The Monkey's Voyage Page 6
I know it sounds suspicious. Where do all the numbers come from? Basically, out of William Diller Matthew’s ear. But the message is not the exact numbers—Matthew admitted that—it’s just Darwin’s old argument that given a very long time (60 million years in this case) a lot of things that seem very, very unlikely will happen. Rats can get to the Galápagos. Monkeys can reach Sulawesi. Pygmy hippos (now extinct) can make it to Madagascar. Actually, according to Matthew, the pygmy hippos didn’t even need a raft. They could have just swum the 300 miles across the Mozambique Strait.
Like many biogeographers, including Darwin and Wallace, Matthew also argued that the nature of an island’s fauna can tell you if the place was ever connected to a continent. He used the mammals of Madagascar as an example. If the island had been connected to Africa at any time, you’d expect to find a large complement of African mammals there. Instead what you see are just a few major groups—lemurs, carnivores, tenrecs, and rodents, plus a shrew that might be introduced, and bats, which can fly there. Also, those mammal groups do not seem to have gotten to Madagascar all at once, but rather, apparently colonized the island one by one at different times. (Matthew probably based this inference on the connections of these groups to African relatives and the fossil records of those relatives.) This pattern of a few colonizations spread out in time is exactly what you’d expect if each group reached the island by chance, overwater dispersal.
Writing before Wegener’s book had been translated into English, Matthew was arguing against the idea of a now-sunken land bridge to Madagascar, not against the island as part of the continental drift story. But the same argument could be made against Madagascar as part of Pangea or Gondwana, and, when Wegener’s theory became widely known, Matthew argued against it, and not just for Madagascar, but in general. Matthew, like Darwin and Wallace, did not completely discount former land connections, but he thought that most modern groups were too young to have been affected by ancient land bridges (to the extent that those even existed). His emphasis was clearly on long-distance dispersal.
To make a long story short, Matthew had a major impact on biogeography through “Climate and Evolution” and other works and through a wide network of students and colleagues. It would be hard to overestimate the networking aspect in Matthew’s case. He became something more than just a respected scientific mentor; one follower, for instance, referred to himself and others as Matthew’s “disciples” and called “Climate and Evolution” “a kind of Holy Writ.” Some of the scientists within his sphere of influence were George Gaylord Simpson, a paleontologist who replaced Matthew at the American Museum of Natural History when Matthew moved to Berkeley in 1927; Ernst Mayr, an ornithology curator at the museum from 1931 to 1953, and later a professor at Harvard; and Philip J. Darlington, also at Harvard. These three were all prominent scientists: Darlington was a well-known beetle expert and the author of widely read books on biogeography, and Simpson and Mayr were arguably the most significant paleontologist and evolutionary biologist, respectively, of the twentieth century. Influence begat influence, and by the 1940s, dispersalism, at least in the United States, was at its height. Hooker’s land bridges had taken precedence for a time, but now Darwin’s ocean crossings again held sway.
Because this dispersalist mode of thinking had originated at the American Museum of Natural History and also had a strong following at Columbia University, it was called by some the “New York School of Zoogeography.” That moniker was coined by none other than Léon Croizat, who seems to have used the phrase as his way of identifying the enemy. Croizat was the most headstrong and divisive character in a field that would soon have no shortage of such people, and what was about to happen in biogeography would not at all resemble the friendly jousting between Darwin and Hooker.
4For a time, Wallace believed in the horizontal movement of continents (that is, continental drift), but he changed his mind in the early 1860s to a belief in fixed continental positions (Parenti and Ebach 2009).
5This was the Austrian geologist Eduard Suess’s theory, which held that the Earth was continuously shrinking as it cooled, producing high and low regions of the crust, like wrinkles in the skin of a drying apple. Suess believed that South America, Africa, and India once had been connected by land, but subsidence had resulted in oceans flooding parts of the giant, conglomerated landmass, thus separating it into pieces. Suess’s grand theory has been thoroughly rejected, but his name for the southern supercontinent caught on (although its boundaries became modified). He had called it Gondwanaland after the region of India where sedimentary rocks from the ancient landmass were found (Oreskes 1988).
6Some have said that Wegener actually got the idea of continental movement from reading Taylor’s paper on the subject, but this claim is at best uncertain (McCoy 2006). Wegener reported that he thought of the idea independently. In any case, since the basic idea had been around for several hundred years, and Wegener’s legacy is based on the volume and quality of evidence he presented, it does not seem to matter how he first came to think of the phenomenon.
7There are some conspicuous gaffes in Wegener’s book as well. For instance, Wegener thought that the rates of continental movement were sometimes orders of magnitude greater than we now know them to be.
8Matthews, on a trip to the Falkland Islands, asked a colleague what he should read about the geology of the area. The reply, according to Matthews, was, “Oh well, you’ve got du Toit [Alexander du Toit’s book on the geology of South Africa], . . . if you don’t believe in continental drift just take out a tape measure and measure the Devonian sections in the Falkland Islands.” Matthews did this and found “they were very much impressively the same as the description [given by du Toit for South Africa], . . . inch for inch they measured up” (Oreskes 1988).
9Oreskes has also pointed out that the lack of an acceptable mechanism does not generally keep scientists from believing in the reality of a phenomenon. For instance, many people were convinced by Darwin that evolution was a reality without being convinced that natural selection was a plausible mechanism. Similarly, scientists accepted the reality of the ice ages without being convinced of a cause for them (Oreskes 1988).
10The title of the paper refers to Matthew’s notion that very long-term climate cycles were critical in shaping evolution and dispersal.
“On several occasions, when the vessel has been within the mouth of the Plata, the rigging has been coated with the web of the Gossamer Spider. One day (November 1st, 1832) I paid particular attention to the phenomenon. The weather had been fine and clear, and in the morning the air was full of patches of the flocculent web, as on an autumnal day in England. The ship was sixty miles distant from the land, in the direction of a steady though light breeze. Vast numbers of a small spider, about one-tenth of an inch in length, and of a dusky red colour were attached to the webs. There must have been, I should suppose, some thousands on the ship. . . . While watching some that were suspended by a single thread, I several times observed that the slightest breath of air bore them away out of sight, in a horizontal line. On another occasion (25th) under similar circumstances, I repeatedly observed the same kind of small spider, either when placed, or having crawled, on some little eminence, elevate its abdomen, send forth a thread, and then sail away in a lateral course, but with a rapidity which was quite unaccountable.”
—Charles Darwin, The Voyage of the Beagle
Chapter Two
THE FRAGMENTED WORLD
THE ICHTHYOLOGIST WHO PLAYED WITH FIRE
In his book The Tipping Point, Malcolm Gladwell likens the spread of products, behaviors, and ideas to epidemics, claiming that, in all these cases, small things, including single individuals, can have unexpectedly large effects. Gladwell describes three sorts of people that often have such disproportionate influence in pushing a trend past a tipping point. First, there are mavens, great collectors of inf
ormation. A maven might be, for instance, the acquaintance who keeps up with the latest news on which foods are good for your heart (this week, it’s coconut oil), or the friend who’s always telling you about a wonderful new hole-in-the-wall restaurant. Second, there are connectors, often extreme extroverts, who have an exceptionally large network of friends, acquaintances, and/or colleagues. Connectors are hubs from which information spreads especially quickly. Finally, there are salespeople, who excel at the art of persuasion. A salesperson might be an actual professional who talks you into buying an especially quiet dishwasher, but is just as likely to be a friend of a friend who convinces you of the evils of fracking.
For the spread of vicariance biogeography that took place from the 1960s into the 1980s, Gareth Nelson was all three of these things. In the mid-1960s, Nelson, an ichthyologist—“Gary” to his friends and colleagues—might not have seemed a likely candidate to become the center of a major scientific movement. The late philosopher and historian of science David Hull described Nelson, who had then just finished his PhD, as “a lanky youngster . . . anything but imposing in his white socks, shirt open at the collar, and brown-and-tan-checked sport coat.” But that “lanky youngster” had or would acquire some special qualities. He was exceptionally well-read within his field, so that he became privy early on to new ideas that few had heard of (a maven); he was a natural networker and became the editor of a key scientific journal (a connector); and, critically, he had a bold, persuasive personality (a salesperson).
Relevant to the art of persuasion, Nelson had no shortage of confidence, no hesitancy when it came to speaking his mind, and, above all, no reflexive deference to authority. In fact, it often seemed that he was fueled by defiance. There was the time, for example, when he and Donn Rosen, another ichthyologist, were discussing something about taxonomy, and Rosen suggested they should wait to see what Ernst Mayr, at the time probably the most famous and influential evolutionary biologist in the world, had to say about the topic in an upcoming book. Typically, Nelson was having none of that. “I’m not about to let Mayr do my thinking for me!” he barked. Defiance isn’t a universally attractive quality, but it seemed to suit the time (the rebellious 1960s) and place (principally New York City, never known as a paragon of politeness). Defiance was part of Nelson’s persuasive appeal.
In short, Gary Nelson—maven, connector, and salesperson all rolled into one—was just the person to become the chief architect for the spread of the ideas of vicariance biogeography. He was the straw that stirs the drink, or, maybe, the guy who fans the flames that end up burning down the building. One could make a case that, without Nelson, the vicariance movement would never have gotten off the ground. Certainly, it wouldn’t have unfolded anything like the way it did.
Nelson’s epiphany came in 1966, not in some splendid isolation in the desert or jungle, but in the library of the Department of Paleozoology at the Swedish Museum of Natural History in Stockholm. Nelson was fresh from completing his doctorate studying fish gill arches at the University of Hawaii and was on a year-long National Science Foundation research fellowship that had him examining fossil fishes in England and Sweden. Browsing the current literature section in the library, he picked up a hefty new publication, a nearly five-hundred-page monograph by a Swedish entomologist named Lars Brundin on the evolution and taxonomy of a group of tiny, swarming flies called chironomid midges.
Brundin had been studying chironomids since the 1940s and through them had become especially interested in taxa that occur on landmasses “separated” by Antarctica. Among other places, chironomids are found in Australia and New Zealand and on the other side of Antarctica, so to speak, in South America. Groups that show such transantarctic relationships are obviously on landmasses separated by oceans, so Brundin, like so many before him, was addressing the great question in biogeography, the question of disjunct distributions. At the time, Nelson had no particular interest in either midges or transantarctic relationships, so it’s an open question why he bothered to pick up the doorstop-like monograph in the first place. Perhaps that’s just what mavens do—they root around to discover things and, sometimes, that act is the spark that sets off the fire.
The bulk of Brundin’s monograph didn’t look like promising reading. It consisted of dry descriptions of the midges, not only the adults but also the freshwater larval and pupal stages. The pupae turned out to be especially important from a taxonomic point of view, and Brundin described them with what might have seemed like obsessive zeal; this was stuff that only a fly specialist—in fact, maybe only a midge specialist—could really get into. However, bookending those descriptions of the midges were sections in which Brundin set forth his approach to studying evolutionary relationships and biogeography and, from that approach, drew strong conclusions about how the tiny flies had ended up on widely separated landmasses. In effect, Brundin was using midges to illustrate some very general methods and concepts, things that could apply to Nelson’s fishes, or any other group, for that matter. The approach was what caught Nelson’s eye; in more ways than one, it was like nothing he had seen before.
First, Brundin used methods developed by the German entomologist Willi Hennig to construct evolutionary trees for the midges. Hennig’s “cladistic” methods involved explicitly identifying each grouping of species or higher taxa (genera, families, etc.) in an evolutionary tree by recognizing the traits that had apparently evolved on the branch leading to the group in question. These shared derived traits (or synapomorphies, in the elaborate jargon of cladistics) identified so-called clades (or monophyletic groups), lineages that traced back to an ancestor not shared by any organisms outside the group. For instance, members of the group Aves have feathers and toothless beaks, shared derived traits that evolved on the branch leading to the common ancestor of all birds. Those traits indicate that Aves is a clade—in this case, the clade that includes the common ancestor of modern birds and all of the descendants of that ancestor. Tangentially, under Hennig’s system, many traditional taxonomic groups were rejected because, although they included the common ancestor of all members of the group, they did not include all the descendants of that ancestor. For instance, the class Reptilia, as traditionally defined, is not an acceptable cladistic group because some descendants of the common ancestor of reptiles—namely, birds and mammals—have been artificially removed from the group. Fishes, amphibians (if extinct forms are included), apes (if we humans are not counted among them), monkeys, invertebrates—none of them are clades, either, and so all are invalid in a cladistic classification system. That fact is not really important to our story, but it was one of the main things about cladism that pissed off traditionalists like Ernst Mayr and George Gaylord Simpson. The cladists, by the way, won this battle, to the point that most taxonomists these days only give formal names to groups they believe are clades.
By identifying all the clades within some group, one is also specifying the pattern of evolutionary branching within that group. Thus, the outcome of applying Hennig’s methods to a set of traits—the pupal characteristics of midges, for example—is a branching diagram, a cladogram, that is a precise illustration of how one thinks taxa are genealogically related to each other. Each branching point in the diagram indicates a speciation event, the splitting of one evolutionary lineage into two. In essence, a cladogram is a representation of a piece of the tree of life.
It all sounds rather simple, like something that should have been figured out in the nineteenth century. (Some people have argued that it was figured out in the nineteenth century.) But the fact is that, in the 1960s, most evolutionary biologists were not operating in this way. They weren’t consistent about identifying shared derived traits (most of them weren’t even clear about the idea and had never heard of the term synapomorphy), and they often ended up drawing evolutionary trees that looked like a river delta with a thick “ancestral” branch (say, reptiles) out of which other branches (say, birds and mammals) arose in a muddy fas
hion. From such river delta diagrams you couldn’t tell exactly how things were related to each other (see Figure 2.1). In the usual reptile-bird-mammal tree, birds just arose out of the great Mississippi-like swath of reptiles in general (or, if the diagram was relatively precise, out of a somewhat smaller swath called diapsid reptiles); it wasn’t at all clear that the closest relatives of birds are actually specific kinds of theropod dinosaurs. The muddiness of these evolutionary diagrams didn’t just stem from a lack of information, but reflected a muddiness of thought. Reading the evolutionary papers and books of that time, one gets the impression that people weren’t often thinking about questions such as exactly which reptiles are the closest relatives of birds. Hennig’s cladograms, in contrast, were perfectly explicit (perfectly Teutonic, it is tempting to say), with a line representing each species lineage or higher taxon intersecting at a particular point with the rest of the lines that made up the tree (see Figure 2.2).11
2.1 Muddy diagrams reflecting muddy thinking: a typical depiction of a vertebrate evolutionary tree before cladistics. Redrawn and modified from Romer (1959).
The logic of cladistics, as expressed by Brundin (who tended to explain Hennig’s ideas more clearly than Hennig himself did), immediately appealed to Nelson, as it has to many others. (I remember my own “conversion” to cladistics in the early 1980s—under the tutelage of my brother Kevin, who is also an evolutionary biologist and was then a “raving cladist”—as one of the most eye-opening scientific experiences I’ve had.) More than just presenting the cladistic method of deciphering evolutionary relationships, though, Brundin was also following Hennig in using cladograms to make inferences about biogeographic history. Even before Darwin, people had intuited that “relationships” (whatever that meant in pre-evolutionary terms) were important in explaining geographic distributions. In a sense, one had to know something about relationships to even know that there was a problem to solve. For instance, the fact that species A lives in Hawaii and species B lives in California only implies an interesting question if one senses that A is fairly closely related to B. (The fact that there are bristletail insects of the genus Neomachilis in Hawaii and California might mean something. The fact that there are fruit flies in Hawaii and condors in California does not.) With the Darwinian revolution came a clearer recognition of the nature of biogeographic questions. If A is part of the same tight evolutionary group as B, then the question is, “Where did the ancestor of these two species live (California, Hawaii, both places, neither place?), and how, in the course of evolution, did one species end up in one area and the other end up in another?” In many cases, the relationship aspect didn’t go much further than that, and biogeographic interpretations were based on other kinds of evidence or assumptions, often debatable ones. In the hypothetical case just given, the conclusion might have been that the group originated in California and later dispersed to Hawaii, based on the not-altogether-airtight assumption that continental species can colonize oceanic islands but not the other way around.