Consider a new allele that has an advantage of 5 percent. In a well-mixed population it would rise to high frequency in about 8,000 years. Among hunter-gatherers like the Bushmen, it might spread 9 miles per generation, on average, and among farmers, about 1 or 2 miles per generation. Since the preponderance of recent evolution seems to have been driven by the changes associated with agriculture, the 1.5 miles per generation you'd expect in farmers would be a good estimate. So in the 400 generations since the birth of agriculture (at twenty-five years per generation), a gene with a 5 percent advantage would have moved out about 600 miles.
Although this way of spreading genes is simple, universal, and easy to understand, it's not the only way, and it's slow. When you run the numbers, it's hard to see how it can carry al- leles as far as they have actually traveled in the time available. There is a similar problem in understanding the spread of oak trees in England. Some 15,000 years ago, back in the Ice Age, oak trees were extinct there, or nearly so—oak trees just don't grow very well under thousands of feet of ice. Possibly a few hung on in protected southern river valleys. Yet today they're all over the island—a fine thing, no doubt, but how did they manage it? Oaks shouldn't spread very fast for the simple reason that the acorn doesn't fall far from the tree.
The answer is that unusual events took place often enough for oak trees to spread faster than would have been possible if all the acorns stayed near their trees. Occasionally a bird carried an acorn far ahead of the main wave of advance and started a new oak forest. An acorn might have floated a long way down a river and sprouted. Maybe some of the first humans who resettled Britain after the ice melted took some acorns with them on a long hunting trip north and dropped a few. The speed of advance is determined more by these rare, long-distance events than by gravity and squirrels.
So although local marriage is a big part of the story, rarer, weirder, more complicated events most likely determined the speed of advance of new favorable alleles in humans.
BARRIERS TO GENE FLOW
Of course, there were a few major factors that blocked or slowed gene flow over the past few millennia.
The Atlantic and Pacific oceans were important barriers. There was very little contact between the peoples of the Old and New Worlds before Columbus. Australia was much easier to reach from Indonesia or New Guinea than the Americas were. There were definitely visiting sailors from Indonesia fishing for sea cucumbers along the north coast of Australia beginning around 1720. There must have been other early contacts, but the amount of gene flow was not large, judging from what we know of Y-chromosome and mtDNA variants in Australian Aborigines.8 The north coast did not attract settlers from Indonesia or Southeast Asia, probably because that coast was un- suited to their forms of agriculture. Moreover, new alleles thatwere adaptive in the context of agriculture may not have had an advantage in Australia: Even if some were introduced, they may not have spread widely.
Deserts mattered. There was plenty of gene flow between North Africa and the other lands surrounding the Mediterranean, particularly after the development of sailing ships, but the Sahara certainly interfered with movement to and from sub- Saharan Africa. The block wasn't absolute. The Sahara was a much friendlier environment in the early Neolithic than it is today. Later, the domestication of the camel favored trans- Saharan trade, while the European and Arab slave trade eventually brought African alleles to other parts of the world.
Still, we know that the amount of gene flow into sub- Saharan Africa was limited in the past, since several local mutations that cause lactose tolerance became common among the cattle-raising peoples in the Sudan and Ethiopia, even though the European version is considerably older. If there had been much gene flow into sub-Saharan Africa back then, the European mutation would most likely have dominated. In fact, if even one person carrying that European allele had successfully settled in the Sudan back in the Bronze Age, he would have had a fair chance of introducing it, as a kind of genetic Johnny Ap- pleseed.The Sahara Desert made such contacts rare, but it may also be that tropical diseases such as malaria and yellow fever interfered, just as they later interfered with European colonization attempts. The local sub-Saharan Africans had enough resistance to those diseases to get by, but outsiders generally did not.
Most mountain ranges affect gene flow but aren't impassable enough to substantially block it. The Himalayas, however, are an exception. Judging from the limited information we havetoday, it seems that India shares a fair number of favored new alleles with Europe and the Middle East, but has mixed far less with China. Since the Himalayas are the tallest mountains on earth, backed by the Tibetan Plateau, it's possible to believe that they greatly reduced gene flow between India and China, and ultimately between east and west Eurasia.
Other people getting in the way must have been one of the most powerful forces slowing gene flow. Many alleles that helped people adjust to agriculture probably came into existence in the Middle East, since agriculture began there. When we consider the ways in which those alleles could have reached western Europe, where agriculture developed later, it's worth remembering that it only takes a few months to walk that entire distance. If that happened often, you'd expect genes to have spread far more rapidly than they actually did. The problem is that would-be long-distance travelers quickly encountered other groups that spoke a different language and had uncongenial customs. Some were enemies, and all were suspicious of strangers. Passing through other groups was very difficult, so long-distance land travel was almost impossible.
In The Third Chimpanzee, Jared Diamond describes this pattern in highland New Guinea, one of the last places in the world to make contact with outsiders. He wrote, "When I was living among Elopi tribespeople in west New Guinea and wanted to cross the territory of the neighboring Fayu tribe in order to reach a nearby mountain, the Elopis explained to me matter-of-factly that the Fayus would kill me if I tried. From a New Guinea perspective, it seemed so perfectly natural and self-explanatory. Of course the Fayus will kill any trespasser: you surely don't think they're so stupid that they'd admit strangers to their territory? Strangers would just hunt their game animals, molest their women, introduce diseases, and reconnoiter the terrain in order to stage a raid later."9 Before outside contact, New Guinea highlanders spent their entire lives within a few miles of their villages, and as far as we know, none had ever seen the sea, which was just 100 miles away. It seems likely the whole world was like this in prehistory.
HISTORICAL PATTERNS OF GENE FLOW
Trade gave people reasons to surmount or bypass those barriers to travel, and it has played an important role in facilitating human gene flow. Although its impact on gene flow was probably not as dramatic as that resulting from conquest and colonization, it was important, particularly after the development of sailing ships. Sailors and barmaids, like traveling salesmen and farmers' daughters, have played a crucial role in recent human evolution.
Trade—more exactly, traders—spread new alleles along every line of communication: along sea routes, up navigable rivers, between villages and market towns. In early times it often took the form of semi-military campaigns, as when Egypt sent trade expeditions to Punt in northern Ethiopia/ Eritrea in order to obtain gold, slaves, ebony, and ivory, or when Mesopotamian kings like Sargon of Akkad sought cedarwood in Lebanon.
Trade connected widely separated civilizations—some of them, some of the time—nearly as far back as we have records. There is evidence of trade between the Indus civilization and Mesopotamia during the Akkadian Empire more than 4,000years ago—Mesopotamian cereals and wool in exchange for wood and ivory from "Meluhha," thought to be the Indus civilization. We have found Indus cylinder seals in Ur and Babylon, and references to a village made up of traders from Meluhha, who eventually became one more component of the local ethnic mix. This is exactly the sort of contact that could have played a significant role in the transmission of new beneficial alleles: The peoples of the Indus civilization probably faced similar selective press
ures to those seen in Mesopotamia, and they were far enough away that gene flow through local marriage would have taken a long time to occur. However, the contact happened early in history, so that any good alleles introduced to the Middle East in this way have had 160 generations in which to spread.
COLONIZATION
One major pattern of migration that affected gene flow was the seeding of colonies along the coasts of the Mediterranean and Black seas by peoples from the eastern end of the Mediterranean. Even as far back as the late Bronze Age, it was easier and cheaper to travel long distances by sea, not least because you could avoid having to fight your way through already-established peoples. Long-distance trade helped pave the way for these colonies. Sometimes they began as trading posts, and trade must have been the basis for the seafaring techniques and geographical knowledge that made the founding expeditions possible.
The colonizers—Etruscans, Greeks, and Phoenicians— came from the Middle East or areas heavily settled by early Middle Eastern farmers. They must have carried many of the new alleles that were adaptive in the context of agriculture, andtheir voyages probably spread many copies of those alleles to the western Mediterranean.
The Phoenicians, a people living in cities along the coast of what is now Lebanon, traded over much of the Mediterranean. Their colonies started out as anchorages and trading posts along their trade routes, but eventually some of them grew into substantial towns. The largest and most important of these colonies was Carthage (Kart-Hadasht) in Tunisia, which eventually became Rome's great rival. Many of those towns still exist today: Palermo and Marsala in Sicily, Cagliari in Sardinia, Tangier in Morocco, and Cadiz and Cartagena in Spain. Phoenician colonies appear to have spread one of the more common versions of beta-thalassemia (a genetic defense against malaria) around the western Mediterranean. Most likely this allele originated in North Africa.10
The Greeks colonized on a larger scale: A single city, Miletus, founded ninety colonies. Some were founded for commercial advantage, some as a refuge for a losing side, and others as a means of getting rid of surplus population. The Greeks founded many colonies in Sicily and southern Italy, which became known as "Magna Graecia."They include such modern cities as Syracuse and Naples. There were also many Greek colonies around the Black Sea, some (like Marseille) in southern France and others as far away as Spain and Libya. It looks as if the Greeks spread at least two characteristic malaria defenses of their own, a different version of beta-thalassemia and a form of G6PD deficiency. (Malaria defenses are probably not the only adaptive alleles that Phoenicians or Greeks transmitted—they are just the ones that have been well studied so far.) Since the recent whole-genome surveys show that many genes have beenunder recent selection and risen to high frequencies, apparently mostly in response to the new conditions that accompanied agriculture, we expect that these colonizations transmitted a number of adaptive alleles.
Recent genetic studies have confirmed a third major colonization in which an eastern group, this time from Turkey, colonized northwestern Italy: the Etruscans. They were a somewhat mysterious people who spoke a non-Indo-European language that we have not yet deciphered. The Etruscans had tremendous influence on Roman art, architecture, and religion. The question of Etruscan origins had long been controversial, with most archaeologists arguing that the culture developed in Italy, although some ancient sources, such as Herodotus, said that it originated in Lydia, a region on the western coast of Turkey. Recent work has shown that some populations in Tuscany have Near-Eastern mtDNA11 and that some distinctive local cattle also have mtDNA characteristic of Middle Eastern breeds,12 confirming an Anatolian origin.
The Etruscans added a healthy dose of Middle Eastern, agriculture-adapted alleles into the Roman mix. We have reason to suspect that those alleles shaped attitudes as well as affecting metabolism and disease resistance. Did they influence Rome's rise to power? It's possible.
YOU HAVE BEEN IN AFGHANISTAN, I PERCEIVE
Military movements also let favorable alleles vault over long distances and geographical barriers. Alexander the Great furnished one of the more dramatic examples. In the course of a remarkable career of conquest (dying undefeated), he marched as far east as Pakistan.
In addition to settling Greeks over much of the Middle East, his legacy included Greek kingdoms that survived for several centuries in Afghanistan and Pakistan. Those kingdoms didn't just influence the artistic development of Buddhism; they also transmitted alleles. Today we see a few Greek Y chromosomes among the Pathans, the dominant ethnic group in Af- ghanistan.13 Of course, we also see some Y chromosomes that are directly descended from Genghis Khan himself in the Pathans' despised neighbors, the Hazara.14 Local marriage could never have spread genes as rapidly as that—but Genghis and Alexander could.
If regional Y-chromosome variants (which as far as we know have no special inherent fitness advantage) could spread that far, you can be sure that any advantageous mutation that had become common in Greece in Alexander's time did as well. Every such allele has had a good chance of becoming common in Afghanistan by the present day. This isn't quite as true for those Mongol alleles, since they've only had a third as long to spread as those of Alexander and his merry men. In these longdistance transfers, the earlier the connection, the more important. And in the same way, large population transfers have a greater effect than small ones.
THE LOST TRIBES
Imperial politics sometimes played an important role in dispersing genes, often in highly unpleasant ways. Forced relocation of peoples has been a standard tactic during times of conflict—whether in ex-Yugoslavia or Chechnya in our day or in Assyria during ancient times. Tiglath Pileser III moved some 30,000 people from what is now northern Syria to the Zagros
Mountains in western Iran in 742 BC, Sargon II displaced about 100,000 Babylonians in 707 BC, and Sennacherib deported another 208,000 in 703 BC. One of the most notorious forced relocations was in 722 BC, when the Assyrians conquered the Northern Kingdom of Israel, destroying its capital and sending its population into exile.
These population transfers were intended not only to punish, but also to break up local elites and traditions, open up strategic areas for occupation, and provide the Assyrian state with labor and soldiers. The number of people forcibly removed from home over three centuries of these policies has been estimated at more than 4 million. Considering the enmity these actions provoked, they may have hastened the fall of Assyria— but they surely spread alleles over most of the Fertile Crescent.
There are other famous examples of forced relocations. The Babylonian Empire defeated Judah, the Southern Kingdom of Israel, in 586 BC and relocated some of the population to Mesopotamia. After the Persian Empire succeeded the Babylonians in 539 BC, Cyrus the Great allowed them to return.
AN ARTHURIAN ROMANCE
The Sarmatians were steppe nomads from the southern Ukraine who spoke an Iranian language. The classical historian Cassius Dio said, "The Sarmatians were a savage uncivilized nation,. .. naturally warlike, and famous for painting their bodies to appear more terrible in the field of battle. They were known for their lewdness. . . . They generally lived on the mountains without any habitation except their chariots. . . . They lived upon plunder, and fed upon milk mixed with the blood of horses."15
They were famous for their heavy cavalry, who fought with lances, longswords, and bows. The Romans had fought them in AD 92 and knew their quality. In AD 175, Marcus Aurelius hired 8,000 Sarmatians into Roman service and sent 5,500 of them to northern Britain. At first they were attached to one of the Roman legions there, Legio VI Victrix, but when their twenty years of service were up, they were settled in a permanent military colony in Lancashire. Apparently they never went home: The colony is still mentioned almost 250 years later.
Imagine that one of those selected alleles we see in the HapMap scans originated far to the east of Britain—perhaps as far east as Kazakhstan—some thousands of years ago, possibly in the Andronovo culture. Then, suppose the allele had
a large selective advantage, and by the time the Sarmatians were fighting for the Romans, it had become common among the Iranian- speaking steppe peoples, among whom it had diffused easily because of their characteristic horse-nomad mobility—but had not yet spread as far as western Europe. If limited to girl-next- door diffusion, it would have required several millennia to reach Britain.
That Sarmatian military colony, however, could have introduced several thousand copies of that hypothetical allele into Lancashire. The Sarmatian cavalrymen were paid well and surely could have managed to raise at least as many children as the average Briton. Starting with an original gene frequency of 0.1 percent in England in the year 175, that hypothetical allele could have a high frequency in the English population by the present time. Trade and war would ensure that the new allele spread effectively over Great Britain, and there was undoubtedly plenty of both, especially war.
Those Sarmatians may have spread ideas as well. They had some very interesting religious beliefs and legends, some of which are preserved among the Ossetians, their descendants in the Caucasus. A number of those legends sound awfully familiar— in particular, the story of a dying warrior who demands that his best friend destroy his sword by tossing it into a lake rather than allowing it to be captured by enemies. The friend can't see throwing away such a beautiful weapon, and twice pretends to have done so—but the sword-bearer, hearing his account, somehow knows that he has not. On the third try, the sword is thrown into a lake and is caught by a woman's hand coming out of the water.
The 10,000 Year Explosion Page 13