The Monkey's Voyage Read online

  At about this time, Nelson took advantage of an odd opportunity to push Croizat’s work. In his role as editor of the journal Systematic Zoology, Nelson had received a manuscript from Croizat criticizing dispersalism and defending panbiogeography. Like almost everything written by Croizat, the manuscript was extreme by academic standards; it was personal, caustic, and blatantly one-sided. Most manuscripts submitted to journals get sent out to a few experts for peer review, but Nelson ended up sending Croizat’s paper to nineteen reviewers, an almost unheard of number. Not surprisingly, most of the reviewers thought the manuscript wasn’t fit for publication and, even after Croizat revised it, the reviews (now reduced to “only” nine) were split between acceptance and rejection. At this point, Nelson made another move that, according to the usual standards for journal editors, was both unusual and bold: he suggested to Croizat that the two of them, along with Nelson’s American Museum buddy Donn Rosen, should collaborate on a paper, using Croizat’s manuscript as a starting point. Nelson likely saw this as a chance to further Croizat’s cause in a way that would end up being more palatable to academic biologists than if Croizat had simply been left to his own unfiltered devices.

  Croizat agreed to the collaboration. The resulting paper, published in 1974 and innocuously entitled “Centers of Origin and Related Concepts,” is something of a mishmash of ideas, but is significant for several reasons. First, it helped introduce Croizat’s use of individual and fundamental tracks to a wider audience. Second, it set forth, in terms no one could miss, the idea that vicariance is the dominant process in generating biogeographic patterns. For instance, the authors concluded that historical biogeography “is to be understood first in terms of the general patterns of vicariance displayed by the world biota.” Finally, the paper brought Croizat’s over-the-top criticism of Darwin and dispersalism into the mainstream of biogeography. “Having failed to dissect these concepts (center of origin, vicariance) to their core,” Croizat et al. wrote, “contemporary zoogeographers founder in a self-created morass of chance hops; great capacities for, or mysterious means of, dispersal; rare accidents of over-sea transportation; small probabilities that with time become certainties; and other pseudo-explanations.” Like Brundin’s work, it was a call to overturn the status quo, but with Croizat’s characteristic lack of restraint. The paper described Darwinian biogeography as “a world of make-believe and pretense” and ended with a final jab at the New York School: “No one well informed of the zoogeography of our times can have an illusion about its manifest disreputability.” Croizat’s vision of a befuddled Darwin and his equally benighted minions was now out there, in a well-known journal, with the names of two established scientists, Nelson and Rosen, to validate it.

  DISCONNECTION

  In his midge monograph and other papers, Lars Brundin had presented vicariance as the most significant reason for disjunct distributions, and he had bashed the dispersalists for their uncritical reliance on rare, chance events. Nonetheless, Brundin still believed that long-distance dispersal was significant. However, the Croizat, Nelson, and Rosen paper embodied a point of view altogether more extreme, in which vicariance was thoroughly dominant and long-distance dispersal was barely worth thinking about. To borrow a term used by Stephen Jay Gould in describing the spread of adaptationism,18 the Croizat et al. paper reflected a “hardening” of vicariance biogeography into a purer form, a form disinclined to tolerate any remnant of Darwinian dispersalism. A few years later, in that same spirit, Nelson came up with a notorious phrase for dispersalism, deriding it as “a science of the improbable, the rare, the mysterious, and the miraculous.”

  This pro-vicariance, antidispersal view was quickly propped up by the spread of two ideas that, depending on one’s point of view, have served ever since as either important assumptions or unfortunate stumbling blocks in biogeography. One was the notion that long-distance dispersal hypotheses cannot be refuted and are therefore unscientific. Consider, for instance, a group (say lungfish, ratite birds, or southern beech trees) found on several Southern Hemisphere continents. The hypothesis that such a distribution was the result of Gondwanan fragmentation would have to be rejected if the evolutionary relationships within the group failed to match the sequence of continental breakup. In contrast, there is no pattern of evolutionary relationships that can definitively refute long-distance dispersal for a particular group. If relationships for a group match Gondwanan breakup, this could just be a coincidence; the distribution might still be explained by a dispersal event that happened to occur in a way that mimics vicariance. More generally, there is no distribution pattern, no matter how scattered, that cannot, in theory, be explained by a set of long-distance jumps. And finally, failure to ever observe chance, long-­distance dispersal by a particular group of organisms does not disprove dispersal explanations for that group; indeed, the very nature of such events makes it highly unlikely that they will be observed. In short, it seemed there was no type of evidence that could really rule out dispersal.

  3.3 Rare and mysterious but true: White-Faced Herons (Egretta novaehollandiae) dispersed naturally from Australia and established a population in New Zealand in the 1940s. Photo by Glen Fergus.

  Right from the start, some people recognized that there was something odd about this notion of dispersal hypotheses as unscientific and therefore not worth considering. For instance, in the mid-1970s, a young ichthyologist named Robert McDowall (who will enter our story again as an older man) pointed out that we know long-distance dispersal happens, so it’s very strange to construct a whole approach to biogeography that ignores dispersal on the grounds that it’s untestable. As an obvious example, McDowall pointed out that White-Faced Herons had colonized his native New Zealand from Australia by long-distance, overwater dispersal in the twentieth century (see Figure 3.3). People knew this because they had recorded the first instances of these birds breeding in New Zealand; the herons had not been established, and then they were—there was nothing hypothetical about it. What good is a “rigorous” science of biogeography, McDowall asked, if it forces us to ignore such events that obviously affect distributions? Nonetheless, the dispersal-is-unscientific argument took root, no doubt in part because Popperianism, which identifies falsification as the hallmark of a scientific hypothesis, was (and remains) the philosophy of choice for many biologists. In fact, in their 1981 book Systematics and Biogeography, Gary Nelson and Norm Platnick claimed that their approach was the union not just of Hennigian cladistics and Croizatian panbiogeography, but also of Karl Popper’s philosophy of science. The implication was that, to be a true scientist, you had to be a falsificationist, and therefore, you couldn’t be a dispersalist.

  The second prop for the pro-vicariance, antidispersal view was the idea that almost all evidence about the ages of taxa was so unreliable as to be essentially worthless. According to the dispersalists of the New York School, many groups were simply too young to have been influenced by continental drift. If these groups showed disjunct distributions that seemed to reflect ancient continental breakup, that was simply an illusion. For instance, Ernst Mayr claimed that most living groups of mammals, birds, and flowering plants had originated long after the separation of South America and Africa, and therefore could not have been affected by that event. But if vicariance was as dominant as Croizat, Nelson, and their followers believed, this argument couldn’t possibly be right. These groups had to be older than they appeared to be.

  The dispersalists based their age estimates in part on the direct evidence of the fossil record, with the earliest fossil specimen of a group taken as a rough indication of the group’s time of origin. Of course, everyone knew that the fossil record was incomplete—Darwin himself had made that point forcefully in The Origin of Species to explain why we don’t have fossils of all the intermediate forms that his theory implies must have existed—but many biologists nonetheless believed that the record was good enough to give ballpark estimates of the ages of many taxa. F
or instance, the fact that the oldest fossils of most orders of mammals date to within the past 50 million years was used to claim that most of these groups originated after the Mesozoic, which ended 66 million years ago. Vicariance scientists argued that these claims were completely unjustified. The oldest fossils of a taxon certainly set a minimum age for the group, but the actual age of origin will almost always be older, and one never knows just how much older. A group with a fossil record going back 49 million years might be 50 million years old or it might be 150 million years old.

  By the 1970s, a new method of estimating ages had emerged, and it too became a target for the vicariance scientists. Specifically, biologists were using genetic differences between species as a way of estimating the time since the lineages in question had split from each other. These early studies assumed a strict molecular clock, that is, they assumed that changes in genes occur at a constant rate. Under this assumption, once the clock’s rate of ticking was calibrated using fossil evidence for any one evolutionary split—for instance, between house mouse and Norway rat—genetic differences between other groups could be easily translated into the time of separation of lineages. Look at the differences in genes (or, back in the 1970s, the proteins encoded by genes) between a human, a chimp, and a gibbon, and you can see that human and chimp separated about 7 million years ago, while the human plus chimp lineage separated from the gibbon about 20 million years ago.

  There were early indications, which have been abundantly verified, that the molecular-clock assumption was wrong, that the clock of genetic changes doesn’t actually tick at a constant rate. Vicariance scientists used this evidence of a faulty timepiece to completely discount any estimation of the age of a group using genetic differences. In fact, if anything, ages based on the molecular clock were considered even more suspect than those based directly on the fossil record, because the clock had to be calibrated using fossils! You had to set the rate of ticking based on a fossil age estimate that could be way off, and then assume that the rate never changed, an assumption that was demonstrably false. Two wrongs couldn’t make a right.

  The skepticism about estimating the ages of groups based on the fossil record or molecular dating analyses is a fundamental part of the extreme vicariance worldview. To a degree, this skepticism is healthy; the fossil record really is incomplete, often incredibly so, and molecular dating, even in its later, more sophisticated incarnations, is based on assumptions that are hard to validate and are often simply wrong. (Whether the ages estimated from such analyses could be as consistently wrong as some vicariance scientists believe is something I will take up in Chapter Six.)

  Nonetheless, there were signs that vicariance biogeography had developed an acute disconnection from reality related to this skepticism about the estimated ages of groups. Often, those signs can be traced back to Croizat and his extreme worldview. I started this chapter with one such disconnect, Michael Heads’s claim that not one of the thousands of native lineages of the world’s oceanic islands are derived from long-distance colonization. Heads is what one might call a panbiogeographic fundamentalist, and his claim about Hawaii and other oceanic islands echoes Croizat’s views, although Heads, in this case, might be even more extreme than Croizat himself. As mentioned earlier, if most Hawaiian lineages are as young as current evidence indicates, then they reached the islands when the configuration of land and ocean was more or less as it is today, which means that they must have traveled vast distances over water to get there. Thus, Heads’s claim against long-distance dispersal can only be true if island lineages are much older than the evidence suggests.

  Another sign of this separation from reality comes at the end of Nelson and Platnick’s book Systematics and Biogeography, and, if anything, is even more bizarre than Heads’s notion about oceanic islands. Nelson and Platnick, in keeping with their Croizatian search for general patterns, perform a thought experiment about the evolutionary history of Homo sapiens: What if the branching order in the phylogenetic tree of humankind matched the history of continental breakup? If that were the case, wouldn’t we conclude that the geographic history of humans was part of the same general process of tectonically driven vicariance that we see in so many other lineages, from chironomid midges to ratite birds to southern beech trees? Wouldn’t it, therefore, also mean that humans are far more ancient than anyone imagines?

  It’s a grand vision for humanity, placing us within the great story of the fragmentation of the world’s biotas through continental drift. It’s an epitome of the Croizatian vision that “Earth and life evolve together.” It’s also completely looney. It reminds me of old science-fiction movies that had cavemen fighting dinosaurs, the movies that well-informed preteens would point to and say, “That’s dumb, people never lived at the same time as dinosaurs.” What Nelson and Platnick were saying was that, if the evolutionary and tectonic patterns matched, one could infer that the human lineage—meaning people as people, not as proto-humans or tree-living apes or any more distant ancestor—extended back 66 million years and more to the Mesozoic, which was . . . the Age of Dinosaurs. Never mind that the intensively studied human fossil record indicates that the genus Homo is only a few million years old. Never mind that the genes of humans and chimps are so similar that they suggest that the common ancestor of these two lineages (which was certainly not a person in the usual sense) existed only 7 million years or so ago. Never mind a whole host of fossils showing that a succession of progressively deeper human ancestors—the first hominids, the first apes, the first monkeys—were not around in the Mesozoic. All that evidence is worthless. Molecular clocks don’t work. The fossil record is hopelessly incomplete. People might have lived with dinosaurs.

  As with Heads’s view of oceanic islands, I find it incredible that Nelson and Platnick actually believed what they said about humans in the Mesozoic. I’m wondering if, in the final stages of writing their book, they were on some kind of “vicariance high” and had a temporary loss of all perspective. To reassure myself that this wasn’t the case, I emailed Nelson and asked him about that section of the book. Would a match between the evolutionary tree of humans and the sequence of Mesozoic seafloor spreading events still make him think that continental breakup might explain human geography? He answered with one word: “Yep.”

  Black and white answers attract us. We want our politicians to tell us they’re 100 percent behind the middle class and small business and our troops overseas, 100 percent against giant multinational corporations and lobbyists and the enemies of democracy. We want to think that we’ll be healthy if we just take that cholesterol drug, happy if we just listen to each other. We experience a record heat wave and want to put it all down to human-caused global warming, or, if we are of another ilk, to blame it on sunspots. Sometimes, things actually are simple and straightforward, but, more often, we’re just filtering out the real complexity. I can’t say I have any real handle on why we do this. Maybe sorting through the complexities makes a person too indecisive and being indecisive gets you selected out of the gene pool. Maybe seeing things in black and white helps us make valuable bonds with other people who share our unequivocal beliefs, often at the expense of some other group: “We are Christians and they are infidels,” or “We are capitalists and they are communists,” or, most generally, “We are good and they are evil.”

  In any case, I have a strong impression that what led vicariance scientists to their extreme views was, at least in part, this attraction to the unambiguous. Many of the tenets of vicariance biogeography can be seen in that light. For instance, we can believe that distributions broken up by oceans reflect a complicated mixture of vicariance and dispersal, or that vicariance is dominant, so let’s believe vicariance is dominant. We can believe that evolutionary relationships and molecular clock analyses and the fossil record should all be used to interpret biogeographic history, or that only evolutionary relationships are critical, so let’s believe that only evolutionary relationships are crit
ical. We can take an eclectic approach and follow the evidence in whatever form it takes, or be strict Popperians who think that only vicariance is falsifiable, so let’s be strict Popperians. Basically, vicariance scientists chose a simple, straightforward approach—using cladograms or, if they were panbiogeographers, tracks—to uncover the workings of a single, general process, namely, vicariance, rather than dealing with messy methods and messy interpretations. They chose black and white over countless shades of gray.

  In doing so, they seemed to think they were finally making a real science out of biogeography. There was Brundin (actually a moderate, as vicariance biogeographers go) saying that the dispersalist approach, which he was trying to replace, “offends a critical mind.” There was Croizat harping on how biogeography had followed the misguided teachings of Darwin for a hundred years, and how Croizat’s own panbiogeography, grounded in incontrovertible facts, “will have powerful influence upon the whole of biological thinking.” There were Croizat, Nelson, and Rosen proclaiming that the way out of the Darwinian morass was, among other things, “to formulate explicit methods of statistical analysis (based on the concept of generalized tracks) that yield unambiguous and repeatable results.” One gets the sense that some vicariance scientists thought they were transforming the mushy science of biogeography into something more like the successful hard sciences of physics and chemistry. It may not be a coincidence that two biologists of the cladistic vicariance school, Dan Brooks and Ed Wiley, actually came up with something called the “non-equilibrium thermodynamics theory” (sometimes called “evolution as entropy”), which placed all of evolution within a framework borrowed from Newtonian physics.