The Monkey's Voyage

by Alan de Queiroz

  All of these lineages and others have been recognized as Gondwanan relicts. However, both geological and biological evidence suggest that, despite appearances, New Caledonia’s history actually parallels that of the Chatham Islands. Geologists now think that New Caledonia was entirely underwater from about 37 to 70 million years ago, a period corresponding to a general subsidence of the area and, later, overthrusting of oceanic crust onto the original continental crust as New Caledonia collided with an island arc. Only marine strata are known from the period of inundation, including layers composed of a type of fine-grained chert that must have formed deep under the ocean.

  As with the Chathams, molecular dating results are in line with this history of submergence and reemergence and, therefore, of mass extinction and recolonization. Many New Caledonian lineages, including the skinks and Araucaria conifers, are estimated to have split from relatives elsewhere only after the “drowning,” indicating overwater colonization rather than persistence from Mesozoic times. Others, such as Amborella, diplodactylid geckos, and troglosironid harvestmen, diverged from relatives before the submergence of New Caledonia, but, critically, there is no compelling evidence that these groups were actually on New Caledonia before the drowning. Such evidence could potentially come either from fossils or from molecular dating showing that branching points within an endemic New Caledonian radiation preceded submergence, but that evidence has not materialized. Within groups studied so far, none of the earliest branching points clearly occurred before the reemergence of land (see Figure 10.5). Thus, although some of these lineages are ancient in the sense of having separated from their nearest living relatives in the Mesozoic, there’s little to show that they’ve been riding on a New Caledonian ark ever since then.51 All could have arrived via overwater dispersal in the past 37 million years.

  10.5 Timetree showing that the earliest branching point within New Caledonian geckos (marked by a black circle) occurred after the period of the island’s assumed submergence. The shaded bar is the 95 percent confidence interval for the age of the branching point in question. Timetree redrawn and modified from Nielsen et al. (2011).

  If the drowning scenario is true, it means that New Caledonia, like the Chathams, is biologically an oceanic island, despite its continental roots. There is still a good deal of resistance to that idea because of those strange and demonstrably ancient lineages such as Amborella and the troglosironid harvestmen; those groups just have to be Gondwanan holdovers, the thought goes, even if that conclusion cannot be demonstrated by fossils or molecules. However, it is telling that similarly unique, ancient lineages inhabit islands that are unequivocally oceanic. For instance, the shrub Lactoris from the Juan Fernandez Islands is the sole living member of a family otherwise known only from Cretaceous to Miocene fossils from various continents; the Hawaiian perennial herb Hillebrandia, in the begonia family, apparently separated from all other living begonias more than 50 million years ago; and the Round Island boas of the family Bolyeriidae split from their nearest living relatives close to 70 million years ago. These ancient lineages must have colonized their young, volcanic island homes by overwater dispersal, subsequently becoming extinct in the source areas. Ancient lineages are often assumed to demonstrate the equally ancient persistence of an area, but the story of New Caledonia and the examples from oceanic islands tell us that, in fact, such an assumption is a leap of faith.

  To complete our tour of Zealandian islands, we will briefly revisit our old friend New Zealand, where some of the first major cracks in the vicariance worldview materialized. We have already seen that the original Gondwanan plants of New Zealand have largely disappeared and, conversely, that almost the entire modern flora, from the southern beeches to the vegetable sheep, has descended from overwater colonists. The story for animals turns out to be similar, although perhaps not quite as extreme; in a recent compilation of molecular dating studies, some three-quarters of New Zealand animal lineages clearly fell into the ocean voyager category, while the rest were possibly, but not definitively, Gondwanan relicts.

  Many of the cases of overwater colonization are not surprising, at least in hindsight; rails, for instance, which have dispersed to dozens of Pacific islands, made it to New Zealand several times, and Bob McDowall’s galaxioid fishes colonized New Zealand at least twice, presumably as immature marine “whitebait.” However, there is one example that, if it holds up, will rank among the most unexpected cases of long-distance, overwater dispersal yet known.

  This case involves the ratite birds, possibly the most iconic example of Gondwanan vicariance. The ratites include living ostriches, rheas, emus, cassowaries, and kiwis, plus several extinct groups, such as the moas of New Zealand and the elephant birds of Madagascar. They are primarily a Southern Hemisphere group (although fossil ratites have been found in Eurasia and possibly North America) and, along with the tinamous, they are clearly the sister group to all other living birds, which suggests ancient origins. Those facts have pointed to Gondwanan breakup as the explanation for their wide distribution across the southern lands. And, of course, ratites are large, flightless birds, which, for many biologists, has closed the case: these birds couldn’t possibly have crossed wide sea barriers, so their distribution must be the result of continental fragmentation.

  10.6 The placement of the flying tinamous within the ratite group suggests several losses of flight among the ratites. That, in turn, raises the possibility that some ratites flew across ocean barriers; for instance, moa ancestors could have flown to New Zealand. Evolutionary tree redrawn and modified from Phillips et al. (2010).

  This apparently clear example of Gondwanan vicariance, however, has started to unravel. A pure vicariance scenario for ratites now seems unlikely because of the lack of agreement between the branching order in the ratite evolutionary tree and the sequence of breakup of the Gondwanan fragments, and because molecular dating studies suggest that some splits in the ratite tree occurred after the separation of the associated landmasses. Furthermore, the case for dispersal is bolstered by another recent, rather astonishing finding that calls into question the whole notion that ratites cannot cross ocean barriers. The new result, convincingly shown in several different phylogenetic studies, is that tinamous, which are heavy-bodied but flying birds, are actually deeply embedded within the otherwise flightless ratite group. What this implies is that the common ancestor of all ratites could fly, and that the capacity for flight was lost several times within the group (see Figure 10.6). Loss of flight has been very common in birds, having occurred in eighteen living families, whereas the alternative in this case, that flightless ratites gave rise to flying tinamous, seems extremely unlikely. For New Zealand, the upshot is that the ancestors of the extinct moas might not have needed an improbable rafting voyage to reach New Zealand; they could have just flown there.

  Even while evidence for the vicariant origins of “obvious” Gondwanan groups like the ratites and southern beeches has fallen apart, biologists still point to ancient lineages, such as the tuatara and the New Zealand wrens (which are likely the sister group to the enormous clade of all other perching birds), as if these were definitive Gondwanan relicts. In general, though, there is no clear evidence from fossils or molecules that these taxa were in place in New Zealand before the birth of Zealandia some 80 million years ago. They were certainly around somewhere, but they could have arrived after Zealandia became an island. However, there are a few exceptions, involving two lineages of mite harvestmen—tiny, short-legged relatives of daddy longlegs—and two lineages of centipedes. All four of these groups have evolutionary branching points within New Zealand that have been dated to around 90 million years ago or earlier. Thus, it appears that members of these groups were resident on what would become Zealandia before it separated from Antarctica/Australia.52 They really do seem to be Gondwanan relicts.

  Ironically, New Zealand, which has been so critical in the deconstruction of the vicariance worldview, might
actually turn out to contain a greater proportion of relict lineages than almost any other Gondwanan island or archipelago (although these lineages are not necessarily the ones that people have often recognized as relicts). It might also be the only Zealandian island group that has any ancient holdovers. Nonetheless, these observations hardly come close to resurrecting “Moa’s Ark.” The message that came first out of studies of fossil plants still holds: in general, the biota of New Zealand is not Gondwanan.

  Finally, we come to Madagascar, the largest and biologically the most diverse of the Gondwanan islands. Like New Zealand, it feels like a piece of some alternate Earth. It has no monkeys, but it is the only place in the world with native lemurs, almost a hundred species of them, from nocturnal, omnivorous, mouse-sized creatures to the diurnal, leaf-eating indri, which can weigh up to thirty pounds. Compared to nearby Africa, Madagascar has relatively few of the major groups of lizards and snakes, but among them are about half of the world’s chameleon species, including some only as big as your thumb, and boas and iguana-like lizards whose closest relatives are in South America. It has three families of birds that are found nowhere else, as well as thirty of the world’s thirty-three species of tenrecs, small mammals that have evolved into shrew-like, hedgehog-like, and even otter-like forms. It has nearly 10,000 endemic plant species. Like many places, it had a surprising number of large vertebrates that were almost certainly driven to extinction by humans. On Madagascar these were not big cats or mammoths, but, among others, a slew of giant lemurs (at least seventeen species) and elephant birds that topped out at over a thousand pounds (and laid the largest eggs of any bird ever known).

  It’s very clear that Madagascar harbors not just a huge number of unique species, but a large set of unique evolutionary radiations, such as the lemurs, tenrecs, and chameleons. These radiations must be fairly old—one-hundred-odd species of lemurs and thirty species of tenrecs do not evolve overnight, or even over 10 million nights—but it doesn’t necessarily follow that they have been on Madagascar since its final separation from another Gondwanan fragment. That final split, when Madagascar detached from India, happened about 84 million years ago, and it wouldn’t necessarily take 84 million years to generate those one hundred species of lemurs and thirty species of tenrecs. The lemurs, tenrecs, and others could have reached Madagascar by overwater dispersal 30 million or 50 million years ago and still had enough time to radiate into the diverse forms we see today.

  In fact, most biologists did not interpret the biota of Madagascar as being predominantly Gondwanan, even during the height of the vicariance movement. This hesitation stemmed from two related observations. First, because Madagascar is so close to Africa, and has been for most of the island’s history, overwater colonization has always seemed plausible, even for groups, like the lemurs, that are not very good at crossing ocean barriers. Second, several well-known Madagascan taxa, such as the tenrecs and the euplerids, a group of small carnivores related to mongooses, are members of African lineages, which is not what one would expect if those groups were Gondwanan relicts. Oddly enough, Africa and Madagascar, despite their current proximity, separated early on, at least 120 million years ago. This means that, under a Gondwanan vicariance explanation, one would not expect especially close ties between those two areas.

  Recent studies show that the traditional, dispersalist view of the Madagascan biota is correct. In a key literature-review study published in 2006, two evolutionary biologists at Duke University—Anne Yoder, director of the Duke Lemur Center, and Mike Nowak, a graduate student interested in plant biogeography—compiled evidence on the evolutionary relationships of many Madagascan lineages, from lemurs and tenrecs to butterflies and baobab trees. They found that the closest relatives of Madagascan groups occurred all over the place; sometimes they were African, sometimes Indian, sometimes Southeast Asian, sometimes Australian, and so on. Crucially, though, Madagascan lineages were connected to African lineages about three times as often as to Indian ones. What Yoder and Nowak’s survey suggests, then, is that the proximity of Africa has indeed allowed a great deal of dispersal from that continent across the Mozambique Channel to Madagascar; in effect, the original India-­Madagascar connection has been overtaken by an Africa-Madagascar connection forged by overwater colonization.

  Yoder and Nowak also looked at molecular divergence times between Madagascan lineages and their closest relatives elsewhere. Out of forty-­five groups, only two had estimated divergence dates as old or older than the separation of Madagascar from India. These results confirmed that most groups originated by overwater colonization. That must be the case if, as the molecular results indicate, most of these lineages—­including the lemurs, tenrecs, and mongoose relatives—didn’t arrive on Madagascar until after it became isolated in the ocean.53

  Subsequent studies have added a striking connection between colonization and ocean currents that reinforces Yoder and Nowak’s conclusions. Models of ocean currents indicate that, from the time of the separation of Madagascar and India until the mid-Miocene, currents flowed eastward from Africa toward Madagascar, whereas since then the flow has been in the opposite direction. That shift in ocean currents, caused by Madagascar’s northward movement into the equatorial gyre, would have made it far more difficult for land vertebrates to reach the island. When the molecular dating results are examined in the light of this change in currents, a pattern jumps out: the frequency of arrival by land vertebrates, almost all coming from Africa, dropped significantly after the mid-­Miocene. Part of that pattern, for instance, is that no land mammals have colonized the island since the reversal in the direction of the current. These results are an example of details validating the general idea that the biota originated mostly by overwater dispersal.

  10.7 Aye-ayes (Daubentonia madagascariensis), bizarre Madagascan primates that use their elongated middle fingers to extract insects from wood, making them ecological analogs of woodpeckers. Aye-ayes are descended from the same overwater colonist ancestor that gave rise to the lemurs. Drawing by Gustav Mützel.

  The fossil record also dovetails strikingly with Yoder and Nowak’s work. The key studies have been headed by David Krause, a paleontologist at the State University of New York at Stony Brook. Krause and various colleagues (one of whom is Scott Sampson, a.k.a. “Dr. Scott the Paleontologist” on the PBS kids’ show Dinosaur Train) have been studying vertebrate fossils on Madagascar since 1993, working in Late Cretaceous strata of the Mahajanga Basin on the northwest coast of the island. There they have unearthed an extraordinary array of forms—theropod and sauropod dinosaurs, bizarre insectivorous and herbivorous crocodiles, sparrow-­sized and vulture-sized birds, marsupials and multituberculates (a diverse extinct group of rodent-like mammals), turtles, snakes, frogs, and freshwater fishes. Many of the specimens are wonderfully preserved; for instance, a skull of the short-snouted theropod Majungasaurus is one of the most complete dinosaur skulls ever found. The Mahajanga Basin has turned out to be one of the world’s great Mesozoic fossil sites.

  At the time these animals died, Madagascar had only recently detached from India. Not surprisingly, the fossil fauna is clearly Gondwanan. That Cretaceous fauna, however, has almost no connection to the island’s modern vertebrate fauna. With the possible exception of one of the turtles, none of the Cretaceous vertebrates are close relatives or potential ancestors of the modern species. In other words, during Madagascar’s long journey through time, virtually all of those old lineages have disappeared. This inference from the fossils agrees with Yoder and Nowak’s molecular divergence dates, which indicate that most of the sampled vertebrates arrived after the Cretaceous. For vertebrates, at least, the old Gondwanan fauna has been replaced by an almost entirely new set of creatures.

  The biota of Madagascar, like that of the whole world, was undoubtedly battered by the end-Cretaceous extinction. Presumably the dinosaurs, which of course are not known to have survived anywhere after the Cretaceous, and many other
lineages on Madagascar didn’t make it through that catastrophe. Unfortunately, after the Late Cretaceous, the known fossil record of the island is pretty much blank until it picks up again in the Late Pleistocene, about 26,000 years ago. Thus, it’s unclear to what extent the earlier biota disappeared in mass-extinction pulses, such as the end-Cretaceous event, versus more slowly, piece by piece over long spans of time. The important point, though, is that it did disappear.

  To sum up, a general message from Madagascar and the other Gondwanan islands is that the loss of many lineages by extinction, on the one hand, and the arrival of many others by overwater colonization, on the other, have both been pervasive in all of these areas since their last separation from the other landmasses of the supercontinent. Whatever species they may have harbored at the time of fragmentation, extinction and oceanic dispersal have since put a massive, defining stamp on their character.