by Brian Fagan
The Pacific, showing locations mentioned in this chapter.
The settlers faced quite a challenge. Rapa Nui is subtropical, which means that the surrounding waters support no coral.2 Fish are rarer than elsewhere in Polynesia; the rainfall is less and soaks quickly into the porous volcanic soils. Fresh water was hard to come by, but the islanders got by, cultivating sweet potatoes, taro, yams, and sugarcane. They brought chickens with them, which they raised in large stone chicken houses, but their diet was high in carbohydrates. Despite the challenges of an isolated environment, as many as fifteen thousand people may have lived on Rapa Nui in about 1600, relying on intensive agriculture and careful water management.
Like all Polynesian societies, Rapa Nui’s had nobles and commoners. Oral traditions suggest that they divided the island into about a dozen kin-based territories, each with its stretch of coastline. The clans vied with one another in building platforms and colossal moiae, stone statues of high-ranking ancestors, which gazed out to sea. The archaeologist Jo Anne Van Tilburg has counted at least 887 moiae on the island, some of them still in the quarry. Over a period of four centuries, the islanders erected moiae and built platforms for them, a task that may have added about 25 percent to the food requirements of the islanders. The ecological cost was devastating. Between first settlement and about 1600, the descendants of the first settlers deforested the entire island of its magnificent indigenous palms. The island population declined by as much as 70 percent during the 1700s.
The settlement of Rapa Nui was the culmination of centuries of canoe voyaging through Polynesia. The ancient voyagers sailed from west to east across the Pacific in stages, from island to island, over many centuries. Many of their passages were no more than 300 miles (482 kilometers), but the voyage from Mangareva to Rapa Nui was about 1,600 miles (2,574 kilometers), a much more formidable proposition, with a tiny, and unknown, target at their destination. Just a few miles off course and they would have missed the island altogether and sailed on into deserted ocean. Remarkably, they also sailed eastward, against the prevailing trade winds, in canoes with limited abilities for sailing to windward. The Polynesians built their canoes with shell adzes and navigated confidently far out of sight of land by the stars, the winds, and ocean swells. To the Spanish explorer Ferdinand Magellan and his successors, the Pacific was a terrifying void that took months to cross, some 8,000 miles (12,874 kilometers) from the Strait of Magellan to New Guinea alone. The seemingly endless ocean was a “sea so vast the human mind can scarcely grasp it.”3 But to the Polynesians, the same ocean was a life-giving world teeming with fish, blessed with islands large and small where they could land, plant the soil, and raise their families. They lived close to its waters, were familiar with its every mood, and traversed its seaways with patience and consummate skill. But, like all of us, they lived, and sailed, at the mercy of powerful climatic forces.
Rapa Nui, Hawaii, and New Zealand were all colonized during the warm centuries, somewhat later than was thought until only a few years ago, which raises an interesting question. What effect did the climatic conditions of the Medieval Warm Period—a time of warmth and drought along much of the eastern Pacific Rim—have on the Polynesians? Was there, in fact, such a climatic event in the Pacific—and is there any significance in the apparent colonization of the most isolated landmasses of Polynesia during these centuries, especially since most long-distance canoe voyaging ended after 1500?
THE MEDIEVAL WARM Period is poorly documented throughout the Pacific; we cannot say with any confidence that it was a universal phenomenon. Only a handful of proxy records cover the past thousand years, most of them from tree rings or from growth rings in tropical coral. As of 2002, there were only three thousand-year-long temperature reconstructions from the southern hemisphere: one from Argentina; another from Chile; and a third from New Zealand. They are enough to tell us that temperatures and climatic conditions varied considerably from one area to the next. As always, climatic conditions were local, even if they were driven by much larger-scale atmospheric and global interactions.
The New Zealand sequence comes from Oroko Swamp on the west coast of South Island, where silver pines and other long-growing species flourish.4 Silver pine wood is virtually impervious to decay, which means that numerous partially fossilized logs survive for tree-ring analysis. Edwin Cook and his research team developed a ring sequence that covers A.D. 700 to 1999, with a high degree of reliability back to A.D. 900. The record, meticulously checked for accuracy, documents two periods of generally above-average temperatures, in A.D. 1137–77 and again in 1210–60. There was a sharp and sustained cold period between A.D. 993 and 1091, the most extreme over the past eleven hundred years. There was rapid cooling after 1500, followed by modern warming.
The Oroko sequence documents temperatures that were between 0.6 and 0.9 degrees F (0.3 and 0.5 degrees C) warmer than the twentieth-century average for the region, and for 1210 to 1260, comparable to warming since 1950.
New Zealand enjoys a temperate climate, very different from that of Polynesia. The geographer Patrick Nunn has drawn on a broad range of climatic proxies, including an oxygen isotope analysis of a stalagmite from New Zealand, and sea level studies from various Pacific islands, to argue that the Medieval Warm Period was warm and dry, with persistent trade winds.5 But he uses flow data from the Nile River on the other side of the world as an indicator of changes in the frequency of El Niño events to argue for a surge in their frequency around 1300. After 1250, Nunn argues, there was an abrupt change to cool, dry times, with increased storminess, that spanned the Little Ice Age, 1350 to 1850. Nunn calls this abrupt changeover the 1300 Event, a rapid event that precipitated a series of cultural and environmental catastrophes. According to Nunn, the 1300 Event saw significant disruptions in human settlement and subsistence, a decrease in voyaging, and greater competition, even endemic warfare throughout the Pacific.
Nunn’s climatic evidence comes, for the most part, from observations of sea level changes and geological sediments that document erosion and flooding. Such climatic proxies are not nearly as accurate as tree rings or high-resolution coral growth data. Corals, in particular, have been the subject of intensive research in recent years and give us a very different picture of the warm centuries. They provide little support for the idea of a catastrophic climatic event throughout the region in 1300.
The paleoclimatologist Kim Cobb first came to Palmyra in 1998. To her delight, she located dozens of uneroded ancient coral heads on the atoll’s west side. Since then, Cobb and her researchers have drilled cores from more than eighty such heads and revealed patterns of ocean temperatures and ENSO events over more than a thousand years. Her spliced-together cores now include samples from the Medieval Warm Period, A.D. 928 to 961, 1149–1220, and 1317–1464, as well as for 1635–1703, the height of the Little Ice Age in Europe. A modern sequence extends from 1886 to 1998.6 The variation in O-18 levels is relatively narrow, except for the tenth and mid-twentieth centuries. The tenth-century cold period was of some significance and appeared relatively abruptly, the coolest and perhaps the driest interval over the past eleven hundred years. The mid to late twentieth century was the warmest and wettest period of the past millennium. If the Palmyra corals are a reliable indicator, then a cool and dry La Niña condition prevailed over much of the eastern Pacific during a significant part of the Medieval Warm Period.
Coral and Climate Change
The annual growth rings from recent coral samples from the central tropical Pacific extend back several centuries, and we know how to read these proxies so accurately that they rival the instrument records of modern times. They provide such invaluable data on past El Niños that climatologists have turned their attention to much older, long-dead fossil corals for data on the more remote past. Unfortunately, such corals are extremely rare, as most samples are fragments that have washed ashore in storms and escaped destruction by later gales. Given their tumultuous history after death, it’s fortunate that any survive
, but most of those that do are no more than a few decades long. They resemble tree-ring samples, and, like them, have to be pieced together into much longer master sequences based on dozens of short sections.
Why are corals such useful climatic indicators? They are formed by colonies of small marine invertebrates known as polyps, which secrete hard fortifications of calcium carbonate. Millions of polyps become fused together in elaborate structures, often with numerous branches. The polyps receive their nutrients from algae that live in these structures. Sunlight penetrates shallow water, shines on the algae, and creates their nutrients through photosynthesis. Corals cannot grow in dark, deeper water and require a narrow range of water temperatures between 77 and 84 degrees F (25 and 28.8 degrees C). They’re fragile, too, usually surviving for no more than a few decades before storms rip them apart and fling them up on beaches, where they are pulverized. Intact, centuries-old coral is found only on islands in the less storm-plagued eastern Pacific with warm water, lush coral growth, and gales strong enough to detach coral heads and wash them ashore, but infrequent enough to leave them intact, undamaged by later bad weather. Few islands have yielded good old coral samples, except the remote and little visited Palmyra.*
The cores drilled into the Palmyra heads record growth rings in the corals, and through their isotopic content, variations in sea surface temperatures as the coral slowly grew. In cooler water, coral incorporates more of the heavier oxygen isotope O-18, whereas in warmer water there are higher counts of lighter O-16. By using high-precision mass spectrometry, paleoclimatologist Kim Cobb was able to measure small differences in the relative level of the two isotopes with remarkable accuracy. A difference of 0.02 percent in the relative level of the two isotopes signifies a temperature change of about 1.3 degrees F (about 0.6 degrees C).
Uranium-thorium dating techniques provided the ages of the dozens of small coral sequences from Palmyra that overlap one another. Uranium-thorium dating, sometimes called thorium-230 dating, is a radiometric dating technique that measures the age of carbonate materials such as coral. Uranium is soluble to some degree in all natural waters, so any coral precipitates minute traces of uranium. In contrast, thorium is insoluble and does not appear in the coral until the uranium-234 in it begins to decay to thorium-230. Uranium-thorium dating calculates the age of a coral sample by measuring the degree to which equilibrium has been restored between the radioactive isotope thorium-230 and its radioactive parent, uranium-234. Thanks to uranium-thorium dating, Cobb was able to piece together the beginnings of an important climatological archive.
The Palmyra cores show how El Niños exercise a powerful influence on Palmyra’s climate, with warmer, wetter conditions during El Niños and cooler, drier conditions during La Niñas. During El Niños, O-18 levels in the coral fall, whereas they rise during La Niñas, providing a surprisingly reliable barometer of climatic events that can be dated accurately with uranium-thorium dates from dozens of short sequences that overlap one another. For instance, ENSO activity was reduced during the twelfth and fourteenth centuries compared with today.
* Palmyra Island is a tiny Pacific atoll over 932 miles (1,500 kilometers) southwest of Hawaii. Fifty small islets with a total area of about 247 acres (100 hectares) form a horseshoe surrounding three lagoons. They stand but 6.5 feet (2 meters) or so above sea level, but the tall trees make the islands visible from 15 miles (24 kilometers) away on a clear day. A platform of coral and sand surrounds Palmyra, named after the American ship of that name, which sought refuge there on November 7, 1802. The Nature Conservancy purchased this near-pristine island in 2000.
Palmyra’s coral sequence is at present unique, except for a handful of more recent coral records. One such sample, from Australia’s Great Barrier Reef, spanning 1565 to 1985, confirms the slightly warmer temperatures at Palmyra between 1635 and 1703, during Europe’s Little Ice Age. The Great Barrier Reef sequence also shows that the tropical Pacific cooled just when the northern hemisphere warmed up during the late nineteenth century. Thus, there is good reason to believe that the Medieval Warm Period was a time of cool and relatively dry conditions over much of the Pacific.
Palmyra is but a tiny speck on a vast ocean, with relatively stable temperatures compared with the more marked shifts observed elsewhere in places like New Zealand. But its climatic history may have much broader significance. Cobb and her colleagues believe that changing sea surface temperature differences from east to west across the Pacific are a key player in global temperature patterns. Under this scenario, the gradient may have been greater during the tenth and twelfth centuries, bringing relatively cool and dry circumstances to this part of the Pacific during the Medieval Warm Period. Such conditions in the tropical central Pacific are very much part of the La Niña pattern, the seesaw opposite of El Niño, and the very same combination of factors brought severe droughts and lower lake levels to the American West and Mesoamerica, even to the Sahel in West Africa on the other side of the world. Furthermore, recent computer modeling relying on numerical experiments and well-established ocean-atmosphere models has replicated a period of more frequent ENSOs in the eastern Pacific during the Little Ice Age (which was warmer there), while the Medieval Warm Period saw reduced El Niño activity and cooler conditions over the same huge region.
Thousands of miles away, on the Pacific shelf off Lima, a core shows reduced ENSO activity during the period A.D. 800 to 1250, while in the Ecuadorian Andes Mountains, a high-altitude site at Laguna Pallca-cocha has provided a twelve-thousand-year record of lake sediments from an area outside the Pacific, but one that is influenced by very much the same set of climatic conditions.7 The cores from here provide evidence for El Niños in the form of marked debris layers caused by high rainfall, so the actual incidence may be much higher. El Niños can be detected as early as seven thousand years ago, but the pulse of events culminated between about 1200 and 1400, at just about the time of the final wave of East Polynesian voyaging.
The Palmyra research has overturned a long-held assumption that climate change in the tropical Pacific had followed that elsewhere, with a well-defined Medieval Warm Period followed by a cooler Little Ice Age. The opposite may have been the case in many areas. Almost certainly there were significant variations between different areas of the Pacific, perhaps measured by latitude. The so-called South Pacific Convergence Zone, a band of low-level wind, cloudiness, and rainfall that extends across the central Pacific from Vanuatu in the west to the Austral Islands in the east, was critical to human settlement. The zone moves and shifts in response to the ENSO seesaw. During El Niños, it moves northeast, while cyclonic activity shifts eastward and is more frequent. The zone shifts southwest during La Niñas, as well as responding to the Pacific Decadal Oscillation, a fifteen- to twenty-year fluctuation in sea surface temperature and rainfall whose warmer phases are associated with stronger and more frequent ENSO events.
With only a handful of meaningful climatic sequences, we still know little about the effects of the Medieval Warm Period on the vast reaches of the Pacific. Whereas in northern latitudes ice conditions or even a warmer shift of a few degrees could have significant effects, in the Pacific both climatic shifts and human settlement were affected more powerfully by the north–south movements of the Intertropical Convergence Zone, and especially by the gyrations of the Southern Oscillation and the El Niño and La Niña events associated with it. These shifts affected human societies in every region of the Pacific and far beyond. Everywhere in the Pacific, the global effects of the prolonged La Niña–like conditions that brought drought to many areas, higher rainfall to others, marked the warm centuries. And when La Niñas gave way to El Niños, as they did at times, and especially in the thirteenth century, the trade winds faltered and Polynesian navigators sailed eastward to new, remote lands.
ON APRIL 13, 1769, Captain James Cook anchored his ship, the Endeavour, in Matavi Bay. Cook’s voyage from Plymouth in England to the heart of Polynesia took eight months, most of them out of sight of l
and. His landfall was a tribute to his remarkable navigational skills in an era when observing longitude was a novelty, but paled into insignificance alongside those of the Polynesian canoe pilots who had traversed the Pacific before him. Many centuries before the Spanish explorer Vasco Nuñez de Balboa gazed out over the Pacific from a “stout peak” on the Isthmus of Panama in 1513, island navigators had made their way across thousands of miles of open ocean to settle the remotest landmasses on earth.
Ever the seaman, Cook admired Tahitian canoes, used for trading, warfare, and long-distance voyaging. He wrote: “In these Proes or Pahee’s as they call them from all accounts we can learn, these people sail in these seas from island to island for several hundred leagues, the Sun serving them for a compass by day and the Moon and Stars by night. When this comes to be prov’d we Shall no longer be at a loss to know how the Islands lying in these Seas came to be people’d.”8 He befriended Tupaia, an expert navigator, who carried with him a mental map of an area of the Pacific as large as Australia or the United States.
Indigenous Polynesian Navigation
When the British explorer Captain James Cook visited Tahiti in 1769, he puzzled over a question that still fascinates scholars: How had the Tahitians colonized their remote homeland? How had humans, with only simple canoes and none of the navigational tools used by Europeans, settled on the remotest islands of the Pacific? Cook met a Tahitian navigator, Tupaia, and asked him how canoe skippers made their way from island to island out of sight of land. Tupaia explained how they used the sun as a compass by day and the moon and stars by night.
