Hominid Migration and Diversification
The Mitochondrial Eve Hypothesis and an Out-of-Africa Origin
Natural and Sexual Selection on Human Behaviors
Our goal today is to gain a sense of time regarding the events leading to the human species. We begin our story in the depths of prehistory when the homonid lineage separated from the other great apes over 5 million years ago. Australopithecus afarensis then gave to two contemporary lineages, the robust and gracile lineages. The robust lineages consisted of two largely herbivorous species (Australopithecus robustus and A. bosei), whereas the gracile lineages consisted of an omnivorous group. The australopithecine lineages date back to 2.6-3.8 million years ago. A crucial event in the evolution of homonid lineages concerns the origins of bipedalism. We realize of course that all the great apes still use a quadrapedal mode of locomotion.
The fossil record speaks loudly on this point. First, functional anatomy of the nearly complete skeleton of Lucy indicates a pelvic architecture that is clearly on the way to bipedalism if not largely bipedal. Second, the most fascinating fossil find concerns the discovery of 3+ million footprints that bellow to an australopithecine. It shows a bipedal set of footprints with a prominent heal mark, large big toe that is in-line with the long axis of the foot (in contrast with ape footprints). Our ancestors clearly evolved bipedalism early. The evolution of bipedialism may have led a morphological constraint to be freed from selection on one use, and allowed this trait to be elaborated for another crucial use. Bipedalism is a key innovation in the evolution of subsequent behaviors, and the evolution of brain size. A key innovation is a trait that evolves and is crucial for the subsequent diversification and evolution of other traits that form the hallmark of a group.
Long hooked fingers were important in locomoting between tree limbs, and a true opposable thumb might have required a shift from arboreal to strictly terrestrial modes of locomotion. In this sense bipedalism is an important requirement for the evolution of opposable thumbs as it releases selection from arboreal functions and allows the development and refinement of the hand for tool use. Our grasping arboreal appendages may have been an exaptation for tool use. An exaptation is a trait that evolved for a different purpose than the current adaptive value of the trait. (For example, feathers may have evolved for thermoregulation resulting in an exaptation for flight -- a light insulating structure which also has aerodynamic properties.) The original use for the hand was in arboreal locomotion, but the function was co-opted for tool use once bipedalism evolved.
The fossil record also speaks loudly on this point as we have an excellent series of skulls with which we can compute cranial volumes. However, as there has also been a general increase in body size from the austrolopithecines (Lucy was sub four feet) to the present, we must correct cranial volume relative to body size. We won't worry about this for the moment as Lucy was just a bit lighter (32 kg) than a chimpanzee (45 kg), it's skull was only modestly larger (450 cc) in volume compared to a chimp skull (350 cc), or compared to the cranial volume (500 cc) of the larger bodied gorilla (120 kg). A pygmy chimpanzee has a brain volume of 350 cc and a body mass of 35 kg. This great ape is the closest in body size to A. afarensis. This means that australopithecus had a brain roughtly 450/340 = 1.3 times larger than great apes or 30% larger.
<I will draw a log-log graph on the board>
The proper way to make such comparisons of brain volume relative to body size is to use allometric plots (log-transformed) of brain size relative to body size. If brain volume scales isometrically with body mass, we would expect that these two traits would be proportional to one another. This would imply a slope of one on a log-log plot of brain size relative to body size. If we were to use body hieght relative to brain volume we would expect these two values to scale with a log-log slope of 3.0 since hieght is a linear metric, whereas brain volume is cubic linear metric (3rd order measure).
A slope of one gives us a useful "null hypothesis" that we can use for comparing the lineages of great apes, the australopithecines, and the other hominid lineages that arose after the australopithecines -- Homo.
The lineages of great apes fall on a line with a slope of 0.34 suggesting that brain size scales with the 1/3 power of body mass.
The lineages of australopithecines fall on a line with a slope of 0.33 which suggests that brain size also scales with the 1/3 power of body mass. Note that the line for australopithecines is higher than the line for the great apes even though both lines have the same slope. This indicates that australopithecines did indeed have larger brains relative to body size compared to the great apes, even though the increase was only 1.3 times larger (see calculations above).
In contrast, the lineages of homo fall on a line with a slope of 1.73. The evolution of brain size in Homo lineages greatly exceeded the slope of 1 which would indicate a proportional increase in brain size with body size. Brain size was evolving rapidly from Homo habilis, the handy man (2.1 million years ago), to H. erectus (~200 years ago), to H. sapiens in the present.
The relative antiquity of the two hallmarks of the human condition, bipedalism and brain size, do differ somewhat in that bipedalisms appears slightly early in the australopithecine lineages relative to the explosion in brain size seen in the lineages of Homo. In addition, the origins of expanding cranial capacity are tightly associated with the evolution of tool use in Homo habilis. These events can be summarized on a phylogeny in which we plot the various cranial traits and use these traits to reconstruct a phylogeny of the great apes, Australopithecus, and Homo.
<character state matrix for brain, and bipedal traits will be plotted>
<phylogeny will be drawn using the principle of parsimony>
A relatively simple developmental explanation for the evolution of large brains concerns the retention of juvenile characters. Chimp infants have a cranial morphology that is remarkably similar to the the modern adult human. Chimps and humans differ in how brains grow after birth. In chimps, the skull (cranial region) does not grow much relative to the face (facial region). In chimps, this results in a greatly elongated facial region, ergo the muzzle of chimps and many apes. In humans, the cranial region continues growing after birth. This growth and proliferation of neural circuits is in a large measure responsible for our greatly enlarged brains. It has been suggested that we remain "juvnile chimps" well into adulthood, and in so doing, acquire greatly enlarged brains. Many of our learned behaviors are acquired during this greatly prolonged, post-birth, growth of the brain. The extended parental care found in humans (up to 18 years) is largely unheard of in the animal kingdom. Recall the ideas on growth and proliferation of neural circuits in vertebrates.
The slope of the developmental allometry, or the log-log plot that describes brain development in apes, corresponds quite nicely to the phylogenetic allometry that describes brain evolution in humans.
<allometry for development of brains will be drawn>
Our next story begins in prehistoric north Africa. A clan of Homo erectus is doing quite well. So well that this clan begins fissioning. We trace the fissioning of these clans through the eons, and we keep track of one tiny part of the genome. A single female from this clan ultimately gives rise to the rest of the world's Homo, which over the course of the ages are evolving into H. habilis, H. neanderthalis, and H. sapiens.
Let's say we could take a snippet of the mitchondria from all of the females in this clan. The mitochondrial genome has a matrilineal inheritance, which is to say we all get our mitochondria from our mother (father's do not supply their progeny with mitochondria from the sperm -- it all comes from the mother's egg. One of these Homo erectus female's mitochondria would have given rise to all subsequent mitochondria in all other homo lineages. This female would be our Mitochondrial Eve. Of course we do not have access to ancient DNA, at least not in these hominid clans.
We only have tissue from extant homonids, so we could sample humans from around the globe comparing all the mitochondria in search of a geographic region or race that had mitochondria that was the most ancestral. In this search we are literally trying to find the outcrop mitochondria for homo -- that mitochondria that has the deepest roots in the phylogenetic tree. If we find an area which possesses such a "root", then all other samples from that area should also root deep into the tree.
This is the logic used by Cahn, Stoneking, and Wilson in a landmark paper on human evolution. They found that a single geographic region possessed such a female -- a region from North Africa. Moreover, all African samples of mitochondria are rooted very deeply in the clade that was drawn from a sampling of human mitochondria from around the globe.
<clade of mitchondrial lineages will be drawn>
They found that the roots from the "most parsimonious tree" that described the phylogenetic relationship of human mitochondria had deep roots in Africa, the next deepest roots were in Asia, followed by European and New World lineages. The search for a common ancestor for all modern humans had deep roots indeed. From fancy versions of a mitochondrial clocks that are termed coalescence times, they computed the time at which all these mitchondria coalesced into a common ancestor -- the Mitochondrial Eve. The time was plotted between 180,000-270,000 years ago, which would correspond with a Homo erectus representative.
The story seemed very nice and it also appeared to unite human ancestry back into the depths of time. This hypothetical clan from North Africa spread from out of Africa to all locations around the globe. Since that time, the lineages could not have experienced much mixing (otherwise we might find lineages of African mitochondrian rooting in Asia, and vice versa. The humans in these regions then evolved in relative autonomy.
Unfortunately, this simple pattern of homo's spread around the globe from an out-of-Africa ancestry was shattered by Alan Templeton and others. They used the data from Cahn, Stoneking and Wilson's paper to erect a different hypothesis -- one in which there has been considerable mixing on genes between many different geographic regions. Cahn, Stoneking and Wilson constructed what they believed to be the most parsimonious tree of human ancestry based upon the mitochondrial genome. However, the number of trees that could be constructed from their data set numbers in the gizillions -- a very big number that would be impossible to search for the single parsimonious tree. There are algorithm's that attempt to find the most parsimonious tree from among the gizillions, however, they are fallible. Indeed, Templeton managed to find an even more parsimonious tree that had a strikingly different topology compared to the tree published by Cahn et al.
<another mitochondrial tree will be drawn>
This tree had asian lineages rooting very deeply amongst the african lineages, european lineages rooting deeply amongst the asian lineages. Gone was the nice picture of strictly deep-rooted African lineages. The new picture of homonid evolution was one in which homo might have evolved in either Africa or Asia -- Eve could have come from two places. Moreover, the picture was also messy enough that considerable exchange of genes could have taken place during the subsequent diversification of homo lineages. This lead to the construction of the Multi-regional hypothesis for the origins of modern homo. The evolution and diversification of modern homo took place across a large geographic region that encompassed Africa, Asia, and Europe. Exchange of genes through migration took place during the 200,000+ years of modern homonid evolution. There is no cradle of origin for the homo lineage -- or so the Multi-regionalists would contend.
The debate rages on as newer and better phylogenies are drawn up that allow for better resolution of the homonid DNA phylogeny. Such information is not restricted to the mitochondrial genome in search of the mitochondrial Eve. The mitochondria of humans was originally used because researchers needed an area of the human genome that evolved quite rapidly. Because the mitochondrial genome does not have elaborate DNA repair mechanisms like the nuclear genome, rates of mutation in some mitochondrial sequences are an order of magnitude higher than nuclear DNA (an exception to this are the genes that code for the Kreb cycle which are located on the mitochondria). Researchers have also begun searching for the Y-chromosome Adam. The Y-chromosome has many degenerate regions that evolve quite rapidly. In addition many nuclear DNA regions are being used, foremost among these is the Major histocompatibility loci (MHC). The MHC evolves quite rapidly because these genes are locked in a coevolutionary arms race with all the pathogens that attack humans. Using many genes will allow us to paint a consistent picture of DNA-based phylogenies that should allow us to support or refute the Out-of-Africa or Multi-Regional hypotheses.