Reconstructing Human Evolution

The preceding discussions of primate evolution are primarily based upon morphological comparisons. Though fossil record is the only direct evidence of past species, there are other sources of inferential information about phylogenetic relationships in the biology of extant primates. Differences can be measured in the configuration, structure, and reactivity of protein structures. Nucleic acids vary in structure and chromosomes differ in their morphology and staining properties.

Since each of these systems is somewhat independent, they offer scientific replication of models of phylogenetic relationship. Agreement among independent systems lends strong credence to the major features of the primate phylogenetic "tree". Since each system and character is subject to its own evolutionary history, minor differences are expected between trees drawn from different systems. It is, therefore, important to measure phylogenetic relationships with a variety of different approaches to establish consensus.(return to outline)

Electrophoresis

Electrophoresis is based on differential movement of proteins through a standardized medium (such as starch or acrilimide gel). A medium of uniform consistency (such as a gel) is prepared with special attention to achieving identical density and pH in each preparation. After a small amount of a protein is placed at one end of the gel, an electric current is passed through the medium, driving the protein molecules along a straight line. Direction and rate of migration of a protein depends upon the size of protein molecule and the net electric charge that the protein molecule bears. After an appropriate amount of time and electric current, the gel is washed with a stain that makes the protein visible and then photographed to record the final position of protein molecules. This technique allows comparison of a protein from different individuals and different species. Direction of the electric current can be altered to produce a two-dimensional pattern of molecular migration called a molecular fingerprint. Variation in the molecular properties of a protein between individuals and between species is presumed to reflect genetic differences. Electrophoresis has been particularly successful in identifying genetic polymorphisms within species.

Immunodiffusion

Immunodiffusion differs from electrophoresis in that no electric current is applied. Serum proteins (donor SP) are isolated by centrifuging a blood sample and removing the relatively clear serum from the upper layers of a centrifuge tube. Antiserum is prepared by injecting a protein sample into a rabbit or chicken. After the animal has had time to produce antibodies to the foreign material, blood is drawn, centrifuged and the clear serum (now containing antibodies to the donor proteins as well as serum proteins of the host animal) is removed to serve as antiserum (AS). Serum protein samples (SP1 and SP2) from two individuals (or two species) and a small amount of antiserum (AS) are placed on a gel. They diffuse outward and encounter each other. A precipitin line is formed in the gel where each diffusing protein encounters an appropriate antiserum. Appearance of this line is enhanced by a stain and the gel photographed to record the results. If the two sample serum proteins (SP1 and SP2) are identical in their antigenic properties, a continuous precipitin line forms. If one of them is more similar in its antigen-antibody properties to the donor SP, a spur forms in the precipitin line between that sample and the antiserum (AS). Complex configurations of precipitin lines, reflect similarities between molecular properties of donor proteins and sample proteins. Although this technique was developed to study cross-reactivity of unmodified serum proteins, it can be applied to purified individual proteins.(return to outline)

Microcomplement fixation

Microcomplement fixation measures the similarity of a donor protein and a sample protein by quantifying the strength of the antigen-antibody reaction. Applied to both unpurified and purified proteins, most of these techniques measure how much anti-sera has to be concentrated in order to achieve a reaction comparable to the reaction seen between the original donor protein and the test antiserum. This difference is expressed as immunological distance (ID). Molecular biologists theorized that ID values were proportional to the number of amino acid differences between proteins.

A shortcoming of immunological comparisons that are used to generate phylogenetic trees is that they measure only the number of differences. They do not distinguish between convergent, derived, or primitive characters. A strength of the immunological measures is that they are comparatively objective, relatively free of observation bias, and results from one protein system can be verified by other unrelated systems.

This technique was used to propose a "molecular clock" that would date the divergence of primate taxa by the ID values. If it is assumed that the average overall rate of molecular evolution is consistent and can date one or more branching points from another source (such as the fossil record), then dates for all of the taxa can be computed proportional to the ID. Sarich (1968, 1971) used an estimate of 70 million years as the time depth for the whole of primate history (from the fossil record) to compute separation times for modern groups of primates.

(from Sarich, 1968 Table 6-7, page 112) Times of Separation Between the Lineages Leading to the Modern Groups _________________________________________________________ Groups Time(MYBP) ________________________________________________________ Homo-Pan-Gorilla 3.5 ± 1.5 Hylobatids-other Hominoids 7± 1 Cercopithecoidea-Hominoidea 22 ± 2 Catarrhini-Platyrrhini 36 ± 3 Anthropoidea -Prosimii 60 (assumed) _________________________________________________________(return to outline)

DNA Hybridization

This technique allows a general comparison of the similarity of nucleotides along two strands of DNA. Heating separates double stranded DNA molecules into two single strands. Association between the two strands is strongest if they are complementary, that is, they bear the same genetic codes. The more nucleotide differences between the strands, the easier they are to separate, and consequently the lower the temperature needed to initiate separation. If heated-dissociated DNA molecules from two species are mixed and allowed to cool, some strands will accidentally associate and bond to a homologous strand from the other species, forming hybrid DNA molecules. When the DNA is reheated, separation or dissociation of the hybrid molecules occurs at an even lower temperature than was required for dissociation of homologous strands. The greater the number of non-complementary base pairs between strands in hybrid molecules, the lower the temperature required to initiate dissociation. Careful temperature measurements provide a crude but efficient way of estimating differences between the DNA of the two species.

The measurement of DNA difference allowed the possibility of computing the depth of divergence of various primate taxa by assuming that the depth of divergence is proportional to the percentage of base pairs that differ between the taxa.

Estimates of time of divergence from DNA Hybridization (from K. D.. Kohne, J. A. Chison, and B. H. Hoyer, published in Sarich, 1971k, Table 3 page 73) _______________________________________________________ Species DNA DNA time of compared difference divergence _______________________________________________________ Human-chimp 2.5% 5 MYBP Human-gibbon 6.1 13 Human-rhesus monkey 10.3 21 Human-capuchin monkey 17.4 36 _______________________________________________________(return to outline)

Amino Acid Sequences

Detailed amino acid sequence data are available for a few proteins for a wide variety of animals. Differences in amino acid sequences are slight under-estimates of differences in the genetic code between species because of the redundancy of codons that code for the same amino acid. When differences in amino acid sequence in a particular protein are found between two species, the number of mutations (changes in nucleotides) can be estimated. Two methods are available for making phylogenetic decisions; 1) minimum mutation distance and 2) minimum path method. Both methods assume relatively uniform rates of mutations and evolution.

The most common and most convenient method is to calculate the minimum mutation distance. The principle of parsimony, i.e., selecting the solution that requires the fewest substitutions, is applied. A way of summarizing information from many different proteins is to combine information and analyze it as if the data represented a single large molecule, a process called tandem alignment.

Another approach, the minimum path method, is to try to recognize convergent, derived, and primitive sequences, and then estimate the most parsimonious evolutionary pathway that produces the observed sequences. Although more laborious, this procedure should produce more accurate phylogenetic trees. A major drawback in p studying long sequences is that one inevitably finds numerous, equally parsimonious, paths. Unfortunately, the greater the number of phenotypes compared, the more equally parsimonious models one can generate. A scientist then, must make an opinion about what the phylogenetic tree should look like to choose among equally parsimonious paths. Thus, trees produced in this way should be considered possible evolutionary pathways. Evidence of their accuracy must be sought elsewhere.(return to outline)

DNA Fingerprinting

Development of techniques for purifying and cloning DNA strands allows production of enough of a particular DNA sample for examination by electrophoretic techniques. Mutations do occur during the cloning process, so unlimited quantities of a particular strand can not be grown. Otherwise mutations would become so numerous that some original characteristics would be lost. Carefully cloned DNA is cut into segments by restriction enzymes that recognize specific base sequences and cut the DNA strand at the site of those sequences. This process cuts the DNA into numerous segments since a "cut" is made everywhere that particular sequence occurs. The resulting DNA segments, restriction fragments (RFs) have different molecular sizes and characteristics that allow them to be separated and visually examined by electrophoresis.

A specially prepared probe, a short radioactive strand of DNA of known base sequence, can serve as a label to identify which RF bears that sequence. The length of particular RFs often varies when DNA samples from different individuals are compared. This restriction fragment length polymorphism (RFLP) reflects variation in the DNA code, and allows a simple but valuable comparison of genetic differences. It also makes it possible to search for probes that can mark the presence or absence of specific genes and provide a laboratory test to identify some genetic disorders. Another application of the technique was to select some highly variable RFLPs that readily serve as unique markers for that individual - a DNA fingerprint. This has immediately been valuable in forensic identification where tissue or fluid samples can be "DNA fingerprinted." Since DNA fingerprints are completely hereditary, they can not be altered or disguised. This technology has already become important in criminal investigations, paternity cases and estate lawsuits. If a person has ever been DNA fingerprinted or if tissue samples are available, they can be positively identified. Although expensive, this procedure also has great scientific value as a way of documenting kinship in animal populations.

It is also worth noting that, because of the complex procedures involved, any human error made in the process will likely result in a "bad" fingerprint, not one mistakenly belonging to someone else.(return to outline)

Mitochondrial DNA

The high rate of mtDNA evolution, thought to be a consequence of less efficient repair mechanisms relative to nuclear DNA, make it an attractive molecule for comparisons between closely related species or intraspecies populations. Its apparent haploid pattern of inheritance (mtDNA is passed only by females to their daughters) makes mtDNA variability more vulnerable to loss than nuclear DNA by population bottlenecks (periods with small breeding populations), founder effects (Wilson et al, 1985), or selection.

Samples of mtDNA are used to estimate divergence times for populations within species. For example, Japanese monkey females usually remain in their native troops for their lives (Sugiyama, 1976). Although male migration disperses nuclear DNA, any mutations in maternal mtDNA are less likely to be passed to other troops. A comparison of Japanese monkeys from four localities produced four types of mtDNA (Hayasaka et al, 1986):

Assuming a substitution rate of 2% per million years, the mtDNA from areas 3 & 4 are estimated to represent 4x106 years of separation. Localities 1 & 2 have been separated from 3 & 4 for about 6x106 years.

Humans have a low mtDNA diversity across geographic areas, suggesting recent dispersal (Ferris et al, 1981). Haseqawa, Kishino, and Yano (1985) using mtDNA sequence data to date the separation of the human and pongid lineages, calculated a divergence time between human and chimpanzee to be 2.7 MYBP! They resolved the conflict between this date and the fossil record of Australopithecus and early Homo by suggesting that humanity is descended from a chimpanzee female that mated with an early human!

Cann, Stoneking & Wilson (1987) used twelve restriction enzymes to study mtDNA from 147 people of various populations to draw a genealogical tree of five geographic populations (Africa, Asia, Australia, Europe, and New Guinea). They concluded that these populations stemmed from one African woman, and that hybridization between the expanding modern stock and archaic humans was absent or minimal. Assuming a divergence rate of 2% to 4% (per million years) for mtDNA, they estimated that this African "Eve" lived 140,000 to 290,000 years ago. Saitou and Omoto (1987) analyzed the same data and were unable to determine whether Africans or New Guineans diverged first. Responding to critics, Stoneking and Cann (1989) revised their model, proposing that the common ancestor of all existing human mtDNA types lived in sub-Saharan Africa between 50,000 and 500,000 years ago.

A major problem with the "Eve" concept is the popularization of the idea that humanity had only one mother (an interpretation not intended by proponents of "Eve"). At any time, humans lived in populations - breeding groups. In the logic of most modern population biologists, human evolution is a process of changing populations, not the sudden appearance of a first human. The mtDNA interpretation ignores nuclear genes, in which every individual bears the contribution of numerous ancestors, not just the one female ancestor represented by mtDNA.

There are other problems represented by the "one mother" tree. It assumes that there is no reverse migration between the groups, an assumption that is historically known to be incorrect. Cann et al. (1987) presumed that the low diversity found in mtDNA could not be the result of lineage extinction since modern population sizes are increasing, the only circumstance that will maintain mtDNA diversity. Any small population or any bottleneck produces dramatic loss of mtDNA diversity. A stable population initiated by n females is likely to trace all its ancestries to a single female mtDNA in 4n generations (Avise et al., 1984). Persistence of a mtDNA lineage is influenced by the number of females bearing that lineage, the variance in the number of daughters produced by a mother, and the growth rate of the population. An error in the assumptions or parameters of the model will produce an erroneous tree.

Analysis of nucleotide sequences should provide better "trees" and more accurate comparisons of divergence. Again, mtDNA is the best studied DNA. R. H. Ward (1991) examined the sequence of nucleotides in a region of mtDNA in three Asian populations- Papua/New Guinea, Japanese, Amerindians and reports a divergence time of 110,000 to 140,000 years. Within each of these groups are subgroups with internal divergence of at least 75,000 years. A similar sequencing analysis of mtDNA was applied to husband-wife pairs of Amerindians from the Nuu-Chah-Nulth of Vancouver Island. About a third of these spouses have identical mtDNA, or differ by only one nucleotide over the 360-nucleotide segment that was studied. The mtDNA of the rest of the spouses is divergent, producing a modal value for sequence divergence of 1.7%, a value that indicates a lineage depth of 50,000 years (Valencia and Ward, 1991).

If these divergence dates are accurate, they augment a rapidly growing body of mitochondrial evidence for great time depth in separation of human mitochondrial lineages within modern populations. On the other hand, many of the tenets of mtDNA analysis are questioned by data from population and cell biology. For example, some primatologists are incredulous that Japanese monkeys could have been isolated in island populations for 6 million years and still exhibit so little evidences of subspecies formation. The evidence from cell biology relevant to hybridizing humans and chimpanzees is that the two species have different numbers of chromosomes, a condition among mammals that would most likely make any viable offspring sterile (mules). The Eve hypothesis appears, at best, to confuse the logic of gene lineages with descent lineages and ignores population biology. There is little evidence that the Nuu-Chah-Nulth have existed as a group for 50,000 years, and it is even less credible that they have maintained a system of marriage that kept maternal lineages intact that long.(return to outline)

Nuclear DNA

There is marked agreement of nuclear DNA trees with those from mtDNA nucleotide sequences for higher primates. Evolution rates of nuclear DNA may have slowed down in the Hominoidea relative to Cercopithecoidea. Most recent estimates from nuclear DNA of divergence dates of human and apes are between 4 and 8 MYBP (Hasegwa, Kishino, and Yano, 1989; Holems, Pesole, and Saccone, 1989) .

Gérad Lucotte (1989) used DNA probes to study the Y chromosome of different human populations and concluded that his phylogenetic tree for the Y chromosome DNA indicated that the Aka pygmies of the Central African Republic have the highest frequency of what he considers the master type from which all other Y chromosome DNAs are derived. In his model, there are about 200,000 years of mutation and evolution to derive all modern variations from the ancestral form.

An enormous data base (thousands of studies and papers) exists for comparisons between human populations for genetic characters. The first bifurcation separates Africans from non-Africans. A second division of non-Africans is into North Eurasian and Southeast Asian groups. Other divisions are relatively shallow in terms of genetic distance.(return to outline)

Chromosomes

One of the older ways of making phylogenetic comparisons among primate species is by comparison of chromosomes. As karyotypes became available for many species, methods were developed for analyzing karyotype variation between species. One early method was to count the number of "arms" that could be observed. If the centromere is at the end of a chromosome, that chromosome is classified as "one armed" (acrocentric). If the centromere is in the middle (more or less), the chromosome is "two armed" (metacentric). Excluding sex chromosomes, the total number of chromosome arms ("the fundamental number") remains relatively stable among closely related species. Fundamental number appears to be a more stable character than counts of chromosome number. Chromosome arms fission and fuse to change the chromosome number, but most serious disruptions or deletions within an arm do not appear to be viable.

Chromosomes contain at least two kinds of chromatin - euchromatin which is thought to be the genetically active component, and the presumed inert heterochromatin. Since these two components have different staining properties, a series of chromatin reactive stains produces visible banding patterns that allow identification of regions within chromosome arms. Identification of specific regions of chromosome arms according to banding patterns makes it possible to identify chromosomal rearrangements and to construct phylogenetic trees based on chromosome evolution.

Chromosomal rearrangements that are not catastrophic may effect fertility more than viability. A hybrid or heterozygote for a particular rearrangement may have lower gamete viability due to the difficulty of aligning mismatched chromosomes during meiosis. In theory then, one would expect homozygotes for a rearrangement to be superior in fertility to heterozygotes. New karyotypes then, are more likely to be established in small populations where chance and higher likelihood of inbreeding promote homozygosity. Presumably deme size and social behaviors (especially mating systems) influence the rate of karyotype change. Chromosome rearrangements, after they are established in a population, might act as reproductive barriers to other adjacent populations and promote speciation.(return to outline)

Non-molecular Phylogenetic Trees

Coevolution between parasites and their hosts provide yet another window to view phylogenetic relationships. Phylogenetic trees of host-specific parasites might be particularly useful since the parasite, usually conservative in its changes, might have a slower rate of evolutionary divergence than the host species, allowing parasite phylogeny to demonstrate a relationship that is obscure in the host phyla. (Dunn, 1966). If one looks at the helminthic parasites of humans and apes for example, Homo and the African apes share 11 of 21 helminth genera (52%), Homo and Pongo 3 of 16 genera (19%), and Homo and Hylobates 2 of 18 (11%). Humanity, even humans who live in Asia, carry helminthic worms of Africa with them. Malaria parasites exhibit a similar pattern. Three of the four human species of Plasmodium have identical or near identical counterparts in chimpanzees and gorillas, but the orangutan and gibbon have different types.

The human malaria varieties that are similar to those of the African apes are not as lethal, suggesting a long co-evolutionary relationship between parasite and host. Indeed, some human genotypes are resistant, the most extreme being one of the alleles of the Duffy system (Fy4) which provides immunity to Plasmodium vivax. Plasmodium falciparum malaria, a more virulent disease, accounts for more than 95% of all human fatalities. P. falciparum is phylogenetically more related to the avian malarias than to ape and human Plasmodium species. Falciparum malaria, being a relatively new human parasite, has not had time for coevolutionary processes to produce a less lethal parasite-host relationship.(return to outline)

Models of Human Evolution

Human evolution features at least two major adaptive radiations. Prior to the first human adaptive radiation, combined genetic, anatomical, physiological, and behavioral data leave little doubt that humans shared a close relationship with ancestors of the African apes. Calibrating molecular clocks is problematical, but it seems likely that common ancestral populations between chimpanzees and human lineages will be found in sediments deposited between 5 and 8 MYBP. The oldest known Australopithecus (at 4.5 MYBP) may not be far from that separation.(return to outline)

The First Human Adaptive Radiation - Bipedal Pongids

Consider a protohominid and selection pressure from injuries as a transition is made from a "pongid" niche to a "hominid" niche. For convenience we can use a chimpanzee niche as our prototype (after Zihlman and Brunker, 1979). Chimpanzees have many of the elements that we expect a protohominid to exhibit. It is largely terrestrial, but like the baboon it still utilizes trees. As our imaginary protohominid increasingly exploits resources on the ground, it retains its dependence upon trees. Unlike the gorilla which has become largely a terrestrial quadruped, the hominid becomes progressively bipedal. As the foot becomes less prehensile and more terrestrial, the brachiating type of upper limb allows the hominid to remain an capable climber (albeit falls probably increased with foot changes). This is a reasonable configuration since the chimpanzee prototype forages with its hands to secure a human-like diet (Goodall, 1986).

A key component to becoming bipedal in a human fashion is the alteration of pelvis, hip joint, knee, ankle, foot, and spine form toward a modern configuration. The pelvis in not a single unit, and must represent a compromise between several sets of functional requirements (Washburn, 1963). Though the birth canal in all primates is constricted, locomotor anatomy puts limits or demands compromises between fetal dimensions and locomotion.

The birth canal of cercopithecoids is challenged by the size of the skull of monkey infants. The pongids however have expanded pelvic basins that accommodate the broader thorax of infants of brachiating species. This relieves some of the selection pressure on neonatal brain size since pongid birth canals are spacious compared to infant skull sizes.

Since bipedalism developed in hominids prior to encephalization, selection was on pelvic locomotor anatomy. There is a functional conflict between demands of bipedalism and the necessity of a birth canal large enough to accommodate the infant's shoulders. A substantial increase in brain size could be accommodated by the birth canal, and further slight increases in neonatal brain size could be accommodated by reducing the face and presenting the head during birth in a different manner to the birth canal. A reduction in muzzle length could be a compromise to the selective pressures associated with encephalization and bipedalism. We now have a protohominid with bipedalism, slight encephalization, and possibly a reduced muzzle that occupies a niche somewhere between the chimpanzee and baboon. That is, it uses the ground like a baboon, but is a far safer climber with its long mobile arms.

What then, would drive our protohominid from the trees? A chimpanzee can climb with objects in its hands, but with difficulty. Our protohominid, with its terrestrially adapted feet, would have an even greater problem. The role of trees would be fundamentally altered in the human niche if the hands were not free for climbing. A possible next step was a continuation in the trend for encephalization. Pelvic anatomy dictated a compromise, that of delaying cranial growth to the postnatal period. Increased postnatal encephalization in turn means increasing dependency of neonates upon the mother for locomotion and support. This dependency could have several consequences. It would be increasingly dangerous for a mother with a neonate to climb. There could be dramatic increases in fall rates among small children. It is possible that increasing encephalization of the protohomonids was a primary component in the selective pressures that took humans toward tree substitutes in behavior and technology. By modern standards our protohominid infants could be quite precocious and still find the trees a dangerous place. Humans are still good climbers, but there is a long period in every child's development during which locomotor skills do not include effective climbing. Unlike objects, a dependent infant can not be discarded to free one's hands for climbing.

A. afarensis, a lineage of relatively terrestrial pongids who were habitual bipeds but may have slept in trees, is a representative of an early human adaptive radiation. Their brains were small (350 cm3), but the pelvis and knee are configured like bipedal humans rather than quadrupedal pongids. Limb proportions are human (Lovejoy (1993). The A afarensis great toe (metatarsal I), which is opposable in pongids, is aligned with the other toes, an adaptation for bipedal walking (Latimer and Lovejoy, 1990a, 1990b). Bipedal locomotion is no longer the facultative bipedalism seen in extant pongids, it is obligate bipedalism - the spinal column, arms, pelvis, legs, and feet are modified to make quadrupedalism difficult. Body size is extremely dimorphic -- males weighing over 60 kg and females about 30 kg. They were apparently part of an adaptive radiation of Australopithecines into the terrestrial pongid niche, a radiation that produced at least two other Australopithecine lineages.

A third lineage derived from this adaptive radiation of terrestrial pongids is characterized by retention of a rather primitive (chimpanzee-like) face and an expanding cranial capacity. The earliest representatives are Homo habilis, thought to be a lineage of bipedal pongids that progressively emphasized tool use as evidenced by discarded tools that litter their landscape. Since Australopithecines overlap H. habilis in time, there is no way of knowing whether Homo was the only tool maker, but it is clear in later Homo sites that tool manufacture and use was a fundamental element of subsistence activities, and is an important character in the second human adaptive radiation.(return to outline)

The Second Human Adaptive Radiation - Tools and Material Culture

Tools were not included in this discussion, but they played an important role from the earliest protohominid (Washburn, 1960). Tool use is one of many things that a chimpanzee (and probably the protohominid) does with its hands. It is evident that tool use confers benefits, and that once our protohominid began to manufacture stone tools, a new set of selection pressures applied. Although less obvious, the manufacture of stone implements is dangerous. Slight errors in skill or lapses in concentration cause injury to hands. Flakes are flung from cores with high energies and dangerously sharp edges. Whatever the functional value of the stone cutting edge, it has a high cost in liability to hand, eye, and lower part of the body. Rapid and dramatic expansion of the hominid cerebrum, potential for manual skills, and enhancement of the previously existing trend toward handedness may be consequences of selection pressures to avoid injury during tool manufacture. Skilled hands not only made safer stone knives, they produced other artifacts, perhaps even something to carry an infant.(return to outline)

Geographic Dispersion

Though Australopithecines were basically an African lineage, Homo apparently evolved relatively quickly into a large brained cultural species (Homo erectus) that colonized Africa, Europe, and Asia, taking Acheulean artifacts with them over most of their range. Crania from various H. erectus are characterized by ranges of cranial capacity that are below, but overlap, those of modern humans. Each continental area is anatomically distinctive, suggesting at least subspecies differentiation due to geographic isolation.

H. erectus regional populations were replaced by H. sapiens with modern ranges of cranial capacity, but retaining many of the robust features of H. erectus. Tool kits (as evidenced by artifact assemblages) change in some geographic areas, but some H. sapiens retain the Acheulean technology of H. erectus. A wide distribution of the transitional anatomy (H. sapiens soloensis) suggests gene flow or migration as the agent of dispersal of archaic H. sapiens rather than independent regional evolution from H. erectus to H. sapiens. However archaic H. sapiens is polytypic with at least two regional subspecies (H. sapiens rhodesiensis and H. sapiens neanderthalensis) whose relationships to other H. sapiens is uncertain.

Anatomically modern humans (H. sapiens sapiens) appear in the fossil record about 100,000 years ago. They overlap in range with Neanderthals, sometimes alternating occupation of the same sites without manifesting the anatomical intermediates that one would expect if gene flow or hybridization occurred. These modern humans are less muscular and have longer legs than Neanderthals, suggesting a more tropical ancestry (Trinkaus, 1984). However, they utilize comparable lithic technologies.(return to outline)

Origins of Anatomically Modern Humans (H. sapiens sapiens)

Currently there are four competing models in discussions of the appearance of Homo sapiens sapiens: MRE, RAE, AES, and RLE.(return to outline)

MRE

The multiregional evolution (MRE) model proposes continuity in each area, especially Europe. H. erectus populations in Africa, Europe, and Asia independently evolved into H. sapiens. A modified multiregional hypothesis suggests that sufficient gene flow occurred across continents to maintain species continuity and geographic distance preserved regional differences as clines (Wolpoff, Wu and Thorne, 1984; Wolpoff, 1989). The MRE model makes these predictions:

1. No single definition of modern humans will apply to all regions due to regional diversity.

2. Anatomically modern form does not necessarily have to be earlier in Africa.

3. Early modern humans in Asia will lack African features since they are not African migrants and are adapted to a different climate.

4. Unique regional anatomical features should be found earlier in the periphery of the different areas, yet in central regions (where gene flow is more likely) will be less diverse.

5. Early modern humans should have some anatomical features derived from the archaic Homo in their region.(return to outline)

RAE

Theories that propose a recent single origin for modern populations and subsequent rapid replacement as they migrate into other areas have been called the "Noah's Ark" hypothesis (Wolpoff, 1989). The "out of Africa" or recent African evolution (RAE) model (Stringer and Andrews, 1988) is one of these. It proposes a sharp break between archaic and modern H. sapiens. In this model, early modern humans originated outside of Europe, probably in Africa, then migrated into Asia and Europe, where they replaced Neanderthals and any other archaic human populations (Bräuer and Rimbach, 1990). In the RAE model, evolution of modern humans was a speciation event. It further proposes that more archaic human populations would not be fertile with modern humans so there are no expectations of regional admixture between modern and archaic forms. RAE makes these predictions:

1. Modern anatomical forms will be found earlier in Africa than other regions.

2. There will be no hybrids or intermediate forms during dispersal due to hybridization.

3. Modern anatomical form existed in Africa by 100,000 YBP.(return to outline)

AES

The Afro-European sapiens (AES) model (Bräuer, 1984) is similar to early versions of the MRE model in that it assumes that archaic humans were a regional development in each area from older H. erectus populations. Gene flow between continents prevented speciation. Modern humans evolved early in southern Africa (100,000 YBP) and gradually moved across Africa, into Europe, and across Asia. Gene flow from these migrants modernized the indigenous archaic human populations, but hybridization permitted local continuity during the process and contributed to regional differences in modern humanity. Identification of African specimens that are early in time yet modern in anatomy rests currently on some fragmentary and controversial Middle Stone Age remains at Klasses River and Border Cave. The AES model makes these predictions:

1. Modern anatomy will be found much earlier in Africa than other continents.

2. Early modern populations in a region will be more "Africanized" than later locally hybridized people.

3. Modern human anatomy should be established in Africa by 100,000 YBP.(return to outline)

RLE

The recent language evolution model (RLE) proposes that modern language abilities and cognitive skills evolved slowly during the history of Homo and achieved contemporary aptitudes after the appearance of modern human morphology (Livingstone, 1969; Isaac, 1972; Klein, 1989). Modern humans are characterized by extraordinary talents for language, art, tool-making, and modification of their habitat. These characters may mark the appearance of our present biology more accurately than chin form or cranial outline. Carving, painting, sculpture, and extremely rapid diversification of lithic technologies began about 40,000 YBP. There is no evidence of earlier innovative cultures that one would expect if modern humans had been present (Clark, 1989). Humans bearing this biological aptitude for language and culture exploded across areas inhabited by more archaic H. sapiens, taking their lithic techniques and burial habits with them, and in a relatively short span of time breached ocean barriers to inhabit every major land mass on the planet. That this dispersal of cultural traits occurred in a few tens of thousands of years implies that all of modern humanity has a relatively recent common ancestry, and that recent regional differentiation into contemporary races is the result of rapid evolutionary change (perhaps strong selection pressures) rather than great time depths for population divergence. There may have been hybridization with previous inhabitants, but those genotypes with modern speech and cognitive skills would have had a strong selective advantage. In this model, the appearance of modern cognitive and speech skills upsets any balance that might have kept archaic human populations in check and produced population growth that released wave after wave of migrants. The RLE model makes these predictions:

1. There will be a substantial time lapse between the appearance of morphologically modern humans and the explosive cultural dispersal characteristic of modern aptitudes.

2. Hybrid or intermediate forms might occur, but they would be quickly swamped by genotypes with modern speech and cultural abilities. Hybridization during colonization would serve to perpetuate some regional gene alleles (such as shovel-shaped incisors in Asia), but true hybrid morphologies should be rare.

3. Migration was not a single event but occurred numerous times as cultural changes de stabilized adjacent populations. Gene flow should produce a pattern of high early variability that diminished with time.(return to outline)

Summary

Regional differences (especially in Asia) among modern humans are reminiscent of archaic populations in that area, but perhaps the strongest evidence for an MRE model is the cultural continuity evident from archaeological assemblages. Some critics of the MRE model, see no convincing intermediate fossils between archaic and modern humans in Europe (Stringer, Hublin, and Vandermeersch, 1984). Others see numerous intermediate anatomies (Jebel Irhoud I, L.H. 18, Dali, Salé,...). Critics argue that if subspecies divergence occurred among H. erectus on different continents, geographic isolation and contrasting environments should have produced even more extreme evolutionary divergence, not convergence in later populations. Advocates of MRE theorize that gene flow occurred between regions at a sufficient levels to prevent speciation and to establish regional morphoclines. Groves (1989) demonstrates that the anatomical characters that produce a similarity between modern Asian peoples and Asian Homo erectus are primitive characters found in other geographic regions and should not be interpreted to fit the MRE model, yet he does not rule out all in-place evolutionary changes in Asia.

Critics of the RAE and AES models point out that these models are built around imprecise dates and fragmentary fossils. Since molecular dates are calibrated by reference to the fossil record, they tend to demonstrate what the modeler expects. The earliest well dated anatomically modern humans, the Skhul/Qafzeh group at 90,000 YBP, are from Western Asia. More and more African sites with possible modern fossils and dates at, or older than 100,000 YBP are being reported.

Did dispersal of humans with modern aptitudes coincide with dispersal of modern morphologies? Perhaps not, if modern morphologies are 100,000 years old. The archaeological record, which reflects human activities and abilities, may be a better indicator of the dispersal of modern biology than chins, brow ridges, tooth size, or cranial dimensions. On the other hand, it is important to remember that many humans of the current millennium had material cultures whose archaeological representation offers little evidence of mental abilities above those of archaic humans (Deacon, 1989). A complex language, folklore, kinship system, music, and all the other components of modern culture are not always evident in non perishable material culture.

Currently there is no consensus for any of these models among paleoanthropologists, although each has convincing advocates. One can only be certain that the continuing search for new data and new techniques will improve the accuracy and completeness of future models. Remember that the choice of which model to accept has substantial impact then on how the species concept is applied to archaic humans (Smith, Falsetti, and Donnelly, 1989 ). If the modern morphotype does not hybridize with archaic forms, the archaic varieties should be elevated to species status.

Whatever the timing of the dispersal of modern genotypes, the molecular evidence documents its major features. The oldest split separates Africans from non-Africans. The fit between linguistic classification and the nuclear gene tree is surprisingly good considering that languages are not gene controlled and that languages are thought to evolve more rapidly than genes.(return to outline)

Home | General | Research | Instruction | World

Table of Contents

15 Aug 2004
Department of Department of Anthropology, College of Liberal Arts , UT Austin
Comments to cbramblett@mail.utexas.edu