Gut Morphology--Conclusion / Clarifying 'Omnivore' vs 'Faunivore'

(Comparative Anatomy and Physiology Brought Up to Date--continued, Part 6E)

A footnote on Chivers and Hladik [1980, 1984]: Human gut morphology

Sussman [1987] describes the analysis of the gut of 6 human cadavers using the measures defined in Chivers and Hladik [1980, 1984]. Analysis of the human gut data using the coefficient of gut differentiation (a measure of gut specialization) placed humans in the frugivore range, along the margin with the faunivore category. However, analysis of the same data using the index of gut specialization (yet another measure of gut morphological specialization) placed humans squarely in the faunivore range.

Note that the frugivore classification above came from using the coefficient of gut differentiation, which is an intermediate result in Chivers and Hladik [1980, 1984], hence presumably less desirable (from a certain analytical viewpoint) than the (faunivore) classification achieved using the end result of Chivers and Hladik [1980, 1984], i.e., the index of gut specialization. Also recall that the term frugivore does not mean or imply that a diet of nearly 100% sweet fruit (as advocated by some fruitarians) is appropriate. Recall that all frugivorous primates eat at least some quantities of animal foods, even if only insects. Thus the result that humans appeared to be frugivores by one measure and faunivores by another suggests a natural diet for humans that includes both animal foods and fruits.

The research of Martin et al. [1985]

Improved methodology. The research of Martin et al. [1985] is very important, for it provides a second, different analysis of the (same) data analyzed previously in Chivers and Hladik [1980, 1984]. The analysis of Martin et al. [1985] includes a number of important features/improvements:

Data for 6 humans, modern Homo sapiens, are included in the analysis.
All comparisons are done with respect to body weight. Weight is more commonly used in allometric analysis/adjustments than is body length. (Chivers and Hladik [1980, 1984] used the cube of body length for their analysis.) Another advantage of performing analysis based on body weight is that it is closely related to the metabolic energy requirements of each animal (Kleiber's Law again).
The research uses gastrointestinal (GI) quotients (explained below), similar to encephalization quotients.
In Chivers and Hladik [1980, 1984], the animals were grouped into dietary categories on an a priori basis prior to quantitative analysis. One can argue that such an a priori classification is inappropriate and introduces bias. In contrast, Martin et al. [1985] allow the data itself to determine the dietary groupings, based on an analysis of gastrointestinal quotients (alone). The groupings are then analyzed for dietary commonality. (In other words, the data themselves here determine the dietary groupings--a critically important point.)
Major axis line-fitting is used in preference to linear regression, as it may have advantages when both dependent and independent variables are measured with errors.

See Appendix 4 for background information on statistical methods used in the paper.

The major points from Martin et al. [1985], in summary, are:

Size scaling adjustments done by body weight rather than length. The raw data is allometrically adjusted (scaled) to adjust for body weight. Martin et al. [1985, p. 66] note:

For instance it is now well established that basal metabolic rate scales to body weight in mammals with an exponent value of 0.75 ["Kleiber's law" (Kleiber 1961; Hemmingsen, 1950, 1960; Schmidt-Nielsen, 1972)], and there is new evidence to indicate that active metabolic rate (i.e., total metabolic turnover in a standard time) also scales to body size with a comparable exponent value (Mace and Harvey, 1982). Hence, one might expect any organ in the body that is directly concerned with metabolic turnover to scale to body size in accordance with Kleiber's law.
GI quotients for 4 different digestive system components (stomach, intestines, cecum, and colon) for each animal in the study are calculated by dividing the component's actual weight by its "expected" weight (the latter projected using Kleiber's Law). Measurement of the size of digestive components was calculated using surface area. A GI quotient greater than 1 means the size is larger than expected; a value less than 1 indicates a size smaller than expected.
Human GI quotient pattern typical of faunivores. Human GI quotients are considerably lower than predicted/expected for all 4 digestive system components measured. Martin et al. report [1985, p. 72] that:

Calculation of gut quotient values has particular interest in the case of the four average surface areas of the gut compartments determined for six Homo sapiens. It can be seen from figs. 1-4 that man has values of less than one for all four gut compartments, most notably with respect to the cecum:

GQ = 0.31; IQ = 0.76; CQ = 0.16; LQ = 0.58
[In the above, GQ is the quotient for the stomach, IQ for the small intestine, CQ for the cecum, and LQ for the colon.]

This is a pattern shared with a number of animals relying heavily on animal food ["faunivores" (Chivers and Hladik, 1980)].
Meaningful dietary groupings based on statistical analysis of GI quotients. A dendrogram, or "tree" diagram, based on statistical analysis of the GI quotients for the different animal species in the study was derived in order to determine meaningful dietary groupings according to similarity of GI tracts. The dendrogram for the study can be found in Figure 11, p. 81 of Martin et al. [1985], and includes Homo sapiens. Humans fall into group A2 in the dendrogram, about which, Martin et al. [1985, p. 82] comment:

Group A can be characterized as containing numerous mammalian species (primates and nonprimates) that include at least some animal food in their diets. Again, there is a separation into two subcategories (A1, A2), the second of which contains most of the mammalian carnivores and only two primate species--Cebus capucinus and Homo sapiens.

Thus the result of the advanced statistical analysis in Martin et al. [1985] is that humans fall into the faunivore--meat-eater--class, yet again. Note also that the Capuchin monkey, Cebus capucinus, is in the same statistical grouping as humans, thereby confirming the remarks in Milton [1987], discussed earlier in this section, that the human and Capuchin monkey gut dimensions are similar.

The research of MacLarnon et al. [1986]

Refinement needed in analytical techniques used in earlier study. The research of MacLarnon et al. [1986] provides an extension and analytical refinement of Martin et al. [1985]. The analytical techniques needed further refinement because:

Clustering technique is labile. The dendrogram clustering technique used in Martin et al. [1985] was found to be labile, i.e., small changes in the data set analyzed could cause substantial changes in the cluster pattern produced.
Non-symmetric transformation used in earlier study. The anti-logarithmic transformation applied in computing GI-quotient values in Martin et al. [1985] was non-symmetric and tended to overemphasize the importance of GI components that were larger than expected.
The multidimensional scaling (MDS: a dimension-reducing data analysis method) plots produced in Martin et al. [1985] were crowded and difficult to interpret.

MacLarnon et al. [1986] addressed the above issues by performing additional analyses of the Martin et al. [1985] data set:

The data were reanalyzed in their original, untransformed logarithmic form. This created another problem, however, as it forced the exclusion (from the analysis) of those animals that lacked a caecum (i.e., some faunivores). [The exclusion was necessary because the logarithm of zero is undefined.]
Primates were analyzed as a separate group, for comparison with the other analyses.
The MDS was repeated using the original logarithmic data.

The results of MacLarnon et al. [1986] can be summarized as follows.

Primates-only analysis. In the dendrogram clustering using the transformed data [Martin et al. 1985], humans (Homo sapiens, #42 in the analyses) and Capuchin monkeys (Cebus capuchinus, #16 in the analyses) appear together in a group that consists of frugivores and frugivore-insectivores. In the dendrogram clustering using the untransformed (logarithmic) data [MacLarnon et al. 1986], humans and the Capuchin monkey occur together, as an apparent outlier in the clustering.
In the MDS analysis using logarithmic data, humans and Capuchin monkeys appear as outliers. In the anti-logarithmic data MDS, humans and Capuchin monkeys appear on the edge of the central cluster of primate species.
MDS analysis of a more complete data set: humans grouped (again) with faunivores. When a more complete data set (primates plus other mammals, excluding those faunivores with no caecum) was analyzed, humans and Capuchin monkeys were grouped with non-primate faunivores in the MDS logarithmic data analysis. This is clearly illustrated in Figure 5 of MacLarnon et al. [1986, p. 302], where humans and the Capuchin monkey appear in the faunivore group. Note that Martin et al. [1985] report similar results using untransformed anti-logarithmic data, though the results are not as clear as in the logarithmic data set.
Wild vs. captive specimens. Further analysis showed that including data on captive primates had only a small impact on the results. The primary effect was to rearrange the frugivores and frugivore-insectivores, which are the most labile species in the analysis.

Conclusions. MacLarnon et al. [1986] conclude that:

The use of logarithmic quotients is preferable to the use of anti-logarithmic quotients in MDS analyses.
MDS analysis techniques are more robust (for the subject data set) than dendrogram-based clustering techniques.
Human GI tract shows possible faunivore adaptations. From MacLarnon et al. [1986, p. 297]:

...[T]his being the case, the new evidence from the approach using logarithmic quotient values (Fig. 1, 3 and 5) is particularly interesting in that it suggests a marked departure of Cebus [Capuchin monkey] and Homo [humans] from the typical pattern of primates lacking any special adaptation for folivory...in the direction of faunivorous non-primate mammals....
5. Use of logarithmic quotient values for clustering purposes suggests that Cebus and Homo possess gastrointestinal tracts that have become adapted in parallel to those of faunivorous mammals, with notable reduction in size of caecum relative to body size. Nevertheless, because of the artificiality of most modern human diets, it cannot be concluded with confidence that the small human sample examined to date reflects any "natural" adaptation for a particular kind of diet. The results obtained so far are suggestive but by no means conclusive.

Thus the research of MacLarnon et al. [1986] suggests, but is not (by itself) conclusive proof, that the human GI tract is adapted for the consumption of animal foods.

Assessing Martin et al. [1985], MacLarnon et al. [1986], and related studies

Further study still needed to confirm results. Although very impressive in their scope and sophistication, the results of Martin et al. [1985] and MacLarnon et al. [1986] that relate to humans specifically are based on limited data (6 individuals). Further study is appropriate before sugrfficient evidence is available to definitely class humans as faunivores or frugivores, based solely on gut morphology.

Some of the reasons for caution regarding the study results are as follows:

Small sample size. Some species used in the study (although there were 180 individuals total from 78 different species) have only a single specimen--i.e., small sample size for those species.
Gut dimensions can vary in response to current diet. The gut dimensions of animals can vary significantly between wild and captive animals (of the same species, of course). Gut dimensions can change quickly (in captivity or in the wild) in response to changes in dietary quality. For information on this topic, consult Hladik [1967] as cited in Chivers and Hladik [1980]; also the following sources cited in Milton [1987]: Gentle and Savory [1975]; Gross, Wang, and Wunder [in press per citation]; Koong et al. [1982];ler [1975]; Moss [1972]; and Murray, Tulloch, and Winter [1977].
Knowledge of human gut variability is limited. More information is needed on how the human gut dimensions vary with diet. Milton [1987, p. 101] notes:

The size of the present-day human small intestine could be an ancient or a relatively recent trait. Indeed, it is not known whether all modern human populations show such gut proportions.
Gut structure and gut transit times. The analyses of Martin et al. [1985] and MacLarnon et al. [1986] are limited to measures of structure, and do not include gut transit times. Note, however, that data on gut transit times for wild animals will likely be difficult to obtain.
Different analytical techniques, different results: the confusion factor. The papers discussed so far in this section provide a range of different analytical techniques, and the results for humans are not consistent across the different techniques. This can be confusing to readers, as one part of the analysis suggests humans are probably faunivores, while another suggests humans might be frugivores.

We have presented the results of these papers in some detail here, so the reader can get a feel for the overall thrust of the analyses. The basic result appears to be that the anatomy of the human GI tract shows what appear to be adaptations for faunivory (consumption of animal foods), regardless of whether humans fall into the faunivore or frugivore class. This leads us to the next paper on gut morphology to be discussed here.

The note of Hladik et al. [1999]

A recent note: Hladik, Chivers, and Pasquet [1999] (referred to hereafter as Hladik et al. [1999]) provides further information on gut morphology. The Hladik et al. [1999] paper must be understood in the context of the 4 major papers that preceded it, as well as the fact that it is a comment on the Expensive Tissue Hypothesis of Aiello and Wheeler [1995], which was discussed in a previous section. The context is important because a fruitarian extremist has already quoted parts of Hladik et al. [1999] out of context, and--in my opinion--misrepresented the meaning thereof.

Humans are frugivores by one measure, faunivores by another. Let's look at some relevant quotes from the note. Hladik et al. [1999] includes a version of a figure from Chivers and Hladik [1980], with the comment [Hladik et al. 1999, p. 695]:

Some of these data are reported here (fig 1), combined with one of the original figures [Chivers and Hladik 1980] illustrating the reduced axis [line fit] for three major dietary tendencies of primates and other mammals (folivore/frugivore/faunivore). The human specimens fall, as expected, on the main axis for frugivores, a gross category corresponding to fruit and seed eaters (or omnivores, when occasional meat eating is practiced).

Examination of figure 1 from Hladik et al. [1999] suggests that it is a version of figure 26 from Chivers and Hladik [1980]. The figure (1) shows the relationship between body size (horizontal axis) and area of absorptive mucosa (vertical axis). Hladik et al. [1999] report that the human specimens fall into the frugivore range. However, Sussman [1987] analyzed 6 human cadavers and found that their intestinal surface areas fell into the frugivore range using one measure, but the faunivore range using another measure (specifically: figure 26 from Chivers and Hladik [1980]; see also figure 9.8 in Sussman [1987, p. 176]).
Thus we have again that the classification of humans by gut morphology is not consistent; humans appear to be frugivores per Hladik et al. [1999], but faunivores by a similar (perhaps the same kind of) measure per Sussman [1987]. If one considers frugivore as a gross category that includes omnivores, then humans might indeed fall into the category, as the sum total of current evidence suggests that humans (and Capuchin monkeys) are (figuratively) where the faunivore and frugivore classes "meet."
Gut surface areas might not support Expensive Tissue Hypothesis. From Hladik et al. [1999, pp. 696-697]:

A specialized carnivorous adaptation in humans that would correspond to a minimized gut size is obviously not supported by our data (fig. 1). The large variations in human diets (Hladik and Simmen 1996) are probably allowed by our gut morphology as unspecialized "frugivores," a flexibility allowing Pygmies, Inuit, and several other populations, present and past, to feed extensively on animal matter...

The first sentence above, re: carnivorous adaptation, must be understood in context: as a comment on the Expensive Tissue Hypothesis. It claims that there is no major change in gut surface areas as the Expensive Tissue Hypothesis suggests. It does not mean there is absolutely no adaptation to faunivory: the major adaptation to faunivory in humans was previously identified as a reduction in size of the caecum and colon, per Martin et al. [1985] and MacLarnon et al. [1986]. The above quote does not contradict the 1985 and 1986 papers.
Humans fail on raw, ape-style frugivore diets, but thrive on faunivore diets. The second part of the above quote is very interesting. Let's consider the raw-food, ape-style frugivore diets that various dietary advocates claim are "optimal" for humans. The fruitarian diet is a diet of predominantly raw fruit, with some leaves and seeds. It is a strict vegetarian frugivore diet. However, extensive anecdotal evidence (the only evidence available) indicates that the fruitarian diet is a massive failure. There are very few long-term success stories on fruitarian diets, and many of the claims of success are not credible.
If we consider an arguably more natural frugivore diet, one of raw fruit and raw meat/animal foods, we have the instinctive eating diet or anopsology. That diet does have a few credible long-term success stories, but only very few. As well, the few long-term success stories known to this writer are of individuals who consume significant quantities of (raw) animal foods (i.e., in caloric terms they are probably closer to faunivore than frugivore).
The point here is the obvious irony: humans might be frugivores, but in fact humans fare poorly and do not succeed, in the long-term, on ape-style frugivore diets. In sharp contrast to the failure of ape-style frugivore diets, the example of the Inuit shows that humans can succeed and even thrive, long-term, on a diet similar to a faunivore diet. This, of course, brings us right back to the key question: should we class humans as frugivores or faunivores?
Insects (or comparable animal foods) a part of natural human diet. Ultimately, Hladik et al. [1999] directly acknowledge that the natural human nutritional requirements call for some animal foods (from p. 697, emphasis below is mine):

There is no doubt our species needs a rich diet to cover large energy expenses, but it requires relatively no richer a diet than many Cebidae and Cercopithecidae feeding on sweet fruits complimented by the protein and fat of a large proportion of insects.

Postscript: humans are where faunivore and frugivore "meet." The Hladik et al. [1999] note reminds us that gut morphology alone probably won't settle the issue of whether humans should be classed as frugivores or faunivores. Indeed, it might not be possible to unequivocally "prove" humans are in one class but not the other. One point is clear, though: gut morphology suggests that humans are where the classes of faunivores and frugivores "meet," i.e., it suggests that animal foods (even if just insects) are a part of the natural human diet.

On the term "omnivore," and misuse of quotes

Definition of "omnivore" a critical point

Out-of-context quotes from the writings of D.J. Chivers are sometimes used by fruitarian extremists in support of their claims that humans are not omnivores, but are instead adapted to a strict vegetarian/fruit-based diet. From the previous material in this section, recall that the research of Chivers and associates suggests that the human gut morphology is similar to that of faunivores, meat-eaters (which suggests that animal foods are part of the natural human diet), and does not support the alleged claim that humans are adapted for a nearly 100% fruit diet, or even a 100% vegetarian diet.

One major factor in understanding the quotes from Chivers regarding omnivores is that he uses the word in a more precise and different way than the most common usage. To understand this point, let's now examine the different definitions of the term "omnivore."

Omnivore: the common definition. The most common usage of the term omnivore is to indicate an animal that includes foods from more than one trophic level in its diet, which is usually interpreted to mean an animal that eats both plant and animal foods. From Milton [1987, p. 93]:

Humans are generally regarded as omnivores (Fischler 1981; Harding 1981). By definition, an omnivore is any animal that takes food from more than one trophic level. Most mammals are in fact omnivorous (Landry 1970; Morris and Rogers 1983a, 1983b)...

Milton [1987] goes on to note that the term omnivore is vague as there is substantial variability in the foods omnivores eat, and the term is not linked to gut morphology. Thus saying a mammal is an omnivore tells one little about the actual diet of the animal.

Omnivore: Chivers uses the term differently. Let's review some of the relevant quotes. From Chivers [1992, pp. 60-61]:

[F]or anatomical and physiological reasons, no mammal can exploit large amounts of both animal matter and leaves, the widely used term "omnivore" is singularly inappropriate, even for primates. Humans might reasonably be called omnivores, however, as a result of food processing and cookery...
Because of cooking and food processing, humans can be described as the only omnivores, but we shall see that their [human] gut dimensions are those of a faunivore.

"Omnivore" a vague term lacking in relevance for GI tract functions. A further relevant quote is from Chivers and Langer [1994, p. 4]; emphasis below is mine:

The concept of omnivory is weakened by the anatomical and physiological difficulties of digesting significant quantities of animal matter and fruit and leaves... animal matter is swamped in a large gut, and foliage cannot be digested in a small gut. A compromise is not really feasible... Humans are only omnivorous thanks to food processing and cookery; their guts have the dimensions of a (faunivore) carnivore but the taeniae, haustra and semi-lunar folds are characteristic of folivores. Among the so-called omnivores, most eat either mainly fruit and animal matter (if smaller) or fruit and foliage (if larger) but not all three.

Thus we note that Chivers appears to define an omnivore as a general feeder with a gut morphology that supports a diet that includes significant amounts of all three types of foods: fruits, leaves, and animal matter. Such a gut morphology is not found in mammals, hence the term is indeed inappropriate for mammals.

Contradictory claims about omnivores: which is correct? Thus we have what appear to be contradictory statements: most mammals are omnivores; no mammal is an omnivore. Which is correct? The answer is that both are correct, because they are using different definitions of the term "omnivore."

Chivers' criticism of the common definition of the term "omnivore" is relevant: it would be better (more precise) to use terms that are linked to gut morphology: folivore, frugivore, faunivore. However, that does not mean that those who are using the common definition are making incorrect or invalid statements. Recall that a definition is simply a convention that people follow. While it is desirable that definitions possess analytical rigor, it is not a requirement that they do so. Hence the meaning of a statment like "chimps are omnivores" or "humans are omnivores" is clear, i.e., the natural diet of humans and chimps includes both animal and plant foods.

A fruitarian extremist has used the difference in definitions of the term "omnivore" to suggest that statements like "chimps are omnivores" are incorrect and irrelevant. Because the meaning of such statements is clear (even to those who support Chivers' remarks), it is my opinion that the fruitarian extremist is engaging here in a blatantly intellectually dishonest word game in an effort to distract attention from the well-known fact that animal foods are a significant (even if small) part of the natural diet of many primates.

More examples of out-of-context quoting by dietary extremists. Those who use (some or all of) the above quotes from Chivers' writings in support of the fallacious claim that humans evolved on a strict vegetarian/fruit diet often neglect to quote the following, from one of the same source articles, e.g., Chivers [1992, pp. 60, 64]:

Exclusive frugivory is practically impossible, because certain essential amino acids and other nutrients are found only in leaves or in animal matter...
Humans are on the inner edge of the faunivore [meat-eater] cluster, showing the distinctive adaptations of their guts for meat-eating, or for other rapidly digested foods, in contrast to the frugivorous apes (and monkeys).

The first part of the above quote indicates how unlikely the bogus claims that humans evolved as fruitarians (on very low-protein diets) really are. The second part of the above quote points out that human gut morphology is more similar to that of faunivores than the fruit-diet frugivores.

Finally, readers should be aware that a more recent paper, Chivers [1998], includes quotes similar to the above. This is mentioned in the event the fruitarian extremist in question here might decide to "update" their argument by utilizing more recent quotes from Chivers than the ones above.

Misuse of other quotes

Speaking of intellectual dishonesty and misuse of quotes, another quote on gut morphology that is sometimes misused by fruitarian extremists is from Elliot and Barclay-Smith [1904], as cited in Stevens and Hume [1995, p. 112]:

There can be little doubt that the human colon is rather of the herbivorous than carnivorous type.

By itself, the above quote may seem straightforward enough. However, now compare the above to a more complete version of the same quotation in full context, from Stevens and Hume [1995, p. 112] (italicized emphasis mine below):

They cited a study of 1,000 Egyptian mummies that indicated that their cecum was considerably larger than that of present-day humans. Therefore, in spite of reservations about deducing function from structure, these investigators concluded that, "There can be little doubt that the human colon is rather of the herbivorous than carnivorous type."

As previously mentioned, human gut dimensions can vary with diet. The diet of ancient Egyptians would likely be much higher in fiber than a modern Western diet of processed foods. See Milton [1987] for more information on this topic.

The above questionable use of a quote is yet another example (in my opinion) of intellectual dishonesty by a fruitarian extremist. The moral of the story here is that checking the references can be critical, particularly where only a few references have been (perhaps selectively) quoted.

Section summary and synopsis

Although by comparative anatomy analysis (alone) the issue is not yet settled, the results of two different statistical analyses of a "large" data set on gut morphology and diet (i.e., the best available scientific evidence) support the idea that animal foods are a natural part of the human diet. That is:

Humans are faunivores or frugivores adapted to a diet that includes significant amounts of animal foods.
The morphology of the human gut does not correspond to that expected for a nearly 100%-fruit frugivore, as claimed by various fruitarian extremists.
Finally, the simplistic analyses of gut morphology found in the various comparative proofs of diet are (badly) outdated.

GO TO NEXT PART OF ARTICLE

(Vitamin B12: Rhetoric, Reality, and Vegan Diets)

Return to beginning of article

SEE REFERENCE LIST

SEE TABLE OF CONTENTS FOR:
PART 1 PART 2 PART 3 PART 4 PART 5 PART 6 PART 7 PART 8 PART 9

GO TO PART 1 - Brief Overview: What is the Relevance of Comparative Anatomical and Physiological "Proofs"?

GO TO PART 2 - Looking at Ape Diets: Myths, Realities, and Rationalizations

GO TO PART 3 - The Fossil-Record Evidence about Human Diet

GO TO PART 4 - Intelligence, Evolution of the Human Brain, and Diet

GO TO PART 5 - Limitations on Comparative Dietary Proofs

GO TO PART 6 - What Comparative Anatomy Does and Doesn't Tell Us about Human Diet

GO TO PART 7 - Insights about Human Nutrition & Digestion from Comparative Physiology

GO TO PART 8 - Further Issues in the Debate over Omnivorous vs. Vegetarian Diets

GO TO PART 9 - Conclusions: The End, or The Beginning of a New Approach to Your Diet?

Back to Research-Based Appraisals of Alternative Diet Lore

Beyond Veg home | Feedback | Links