The following is taken from an article by Christopher B. Ruff, Climatic Adaptation and Hominid Evolution: the thermoregulatory imperative published in Evolutionary Anthropology: issues, news, and reviews, Vol. 2, Number 2, 1993.

During the past several years there has been a resurgence of interest in the effects of climate on hominid morphological variation and evolution. This interest has been spurred, at least in part, by the discovery of new hominid fossil remains. For example, discovery of the Laetoli footprints and Hadar fossils during the 1970s confirmed the large time lag between the development of bipedalism and the increase in brain size in hominids and contributed to the development of Falk's thermoregulatory "radiator" theory of brain evolution in Homo. The discovery in 1984 of the juvenile Homo erectus skeleton, KNM-WT 15000, together with the discovery a decade earlier of the Australopithecus afarensis A.L. 288-1 ("Lucy") skeleton, demonstrated conclusively that general body morphology differed dramatically between early African H. erectus and Australopithecus. This has prompted exploration of possible thermoregulatory mechanisms that could have led to such variation in body shape and the implications of these mechanisms with regard to the ecology of early hominids. 25 The evidence, both theoretical and empirical, suggests that climatic adaptation was and is a pervasive factor in hominid evolution, influencing not only body form, but also other biological and behavioral characteristics, including brain size.

ECOGEOGRAPHICAL RULES

If one drops from the same height a piece of paper and another object having the same weight but a more compact geometry - say, a paperclip - the more compact object hits the ground first. The reason it does so is because its surface area is smaller, so that it offers less air resistance on the way down. The same general kind of principle underlies the so-called climatic, or ecogeographic "rules" relating variations in body form to climatic gradients. Changes in the size or geometry of a body can change markedly its ratio of surface area to mass (weight). This, in turn, affects temperature regulation: a large surface area promotes loss of heat through increased radiation, convection, and evaporation, whereas a small surface area promotes heat retention. (This general observation will be obvious to anyone who has sat down to breakfast and noticed how quickly waffles, with their increased surface area, cool down when compared to pancakes!).

Bergmann's and Allen's rules, named after the nineteenth-century scientists who first clearly formulated them in print, define two aspects of the relationship between surface area and body mass. Bergmann's rule states that within a warm-blooded polytypic species (i.e., one that varies in morphology over its range), populations inhabiting colder regions tend to be larger than those in warmer regions. This is because, all else being equal, large body size in itself leads to a lower ratio of surface area to body mass. For example, the surface area of a sphere with a radius r equals 4(pie)r2, while its volume equals 4/3(pie)r3. Thus, the ratio of surface area to volume (equivalent to mass) is 1/(3r). The fact that r is in the denominator means that as r increases, the ratio of surface area to body mass decreases, just as a function of size increase.

Allen's rule states that under the same general conditions as those pertaining to Bergmann's rule, populations in colder regions are characterized by relatively shorter extremities (limbs, tails, ears, etc.) than those in warmer climates. Extremities, by their nature, are smaller than the body as a whole, as well as more drawn out and cylindrical, or flattened, which increases their surface area relative to their mass (as in the example of the paper and paperclip). The literature of the last several decades cites many examples of mammalian and bird species, as well as groups of closely related species, that vary in body size and shape across their ranges in accordance with these rules. I have recently discussed some of the possible objections to the rules and how they can be reconciled with the available empirical evidence.

BODY SHAPE IN LIVING HUMANS

As I noted earlier, anthropologists and others did not hesitate to apply such ecogeographical rules to humans, particularly by the middle of this century, when considerable metric data had been collected. Since that time, one of the most frequently illustrated comparisons of human body form to appear in biological anthropology texts and other publications is that between an African Nilotic and an arctic Eskimo. The left half of Figure 1 shows a tracing of what is, perhaps, the most well-known of these illustrations. The ratio of approximate surface area to body mass of each body form is shown below its outline.

The body of the Nilotic, which is more linear than that of the Eskimo, clearly has a greater ratio of surface area to body mass. However, as Schreider emphasized, it is misleading to limit the analysis to this particular contrast. The third outline in Figure 1 is that of a living Pygmy, drawn to the same scale as the Nilotic and Eskimo. Pygmies are not long and linear like Nilotics but, as Coon et al. recognized, their small size also gives them a high surface area relative to their body mass.

On the right side of Figure 1 is another outline of the same Pygmy enlarged to the stature of the Nilotic. This makes the nonlinear, bulky build of the Pygmy even more obvious. At this size, his ratio of surface area to body mass falls below even that of the Eskimo. In fact, populations having this general body size and shape are rarely, if ever, found living in the tropics. Instead, as I have shown elsewhere, taller tropical populations are also relatively more linear, because body breadth remains almost constant regardless of stature. For example, male Nuer, the tallest living Africans, average 48 cm taller, but less than 3 cm wider in body breadth than female Eastern Pygmies, the smallest living Africans. In these and other comparisons, the dimension I have used for body breadth is bi-iliac breadth, the maximum width across the iliac blades of the pelvis. This dimension has several advantages as a general measure of body breadth, including comparability between living and skeletal samples.

Taking the analysis one step further, if the body is modeled as a cylinder, it can be shown theoretically that as long as the breadth of the cylinder (body) is held constant, stature can vary freely while still maintaining the same ratio of surface area to body mass. Conversely, increasing the absolute breadth of the cylinder will always reduce the ratio of surface area to body mass, regardless of stature. Geographic variation among modern human populations fits the theoretical model well: not only do populations living in the same climatic zone have similar body breadths, regardless of stature, but populations living in cold climates have, on average, absolutely wider bodies than do those in the tropics, regardless of variation in stature. In fact, the correlation between latitude and absolute body breadth in a large number of modern populations is better than 0.90. The partial correlation between absolute body breadth and latitude, holding stature constant, is about the same. Nutritional factors are unlikely to explain this geographic variation.

BODY SHAPE IN EARLIER HOMINIDS

Because it is so rare in paleoanthropology that enough of a single skeleton is recovered to allow any certainty in the reconstruction of the individual's body size and shape, such reconstructions can vary widely. For example, robust australopithecines were pre- viously thought to weigh about 70 to 90 kg. Now, based on new skeletal material and new methods of analysis, they are believed to have been only about half as large.

Over the past two decades, though, two extraordinary partial skeletons of earlier (pre-sapiens) hominids have been discovered that allow fairly confident assessment of their body sizes and proportions. A.L. 288-1 ("Lucy"), an adult female Australopithecus afarensis from Hadar, Ethiopia has been dated to about 3 million years ago and KNM-WT 15000, an 11 to 1 2-year-old male Homo erectus from West Turkana, Kenya, is dated to about 1.5 million years ago.

The skeletal outlines of these two individuals are shown in Figure 2. In the middle of the figure, A.L. 288-1 is depicted at actual size relative to KNM-WT 15000 and, on the right, expanded to the same height as KNM- WT 15000. What is most obvious on first comparison of these outlines is the tremendously greater stature of KNM- WT 15000. This difference is especially striking when one realizes that he was only 11 to 12 years old at the time of his death and would probably have grown another 15 to 20% by adulthood, perhaps reaching a height of about 185 cm (6 feet, 1 inch). However, his body (bi-iliac) breadth, after reconstruction of his pelvis, is actually less than that of A.L.288- 1. Even at adulthood, his body breadth probably would have been only marginally greater than that of A.L. 288-1, even though he would have been about 78 cm (31 inches) taller. Thus, KNM-WT 15000 was not only much taller, but also much more linear than A.L. 288-1.

The estimated ratio of surface area to body mass for both fossils is similar to that prevailing among living tropical populations (compare Figs. 1 and 2). In contrast, if A.L. 288-1 is expanded to the same height as KNM-WT 15000, but her body proportions are maintained (Fig. 2), the estimated ratio of surface area to body mass falls well below the range for modern tropical populations. This is similar to what we saw when the hypothetical Pygmy was enlarged to the stature of a Nilotic (Fig. 1). As discussed earlier, this body morphology is maladaptive for tropical climates.

Figure 1

Figure 2

Less complete hominid fossils support the same general contrast in body shape between small early australopithecines and larger, later Homo. These include the Australopithecus africanus STS 14, which was about the same size and shape as A.L. 288-1, as well as several other early Homo individuals. The average estimated stature of six African Homo erectus individuals dated to between 1.7 and 0.7 million years ago is 170 cm, tall even by modern standards. Available pelvic remains of other early Homo are similar to the pelvis of KNM- WT 15000, suggesting that they also had relatively narrow body breadths.

All of this strongly suggests the presence of similar thermoregulatory constraints on the body shape of earlier hominids and modern humans. With an increase in body size from early Australopithecus to Homo erectus, the same body shape could not be retained along with a sufficiently high ratio of surface area to body mass. Thus, populations of large early Homo living in the tropics necessarily developed relatively linear bodies, despite the fact that a wider, australopithecine-like body would have had definite advantages, including easier delivery of a large- headed baby and more efficient balancing of the body over the hip joint during walking. In some ways the final body form of Homo erectus can be viewed as a compromise between these various factors.

Ecological Implications

It is interesting and instructive that all present-day populations exhibiting the extreme linearity of body build illustrated by the Nilotic in Figure 1 inhabit not only hot environments, but also relatively open, dry environments, such as savannah grasslands. As Wheeler has demonstrated theoretically, a tall linear body is a distinct advantage when moving about in the open during the day. Relative to its mass, such a body leads to less heat gain from the sun, particularly near mid-day, and greater convective heat loss from the body, particularly in the morning and late afternoon. In contrast, in a closed, forested environment with little direct sunlight and little air movement, this kind of physique loses these advantages. In addition, the usefulness of a relatively large surface area for evaporative cooling by sweating is decreased in a humid environment. Thus, given the fact that heat production is related to body size, the best way to avoid overheating under such conditions may be to limit body size itself. This is one interpretation of why present-day Pygmies, whether in Africa or elsewhere, are universally found in rainforest environments .

These considerations make it likely that African Homo erectus was limited in distribution to relatively open, at least semi-arid environments, for these are where its physique would have been most adaptive. Smaller hominids, including Australopithecus, could have inhabited either closed and wet or open and dry environments. However, given the foregoing considerations as well as factors such as their relatively small stride length and difficulty in covering long distances on foot, it seems most likely that they were limited to relatively closed, wet environments. This scenario of a shift from a more closed, wet to a more open, dry environment from Australopithecus to Homo erectus is consistent with evidence from other parts of the skeleton, such as the nasal region. It may also have implications regarding these hominids' food procurement behaviors.

Later Hominids

We know that Homo erectus spread from Africa to other areas of the world more than one million years ago but, unfortunately, skeletal remains from the great majority of this period are so scanty that we have little idea of the body size and shape that characterized these populations. It is not until archaic Homo sapiens of the past 100,000 years that we begin to have enough evidence to reconstruct general body proportions.

The adult male Kebara 2 specimen from Israel, dated to about 60,000 years ago is the only Neandertal pelvis that can be completely reconstructed, with mirror imaging of its right and left sides. The bi-iliac breadth of this specimen is very wide, more than 5 cm larger than either A.L. 288-1 or KNM-WT 15000 extrapolated to adult size. In fact, this body breadth is at the extreme range of variation for modern humans. In both size and body proportions, the Kebara 2 individual was similar to a modern-day Eskimo, having an estimated ratio of surface area to body mass of 247 cm2/kg (compare with Fig. 1).

The Kebara remains are consistent with those of other Neandertals, including "Classic" Neandertals such as La Chapelle-aux-Saints 1, indicating a relatively wide body coupled with at least moderately large size. As noted earlier, this body morphology, which among modern humans is always associated with cold environments, lends strong support to the view of Neandertals as "cold adapted." It also indirectly supports the position that Neandertals originated and developed in higher latitudes, but made periodic incursions into more temperate regions such as the Middle East. Interestingly, human populations directly following Neandertals in Europe and more "modern" populations interspersed in time with Neandertals in the Middle East are not characterized by these body proportions but, instead, they have moderate to linear forms. This supports the concept of a more equatorial origin for these populations with subsequent northward migration into Europe, albeit with some possibly significant gene flow with at least some resident earlier "archaic" populations (i.e., Neandertals).

LIMB PROPORTIONS

Another morphological characteristic of modern humans that has been shown to be strongly correlated with climate is relative limb length. In accordance with Allen's rule, populations inhabiting warmer climates have relatively long limbs, whereas those in colder climates have relatively short limbs. This has been demonstrated in living humans in a variety of ways, including limb length over body weight, arm span over stature, trunk length over stature, and lower limb length over trunk length. The differ ence between the limb proportions of Africans and Europeans, at least, can not be explained on the basis of nutrition. Unfortunately, it is extremely diffi cult to quantify this proportional difference in skeletal remains, particularly given the incomplete re mains characteristic of the vast majority of fossil hominids. Even A.L.288-1 preserves only a few vertebrae and no complete tibiae, making both trunk length and lower limb length difficult to assess precisely. KNM-WT 1500 preserves more of these elements. However, because he is a juvenile, the epiphyses, or endplates of his vertebral bodies, had not yet fused, making it difficult to estimate his trunk length. However, for several reasons which I have discussed elsewhere, it is reasonable to use the ratio of the distal (lower) to proximal (upper) long bone segment of each limb to evaluate the length of the limb relative to the body or trunk. These crural (lower limb) or brachial (upper limb) ratios have long been noted to show the same geo graphic variability as limb length to body size indices, with high ratios occurring among tropical populations and low ratios among temperate to arctic populations. Another, more statistically valid way to compare the relative lengths of distal to proximal limb segments is to regress the length of one element against that of the other. This has been done in Figure 3 for the tibia against the femur of modern humans, KNM WT 15000, and Neandertals. The modern humans have been divided into a tropical group (native Africans, Australian aborigines, and Melane sians) and a higher latitude group (Europeans, American Whites, and Eskimos). The tropical group clearly has a relatively longer tibia than the higher latitude group. Furthermore, KNM-WT 15000 falls above the modern tropical line, indicating a very high ratio of tibia length to femur length. I have shown elsewhere that the immaturity of KNM-WT 15000 should not affect this particular comparison. Neandertals, in contrast, have relatively short tibiae, falling below the modern higher latitude populations. (A.L. 288-1 could not be included in this analysis because her one preserved tibia is missing a significant portion of its shaft.)

Figure 3. A comparison of modern human populations and some selected fossil specimens comparing lengths of the tibia of the upper arm and the femur in the upper leg. This is a statistical regression line plot.

Thus, where it is possible to evaluate limb length proportions, earlier hominids that lived in hot climates (KNM- WT 15000) are found to have hyper-tropical proportions, whereas those that lived in colder climates (Neandertals) have hyper-arctic proportions. This is similar to what I found regarding the proportions of body breadth and height: KNM-WT 15000 is extremely linear relative to modern humans and Kebara 2 is extremely stocky. These results are not too surprising, for it can be expected that cultural buffering against climatic stress would have been less effective in earlier hominids than in modern humans. In other words, biological adaptation to climate in earlier hominids was probably more important than it is today.

BROAD IMPLICATIONS

The finding that the body form of earlier hominids, like that of modern humans, appears to follow ecogeographical climatic rules, suggests various corollaries. First, it probably is not appropriate to characterize a species as having a single "typical or "average" body form, if that species spans a wide geographic range. This applies to both Homo erectus and Homo sapiens, whose remains have been found from tropical to at least cold temperate climatic conditions. In this regard, it is interesting that the Homo erectus fossils from Zhoukoudian, China, indicate a shorter stature than those of Homo erectus from Africa, including KNM-WT 15000 and others.

Second, thermoregulatory factors should be considered when interpreting variation in body form among fossil hominids. For example, the body forms of A.L. 288-1 and KNM-WT 15000, although strikingly different in many ways (Fig. 2), are similar in terms of body heat regulation. Some observed differences between the pelves of taller and shorter australopithecines may also be the result, at least in part, of thermoregulatory constraints on body breadth.

Third, when deriving estimates of stature or body weight from fragmentary fossil hominid remains, these systematic differences in body proportions should be taken into account. For example, stature estimates for KNM-WT 15000 based on his long bone lengths should obviously take into account his relatively long limbs. Thus, European-based formulae make him too tall, while the most reasonable estimates of his stature are derived from modern African populations.

I emphasize the point that this general recommendation is not based on any special racial or "ethnic: affilition between these modern groups and African Homo erectus (contra Feldesman, Kleckner, and Lundys.) Rather, it is based on a thermoregulatory similarity resulting from these groups' similar climatic environment, which is shared, for example, by such distantly related modern groups as Australian aborigines. In fact, this is precisely the problem that earlier biological anthropologists encountered when they attempted to interpret morphological variation in broad racial terms. The patterning of geographic variation in morphological features among living and fossil hominids will depend on the particular environmental clines relevant to those features. In other words, the study of morphological variation and its significance, at least for those characteristics that are easily modified by the environment, should be couched in physiological rather than taxonomic terms.

Finally, thermoregulatory constraints on general body shape may have had even wider significance during human evolution. If body breadth was constrained during the increase in body size from Australopithecus to Homo erectus, this would also have constrained the size of the pelvic outlet for birth. Based on the obstetric dimensions of the KNM-WT 15000 pelvis extrapolated to the size of an adult female of his population and on his cranial capacity, it seems likely that this individual was much more similar to modern humans than to apes with regard to his degree of development at birth and his subsequent pattern of brain growth. That is, he would have been born with a relatively small, undeveloped brain and in a helpless state, a condition referred to as secondarily altricial. In contrast, australopithecines probably did not follow this growth pattern, but were more similar to great apes. This difference may have many broad implications regarding postnatal infant care, social organization, and cultural complexity in early Homo.

It is highly likely that the shape of the locomotor skeleton - i.e., the bones of the lower limb - was directly influenced by the relative narrowing of the body in Homo erectus. In particular, it has long been noted that the femur of many Homo erectus fossils has a highly distinctive shape, one that seems to indicate relatively large bending forces in one plane. These forces may be a natural consequence of the restructuring of the pelvis that occurred with the increase in the body size of early Homo and the attendant changes in muscle positioning and action about the hip joint. Currently I am further examining this issue by measuring osteological samples of modern East Africans, who share the same general linearity of body shape. In collaboration with others, I am also contrasting these results with those obtained for modern humans of very different body shape (e.g., Eskimos and Aleuts). It is only by studying such natural variability among modern human populations within a physiological framework that we will better understand the significance of morphological variation in earlier hominids.

RETURN