To appear in The Evolution of Primate Nervous Systems (Eds: Todd M. Preuss & Jon H. Kaas), Vol. 5, Evolution of Nervous Systems, editor-in-chief Jon Kaas et al., Elsevier - Academic Press.
Constraints on brain size: The radiator hypothesis
Dean Falk
Florida State University
I. Synopsis Of Radiator Hypothesis
A. Temperature sensitivity and selective brain cooling in humans (a controversial topic)
1. Role of emissary veins (Michel Cabanac et al.)
a. Patterns of cranial blood flow in apes, hominids, and contemporary people
i. Relationship of vascular pattern to brain size
B. Significance of bipedalism
1. Interaction of bipedalism with climate/solar radiation (Wheeler 1988)
a. Two lineages of hominids (Paranthropus, Australopithecus-Homo)/two ways to drain cranial blood (Falk & Conroy 1983, Tobias & Falk 1988)
II. Significance for Understanding Human Encephalization
A. The cranial radiator was a ‘prime releaser,’ not a ‘prime mover’ of human encephalization
1. The ‘prime mover(s)’ of encephalizaton and cognitive evolution must be sought elsewhere
III. Recent Evidence Bearing on the Theory
A. New data are consistent with the radiator hypothesis
1. The ‘takeoff’ in hominid brain size may have begun earlier than previously believed (Falk et al. 2000)
2. New findings regarding early hominid habitats
3. New data for frequency of enlarged occipital/marginal venous sinuses in Paranthropus (Falk et al. 1995, Falk 2004)
4. Anatomical demonstration of the radiator (Zenker & Stefan 1996)
IV. Summary and Conclusions
A. Physiological data such as those incorporated in the radiator hypothesis are beginning to inform paleoanthropology
Australopithecus a genus of early hominin with many species, one of which probably gave rise to the human lineage; sometimes called gracile australopithecines.
emissary veins veins that pass through apertures (foramina) in the cranial wall and establish communication between the sinuses and veins inside the braincase and the veins external to it.
hominin humans and their early bipedal ancestors (excludes apes), what ‘hominid’ used to mean.
occipital/marginal (O/M) sinus route that delivers venous blood to the vertebral plexus of veins and causes a groove on the occipital (O) bone and along the margin (M) of the foramen magnum in some hominins.
Paranthropus a rugged-looking genus of early hominin that was not ancestral to humans, also called robust australopithecines
selective brain cooling a natural mechanisms that enables mammals to maintain brain temperature below that of the rest of the body during states of hyperthermia.
transverse/sigmoid sinuses cranial venous sinuses that drain blood from the back of the cranium to the jugular veins in most humans.
vertebral plexus of veins network of veins surrounding the spinal column that drains significantly more cranial venous blood from upright than supine humans.
Abstract
The radiator hypothesis focuses on differential effects that natural selection had on the vascular physiology of hominins that occupied separate niches and the consequences for brain evolution in Paranthropus, Australopithecus, and Homo. Comparative studies of emissary foramina suggest that the network of cranial veins of the australopithecine ancestors of Homo began to increase in complexity in response to temperature constraints associated with occupying niches in thermally stressful habitats. This venous network subsequently became elaborated in Homo in conjunction with increasing brain size. Because physiologists have shown that emissary veins participate in selective brain cooling by delivering blood cooled at the head’s surface into the braincase, the radiator hypothesis postulates that the elaborate network of cranial veins contributed to the regulation of brain temperature as it co-evolved with brain size during hominin evolution. The radiator hypothesis has been controversial for two basic reasons: First, it has implications for hominin phylogeny, which is always a contentious subject in paleoanthropology. Second, the subject of selective brain cooling in humans has been controversial within the physiological community. New findings are discussed that bear on selective brain cooling and the radiator hypothesis.
I. Synopsis of Radiator Hypothesis
A. Temperature sensitivity and selective brain cooling in humans
The human brain is an exquisitely heat-sensitive organ. According to noted vascular physiologist, Mary Ann Baker:
“A rise of only four or five degrees C above normal begins to disturb brain functions. For example, high fevers in children are sometimes accompanied by convulsions; these are manifestations of the abnormal functioning of the nerve cells of the overheated brain. Indeed, it may be that the temperature of the brain is the single most important factor limiting the survival of man and other animals in hot environments.”
Humans lack the special network of arteries and veins (rete mirabile) that helps regulate brain temperature in numerous carnivores and ungulates. The differentially enlarged human brain is especially sensitive to hyperthermia because its high cerebral (compared to resting) metabolic rate generates a relative abundance of heat. One way in which human brain temperature is regulated is through arterial blood that is delivered into the cranium, which is cooler than the brain it supplies. As the arterial blood circulates it removes heat from the brain, so that venous blood exiting the braincase is warmer than the arterial blood supplying it.
1. Role of emissary veins
Whole-body cooling takes place when arterial blood is cooled through the effects of evaporation of sweat from the body’s surface, a process that also contributes to regulation of brain temperature via its arterial supply. Michel Cabanac and Heiner Brinnel proposed an additional mechanism for selectively cooling the brain under conditions of intense exercise that results in hyperthermia. Because experimental evidence revealed that blood flows out of the cranium through the mastoid, ophthalmic and parietal emissary veins in hypothermic subjects but into the braincase in hyperthermic subjects, Cabanac and Brinnel reasoned that venous blood that is cooled at the head’s surface through the effects of evaporation on dilated veins is selectively delivered into the braincase under, and only under, conditions of hyperthermia (oral temperature of 37.6oC + 0.18o). The authors noted that innumerable, microscopic emissary veins exist in humans, and demonstrated (by massaging a cadaver’s skullcap) that blood is capable of flowing through this network from the outside of the skull to the diploic veins within the cranial bones and then to the inside of the braincase.
The three emissary veins that were used to record direction of blood flow are located at dispersed points of the network that supplies the entire skull: at the face (ophthalmic), behind the ear (mastoid), and at the top back part of the skull (parietal). (See Figure 1.) Cabanac and Brinnel concluded that when blood flows into the braincase in these three emissary veins, it also does so in the innumerable tiny veins that comprise the entire network. According to this hypothesis, venous blood cooled at the head’s surface under hyperthermic conditions flows into the braincase over a disperse network of tiny veins (the cranial radiator). This is a selective brain cooling mechanism that serves to keep brain temperature in check. Cabanac and Brinnel’s hypothesis became controversial among physiologists who claimed that existence of an anatomical network of cranial veins capable of delivering cooled blood into the braincase was speculative. This point will be returned to in Section III.
Figure 1. Venous sinuses and parietal (P) and mastoid (M) emissary veins in humans. SS, superior sagittal sinus; T, transverse sinus; S, sigmoid sinus; V, vertebral plexus of veins (copyright Dean Falk).
a. Patterns of cranial blood flow in apes, hominins, and contemporary people
All scorable skulls of robust australopithecines (Paranthropus) show traces of an unusual enlarged sinus at the lower backends of their skulls that routed venous blood from their heads, called the occipital/marginal (O/M) sinus. Curiously, only one other hominin shared this high frequency, namely Australopithecus afarensis from Ethiopia (the group to which the famous fossil ‘Lucy’ belongs). Neither living people nor apes manifest enlarged O/M sinuses to any appreciable degree (Figure 2). Instead, apes and humans drain blood from their heads through another route, the transverse/sigmoid sinus system (Figure 2). (Paranthropus skulls may or may not reproduce one or both sides of the transverse/sigmoid route in addition to one or both sides of an enlarged O/M system.) Because the enlarged O/M sinuses that are infrequently found in human cadavers deliver blood to a network of veins around the spinal column (called the vertebral plexus of veins), this is likely to have been an important drainage route for the enlarged O/M sinuses of early hominins.
Figure 2. Occipital views of typical cranial venous sinus systems in modern humans (left) and fossil robust australopithecines (Paranthropus) and earlier hominins from Hadar, Ethiopia (right). In humans, venous blood exits the skull through the transverse (T)-sigmoid (S) sinuses, which deliver blood to the internal jugular veins. This system may or may not be present in robust and Hadar australopithecines, in which a significant amount of blood is always drained through an enlarged occipital (O)-marginal(M) sinus system that communicates with the vertebral plexus of veins. The O/M system is always present on one or both sides of the latter (copyright Dean Falk).
In order to explore the evolution of cranial blood flow, data were collected on the frequencies of O/M sinuses and mastoid and parietal emissary foramina from skulls of African apes, fossil hominins, and extant humans. Apes have low frequencies (appearing in < 25% of specimens) for all three traits. Emissary foramina of Paranthropus are equally infrequent but all scorable specimens to date (n=11) have enlarged O/M sinuses. Humans, on the other hand, have high frequencies of both emissary foramina but very low frequencies for enlarged O/M sinuses. As discussed below, the non-robust hominins that preceded Homo show decreasing frequencies of O/M sinuses and increasing frequencies of parietal and mastoid foramina over time.
i. Relationship of vascular pattern to brain size
Figure 3 shows the frequencies of enlarged O/M sinuses and emissary foramina plotted for African apes and various hominins along with their mean cranial capacities (which approximate brain size) expressed as a percentage of the average capacity for extant Homo sapiens. The graph reveals that emissary foramina increased in frequency in Homo as brain size enlarged over time. These data and experimental evidence that mastoid and parietal emissary veins participate in selective brain cooling in hyperthermic people suggest that the wider network of veins that incorporates the two emissary veins increased in size and complexity as cooling requirements became more demanding due to the ongoing enlargement of an evolving brain. Because this network of veins acts to keep brain temperature within check much as an automobile’s radiator regulates engine temperature, it has been dubbed a cranial radiator.
Figure 3. Frequencies of mastoid (blue diamonds) and parietal (green triangles) emissary foramina and enlarged O/M sinuses (black squares) plotted against mean cranial capacities (CC) expressed as percentages of the 1350 cubic centimeter mean for modern humans (red circles). Data from Falk (1986), Tobias & Falk (1988), Falk et al. (1995), Falk et al. (2000). Dashed lines indicate fossils that have not been scored; robust = Paranthropus, gracile = Australopithecus africanus. Chimpanzees are included for comparative purposes; hominins are arranged in approximate chronological order from left to right. In Hadar and robust specimens, O/M is fixed; in the latter emissary foramina occur in low apelike frequencies. The reverse is true for the Australopithecus (gracile)-Homo lineage in which O/M frequencies fluctuate around those of apes while those for emissary foramina increase through time in conjunction with brain size (copyright Dean Falk).
B. Significance of bipedalism
Experimental evidence reveals that the vertebral plexus of veins around the spinal column receives significantly more cranial blood when living humans are upright than when they are supine. The enlarged O/M sinuses of an early biped, A. afarensis, presumably delivered blood to their vertebral plexes and have therefore been correlated with the development of upright walking (bipedalism) in that species or its ancestors. Because of their shared enlarged O/M sinuses, it was suggested in 1983 that A. afarensis was ancestral to more recent Paranthropus (or that the two groups shared a common ancestor) rather than to Homo. The proponents of Lucy as “the mother of us all” did not take kindly to this suggestion, which nevertheless found support in 1986 with the announcement of the discovery of a remarkably early Paranthropus specimen, the Black Skull dated to around 2.5 mya (million years ago), which appeared to be a probable descendant of A. afarensis.
Humans do not usually have enlarged O/M sinuses, and appear instead to utilize a network of tiny pathways to deliver blood to their vertebral plexes when they are standing or walking. The finding that blood drains preferentially to the vertebral plexus (instead of to the jugular veins) in upright but not supine postures (due to hydrostatic and hemodynamic factors) is well documented and has important implications for patients undergoing surgery of the head and neck. The named emissary veins at the back of the head contribute to the routes that drain to the vertebral plexus and, fortunately for paleoanthropology, these veins go through identifiable foramina that may be observed in fossil skulls (Figure 3). To summarize: Both an enlarged O/M sinus system and the cranial radiator channel blood toward the vertebral plexus of veins surrounding the spinal column when individuals are bipedal, but only one of these (the radiator) is involved in selective brain cooling.
1. Interaction of bipedalism with climate/solar radiation
According to Pete Wheeler, hominins that foraged for food during the heat of the day initially reduced their risk of hyperthermia by behavioral means, namely by maintaining bipedal stances that reduced the amount of body surface exposed to direct sunlight, which allowed them to occupy a noon-day scavenging niche. By the time bipedalism was fully achieved, however, hair had reduced on body surfaces that were no longer exposed to direct radiation from the sun at its zenith, and the newly naked skin and an increased number of cutaneous sweat glands facilitated evaporation and therefore whole-body cooling. Wheeler also suggested that such cooling released a constraint on brain size in Homo. Add to this the elaboration of the network of emissary and other small veins that participate in selective brain cooling and you have the radiator hypothesis!
a. Two lineages of hominins (Paranthropus, Australopithecus-Homo)/two ways to drain cranial blood
A variety of patchy environments were available to hominins living in Africa during the Plio-Pleistocene, and evidence from dentition and cranial blood flow shows that robust and non-robust (gracile) australopithecines occupied different niches. Paranthropus was a vegetarian who had no need to engage in hunting of game during the heat of the day and therefore no need of a cranial radiator that would kick-in during intense exercise. This hominin and its earlier A. afarensis relatives from Ethiopia relied on an enlarged O/M sinus system to drain blood from the cranium. Gracile australopithecines and early Homo, on the other hand, were more eclectic in their diets and, like living chimpanzees, had a taste for meat. By around 2 mya, late Australopithecus and/or some of its early Homo descendants had begun scavenging and hunting on the hot African savanna, and cranial radiators began to evolve.
II. Significance for Understanding Human Encephalization
A. The cranial radiator was a ‘prime releaser,’ not a ‘prime mover’ of human encephalization
During the early development of the network of radiator veins, thermal constraints that had previously kept brain size in check were released by the emergence of selective brain cooling in hominins that were living in niches associated with exposure to intense solar radiation. Consequently, brain size began to increase in late non-robust australopithecines and their Homo descendants, but remained forever static in Paranthropus. The radiator network of veins continued to increase along with brain size during most of Homo’s evolution, and today the human brain averages three times the size one would expect for a nonhuman primate of equivalent body size. The radiator network of veins is not hypothesized to have been the ‘prime mover’ of human brain evolution, i.e., the one trait whose selection was primarily responsible for brain evolution. (Prime mover candidates include hunting, tool production, work, warfare, throwing, language, and Machiavellian intelligence.) Instead, the radiator is viewed as an underlying and dynamic mechanism that helped regulate brain temperature and, as such, released thermal constraints that would otherwise have kept brain size in check in Homo, as they had in Paranthropus. The radiator is therefore best viewed as a ‘prime releaser’ (or one of several coordinated physiological releasers), not a prime mover of human brain evolution.
1. The prime mover(s) of encephalizaton and cognitive evolution must be sought elsewhere
The radiator hypothesis is purely mechanistic and does not address specific behaviors that may have been favored directly by natural selection as the prime mover(s) of hominin brain evolution. Of all the candidates that have been suggested as prime movers, humanlike language may be the most likely because (a) it is not found in other animals and (b) it is germane to many of the advanced cognitive abilities that characterize Homo sapiens.
III. Recent Evidence Bearing on the Theory
A. New data are consistent with the radiator hypothesis
1. The ‘takeoff’ in hominin brain size may have begun earlier than previously believed.
Since the radiator hypothesis was first published in 1990, major changes have occurred in the understanding and interpretation of the hominin fossil record that pertain to brain evolution. Revised cranial capacities for numerous australopithecines and new data for hominins from Dmanisi, Republic of Georgia contradict the traditional view that brain size ‘took off’ around 2.0 mya in Homo and suggest, instead, that brain size may have begun to increase considerably earlier in the Australopithecus ancestors of Homo. Such a paradigm shift fits with the radiator hypothesis (figure 3), and newly discovered skulls offer the opportunity to collect more data pertaining to cranial blood flow.
2. New findings regarding early hominin habitats
The radiator hypothesis incorporated a controversial assumption that early Homo and its direct australopithecine ancestors occupied patchy savanna niches that exposed individuals to risk of hyperthermia from intense solar radiation, and that other hominins including Australopithecus afarensis and Paranthropus occupied more wooded, less thermally-challenging niches. The assumption that different early hominins occupied different habitats has received recent support from discoveries of 6-7 million year old hominins (Sahelanthropus, Orrorin) that are thought to have lived in forested or wooded habitats rather than savannas. Although it now looks as if bipedalism may not have originated in savanna habitats as previously thought, by around 2 mya non-robust hominins were ‘standing tall and staying cool’ while adding more meat to their diets by scavenging and hunting in thermally stressful savannas. Such a shift to ‘higher quality’ (and easily digestible) foods at this time is incorporated in the expensive-tissue hypothesis of Leslie Aiello and Pete Wheeler that postulates a tradeoff between metabolically-expensive guts and brains, with the former decreasing as the latter increased in mass during hominin evolution. These new data are consistent with the radiator hypothesis’ basic assumption that a cranial radiator developed when hominins began to occupy thermally stressful habitats, thus releasing a thermal constraint that had previously kept brain size in check.
3. New data for frequency of enlarged occipital/marginal venous sinuses in Paranthropus
According to the radiator hypothesis, an enlarged O/M sinus occurs in 100% of scorable robust australopithecines. This assertion has met with occasional claims that a particular Paranthropus specimen lacks this feature, but in these few cases the specimens have not been complete enough to score (per guidelines published by Tobias and Falk). When enlarged O/M sinuses were first tallied, seven Paranthropus specimens were scorable and all seven had the feature. Today, 11 out of 11 scorable Paranthropus specimens have the trait, which lends strong support to suggestion that enlarged O/M sinuses are, indeed, fixed in Paranthropus.
4. Anatomical demonstration of the radiator
The radiator hypothesis’ basic assumption that humans possess an extensive network of tiny veins that participate in selective brain cooling was based on physiological and frequency data collected from a few named emissary veins. Although this assumption was controversial, in 1996 Wolfgang Zenker and Stefan Kubik demonstrated an anatomical basis for a convection process of cooling the brain by evaporation of sweat from the head (Figure 4). Transcranial heat exchange may thus occur between cool venous blood, the CSF (cerebral spinal fluid), and thin-walled subarachnoid and pial arterial branches before they enter the brain. As the authors note, the suggestion that the “unprecedented vascular bed” shown in Figure 4 may transmit temperature changes to the cerebrospinal compartment that, in turn, selectively cools the brain is “open to physiological verification or falsification.”
Figure 4. Corrosion cast of cranial blood vessels. Parietal bone preserved. Note vessels arising from the dural vascular rete and entering the bone (dv). Photograph reproduced with permission: Wolfgang Zenker & Stefan Kubik (1996). Brain cooling in humans - anatomical considerations. Anatomy and Embryology 193, 4, figure 4, copyright Springer-Verlag.
IV. Summary and Conclusions
A. Physiological data such as those incorporated in the radiator hypothesis are beginning to inform paleoanthropology
Traditional hypotheses about hominin brain evolution have often included suggestions regarding single behaviors or traits that may have been the prime mover(s) of cognitive evolution targeted by natural selection. Although entertaining, such prime-mover hypotheses are extremely speculative and do not lend themselves readily to testing. Physiological hypotheses such as those regarding a cranial radiator and expensive tissue are beginning to emerge in paleoanthropology, and these lend themselves more easily to scientific testing. The radiator and expensive-tissue hypotheses should be viewed as complementary rather than competing. Both incorporate data that suggest the direct ancestors of humans increased their exploitation of resources in open African habitats; and both identify physiological constraints that, when released, permitted brain size to increase.
References
Aiello, L. C. and Wheeler, P. (1995). The expensive-tissue hypothesis: the brain and the digestive system in human and primate evolution. Current Anthropology 36, 199-221.
Baker, M. A. (1979). A brain-cooling system in mammals. Scientific American 240, 130-139.
Blumenschine, R. J. (1987). Characteristics of an early hominid scavenging niche. Current Anthropology 28, 383-407.
Boyd, G. I. (1930). The emissary foramina of the cranium in man and the anthropoids. Journal of Anatomy 65, 108-121.
Cabanac, M. and Brinnel, H. (1985). Blood flow in the emissary veins of the human head during hyperthermia. European Journal of Applied Physiology 54, 172-176.
Caputa, M. (2004). Selective brain cooling: a multiple regulatory mechanism. Journal of Thermal Biology 29, 691-702.
Falk, D. (1986). Evolution of cranial blood drainage in hominids: Enlarged occipital/marginal sinuses and emissary foramina. American Journal of Physical Anthropology 70, 311-324.
Falk, D. (1990). Brain evolution in Homo: The “radiator” theory (target article, open peer commentary and author’s response). Behavioral and Brain Sciences 13, 333-381.
Falk, D. (2004). Braindance, revised and expanded edition. University Press of Florida, Gainesville.
Falk, D. (2004). Hominin brain evolution – New century, new directions. Collegium Antropologicum 28 in press.
Falk, D. and Conroy, G. C. (1983). The cranial venous sinus system in Australopithecus afarensis. Nature 306, 779-781.
Falk, D., Gage, T. B., Dudek, B. and Olson, T. R. (1995). Did more than one species of hominid coexist before 3.0 Ma?: Evidence from blood and teeth. Journal of Human Evolution 29, 591-600
Falk, D., Redmond, J. C. Jr., Guyer, J., Conroy, G. C., Recheis, W., Weber, G. W., and Seidler, H. (2003). Early hominid brain evolution: A new look at old endocasts. Journal of Human Evolution 38, 695-717.
Hauser, G. and De Stefano, G. F. (1989). Epigenetic Variants of the Human Skull. E. Schweizerbart’sche Veerlagsbuchhandlung, Stuttgart.
Tobias, P. V. and Falk, D. (1988). Evidence for a dual pattern of cranial venous sinuses on the endocranial cast of Taung (Australopithecus africanus). American Journal of Physical Anthropology 76, 300-312.
Vekua, A., Lordkipanidze, D., Rightmire, G. P., Agusti, J., Ferring, R., Maisuradze, G., Mouskhelishvili, A., Nioradze, M., Ponce de Leon, M., Tappen, M., Tvalshrelidze, M., and Zollikofer, C. (2002). A new skull of early Homo from Dmanisi, Georgia. Science 297, 85-89.
Wheeler, P. E. (1988). Stand tall and stay cool. New scientist 12 May, 62-65.
Zenker, W. and Kubik, S. (1996). Brain cooling in humans – anatomical considerations. Anatomy and Embryology 193, 1-13.