The Monkey's Voyage

by Alan de Queiroz

  Not long after returning from São Tomé, Measey was standing in the lunch line of the canteen at the IRD in Paris and got to talking about his trip with the director of the IRD’s Bondy office, Alain Morlière. Measey (now speaking fractured French instead of fractured Portuguese) started telling Morlière about amphibians dispersing to the islands and, in particular, about the cobra bobo and the curious fact that no close relative of this species was known from the nearby West African coast. The nearest place where another specimen of the genus Schistometopum had been found was the eastern part of the Congo River drainage, closer to the Indian Ocean than the Atlantic. Measey was explaining that it was hard to imagine how a burrowing amphibian could get from the eastern Congo Basin to the Gulf of Guinea islands. Morlière is an oceanographer and, coincidentally, had done research on Rolas, a small island off the south coast of São Tomé. It’s simple, he said, the plume of water coming out of the Congo River delivers lots of debris right to the Gulf of Guinea. I picture Measey’s pupils suddenly dilating at that point. Fantastic! he thought. Here was the explanation, or at least the beginning of an explanation, to the mystery of how the cobra bobo and other species crossed to São Tomé and Príncipe. Natural rafts, perhaps clots of earth and vegetation dislodged from the banks of the Congo, could have been carried downriver to the ocean, then drifted north with the prevailing current to eventually make landfall on the Gulf of Guinea islands.

  One can imagine Measey running to his office and immediately getting to work on a paper about the dispersal of amphibians to the islands. That’s not what happened, though. Instead, he sent an email to some oceanographers who Morlière said would know more about the Congo plume and the currents off the west coast of Africa. Then he waited for the responses. Nothing happened. Nobody replied, probably because oceanographers, by and large, couldn’t care less about how the cobra bobo and a bunch of frogs ended up on São Tomé and Príncipe.

  Measey, stymied for the time being, went back to thinking about the effect of caecilians on their soil-dwelling prey.

  SEEING THE UNEXPECTED

  On the other side of Africa, a half-Spanish, half-German biologist named Miguel Vences was studying frogs on Madagascar, the Seychelles, and other islands in the Indian Ocean. Vences is one of those scientists who’s involved in so many projects that one wonders when he finds time to sleep. Among many other things, he and his colleagues have discovered something like three hundred new species of frogs on Madagascar in the past twenty years. (Apparently he does have his limits, though, because so far only about a hundred of the new frogs have been formally described in scientific papers.)

  For his doctoral dissertation at the Rheinische Friedrich-­Wilhelms-Universität in Bonn, Vences studied the geographic origins of Indian Ocean frogs. When he began the project, he was thinking what almost any biogeographer would have thought about frogs on the Gondwanan islands of Madagascar and the Seychelles: frogs can’t cross salt water, so these island species must be Gondwanan relicts. His aim was to work out the details of the vicariant origins of the frogs. Overwater dispersal wasn’t even in his thoughts. Then a funny thing happened, the same thing that has happened to many biogeographers over the past twenty years or so: he started looking at the frogs and sequencing their DNA—that is, he started gathering data—and the frogs told him a different story. Actually it was several stories, but all with the same basic conclusion.

  8.3 Miguel Vences looking for frogs in 1991 on the first of many trips to Madagascar. At the time, Vences’s thoughts on biogeography revolved around vicariance, but that would change. Photo by Katharina Wollenberg-Valero.

  Vences’s most straightforward study didn’t focus on frogs on former pieces of Gondwana, but on the island of Mayotte, known for its white sand beaches, turquoise seas, and French ambience (Mayotte is a territory of France). The island lies some 190 miles west of the northern tip of Madagascar and, geologically, is part of the Comoros Archipelago. The Comoros are volcanic, with successive islands probably forming as the Somali tectonic plate slid eastward over a “hotspot,” a plume of magma rising up from the Earth’s mantle (the same kind of process that formed the Hawaiian Islands.) All of the Comoros are thought to have been formed in the past 9 million years, and, as with São Tomé and Príncipe, there is no evidence that they have ever been connected to a continental landmass. In this case, the point to remember is that they have never been connected to the continental island of Madagascar.

  Mayotte has two species of frogs, and, since amphibians are not supposed to be able to get to oceanic islands on their own, it was assumed that they had been introduced. In fact, biologists describing the Mayotte frogs had concluded that they were Boophis tephraeomystax, a pale gray-green treefrog, and Mantidactylus granulatus,38 a nondescript little brown frog, both of them common Madagascan species. Maybe they had arrived on Mayotte as accidental stowaways on a boat from Madagascar. Or maybe some kids had brought them over as pets and later released them.

  When Miguel Vences and his colleagues examined the Mayotte frogs, though, they immediately found a problem with this story: the frogs didn’t actually look like the species they were supposed to be. The Boophis on Mayotte was larger, had eyes of a different color, and had bumpier skin than its Madagascan counterpart. Mayotte’s Mantidactylus was even more distinct. Compared to the “same” species on Madagascar, it was smaller and had a single white vocal sac rather than two blackish ones. Its mating call was especially different, sounding more like the rapid “scritch-scritch-scritch-scritch” of a Bank Swallow than the measured croak of the Mantidactylus granulatus on Madagascar. The differences in vocal-sac anatomy and mating calls were especially telling: female frogs are not attracted to males with the wrong call, so frog populations with dissimilar calls are almost always different species. The two Mayotte frogs were also unlike any of the hundreds of other frog species on Madagascar. So, even by looking at fairly obvious physical traits and behavior, Vences was thinking the Mayotte frogs were species unique to that island.39

  Vences and his colleagues also sequenced several genes—one from the nucleus and three from the mitochondrion—of the Mayotte frogs and forty-five related Madagascan species. When they constructed an evolutionary tree from these gene sequences, they found that the Madagascan species that were supposed to be the same as the Mayotte frogs are apparently not even the closest relatives of the latter. More significantly, they also found that the Mayotte species are genetically distinct from all the other species they sampled. Both anatomy and genes thus pointed to the same conclusion: the two frog species on Mayotte are indeed new species only found there. The evolutionary tree clearly shows that their ancestors came from Madagascar (and not, for instance, from mainland Africa), but the ocean voyages those ancestors made to Mayotte must have happened long ago, long enough to allow the current differences in anatomy and gene sequences to evolve. The upshot of all this is that these voyages could not have been made with the help of people, because they happened well before anybody knew how to make a boat. In fact, these journeys almost certainly took place before there were any people at all, if, by people, we mean members of Homo sapiens. In short, batrachians, Darwin’s hopeless ocean voyagers, had made at least two natural passages (one for Boophis and one for Mantidactylus) between Madagascar and Mayotte.

  The story of the Mayotte frogs reminds me, oddly enough, of a quail that Tara and I raised some years ago. Motivated, maybe, by some stirring of parental instincts, we bought a dozen or so mail-order California Quail eggs and put them in an incubator meant for reptiles. Predictably, most of them didn’t develop, but we successfully raised one chick, whom we named Fiver. One day our friend Janet Bair, an accomplished birdwatcher, was at our house, and we brought out Fiver the California Quail, who by then was grown up and tame and would entertain people by hopping on their shoulders and pecking food off the table. As I’ve mentioned, I’m a birder, and I think of myself as being, not an expert, but at least competent,
to the point of having led birdwalks for the Audubon Society and other organizations from time to time. Anyway, Janet took one look at the bird and said, “That’s not a California Quail.” I looked at Fiver, as if for the first time, and realized she was right—our “California Quail” had a white throat and a big rusty stripe on the side of his face, coloring that no bird of that species would have. Checking the field guides, we decided that Fiver was probably a cross between a Gambel’s Quail and a Northern Bobwhite, which wasn’t too unusual for a bird hatched out of an egg from a quail farm.

  The point of this story is that, having been told that we had California Quail eggs, I simply accepted that Fiver was a California Quail, despite the obvious evidence to the contrary and my years of birdwatching experience. In the same way, I think the biologists who originally described the frogs of Mayotte were blinded by the notion that amphibians could not have reached that volcanic island on their own, and that, therefore, people must have brought them there. I suspect that these biologists saw them as species from Madagascar because that was what they expected to see, despite the obvious physical traits declaring that the Mayotte frogs were distinct.

  Being blinded by preconceptions is such a common thing, not just in science but in all human affairs, that you can hardly blame someone for falling victim to it. (Maybe that’s why I’m not too embarrassed to tell the story of Fiver.) In dealing with the history of poor dispersers, though, we have stumbled into a particularly egregious area, a kind of country of the blind. Rule out even the possibility of oceanic dispersal from the start, and one must then force the evidence to fit one of the remaining alternatives, either human introduction or vicariance (involving some kind of former land connection), or, in a bygone time, separate creations of species. In a sense, when Darwin did his seed experiments, he was thinking about the need to overcome this very preconception, the notion that living land organisms cannot cross wide stretches of ocean. That battle, in slightly different guise, continues today.

  Along with the case of the Mayotte frogs, Miguel Vences and his colleagues discovered two other instances of amphibians naturally crossing seas to colonize Indian Ocean islands, both involving treefrogs in a family called the Hyperoliidae. In these cases, the island species were known to be quite distinct, so nobody thought they had been introduced. Instead, the accepted explanation for how they arrived on islands was continental drift; that is, the ancestors of these frogs were supposedly on the landmasses in question before those areas became islands.

  Vences addressed these cases through molecular dating using relaxed clock methods, that is, ones that do not assume that the clock ticks at a constant rate. In one case, he found that a lineage of treefrogs on Madagascar had split from its closest relatives in Africa sometime between 19 and 30 million years ago. In the other, a treefrog species on the Seychelles had split from its closest relatives on Madagascar between 11 and 21 million years ago. The ages of these branching points in the frogs’ evolutionary trees are not even close to being old enough to be explained by continental drift: the landmasses of Africa and Madagascar have been separated for about 130 million years, and Madagascar and the Seychelles have been separated for more than 80 million years. In other words, by the time the two frog species in question reached Madagascar and the Seychelles, those landmasses had long been islands.

  Vences, who had begun his research in the Indian Ocean thinking only about vicariance, had come up with four separate instances of frogs crossing to and from islands. As with Matt Lavin and his bean plants, the evidence—in the frog studies, meaning anatomy, behavior, and DNA sequences—had forced Vences to change his mind.

  OF FROGS AND FLOATING ISLANDS

  Vences and his colleagues reported their results in a 2003 paper called “Multiple Overseas Dispersal in Amphibians” published in the Proceedings of the Royal Society of London. I remember that paper well—the discovery that frogs had dispersed to and from Indian Ocean islands not just once but four times was one of the key findings that made me think that something unusual was happening in the science of biogeography. At the time, we were living in Ely, a copper-mining town in eastern Nevada and a place that, for one reason or another, often had me thinking about long-distance journeys. For starters, Ely is one of the most isolated communities in the lower forty-eight states, which meant that, when Tara and I needed to take a plane flight, we usually had to drive at least three and a half hours just to get to an airport, the nearest large one being in Salt Lake City. Also, much of the human history of the area involved almost inconceivably strenuous journeys. For instance, forty miles north of Ely is Schelbourne, the site of an old Pony Express station, and seventy-five miles farther north lies the Hastings Cutoff, the wagon train “shortcut” that delayed the Donner Party long enough to eventually get them trapped in the snows of the Sierra Nevada. And, finally, the Basin and Range landscape, stark and expansive, with its seemingly endless rhythm of valley after mountain range after valley after mountain range, made me think about the long-distance dispersal of plants and animals.

  I spent a lot of time hiking and, occasionally, awkwardly cross-­country skiing in two nearby ranges, the Snake Range and the Schell Creek Range, high mountains built of sand and shells laid down in shallow seas 500 million years ago. Having grown up taking many trips to the classically scenic Sierra Nevada, I at first found these Great Basin mountains, with their scruffy vegetation, few lakes, and no rivers, pale in comparison, but I soon came to love their subtle beauty. And as I wandered in these ranges, separated from each other and from the mountains east and west by desert valleys, I thought about mountains as ecological islands, as countless biologists had before me (although not many of them did their thinking in eastern Nevada).

  I especially thought about that when I hiked above 11,000 feet or so, where gnarled old bristlecone and limber pines and mountain sagebrush give way to the ground-hugging plants of the alpine tundra. These tundra environments, widely separated from each other by much hotter, drier environments, are the most island-like of all the altitudinal zones in the mountains. How do alpine plants and arthropods—especially plants that have no special means for dispersing their seeds, and arthropods that cannot fly—get from one of these tundra islands to another? The old story is that alpine species moved by normal dispersal among the mountain ranges during the ice ages, when the valleys that they had to cross were much cooler and wetter, and therefore more alpine-like, than they are today. Vences’s study, however, along with my own research with Robin Lawson on the Baja California garter snakes and some other works I had encountered, were making me wonder: maybe biologists had underestimated the ability of alpine species to travel long distances across inhospitable terrain, just as they had underestimated the ability of frogs to cross salt water.

  In any case, living in such isolation, I was very glad that Al Gore had invented the Internet (although our “high-speed” connection would break up if rain or snow was falling between our house and the mountaintop transmitter), and that I could go to the website of the Proceedings of the Royal Society of London and download any article I wanted from the journal. I read about the frogs of Mayotte and wondered how a dwarf alpine paintbrush could get from the top of the Snake Range across the desert to the top of the Schell Creeks.

  In Paris, a place that, unlike Ely, could reasonably be considered the heart of the civilized world, John Measey also read Vences’s article. Measey was convinced by Vences’s evidence, but he was also struck by a sentence in the paper’s introduction that was meant to emphasize the uniqueness of the results. That sentence read: “No endemic amphibian species are known from truely [sic] oceanic islands.” (“Endemic” means found only in the area in question.) Vences and his colleagues had mentioned oceanic islands, like Hawaii, that have only nonnative amphibians, and also islands, like those of the Sunda Shelf (Borneo, Sumatra, and others), that have endemic amphibians but were once connected to continents. They never mentioned the amphibians of Sã
o Tomé and Príncipe. Bob Drewes was right: the Gulf of Guinea islands really were overlooked biological treasures; even Miguel Vences, an expert on island amphibians, had never heard of the cobra bobo and the other endemic amphibians of São Tomé and Príncipe.

  Measey and Drewes briefly considered writing a paper that would have pointed out that Vences and his colleagues were wrong about the lack of known endemic amphibians on oceanic islands. The paper would have been short, basically just saying that several endemic amphibian species from the oceanic islands of São Tomé and Príncipe had been described beginning in the late 1800s, and that the ancestors of those species probably arrived by oceanic dispersal. But Drewes knew Vences and liked him, and he was a little uncomfortable with the idea of writing a paper that took Vences to task, however mildly. Instead, Measey and Drewes ended up recruiting Vences into a more involved project on the Gulf of Guinea amphibians. I’m guessing they didn’t have to twist his arm. Vences was by now the “king of amphibian oceanic dispersal,” so the new study was right up his alley. Plus, he was the guy who had discovered three hundred new species of frogs while being involved in dozens of other projects at the same time; if “overextended” was in his vocabulary, he probably thought it was a state of grace.

  The new study focused not on the cobra bobo, but on a fairly generic-­looking green and brown frog called Ptychadena newtoni, one of the species that Measey had barely noticed in passing while he was looking for the bright yellow caecilians on São Tomé. The goal was to figure out if P. newtoni was indeed a species unique to São Tomé by comparing its DNA to that of other Ptychadena. Bob Drewes will tell you that P. newtoni is exciting because it’s the largest species in the genus Ptychadena and is, therefore, an example of the tendency for island organisms to become either giants or dwarfs (although, topping out at three inches long, P. newtoni is only a giant among smallish frogs). However, the main reason they chose to study P. newtoni was more practical and prosaic: between them, Vences and Drewes already had mitochondrial DNA sequences or tissue samples from most species of Ptychadena, including the ones that might conceivably be the same as P. newtoni.