12.2 A hint of the great diversity of caviomorph rodents. From top to bottom: Brazilian porcupine (Coendou prehensilis), plains viscacha (Lagostomus maximus), capybara (Hydrochoerus hydrochaeris). Drawings by Gustav Mützel.
Numbers cannot tell the whole story of biological influence, but here are some that must indicate a massive impact: In South America today, there are 124 described species of monkeys, 219 species of caviomorphs, and 330 species of sigmodontines, a total of 673 species. Together, these three groups make up 73 percent of all living, native South American mammal species, excluding bats and purely aquatic species. In terms of individual animals the percentage is undoubtedly much higher, because the caviomorphs and sigmodontines include many of the most abundant species on the continent. These numbers, in turn, imply countless others—all the fruits and leaves and insects eaten by monkeys; the seeds destroyed or dispersed by agoutis; the square miles of ground undercut by the burrows of tuco tucos;63 the furry meals for various species of weasels, cats, dogs, snakes, hawks, and other predators; and the monkey and rodent bodies scavenged by birds and mammals and decomposed by various invertebrates, bacteria, and fungi.
It is hard to say exactly what effects these and other ecological interactions have had on the evolutionary history of the South American biota, but those effects must be vast. A very clear impact comes from the observation that the monkey, caviomorph, and sigmodontine radiations essentially encompass the evolution of many other species, namely, parasites that depend on these vertebrates as hosts. As an especially striking example, consider a group of nematodes, related to human pinworms, that are intestinal parasites of New World monkeys (and presumably give monkeys itchy butts just as pinworms do in humans). It turns out that the evolutionary tree of these nematodes largely mirrors the evolutionary tree of the monkeys; nematode lineages tend to branch in concert with the branching of their monkey host lineages. That pattern of one tree matching another indicates that the evolutionary history of the parasites has been dictated by the history of the hosts; to a degree, the parasite acts as if it were simply part of the host. For our purposes, what this phenomenon of trees mirroring trees means is that, if there were no monkeys in the New World, the entire associated nematode tree also would not exist. More generally, even for parasites that do not evolve in such lock-step with the species they infect, the hosts are still critical to the evolution of the parasites. Thus, any South American parasite that requires a monkey, a caviomorph, or a sigmodontine as a host—and there are many such parasites, from protists and fungi to tapeworms, chewing lice, and botflies—owes its existence to those host lineages and, therefore, to the fact that the ancestors of these mammals were able to colonize the continent in the first place. In addition, many of these parasites have complex life histories that require a mammal host for part of the life cycle and another, completely unrelated animal for another part of the cycle. And this means that the influence of the mammals expands outward through the parasites to other kinds of animals, all part of the cascade of ecological effects.
The cases of monkeys and rodents are striking, but the importance of overwater dispersal in the history of the South American biota is not just about land mammals. Similar histories of colonization and subsequent radiation—a single branch becoming a sizable evolutionary tree—apply to other vertebrate taxa as well (see Figure 12.3). From my own somewhat serpent-centric point of view, an especially memorable example involves a group of snakes called the Xenodontinae, whose common ancestor apparently rafted from North America to South America sometime between about 12 and 28 million years ago, when those two continents were still separated by an expanse of ocean hundreds of miles wide. That single branch persisted and eventually gave rise to a great diversity of forms, including terrestrial and arboreal species that kill prey using a combination of constriction and envenomation; small snakes that specialize in eating reptile eggs; large species, called false water cobras, that search for fish by probing with their tails in aquatic vegetation; and hog-nosed snakes with upturned snouts and with color patterns that mimic the warning coloration of highly venomous coral snakes. When one thinks of South America, the snakes that come to mind are boa constrictors and anacondas, lineages that have been on the continent since it broke away from the other fragments of Gondwana. However, the Xenodontinae, with nearly three hundred living South American species, is actually the largest evolutionary radiation of snakes on the continent, much larger than any group of Gondwanan snakes.
12.3 More single branches leading to substantial trees: some nonmammalian vertebrate groups that reached South America by overwater dispersal, as indicated by molecular timetrees. In the list below, if only one reference is given for a taxon, that reference includes the molecular dating analysis and a tally of the number of species in the group. If two references are given, the first is the molecular dating study and the second provides the number of species. Rhinella toads, Pramuk et al. (2008), Frost (2011); phyllodactylid geckos, Gamble et al. (2011); Mabuya skinks, Whiting et al. (2006); Amphisbaenidae, Vidal et al. (2008), Gans (2005); Turdus thrushes, Voelker et al. (2009). Pramuk et al. (2008) did not conclude that Rhinella reached South America by oceanic dispersal; however, the authors estimated that its separation from North American relatives took place between about 30 million and 50 million years ago, which would indicate overwater colonization. For references on the xenodontine snakes, see the endnotes for Chapter Twelve. Photo of Rhinella alata by Brian Gratwicke.
The xenodontines also are instructive in that they make up part of an ecological intersection of colonizing lineages in the New World. Specifically, many of these snakes eat lizards, small mammals, and/or toads (the hog-nosed coral-snake mimics are toad specialists), and this must mean that often a snake descended from an overwater colonist ends up eating another vertebrate that is also descended from an overwater colonist. That kind of colonist-meets-colonist interaction underscores the depth of the influence of oceanic dispersal on the South American biota. And if we add plants to the discussion, such encounters become truly rampant. Although the geographic origins of most South American plants have not been worked out, there are certainly hundreds, and probably thousands, of species derived from overseas immigrants, and these plants must have ecological connections with tens of thousands of other species, including many of the vertebrate overseas colonists. For instance, sigmodontine rodents of the genus Oecomys eat the fruits of Renealmia alpinia (Zingiberaceae), part of a group of some sixty species of gingers whose common ancestor came from Africa in the Miocene or Pliocene; at least five kinds of South American monkeys feed on the flowers of the rainforest tree Symphonia globulifera (Clusiaceae), a recent immigrant from Africa; and Turdus thrushes eat the fruits (and presumably disperse the seeds) of Ocotea (Lauraceae) and Miconia (Melastomataceae), plant lineages descended from species that arrived over water from North America within the past 40 million years.
All of this is to say that there is nothing subtle about the effect of natural, overwater colonization on the history of life in South America. It is clear that such colonists have given rise to a large part of the continent’s biota and that this must have entailed an enormous cascade of impacts on other species. Ripples, domino effects, the fire that transforms the landscape, all of those metaphors and others apply. The cumulative effects are hard to even imagine, but it is safe to say that, without those colonists, the modern biota would feel very alien, both in what it would have and what it would lack. And if we can extrapolate from South America to other regions, it must be that the living history of the entire planet has been deeply influenced by ocean crossings and other long-distance colonizations. That is an inescapable conclusion, a tangible and general message, from the recent flood of biogeographic case histories.
PIKAIA AND THE NATURE OF DEEP HISTORY
If we accept that chance colonizations have had clear and profound effects, as it seems we must, what larger meaning can we take from this? What does this conclu
sion say about the nature of the history of life in general? What does it tell us, on a higher, conceptual level, about how the world came to be the way it is?
To answer these questions, we need to think about the linked ideas of contingency and unpredictability as they apply to history. In this context, contingency refers to the property of events being dependent on prior events: D happened because C happened; C happened because B happened; and B happened because A happened. To put it in the form of the usual sort of historical example: in 1972, South Dakota senator George McGovern ran for president on an antiwar platform (event D), in response to the continuing Vietnam War (C), which was itself caused by (among other things) escalating fear within the United States of Soviet communist influence (B), which was a result of (among other things) the occupation of much of Eastern Europe by the Soviet Union during the Second World War (A). (Incidentally, for this particular case, if you keep going back in time, you’ll eventually reach Spanish ships bringing potatoes across the Atlantic, leading to the rise of Russia.) Eliminate A or B or C, and D never happens; thus, D is contingent upon A and B and C.
The observation that history has this property of contingency is a truism (how could it not have this property?). Similarly, at least for human history, it is uncontroversial that at least some events in any sequence are unpredictable, which means that the course of history in general must be unpredictable; destiny might pivot on the mood of a head of state at some critical juncture (do the negotiations continue or do we launch the missiles?), or even on whether a child runs into the street at exactly 2:13 on a Tuesday afternoon, causing a car to swerve and crash into a telephone pole, making the driver miss an engagement where she would have met her future husband, etc., etc., etc. That sort of unpredictability is the basis for an entire cottage industry of “what if?” novels, short stories, TV episodes, and movies that depend on the notion that things might have turned out very differently (for Peggy Sue or Jean-Luc Picard or the entire human race) given some small and entirely plausible alteration of events. Even among many academic historians, such “what if?” scenarios are considered legitimate exercises, an acknowledgment that there truly is very little separating what actually has happened from an infinite number of radically different alternative histories.
This notion that apparently minor events can cause profound changes down the line, although widely accepted for human history, has not been so obvious for evolution, particularly when it comes to the large-scale history of life. Could some seemingly insignificant event hundreds of millions of years ago—the survival or demise of a few individuals of a single species, for instance—lead through long chains of cause and effect to great evolutionary success for some groups and the extinction of other groups, that is, to massive consequences for the overall form of biological diversity? Or, alternatively, is that large-scale history more like water running through deep, narrow channels, such that minor events have little effect on the long-term outcome, analogous to pebbles that fall into those channels without diverting the flow?
The late paleontologist and evolutionary theorist Stephen Jay Gould argued vehemently for the former, particularly in Wonderful Life, his 1989 book on the Cambrian fossils of the famous Burgess Shale Formation in the Canadian Rockies. That book still stands as the most detailed and widely read explication of contingency and unpredictability in evolution, so it is worth considering what Gould said and, in particular, how he reached the conclusions he did.
Life’s history, Gould claimed, is dominated by events that have produced massive effects that could not have been anticipated. Importantly, many of these events would have appeared small in magnitude and thus insignificant when they happened. Thus, evolutionary history on the grand scale is not only contingent, but thoroughly unpredictable, because of its sensitivity to tiny perturbations. Replay the tape of life over and over, Gould suggested, “altered by an apparently insignificant jot or tittle at the outset,” and the result would be entirely different each time.
In Wonderful Life, Gould used the Burgess Shale fossils, representing the rapid early diversification of animals known as the Cambrian Explosion, as his centerpiece example of the contingent and unpredictable nature of history. He followed Harry Whittington, Simon Conway Morris, and Derek Briggs—experts on the Burgess specimens—in suggesting that the Cambrian Explosion gave rise to a great array of very distinct anatomical types, most of which disappeared shortly thereafter. According to Gould, the decimation of most of these early forms dictated the course of all subsequent animal evolution or, to put it another way, the history of animals was contingent on exactly which forms survived that early winnowing.
Gould then went on to argue that the set of surviving lineages was unpredictable. For instance, he asks us to consider a fairly unassuming Burgess fossil called Pikaia, once considered a polychaete worm, but recognized by Conway Morris and subsequent researchers as an early member of our own phylum, the chordates. Gould notes that there was nothing special about Pikaia to indicate that it would be one of the survivors of the early decimation. Replay the tape of life again, he suggests, making some slight change in initial conditions, and perhaps Pikaia is one of the losers this time around. And if Pikaia was a direct ancestor of all subsequent chordates, then a substantial portion of the tree of life, including ourselves and all other vertebrates, would never come into being in the replay.
My gut reaction on reading Wonderful Life years ago was that Gould was probably right. Perhaps I was thinking in human terms—how the course of a life could depend on that child running into the street or a thousand other events that, if the tape were replayed, might easily turn out differently. However, at the same time I was dissatisfied with Gould’s examples of the unpredictable. So if Pikaia went extinct without producing any descendants, then the course of life on Earth would be radically changed. That I could easily accept (having no religious or other beliefs to dictate to me that humans are either necessary or inevitable). But what kind of perturbation, what “apparently insignificant jot or tittle at the outset,” would suffice to make it so? Would an underwater avalanche that buried a few of these worm-like chordates push the species toward an early exit from history, or would some much larger deviation be necessary? Perhaps there was, in fact, something special about Pikaia, some working of its muscles or gills or rudimentary brain that earmarked it for survival, even if, 500 million years later, nothing of the kind is apparent to us. (After all, most of what went on 500 million years ago is not apparent to us.) If that were the case, maybe Pikaia was destined to make it through the period of decimation in any replay of the tape.
The point here is not to suggest that Gould was wrong about the unpredictability of life’s history. The problem, as I see it, was in how he reached this conclusion. Gould claimed that the early decimation of most of the incredible early diversity of animal life, on the one hand, and the survival of a few lineages, including Pikaia, on the other hand, is a supreme example of the contingency and unpredictability of history, but actually it isn’t, or, at least, it’s not a convincing one. We really have no idea why some lineages survived the decimation and others did not, so it doesn’t follow from this case that the set of survivors was unpredictable and that a hypothetical replay of the tape would produce an entirely different outcome.64 Gould brought up many other cases, too, from the extinction of the bizarre set of early multicellular organisms known as the Ediacaran fauna to the origins of Homo sapiens, but they were similarly unsatisfying. Wonderful Life is in many ways a brilliant book—it’s deep, erudite, and entertaining, all trademarks of Gould’s writing in general. But his argument in this instance could have used some better examples.
ARCHETYPES OF THE UNEXPECTED
Biogeography, especially now, with the evidence from molecular timetrees, can provide those better examples of the unpredictable. In fact, if I’m even close to correct about the frequency of chance ocean crossings and other long-distance colonizations, we are up to our e
yeballs in cases that Gould could have used in Wonderful Life. The transoceanic journeys of monkeys, rodents, lizards, crocodiles, sundews, araucarian conifers, southern beeches, and countless others were all rare, chance events, the kinds of events that, for any particular time and place, one would not expect to happen. Replay the tape of life again, altering some “apparently insignificant jot or tittle,” and none of those colonizations occur (but others, absent from the history we know, do take place).
For instance, consider a reasonable scenario for monkeys colonizing South America from Africa. Start with a group of African monkeys lounging in a tree by a major river, a river similar to the modern-day Congo or Niger. Heavy rains have been falling for days, and the resulting floodwaters undercut a large chunk of the bank, including the monkeys’ tree, and the whole piece falls into the water and is swept away. This natural raft is carried many miles downriver, eventually reaching the ocean, where it gets caught up in a westward-flowing current. On the ocean, the monkeys eat anything edible they can find on their small, floating island. When it rains, they drink the water that briefly pools up. Weeks later, the raft, now waterlogged and barely buoyant, is grounded on the South American coast, and the few surviving monkeys, scrawny and dehydrated, splash onto the beach and disappear into the adjacent forest. They stay together, feeding on unfamiliar but edible fruits and insects, eventually mating, giving birth, and raising young. A thousand years later, their descendants form a substantial population that will ultimately give rise to the entire radiation of New World monkeys.
The Monkey's Voyage Page 35
