The Monkey's Voyage Read online

  2.2 Evolutionary clarity: a cladogram of tetrapod vertebrates, showing the precise relationships among the different groups.

  Hennig’s insight was realizing that having precise knowledge about relationships—the cladogram with its exact representation of the evolutionary branching pattern—could profoundly affect how one explained geographic distributions. In retrospect, some of his methods were not entirely sound, and Brundin unfortunately followed closely in Hennig’s footsteps and made the same mistakes. But the general idea was right, and it remains at the core of understanding the geographic history of living things: to do historical biogeography, you need to know how different species are related to each other, and the more precisely you know it, the better.

  That link between cladistics and biogeography was a great insight that Nelson took from Brundin. In Brundin’s monograph, Nelson saw cladograms of midges, with lines connecting each taxon to the area (southern South America, New Zealand, Australia, and southern Africa) in which that group was found. To the uninitiated, these figures look like some kind of indecipherable electrical wiring diagram, but in them Brundin had found a clear pattern. That pattern had three parts to it. First, every group of midges found in New Zealand was most closely related to a group in South America or in South America plus Australia. Second, Australian groups were typically placed within some larger South American group. The third part is actually a corollary of the first and second, but is worth emphasizing on its own: Australian and New Zealand midges were never each other’s closest relatives.

  2.3 Two architects of the vicariance biogeography “revolution”: Lars Brundin (left) and Gary Nelson in Stockholm in 1988. Nelson was converted upon reading Brundin’s monograph on midges. Photo by Christopher J. Humphries.

  From a traditional “fixed-continent,” dispersalist point of view, this pattern didn’t make much sense. If chironomid midges had spread to the various areas by dispersing over oceans, one would expect the taxa on New Zealand and Australia—two landmasses quite close to each other—to be, more often than not, each other’s closest relatives. Instead, these groups were always connected to South American ones. In other words, the midges really did seem to show transantarctic relationships in a very specific sense. But none of the landmasses inhabited by the midges are connected to Antarctica, so how did this pattern come to be? Was it possible that those areas had all been linked to each other through Antarctica at some time in the past, allowing midges to move easily from, say, southern South America to New Zealand? According to Brundin, not only was that possible, it was exactly what the latest geological evidence showed.

  FROM THE BOTTOM OF THE OCEAN

  When we left our discussion of geology, Alfred Wegener lay buried in the Greenland ice (where the ground beneath his frozen remains now lies several feet farther from Europe than when he died) and continental drift was a theory in need of a mechanism. Wegener had suggested centrifugal and tidal forces that had the continents plowing through oceanic rock like giant ice-breaking ships, but almost nobody believed these notions, and with good reason. Even so, there was strong evidence that the continents had moved somehow, and, because of that evidence, Wegener’s theory continued to have advocates through the 1930s, 1940s, and 1950s.

  One of those advocates was Arthur Holmes, a British geologist best known for studies of radiometric dating that indicated the Earth was at least several billion years old, at a time when many geologists thought the correct age was only about 100 million years. Holmes’s interest in radioactivity led him to propose that radioactive decay in the Earth’s mantle would give rise to convection currents involving mass movements of the liquid magma. If a liquid is heated gently from below, a temperature gradient is set up, but the liquid itself doesn’t move en masse. However, if more and more heat is applied, at some point a threshold is passed, and the liquid begins to circulate, hotter liquid rising from the bottom and then falling as it cools. Holmes realized that such convection currents in the mantle, driven by the heat from radioactive decay, could provide a mechanism for continental drift. He envisioned the hottest magma rising underneath a continent and, upon reaching the crust, sliding laterally in opposite directions, pulling the continental crust apart as it did so. The split halves of the continent would then move away from each other, riding on the mantle currents like boxes on a conveyor belt. The problem of continents having to push through the ocean floor was circumvented by descending magma pulling oceanic crust down with it, making room for a continent to move over the top.

  If “convection in the mantle” sounds familiar, that’s because most geologists today believe Holmes was right about the driving force for continental movement. Other aspects of Holmes’s theory are now known to be wrong; for example, he didn’t envision magma actually creating new seafloor, as we know it does, and he thought that convection currents would generally rise at the Equator and sink at the poles, which isn’t the case. However, given the state of knowledge at the time, his proposed mechanism for continental drift made sense; unlike Wegener’s, the mechanism he proposed actually seemed strong enough for the task, and it didn’t require having the lighter material of continents plowing through the dense basalt of the ocean floor.

  Holmes first published his theory in 1928 and later included it as a small part of his textbook Principles of Physical Geology, published in 1944. From his radiometric work, he had become a well-known geologist, and his book became something of a classic. But the theory of mantle convection didn’t have much impact. It was plausible, but there was little evidence to suggest that it was right, and why would other geologists believe something new and radical without much to go on? Maybe for the same reason, Holmes had a hard time even convincing himself. In 1953, twenty-­five years after his first paper on the convection theory appeared, he admitted, “I have never succeeded in freeing myself from a nagging prejudice against continental drift.”

  In 1944, when Holmes’s textbook was published, Harry Hess was somewhere in the Pacific. Hess, a Princeton geologist, had become a naval reserve officer several years before the war to help obtain gravity recordings in the Lesser Antilles that required the use of a navy submarine. As a result, he had to report for duty the day after the attack on Pearl Harbor, and he eventually found himself on a transport ship, the Cape Johnson, making a circuit that would have appealed to Darwin and Wallace in more peaceful times—the Mariana Islands, the Philippines, the Admiralty Islands, Guam, the Carolines.

  Even at war, Hess didn’t forget about geology. When he had free time ashore, he collected rocks. On board, he was in charge of a fathometer, an echo-sounding device used to help make beach landings, and he kept the thing switched on almost all the time, even when the ship was far out at sea. In this way he was able to gather quite a bit of data on the topography of the ocean floor. From Iwo Jima, where the Cape Johnson survived attacks from Japanese planes as well as its own side’s “friendly fire,” he wrote to one of his former students about the soundings he had collected: “Have been able to get about a dozen or more traverses across deeps and can outline their course pretty well from Iwo to Palau. Have four across the Mindanao deep, too. We filled in a lot of blank spots on the charts.”

  One thing that stood out in the soundings were many isolated seamounts—extinct, submerged volcanoes that Hess termed guyots after a nineteenth-century Princeton geographer, Arnold Guyot. These would turn out to be important, but Hess didn’t see why just yet; like Darwin, he came back from his long ocean journey with some interesting observations but no grand theory. But Hess’s mind was turned toward the depths now, toward the floor of the ocean.

  That was the right place to look and soon it would be the right time to be looking. In the 1950s, there was an explosion of new information about the nature of the ocean floor, not only its topography, but also its magnetic, seismic, and heat flow characteristics. Reams of sounding records eventually showed a giant, more or less continuous network of submerged moun
tain ranges, nearly 50,000 miles long, running through all the world’s oceans. The range that traced a sinuous path down the middle of the Atlantic, from the Arctic Ocean to the latitude of the southern tip of South America, was especially well-studied, and turned out to have some unexpected properties. Running down the center of this Mid-­Atlantic Ridge was a valley, suggesting an ocean-long rift in the Earth’s surface. That impression was reinforced, to some at least, by seismic recordings indicating that the crust in the area of the ridge was especially thin, along with temperature measurements showing it had an especially high rate of heat flow. It looked like magma was rising to the surface at the ridge. Another odd observation was that nothing in the oceans seemed to be very old; when geologists sampled rocks and associated fossils from the deep sea, from oceanic islands, or from Hess’s guyots, they were always of Cretaceous age or younger.

  By 1960, Hess had put all the facts together and had written a paper outlining a new theory. That paper, “History of Ocean Basins,” was published, after some delays, in 1962.12 It built on Holmes’s idea of mantle convection and incorporated the new information about the ocean floor to develop a theory of seafloor spreading that students today would easily recognize as part of the modern geological worldview. Hess followed Holmes in suggesting that convection was driving magma on great cyclic paths from deep in the mantle to the crust and back again, but the new observations allowed Hess to flesh out the theory. The mid-oceanic ridges were where “cells” of magma rose up from the mantle. From there, magma would spread out in both directions from the ridge, cooling as it went, creating new ocean floor, and pushing continents apart, as with South America and Africa on opposite sides of the Mid-Atlantic Ridge. Where the new ocean floor met a continent, it would be impelled (by the force of the magma coming up behind it) downward, back into the mantle (in other words, it would be subducted, although Hess didn’t use that term).

  Hess wasn’t a bold, outspoken scientist in the manner of Wegener, and, in “History of Ocean Basins,” he was sometimes circumspect in his language. “I shall consider this paper an essay in geopoetry,” he wrote, almost disparagingly, in the first paragraph. Still, he knew he was onto something big. In a later section, he starts with a conjecture and turns it, between sentences, into an assertion: “If it [mantle convection] were accepted, a rather reasonable story could be constructed to describe the evolution of ocean basins and the waters within them. Whole realms of previously unrelated facts fall into a regular pattern, which suggests that close approach to a satisfactory theory is being attained.” He had constructed that “reasonable story” and it did indeed explain many “previously unrelated facts.” Why are there no ancient rocks in the deep ocean? Because the ocean floor is part of a giant cycle of creation and destruction, and all of the really old ocean crust has long since been returned on this “conveyor belt” back to the Earth’s mantle. Why are sediments on the ocean floor relatively thin, not at all what one would expect from several billion years of erosion and deposition, and why are they thinnest of all at the mid-oceanic ridges? Because, again, the ocean floor is nowhere very old, and is, geologically speaking, brand new at the spreading centers. Why is there a ring of volcanoes encircling the Pacific? Because ocean crust thrust down into the mantle is melted into magma, which sometimes makes its way back to the surface. Why do ocean ridges often run right down the middle of an ocean? Because the new crust forming at the ridges flows out equally in both directions.

  The theory of seafloor spreading also explained Hess’s guyots, although one has to read between the lines of his 1962 paper to see this fully. A guyot begins as a volcano formed at a mid-oceanic ridge, with its top above the water. At some point the volcano is no longer active, no longer building itself up, and its emergent peak gets flattened by erosion. From its place of origin, the guyot rides the “conveyor belt” out from the ridge, sinking along the way as the crust cools and subsides. So, when a lot of these volcanoes have formed, the result is a series of flat-topped seamounts, with those farthest from the spreading center also the most deeply submerged. Without the theory of seafloor spreading, guyots are unexplained curiosities. With the theory, they suddenly make sense, another piece of the great puzzle falling into place.

  For biogeographic purposes, the critical point is that Hess, building on the work of Holmes and others, had basically solved the problem of how landmasses can drift apart. Although Hess never mentioned Wegener in his paper, it is clear that Wegener’s problematic mechanisms for drift—or, perhaps, people’s reactions to those mechanisms—were on his mind. From that perspective, the key was what happens at the leading edge of a moving continent: when continental crust meets oceanic crust, the latter, being more dense, is thrust down toward the mantle. When continent meets continent, the crust gets folded up, like two rugs pushed against each other, as when India collided with Asia to form the Himalayas. Like Holmes, Hess didn’t have continents plowing implausibly through the dense volcanic rock of the ocean floor. And Hess made sure to mention that fact more than once, distancing himself from Wegener.

  Hess’s theory of seafloor spreading is the basis of what we now call plate tectonics, each plate being a region of the crust and upper mantle that moves as a unit. For instance, there is a South American Plate bounded in part by the Mid-Atlantic Ridge to the east and the Pacific edge of the continent to the west, where the Nazca Plate is being subducted. The theory of plate tectonics certainly wasn’t finished by Hess (nor is it anything like complete today). Still, if Hess didn’t erect the whole building, he at least laid the foundation and put up a couple of walls.

  One might think that geologists would have rushed to embrace this new theory that finally provided a reasonable mechanism for continental drift and explained many of the odd new facts about the ocean floor. But that isn’t what happened. Hess wasn’t considered an auto-intoxicated dreamer like Wegener; instead, his theory was mostly just ignored, at least for a time. Still, “History of Ocean Basins” was an indication that the old view of fixed continents and oceans was getting creaky, even in the United States, where that idea was most deeply entrenched. The information that was piling up was about to make the old edifice collapse.

  Here is one version, the textbook version, about what happened after Hess’s paper was published: In 1962, Fred Vine, a beginning graduate student at Cambridge, took an interest in seafloor spreading—partly because he saw Hess give a talk about it—and began looking for a connection between Hess’s theory and some new data on the magnetic properties of the ocean floor. In particular, what Vine and his adviser, Drummond Matthews, were investigating was a peculiar magnetic pattern on opposite sides of oceanic ridges in the Atlantic and Indian Oceans.

  As magma cools, iron-containing particles within it align themselves with the Earth’s magnetic field. Oddly, in some volcanic rocks, the particles are aligned in the opposite direction of the current magnetic field, an indication that the direction of the field has flipped back and forth through Earth’s history. In the recordings that Vine and Matthews were examining, the rocks on the mid-oceanic ridge had the orientation of the current magnetic field, but next to the ridge on either side were areas of rock with the opposite orientation, and then, outside those areas, more rock with the current orientation. Vine and Matthews realized that this pattern was exactly what one would expect if new ocean floor was being formed at the ridge and then flowing outward on either side: the rock at the ridgetop, having just recently solidified, showed the current magnetic field orientation, while the alternating magnetic orientations going outward recorded the formation of older rocks and the flip-flopping of the direction of the field through time. Vine and Matthews published their findings in Nature in 1963, their beautiful insight providing a striking demonstration of the reality of seafloor spreading.13 Now, finally, geologists were converted en masse to a belief in plate tectonics.

  As far as I know, everything about that story is true, except for one thing: the mass conversion
didn’t occur just yet. Interviewed years later, Vine said of the paper’s reception, “It was the classic lead balloon, . . . one can’t overemphasize that.” He chalked up the negative reaction in part to the fact that the symmetric pattern around the ridges wasn’t as clear as it might have been. Vine also realized that he and Matthews were relying on two assertions that at the time were not widely accepted: the reversal of the Earth’s magnetic field through time, and the notion that the field’s orientation could be read in the rocks of the ocean floor. If people didn’t believe those two assertions, they surely weren’t going to believe the radical conclusions Vine and Matthews had drawn from them.