat intake coincided with an increased ripe fruit availability, and 3. the amounts of fat in the diets were very low, even at peak consumption levels; peak was only about 8.5% lipid, and the average annual intake was only about 2.5%. As a point of reference, humans do not need more than 3-5% fat on a dry matter basis in their diet, enough to provide the one essential fatty acid and the fat soluble vitamins (RDA, 1989). Modern, westernized humans consume 15-25% fat on a dry matter basis (usually referred to as 30-45% of calories consumed) (RDA, 1989), far in excess of need or recommendation (Butrum et al., 1988).
Dietary crude protein (CP) content showed one of the more dramatic differences between the monkey (with an annual average around 15% CP) and chimpanzee diets (annual average 9.5% CP) (Figure 10). There were no significant differences among the monkeys but the difference between the chimpanzees and the monkeys was significant. There was no significant seasonal variation. As a point of reference, adult humans need only about 9.5% protein on a dry matter basis (RDA, 1980), so the chimpanzees' plant diet was probably not deficient in protein, simply lower in content than the monkeys.
The high protein content of the monkey diet was the result of eating leaves as a fall-back food (Conklin-Brittain et al., submitted), given that leaves were generally quite high in protein. The chimpanzees, on the other hand, consumed pith as a fall-back food, which was on average low to moderate in protein. We believe the chimpanzees ate piths because they were high in one of the more easily digested fiber fractions (hemicellulose), therefore contributing to their carbohydrate and energy intake (Wrangham et al., 1991).
Considering the carbohydrate intake, for the water-soluble carbohydrate or simple sugars, there was pronounced and significant seasonality in the simple sugar content of the chimpanzee diet (Figure 11). This peak coincided with peak ripe fruit availability. The monkeys did not take advantage of the fruiting peak to increase their sugar intake, and there were no significant differences among the monkeys.
Total nonstructural carbohydrates include the simple sugars, starch and most soluble fibers (Figure 12). Soluble fibers include pectins, gums, beta-glucans, and other non-starch polysaccharides. Humans can digest (via hind-gut fermentation) 0-100% of soluble fibers (Bourquin et al., 1996; Eastwood et al., 1986; Siragusa et al., 1988), depending on the fiber. We assume that because chimpanzees have a much larger capacity for hind-gut fermentation (Milton and Demment, 1988), they can digest (via fermentation) most soluble fibers, so we included them in this category of potentially digestible carbohydrates. Again the chimpanzees showed significant seasonality, increasing their digestible carbohydrate intake when given the opportunity. This is important because the most healthful human diet is believed to be one based on complex carbohydrates with small amounts of fat and protein (Butrum et al., 1988).
Neutral-detergent fiber is also referred to as total insoluble fiber or plant cell wall. Figure 13 showed the most surprising similarity in diets among all four primates; they all consumed diets at the same insoluble fiber level, around 32% throughout the year. There were no significant seasonal differences. This was surprising because the monkeys were small, about 10-20% the body weight of the chimpanzees. In general smaller animal species consume diets lower in fiber than those of larger animals (Cork, 1994; Demment and Van Soest, 1985; Parra, 1979). We do not know whether to consider the monkeys' diets as high in fiber or the chimpanzee diet as low, though we suspect the latter (Conklin- Brittain, et al., submitted).
To summarize: all of our study species consumed a low fat, high fiber diet compared to humans. The chimpanzees' diet was higher in digestible carbohydrates when there was an increase in ripe fruit availability. In addition, the chimpanzees maintained a fairly low and constant protein intake, due to their focus on fruit, with pith as a fallback food.
The simple averages for chemical composition of different food items used in calculating the nutrient contents of these primate diets are summarized in Table 1.
|ripe fruit||unripe fruit||leaf||seed||pith||flowers|
|n = number of species included, CP = crude protein, WSC = water-soluble carbohydrates, NDF = neutral-detergent fiber|
No plant part was high in lipid. Seeds were the highest, but 8.4% is not very high. As expected, the leaves had the highest crude protein. The leaf fraction was mostly young leaves, while piths on average contained less than half that amount of protein. Flowers were surprisingly high in protein, but eaten only sporadically. The average ripe fruit had exactly the same protein content as the annual average intake for the chimpanzees. Ripe fruit was the sweetest, and piths were the second sweetest. Ripe fruit was also the lowest in fiber, but 33.6% is quite high for a fruit pulp. Not coincidentally, 33.6% is also the annual average fiber content of the chimpanzees' diet.
It is helpful as a point of reference to compare these numbers to domestic fruit and vegetables (Table 2). These results are from our laboratory also, using the same methods as for the wild plants. All values are on a percentage dry matter basis. Wild fruit was much lower in water and sugar than either domestic fruit or vegetables and much higher in fiber.
As a prerequisite to considering the Australopithecus diet, we will briefly discuss hunter-gatherer diets, modern but traditional human diets, and the minimum nutrient requirements of humans.
Modern humans do not have high protein or fat requirements, as already mentioned. The value of 9.5% CP in the chimpanzee diet in our study is consistent with the prediction by Oftedal (1990) that all primates should have relatively low protein requirements because they have slow growth rates compared to other mammals (Case, 1978). Although a need for protein or fat is often assumed to explain increasing amounts of hunting throughout hominid evolution, primates do not have metabolic demands for high levels of protein or fat.
Eaton et al. (1988) proposed an ancient hunter-gatherer diet in the book The Paleolithic Prescription. We can now evaluate their hypothetical diet in the light of what we have just learned about the chimpanzee diet and with what is known of modern human nutrient needs. The hunter-gatherer diet Eaton et al. proposed contained 35% meat and 65% (wild) plant foods. Table 3 presents the results of the Paleolithic Prescription model diet, with some additional calculations:
|Grams||Conversion||% of total energy||% of total grams consumed|
|Protein||250||x 4 kcal/g] / 3000 =||33||30.9|
|Fat||70||x 9 kcal/g] / 3000 =||21||8.6|
|Carbohydrate||340||x 4 kcal/g] / 3000 =||46||42.0|
The format of the third column, % of total grams consumed, is comparable to the way we have been reporting the chimpanzee diet. The protein intake is almost 4.5 times higher than required by humans, so we can not make meaningful comparisons there. The fat intake is also high compared to the chimpanzee diet and to that required by humans, even though the authors used nutrient values from wild game meat instead of domestic meat, so the fat intake is moderate compared to a modern human diet.
The fiber content is the interesting point for comparison. The content is about half of that in the chimpanzee diet. This results from 35% of the plant component being replaced by meat, in effect diluting the fiber content of the diet. We have seen that a wild herbivore diet, such as the chimpanzee in Kibale Forest, is high in fiber because wild foods are high in fiber (Table 1). In order to dilute that fiber level further, a new source of food must be found that is low in fiber. Meat is guaranteed to reduce the fiber content of a diet considerably and to be fairly easily digested.
Nevertheless, because wild vegetation is high in fiber, the Paleolithic hunter-gatherer diet was still assumed to contain 150 g of fiber from its 65% plant component, a huge intake by modern standards. Westernized diets normally include only 10-20 g of fiber per day (Johnson and Marlett, 1986; Georgiou and Arquitt, 1992), although the National Cancer Institute recommends 35 g (Bourquin et al., 1996). Consequently, it is useful to consider a traditional, nonwesternized modern diet from Zaire where the only domesticated component of the diet is cassava, a tuber very low in fiber (Pagezy, 1990). The rest of the diet is wild, either game or wild plant food (Table 4).
A. Rainy Season in the Village
|Grams||Conversion||% of total energy||% of total dry grams consumed|
|Protein||38.5||x 4 kcal/g] / 2000 =||7.7||8.3|
|Fat||60.9||x 9 kcal/g] / 2000 =||27.4||13.1|
|Carbohydrate||325.3||x 4 kcal/g] / 2000 =||65.0||69.8|
B. Dry Season in the Camp
|Grams||Conversion||% of total energy||% of total dry grams consumed|
|Protein||87.6||x 4 kcal/g] / 2000 =||17.5||18.4|
|Fat||24.2||x 9 kcal/g] / 2000 =||10.9||5.1|
|Carbohydrate||325.3||x 4 kcal/g] / 2000 =||65.0||68.5|
Fat intake levels in "B", the hunting camp, were at the 5% level referred to previously as being adequate for humans. The men in camp were eating more game and fish, about 20% of dry matter intake, but their diet was still dominated by cassava in these comparisons. Fiber levels have dropped from 33.6% NDF for chimpanzees to 18.5% for an ancient hunter-gatherer diet to about 9% for these modern but traditional diet containing meat but dominated by a low fiber root crop, cassava. We suggest that this pattern of fiber reduction represents what happened in the transition from a chimpanzee-like ancestor to modern humans.
With this information regarding the plant part of a chimpanzee diet and the progressively lower fiber diet of later hominids, what can we project regarding the Australopithecus diet? Hatley and Kappelman (1980) describe how Australopithecus may have added underground storage organs (roots, tubers, rhizomes, etc) to their diet, a food source underexploited by apes, past and present. A similar scenario is also described by Wrangham and Peterson (1996). Underground storage organs are also more common in woodland areas, the habitat Australopithecus were living in, as opposed to rain forest, were chimpanzees live (Hatley and Kappelman, 1980).
To evaluate the nutritional feasibility of Australopithecus consuming underground storage organs, we examined the macronutrient content of wild roots and tubers. Assuming that Australopithecus evolved from a frugivorous ape like a chimpanzee, what would have been the nutritional consequences of consuming roots as a fallback food instead of pith? We have already seen (Table 1) that pith contains 44% fiber. And we also know that neither chimpanzees nor humans need a high protein intake (RDA, 1980; Conklin-Brittain et al., submitted), but that complex carbohydrates figure importantly in chimpanzee selectivity (Wrangham et al., 1991). Consequently, what would wild roots and tubers add to the diet that piths cannot provide? The comparison of the nutrient contents of 28 wild roots and tubers (Malaisse and Parent, 1985; Vincent, 1984) with 19 domestic roots and tubers (Leung et al., 1968) is shown in Table 5. The pith from Table 1 and the chimpanzee diet averaged over the year are also listed below for comparison.
|wild, n = 28||8.3 ± 8.8||4.0 ± 4.3||65.9 ± 16.4||16.2 ± 8.5|
|range||1.0 - 39.2||0.3 - 19.8||17.4 - 92.5||3.8 - 38.0|
|domestic, n = 19||7.0 ± 3.4||1.7 ± 4.6||82.1 ± 10.2||10.8 ± 14.0|
|range||1.4 - 14.4||0.3 - 20.0||53.8 - 92.9||2.3 - 63.3|
|pith, n = 19||9.6 ± 6.2||1.2 ± 1.6||31.9 ± 13.8||44.1 ± 14.3|
|range||1.4 - 20.4||0.2 - 7.0||11.1 - 68.9||24.5 - 79.3|
|Chimpanzee diet||9.5 ± 3.1||2.5 ± 2.1||38.8 ± 7.6||33.6 ± 4.5|
|range||5.3 - 15.4||0.6 - 8.2||31.1 - 48.5||20.7 - 37.6|
Twenty-five of the wild roots and tubers were Zambezian edible plants from a woodland habitat, and three roots were from the Hadza, in a savanna environment. The domesticated root and tuber data were of tropical origin.
On average the wild roots and tubers would be marginal in protein, but there is quite a range of values, offering the possibility for selection. Fat is probably adequate. The fiber values were determined using the very out-dated method of crude fiber analysis that always underestimates fiber content. We multiplied all the crude fiber values by 2, which is a conservative correction value. Three would be the highest reasonable correction factor (Van Soest, 1994), and would have resulted in fiber values of about 24% instead of 16%. Nevertheless, 24% is considerably lower than what the chimpanzee diet in our study contained, and much lower than the 44% in piths, the chimpanzee fall-back food. Consequently the inclusion of roots or tubers in the diet would decrease the fiber intake of a potential consumer and increase the nutrient density of the diet, especially the carbohydrate intake.
In summary it appears that underground roots and tubers would make an important nutritional addition to the diet of Australopithecus, who might have been able to live exclusively on roots and tubers during short periods of above-ground food scarcity. Furthermore, the dental and microwear patterns exhibited by Australopithecus are compatible with the additions of roots to a chimpanzee-like diet (Hatley and Kappelman, 1980; Grine and Kay, 1988). They would not have needed additional protein supplement to top-up their protein intake to safe levels. In addition, the lower fiber values would improve the quality of their diet. This does not imply that a need to decrease fiber in the diet was a driving force in the evolution of the hominid diet. However, with the serendipitous addition of underground storage organs to the Australopithecus diet and the resulting increase in the nutrient density of the diet, the stage was set for Homo to further reduce fiber levels and further improving the nutrient quality of their diet.
In Figure 14 the lipid intake levels are in comparison to modern western diets. Protein needs or requirement levels are in relation to modern western diets; western diets provide much more protein than needed. The fiber intake levels are in relation to each other; all of these diets provide more fiber than what is recommended for modern humans (35 g/day).
|lipid intake||protein needs||fiber intake|
|modern, traditional diet||low-moderate||low-moderate||lowest|
Bourquin LD, EC Titgemeyer, and GC Fahey (1996): Fermentation of various dietary fiber sources by human fecal bacteria. Nutr. Research 16:119-131.
Butrum RR, CK Clifford, and E Lanza (1988): NCI dietary guidelines: rationale. Am. J. Clin. Nutr. 48:888-895.
Case TJ (1978): On the evolution and adaptive significance of postnatal growth rates in the terrestrial vertebrates. Quat. Rev. Biol. 53:243-282.
Conklin-Brittain NL, RW Wrangham, and KD Hunt (submitted): Dietary response of Chimpanzees and Cercopithecines to seasonal variation in fruit abundance. II. Macronutrients. Int. J. Primatol.
Cork S (1994): Digestive constraints on dietary scope in small and moderately-small mammals: how much do we really understand? In Chivers DJ and P Langer (eds): The Digestive System in Mammals: Food, Form and Function. Cambridge, Cambridge University Press, pp. 337-369.
Demment MW, and PJ Van Soest (1985): A nutritional explanation for body-size patterns of ruminant and nonruminant herbivores. Am. Nat. 125:641-672.
Eastwood MA, WG Brydon, and DMW Anderson (1986): The effect of the polysaccharide composition and structure of dietary fibers on cecal fermentation and fecal excretion. Am. J. Clin. Nutr. 44:51-55.
Eaton SB, M Shostak, and M Konner (1988): The Paleolithic Prescription: A Program of Diet and Exercise and a Design for Living. N.Y., Harper & Row, Publishers, pp. 79.
Georgiou CC, and AB Arquitt (1992): Regional differences in food and nutrient intakes of college women from the United States: Oregon and Oklahoma. Ecol. Food Nutr. 28:251-260.
Grine FE, and RF Kay (1988): Early hominid diets from quantitative image analysis of dental microwear. Nature 333:765-768.
Hatley T, and J Kappelman (1980): Bears, pigs, and plio-pleistocene hominids: a case for the exploitation of belowground food resources. Human Ecol. 8:371-387.
Johnson EJ, and JA Marlett (1986): A simple method to estimate neutral detergent fiber content of typical daily menus. Am. J. Clin. Nutr. 44:127-134.
Leung W-T W, F Busson, and C Jardin (1968): Food Composition Table for use in Africa. Rome, Food and Agriculture Organization (FAO) of the United Nations, and Washington, D.C., U.S. Department of Health, Education, and Welfare.
Malaisse F, and G Parent (1985): Edible wild vegetable products in the Zambezian woodland area: a nutritional and ecological approach. Ecol. Food Nutr. 18:43-82.
Milton K, and MW Demment (1988): Chimpanzees fed high and low fiber diets and comparison with human data. J. Nutr. 118:1082-1088.
Oftedal OT (1991): The nutritional consequences of foraging in primates: the relationship of nutrient intakes to nutrient requirement. Phil. Trans. R. Soc. Lond. B 334:161-170.
Pagezy H (1990): Seasonal variation of food supply in the Lake Tumba region of Zaire. In Hladik CM, S Bahuchet, and I de Garine (eds): Food and nutrition in the African rain forest. Paris, Unesco/MAB,. pp. 37-42.
Parra R(1979): Comparison of foregut and hindgut fermentation in herbivores. In Montgomery, GG (ed): The Ecology of Arboreal Folivores. Washington, D.C., Smithsonian Institute Press, pp. 205-229.
Pilbeam D (1996): Genetic and morphological records of the Hominoidea and Hominid origins: a synthesis. Mol. Phyl. Evol. 5:155-168.
RDA (1980): Recommended Dietary Allowances, 9th revised edition. Washington, D.C., National Academy of Sciences.
RDA (1989): Recommended Dietary Allowances, 10th revised edition. Washington, D.C., National Academy of Sciences.
Siragusa RJ, JJ Cerda, MM Baig, CW Burgin, and FL Robbins (1988): Methanol production from the degradation of pectin by human colonic bacteria. Am. J. Clin. Nutr. 47:848-851.
Teleki G (1981): The omnivorous diet and eclectic feeding habits of chimpanzees in Gombe National Park, Tanzania. In Harding RSO, and G Teleki (eds): Omnivorous Primates: Gathering and Hunting in Human Evolution. N.Y., Columbia Univ. Press, pp. 303-343.
Van Soest PJ (1994): Nutritioanl Ecology of the Ruminant. Ithaca, N.Y., Cornell University Press.
Vincent A (1984): Plant foods eaten by the Hadza of northern Tanzania. World Archaeol. 17:131-148.
Wrangham RW, NL Conklin CA Chapman, and KD Hunt (1991): The significance of fibrous foods for Kibale Forest chimpanzees. Phil. Trans. R. Soc. Lond. B 334:171-178.
Wrangham RW, NL Conklin-Brittain, and KD Hunt (in press): Dietary response of chimpanzees and cercopithecines to seasonal variation in fruit abundance: I. Antifeedants. Int. J. Primatol.
Wrangham R, and D Peterson (1996): Demonic Males: Apes and the Origins of Human Violence.
London, Bloomsbury Publ. Plc., pp 49-62.