A message that comes out of the new evidence for long-distance dispersal is that, although there are many useful generalizations about such events, one should not mistake these for laws that cannot be broken. Darwin himself made this mistake, discerning that mammals are not very good at crossing ocean barriers, but then going too far in declaring that “no terrestrial mammal can be transported across a wide space of sea.” More generally, the message is that we should not be too wedded to our preconceptions. There is evidence about improbable events in deep history, and we need to follow it to wherever it leads. This brief interlude in Hawaii provides two final illustrations of such evidence leading to the unexpected. More importantly, these examples serve as a point of entry for discussing why views on chance dispersal have oscillated from Darwin’s time on, and, especially, why the latest shift in thinking might be fundamentally different from earlier ones.
Regarding bristletails and their poor dispersal abilities, it turns out that there are native species on quite a few volcanic islands, such as the Azores, Lord Howe, and São Tomé (where they may fall prey to the ocean-voyaging amphibians studied by John Measey), islands that they must have reached by overwater dispersal. Also, one bristletail genus, Neomachilellus, suspiciously contains dozens of species in the New World tropics and just two in West Africa, a geographic pattern that may reflect a recent eastward crossing of the Atlantic. In short, bristletails have likely negotiated ocean barriers more often than one might have guessed based on their physical characteristics and behavior.
The evolutionary connections of the Hawaiian bristletails suggest that they have also failed to follow another generalization about dispersal and colonization. The generalization is that nonaerial animals—that is, ones that can neither fly well nor use silk to float by “ballooning”—can only reach Hawaii from the west, not from the New World. The thought here is that aerial animals, such as birds, butterflies, and ballooning spiders, might make the trip from the Americas in a matter of days, either under their own power or blown by storm winds, whereas nonaerial ones would require an almost impossibly long rafting voyage, because there are no islands between the Americas and Hawaii to use as waystations. However, there are many islands in the western Pacific, and these would allow nonaerial animals from, say, Asia or Polynesia to reach Hawaii by a series of shorter rafting “hops.” A 2012 compilation of phylogenetic studies shows very clearly the pattern expected under this reasoning: out of twenty animal lineages that apparently arrived from the west, nine were aerial and eleven were nonaerial, whereas, out of thirty-seven lineages that came from the New World, all were aerial.
Bristletails might be the first exception to this generalization. In the early 1990s, a German entomologist named Helmut Sturm, who probably knows more about the bristletail order than anyone else in the world, examined the anatomical traits of the Hawaiian species—the shape of their second pair of eyes, the number of water-absorbing vesicles on their abdomens, and the like—and concluded that they were most closely related to a species called Neomachilis halophila, from the west coast of North America. Recent molecular evidence supports Sturm’s conclusion: DNA sequences from our Hawaiian specimens and from various continental species strongly confirm the close connection of the island forms to N. halophila. The obvious interpretation of these results is that the Hawaiian bristletails originally came from North America, presumably by rafting.
A monumental rafting voyage might not be as implausible as it at first sounds for Neomachilis. To begin with, the mainland species, N. halophila, is found among rocks around the high tideline, where the chances of being swept out on a wave must be great. Furthermore, Sturm envisioned a mechanism that might get around the problem of delicate, erratically jumping insects surviving such a voyage: he suggested that they made the passage not as adults, but as eggs attached to driftwood. To bolster this argument, he noted that bristletail eggs in general take a long time to develop, and that N. halophila eggs, in particular, are resistant to salt water. (In fact, one investigator has claimed that the eggs of some bristletails are even resistant to chemicals used to preserve specimens.)
Sturm’s “how-possibly” argument seems reasonable, but he might have gotten the direction of the journey backward. Specifically, the DNA sequences we’ve obtained so far suggest that the North American N. halophila might actually fall, in an evolutionary sense, within the Hawaiian bristletail group. We clearly need more data before we can say anything definitive, but if that result holds up, it would indicate, not a North America to Hawaii colonization, but the reverse, that is, island bristletails establishing themselves on the continent. Such an island-to-continent colonization would, of course, require a long unbroken journey across the eastern Pacific, just as the assumed continent-to-island colonization would. In addition, it would go against a very general and well-known idea about island organisms, an idea that is sometimes treated as a biogeographic rule.
The rule is this: Islands, especially small, remote islands like the Hawaiian chain, are evolutionary dead ends. Many groups colonize such islands and flourish on them, but once they have evolved into distinctive island forms, they do not go back and successfully establish themselves on continents. Islands are like evolutionary black holes; lineages go in, but they don’t come out.
Two main arguments have been given to explain this islands-as-dead-ends rule. First, island species—especially ones on remote islands like Hawaii—are thought to suffer from a kind of evolutionary degeneracy, because, unlike continental organisms, they do not have to compete with a large array of other species. The notion here is that natural selection is relaxed on islands, producing an assemblage of “slacker” species that would never survive in the dog-eat-dog (or dog-outgrow-dog, or, ultimately, dog-outreproduce-dog) environments of continents. Island plants that have lost their defensive spines or thorns, and island animals that show no fear of terrestrial predators, are obvious examples, although “relaxed” features aren’t always so blatant. The second argument is that oceanic islands, in particular, are tiny compared to continents, so they typically do not harbor very large populations from which “chance” colonists might be drawn. This effect probably explains, at least in part, why natural invasions recorded in recent times have usually been by species that are both widespread and abundant in their native areas; those species are the ones that contain especially large numbers of individuals.
These arguments against island-to-mainland colonizations sound reasonable if one believes that long-distance dispersal of any kind is exceedingly rare in general. However, if, through deep time, such dispersal events are fairly common, then maybe we should expect that at least some island lineages have broken the rules and succeeded in colonizing continents. In any case, we should let the evidence sort things out. For Neomachilis, the case for an “out of Hawaii” dispersal is tenuous, but there are many other Hawaiian groups, and some of them have been studied far longer and far more intensively than the bristletails.
In the scruffy forest by the Nuuanu Pali, between sweeping mossy boulders for bristletails, Steve Montgomery and I got to talking about ocean crossings. We were in Hawaii, after all, so it was a pretty obvious thing to bring up. Not surprisingly, Steve knew a fair amount about the subject, and he mentioned the case of Hawaiian fruit flies, a group that probably has been studied more than any other Hawaiian evolutionary radiation and perhaps more than any island radiation anywhere.
Steve was talking in particular about the work of a Berkeley entomologist named Patrick O’Grady (who also happened to be the person who had told me to contact Steve for advice on where to find bristletails in Hawaii). One of Patrick’s scientific missions is to construct an evolutionary tree for all Hawaiian fruit flies, which might sound like a modest ambition, but in fact is a large and unending task, because the group includes close to six hundred described species, a backlog of several hundred more that are known to exist but haven’t yet been formally described, and new forms being disc
overed every year. Beyond this, the latest research by Patrick and his colleagues indicates that piecing together the Hawaiian fruit-fly tree is actually a much larger and more complex undertaking than even the fly experts imagined.
The big complication is that not all Hawaiian fruit flies live in Hawaii. The fruit-fly evolutionary tree implies a single colonization of the archipelago leading to a large proliferation of flies—the hundreds of species that have been the focus of Patrick’s research. However, it also indicates, unexpectedly, that several lineages of Hawaiian fruit flies in the genus Scaptomyza have “escaped” to other areas without the help of humans. Some of these travelers ended up on other warm-weather Pacific islands, such as the Marquesas, which maybe is not so surprising; after all, Hawaiian flies might be expected to do well on other islands with similar climates and a similar lack of a full continental array of species. A slacker species from Hawaii, the thought goes, ought to be able to hold its own among the comparable evolutionary degenerates on other oceanic islands.
The weird thing is that some of the Hawaiian escapees have given rise to flies that live not on islands, but on continents. They have been found in Australia and Africa, for instance, and, by the time Patrick and his colleagues are done, it is likely that some flies on every continent except Antarctica will be found to have Hawaiian ancestors. The meek species of islands are not supposed to inherit the Earth, but Hawaiian fruit flies have apparently taken a serious step toward populating much of the planet. In fact, the mainland Scaptomyza flies have become enough of a presence that they have also colonized numerous islands, replaying the original colonization of Hawaii by continental flies. Thus, a partial history of successful fruit-fly dispersal goes something like this: from some as yet unidentified continent to Hawaii, back to continents, then back to islands (many times), as well as a bunch of jumps between islands. When it comes to ocean crossings, the fruit flies are the insect equivalent of Matt Lavin’s bean plants.
Why have Scaptomyza flies been able to cross ocean barriers so readily? Patrick thinks there are a couple of key factors. One is that members of this genus are resistant to desiccation compared with other fruit flies and with flies in general, a good property for surviving a transoceanic voyage, whether on a raft or on storm winds. Second, Scaptomyza flies tend to be generalist feeders, meaning they can subsist on many different kinds of flowers or fruits; thus, relative to more specialized species, they are more likely to find plants they can eat wherever they happen to end up. Finally, because they have short generations (even for flies), they can build up large populations quickly, perhaps making them less subject to the random “drunkard’s walk” off the cliff of extinction.
In any case, however they managed to do it, these fruit flies add an especially striking example to the long list of “things that shouldn’t have happened but did.” Degenerate island lineages are not supposed to establish themselves on continents, but Scaptomyza flies have done exactly that, probably several times. In fact, they’re apparently only one of several island groups that have managed this trick: DNA-based evolutionary trees also indicate, among other cases, that aquatic beetles in the genus Rhantus successfully dispersed from New Guinea to both Australia and Eurasia; that Anolis lizards went from the West Indies to Central or South America and, independently, to North America; and that monarch flycatchers traveled from Pacific islands to Australia.
These surprising events have been revealed by scientists like Patrick, who were gathering evidence and following where it led them instead of just believing the prevailing dogma. John, Cheryl, and I are attempting to do the same with the Hawaiian bristletails, as are many other researchers studying other groups, not only in Hawaii but all over the world. This is not to suggest that Patrick or any of us deserve a pat on the back; we’re all just doing what scientists are supposed to do, namely, focusing on the evidence. That sounds like such a simple, not to mention essential, thing to do. However, in the field of historical biogeography, it often hasn’t worked that way.
BEAUTY, TRUTH, AND EVIDENCE
Say that we could pluck four well-known biogeographers representing different time periods from the past 150 years, put them in a room together, and pose to them this question, “How is it that there are monkeys on both sides of the Atlantic Ocean?”
Our first biogeographer is none other than Charles Darwin, who hems and haws a bit because he doesn’t think mammals have much chance of surviving a long ocean journey, but finally settles on the idea of monkeys crossing the Atlantic on an enormous raft. This “chance dispersal,” though it appears unlikely, seems more plausible to Darwin than the obvious alternative, some sort of transatlantic land bridge. “It shocks my philosophy to create land,” he says, “even more than to imagine an ocean-crossing primate.”
Charles Schuchert, a Yale invertebrate paleontologist snatched from the early twentieth century, is the next to speak. Schuchert says that even invertebrates with larvae that drift in the open sea rarely traverse oceans, so the idea of a monkey making such a journey and establishing a new population is absurd, raft or no raft. Schuchert is a diehard land-bridge advocate, and his explanation of choice is a wide swath of continental crust stretching between the westward-projecting hump of Africa and the eastward-projecting hump of Brazil. In support of this hypothesis, he notes that the largely volcanic islands of Tristan da Cunha, St. Helena, and Ascension all contain some granite, which he claims is evidence of continental origins. “Those granitic pieces,” Schuchert pronounces, “are the ancient remnants of the land bridge.”
Then, from the middle of the twentieth century, comes George Gaylord Simpson, the prominent American Museum of Natural History mammal paleontologist and disciple of the great dispersalist William Diller Matthew. Simpson doesn’t like Schuchert’s hypothetical land bridge and seems to have less of a problem with overwater dispersal from Africa than Darwin did (although he wonders whether early monkeys colonized South America from North America across an ocean barrier not so wide as the Atlantic). He pulls out a pen and jots down some calculations to show that an event with a minuscule probability of occurring in any given year, such as monkeys rafting across 1,800-plus miles of ocean, might yet be reasonably likely given a long enough period of time. Monkeys, he notes, have been around for something like 50 million years, which makes a single, successful Atlantic crossing not so far-fetched. Darwin carefully follows Simpson’s argument and nods approvingly; he’s had the same thoughts, but couldn’t do the math.
Our next biogeographer is Gary Nelson, the Gary Nelson of the 1970s at the American Museum, although, frankly, he might just as well be the Gary Nelson of today (who lives in Melbourne, Australia, and is as vehement as ever when it comes to cladistics and vicariance). Nelson drops a bombshell on the others: the geological evidence for continental drift. There’s some rumbling over this, but when Nelson brings up the alternating magnetic stripes indicating seafloor spreading (after explaining the necessary background), the others grow quiet. (Schuchert, an outspoken critic of Wegener’s theory, seems less than thrilled with the revelation.) Nelson goes on to sketch some crude maps that show the opening of the Atlantic Ocean through time and scribbles a cladogram of primates on top of the last map. (Darwin, who included an evolutionary tree as the only figure in The Origin of Species, seems to intuitively understand the importance of the cladogram more than Schuchert or even Simpson does.) “It’s all about the fragmentation of ancestral biotas,” Nelson says, “and not just for monkeys, for everything.”
Simpson seems a bit discombobulated on seeing the evidence for continental drift, but he regains his composure and asks Nelson when the Atlantic first opened. On hearing that it was about 100 million years ago, he says that monkeys didn’t even exist then, so the whole business of continental drift is irrelevant in this case. This Simpson from midcentury is meeting Nelson for the first time, but seems to instinctively distrust the brash young man. Meanwhile, Darwin is sitting back, silent, with one hand sc
ratching through his beard. In The Origin, he spent a good deal of time arguing that one doesn’t see much evidence of gradual evolution in the fossil record because that record has so many gaps, and now he’s wondering if Simpson might be relying too much on fossils. “Maybe monkeys were in existence far earlier than we know,” Darwin thinks to himself. “However, I greatly doubt it.”
As Nelson and Simpson continue arguing, a door opens and a fifth biogeographer walks in, this one nameless and from the present day. “I think you people need to hear about relaxed molecular clocks,” she says.
The Monkey's Voyage Page 31