The Monkey's Voyage

by Alan de Queiroz

  6.2 Molecular estimates compared to fossil estimates for the same evolutionary branching points. The fossil estimates are ones considered to be relatively reliable. Mya = millions of years ago. Redrawn and modified from Hedges and Kumar (2009b).

  For points older than 400 million years ago, a glance at the graph shows that something odd is happening. These points represent events such as the separation of animal phyla or groups of phyla, and here the molecular estimates are consistently older than the fossil ones, in some cases far older. Some scientists argue that this discrepancy reflects an extremely poor fossil record pertinent to very ancient branching points, whereas others claim that it is caused by a one-time, universal shift in the process of genetic change that happened early in the history of multicellular animals. However, for our purposes, that part of the graph does not nullify the importance of molecular dating. For one thing, all of the pertinent biogeographic studies in this book involve groups and events much younger than that 400-million-year threshold; at that time, Pangea hadn’t even formed, and the beginnings of the breakup of Gondwana were still far off in the future. Also, if the fossils are actually correct, the anomaly in the graph represents a tendency for molecules to produce age estimates that are too old; and, as indicated above, a bias in that direction would lead to rejection of recent dispersal explanations, not their widespread acceptance. There are a couple of other notable cases, involving early evolutionary radiations of birds and mammals, in which the fossil and molecular age estimates conflict, but the molecular estimates are, if anything, too old in these cases as well. The bottom line is that, if this bias toward overestimating ages really exists, yet it turns out there is still strong evidence for recent dispersal, we should be even more confident in that conclusion.

  MESSY, USEFUL SCIENCE (OR WHAT WOULD DARWIN DO?)

  Molecular clocks are easy to find fault with and hard to love. Hence all the pejoratives thrown at these studies—the comments that they’re “pretty suspect,” or a futile “molecular dating game,” or, simply, “bullshit.” In some cases, the criticisms seem justified. Even now, one sees cringe-­inducing studies based on very small amounts of DNA sequence, or using only one or a few untrustworthy calibration points. However, since the early days of the molecular clock, when Zuckerkandl and Pauling were thinking that the clock ticked at a constant rate, there have also been vast improvements in methods, not to mention an explosion of DNA sequence data to which the new methods can be applied. Moreover, the overall agreement between molecular age estimates and good fossil-based ones indicates that the approach is reasonable in general, even if some particular studies are not. And, finally, if there is a bias in molecular dating, it seems to be toward giving spuriously old ages, which will, if anything, lead to underestimating the frequency of oceanic dispersal. It is all too easy to find bad molecular dating studies—the disease-ridden trees—and to extrapolate that the entire enterprise is rotten. But that is a mistake: for biogeography, we need to pay attention to the message from the whole forest.

  The idea that we should believe the forest relates to the point I made in Chapter Three about vicariance biogeographers wanting a black-and-white science. This business of molecular dating is messy; it’s thoroughly statistical, meaning it gives one probabilities, not absolutes, and it relies on assumptions about calibration points and the way that DNA sequences evolve that can be hard to justify for any particular case. There is a certain kind of mind that revolts against that sort of messiness and finds no solution other than to dismiss the whole practice. If some molecular dating studies are obviously botched and all are based on unverified assumptions, why should we believe anything that comes from this approach? These “anti-dating” scientists never reach the point where they can see that the picture as a whole is highly informative. They never see the forest.

  More generally, I would also suggest that it is the business of evolutionary biologists and other historical scientists to draw conclusions from obviously flawed and potentially misleading information. That is simply the nature of historical evidence, especially the kind that bears on ancient events. Evidence degrades through time, and over millions of years, most of it disappears completely. Thus, we are forced to try to unravel deep history by examining incomplete relics, like a fragment of a fossil jawbone, or indirect evidence, as when we use the DNA sequences of living species to build evolutionary trees. Quite often, it is only through the convergence of inferences from different sources that we are persuaded that some evolutionary scenario is correct. For instance, we might become convinced of a vicariance explanation through the dovetailing of support from fossils, DNA, and geology.

  Given the nature of historical evidence, it is not surprising that good evolutionary biologists often are notable sponges, sifters, and synthesizers of information. As the archetype, we have Darwin, who absorbed encyclopedic knowledge of taxonomy, animal breeding, anatomy, embryology, geology, the fossil record, and (of course) biogeography to construct his argument for evolution. No one piece of evidence was entirely convincing, and some of it, such as the sudden appearance of many animal phyla in the fossil record, seemed to contradict his views, but he didn’t let that blind him to the bigger picture. More recently, evolutionists like Ernst Mayr, George Gaylord Simpson, and Stephen Jay Gould, among others, have followed in that tradition. Those three were all great sponges, sifters, and synthesizers.

  These thoughts make one wonder what Darwin would have made of the ages of lineages estimated from DNA sequences. I’m sure he would have recognized that molecular dating evidence—like all evidence that bears on history—is not entirely to be trusted. However, I also believe that, as he looked into the subject further, he would have seen that this dating evidence has the potential to be illuminating, and that to believe it is grossly misleading is unjustified. And then he would have done what he did with so much other information, from the variation in pigeon breeds to the anatomy of eyes: he would have used it.

  26Dimetrodon is the most famous of these older sailfin reptiles, well known to children who play with plastic prehistoric animals.

  27There are other fossils almost as old from both the crocodile and bird sides, so the argument does not depend entirely on Arizonasaurus (Benton et al. 2009).

  28The widely used program BEAST (Drummond and Rambaut 2007) can incorporate age uncertainty by allowing the user to enter the calibration point as a curve that, across a range of ages (e.g., 240 to 250 million years ago, in the above example), specifies the probability that a given age is the actual age of the branching point.

  29The clock also runs at vastly different rates in different genes or parts of genes, but those problems are easier to deal with than the variability among lineages.

  30A caveat is that relaxed clock methods may produce less accurate ages than constant clock methods in cases where the constant clock assumption actually holds.

  Researchers tracked three estuarine crocodiles (Crocodylus porosus) by satellite as the animals moved at least 34, 255, and 366 miles in the coastal waters of northeastern Australia, the fastest of the three covering more than 14 miles per day. Crocodiles have very limited endurance, but the movements of these three were consistently aligned with the direction of ocean currents, suggesting that they were simply going with the flow, like giant, scaly message bottles. In one case, when the current slowed sharply, the crocodile in question moved ashore for two days, and only went back into the sea when the current picked up again.

  The researchers noted that estuarine crocodiles might travel vast distances in this way, as they are enormous animals, have low metabolic rates, and can excrete excess salt through glands in their nasal passages. A 22-pound C. porosus can last up to four months in seawater without eating, which implies that an adult weighing 200 to 400 pounds could survive an ocean voyage of many months. Fossil and molecular genetic studies indicate that, in fact, crocodiles have dispersed across wi
de ocean barriers several times in their history. Within the past several million years, for example, one species apparently crossed the Atlantic from Africa, giving rise to the four species of New World crocodiles.

  Section Three

  The IMPROBABLE, the RARE, the MYSTERIOUS, and the MIRACULOUS

  Chapter Seven

  THE GREEN WEB

  THE LOST WORLD, FOUND

  If you get stranded up there, whatever you do, don’t try to climb down. That’s basic advice for biologists working on the summits of tepuis, the isolated, table-top mountains that dot the savanna of southern Venezuela and nearby parts of Brazil, Colombia, Guyana, and Suriname. Tepuis—“houses of the gods” in the language of the local Pemón Indians—are the monolithic remnants of a sandstone plateau that formed well over a billion years ago, long before the breakup of Gondwana or the formation of Pangea. Much more recently, over the past 70 million years, wind and water have been grinding down the plateau, with only those parts that were capped with more resistant rock remaining as these towering flat-topped mountains, their sandstone cliffs now rising thousands of feet above the surrounding country. On the tops of the tepuis are misty, dripping landscapes of dwarf forests; streams running through pink sand; rock formations eroded into weird convolutions; and fields of plants with naked stalks and tufted heads, like something out of Dr. Seuss. “Otherworldly” is a word that crops up often in descriptions of these places. Adventurous tourists and scientists can get dropped off by helicopter onto tepui summits, but, not uncommonly, clouds or fog will prevent the scheduled pickup. At that point, the thing to do is wait, even if the delay drags on for days, because trying to find a way down the cliffs is likely to get you lost, or stuck halfway down, or inadvertently airborne.

  The otherworldliness and inaccessibility of tepuis are exactly what draws biologists to them. Those cliffs that keep people from climbing up or down are also a barrier for other living things and, to many biologists, this has implied that the organisms living on the tops of the tepuis have been evolving in prolonged isolation, perhaps for many millions of years. This idea of tepuis as areas populated by relicts took root in the Western world in the 1860s, before Europeans had even climbed them. In the early 1900s, when Arthur Conan Doyle wrote The Lost World, famously imagining a tepui filled with dinosaurs, pterodactyls, ape-men, and other prehistoric creatures, he wasn’t inventing a view of tepui biotas so much as taking the accepted notion to a romantic extreme. And it’s a notion that persists, albeit without the megafauna. For instance, in the late 1980s, Vicki Funk and Roy McDiarmid, Smithsonian biologists who had explored tepui summits (and gotten stranded on one for ten days), speculated that tepui organisms might be “remnants of an ancient biota that dates back millions of years to a time when South America and Africa were a single land mass.” In other words, tepui biotas are like isolated pieces of old Gondwana, reminiscent of what people have imagined for islands such as New Zealand and New Caledonia. Along the same lines, Jesús Rivas, a Venezuelan herpetologist, said of the tepuis, “They are like a place where time stopped.”

  But, then again, maybe not.

  On the boggy plateau of the tepui called Cerro de la Neblina—“Mountain of the Mists,” a name that would fit almost any tepui—lives a miniaturized plant that eats animals. The species is Drosera meristocaulis, and it’s a pygmy sundew, a plant that catches insects in sticky drops exuded from tiny stalks that cover its red, spoon-shaped leaves. On the tops of tepuis, it turns out, carnivorous plants like D. meristocaulis are a dime a dozen; tepui soils are infertile, which makes them good places for plant species that can supplement nutrients from the soil by trapping and digesting insects and other small creatures. Some tepui plants may even specialize in eating protozoans, which they trap in corkscrew-shaped structures on root-like underground leaves. For our purposes, however, what is ­really significant about D. meristocaulis is not what it eats, or anything else it does, but where it came from.

  Drosera is a big genus, with some two hundred species, concentrated in the Southern Hemisphere, but with at least some species on every continent except Antarctica. Find a nice bog, and there’s a good chance some species of sundew will be living there. Despite their wide distribution, though, sundews, because of the properties of their seeds, are not considered great long-distance dispersers. Their seeds don’t float and are killed by salt water; they don’t have wings or tufts for drifting on the wind, or hooks for latching onto fur or feathers; and they aren’t enclosed in fruits designed to be eaten by birds or other animals that might transport the seeds to some faraway place. Basically, sundew seeds just fall to the ground or get blown a short ways, usually ending up close to the parent plant. Over millions of years, members of the genus obviously have moved great distances, as their worldwide distribution indicates, but the assumption is that they accomplished this in small steps, not giant leaps.

  In the evolutionary tree of sundews, one can see the signature of their limited powers of dispersal. Specifically, the sundew phylogeny strongly shows geographic structure, a fancy term that just means that closely related species tend to occur near each other.31 So, for instance, a western Australian sundew’s closest relatives are likely to be other western Australian sundews, and a southern African sundew’s closest relatives are usually other southern African sundews. However, there are some conspicuous exceptions. For instance, two species found in the eastern United States fall within a group of South American species, indicating that their ancestors dispersed from that continent.

  The weirdest exception to the rule of geographic structure in sundews, though, is the Cerro de la Neblina sundew, D. meristocaulis. There are many other sundew species in South America, including some that live near the tepui region, but, according to a recent DNA sequence study, none of these are the closest relatives of D. meristocaulis. Instead, D. meristocaulis is firmly placed by the molecular data within a group of Australian sundews. These Australian species also happen to be miniaturized, like D. meristocaulis, and they share unusual pollen and leaf-hair features, among other things, with the tepui species. In other words, both DNA and anatomy indicate a close evolutionary relationship between D. meristocaulis and the Australian sundews.

  South America and Australia were both once part of Gondwana, so one can imagine a continental drift scenario to explain the connection of a Neotropical sundew to Australian species. However, molecular dating makes that explanation extremely unlikely, if not impossible. According to molecular age estimates, D. meristocaulis split from its Australian relatives only within the past 12 to 13 million years, long after the breakup of Gondwana, and its ancestors therefore must have reached South America by oceanic dispersal. The tepui sundews are certainly now isolated from any closely related species, but they are by no means ancient relicts. Measured by the span of deep time, they are yesterday’s immigrants.

  7.1 Long-distance voyager: the sundew Drosera meristocaulis from the Cerro de la Neblina tepui. Photo by Fernando Rivadavia.

  The conclusion that a sundew in some form, whether seed or whole plant, with no obvious means of overwater dispersal, managed to travel from Australia, either east across the Pacific or, more circuitously, west across both the Indian and Atlantic Oceans, and, on top of that, negotiated the towering cliffs of Cerro de la Neblina to end up on the plateau of the tepui, seems, to use a stock phrase from Darwin, “absurd in the highest degree.” The authors of the sundew study suggested that Darwin’s old mechanism of seeds stuck to the feet of a bird was one possible means of transport, although they also pointed out that there is no known avian migration pathway connecting Australia to northern South America. Fluky events of one sort or another can take birds out of their normal routes, however, leading to what birdwatchers call “accidental” occurrences. The sundew story reminded me, for instance, that I (and several dozen other birders, armed with a thicket of spotting scopes) once saw a Fork-Tailed Flycatcher, a species that normally gets no farther north t
han tropical Mexico, sallying for insects over a farmer’s field near Rochester, New York. This bird probably had been migrating from southern to northern South America and missed some cue to stop, overshooting its mark by several thousand miles. A mistake like that could lead to a bird transporting seeds outside of any normal migration route, and even, perhaps, to a site on the other side of the world.

  Admittedly, this argument has gotten us into the realm of nothing-that-you-would-want-to-hang-your-hat-on. But maybe that’s part of the point—that even when no likely mechanism presents itself, nonetheless we often must infer that long-distance dispersal did occur. Rare, mysterious, and miraculous things do happen, even if we may never know exactly how. (I’ll take that point up in more detail in Chapter Eight.) The main point here, however, is that if a sundew can voyage thousands of miles, mostly over the ocean, from Australia to the top of a tepui in northern South America, then we should also expect many other plants, some of which have better means of traveling long distances, to have made similar journeys. We should perhaps expect that, if such plant dispersal events were traced on the globe, we would end up with an amazing tangle of lines going every which way, a green web connecting the landmasses of the world.

  FIFTY-ONE BEAN PLANTS

  I was talking on the phone with a botanist named Matt Lavin, a professor at Montana State University, about the biogeography of plants. Lavin and I have never met, but I know from photos that he’s thin, with angular features and slightly unkempt, graying hair. (In the course of our conversation I found out that we know some of the same people and even overlapped at Cornell for a year, while I was a grad student and he was a postdoctoral researcher. We were in different departments, but undoubtedly went to some of the same talks, so we had probably seen each other without knowing it, which might seem surprising, but, in the small world of evolutionary biology, probably shouldn’t.) For an academic, Lavin’s speech is relaxed and down-to-earth, with the hint of a cowboy drawl. He tends to leave the “g”s out of his “ing”s. When I asked him how he got interested in biology, he said, “I grew up all over the West because my dad was a Forest Service guy, and he moved around every couple years, so we lived all over southern Idaho, western Wyoming, northern Nevada, so kind of livin’ in sagebrush country, gettin’ out in it, . . . and adjacent forest areas, and learnin’ plants.” That upbringing explains the rural western shadings in his voice and, maybe, the fact that a big part of his research deals with sagebrush communities. Since I live in Nevada and like getting out in the wilds too, our conversation veered into talk of out-of-the-way places in the West. I asked him, for instance, how he ended up writing a flora of the Sweetwater Mountains, an eastern outlier of the Sierra Nevada, where I’ve collected alpine plants and arthropods. (Our shared familiarity with this obscure mountain range struck me as more unexpected than our overlapping time at Cornell.) What I really wanted to discuss with him, however, was the other major part of his research, dealing with the worldwide biogeography of legumes, that is, plants in the bean family.