One of the strange things about primate evolution is the arrival of anthropoid monkeys in South America sometime during the Oligocene. South America was an island continent at the time, like Australia, filled with marsupials (even marsupial carnivores) and basal placental mammals like the sloths. And then, all of a sudden, primates showed up. It's the closest thing in biogeography to a spaceship landing and dropping off an alien race.
There is one inescapable conclusion: Thirty-five million years ago, a bunch of ancient monkeys got on a raft and sailed to South America. That's the date that comes from molecular comparisons (e.g., Schrago and Russo 2003). The earliest fossil monkeys in South America are of Late Oligocene age (Branisella boliviana), but they anatomically resemble Late Eocene monkeys from Africa (Takai et al. 2000), suggesting an earlier arrival.
A substantial debate arose over the source for these monkeys — did they arrive by raft from Africa, or from North America? Africa was farther away, but definitely had monkeys, and the South Atlantic was substantially narrower during the Late Eocene-Early Oligocene than today. North America was closer, but there is no clear evidence of anthropoids there until after the Isthmus of Panama arose 5 million years ago. So Africa seems more likely. The most likely (while still improbable) scenario is a large logjam of vegetation floating down one of the major African rivers and out to sea, with monkeys on top. They needed to survive for anywhere from a few weeks to a few months, without drinking too much sea water, to make it.
It would all seem to incredible to believe, except we know that it happened twice. Not only primates, but also caviomorph rodents got to South America during the Late Eocene or Early Oligocene (Poux et al. 2006). That includes capybaras, guinea pigs, chinchillas, and New World porcupines. Their radiation appears to have been earlier, based on earlier fossil diversification and an earlier molecular divergence date, although as yet not significantly so.
And now, a new report from Blair Hedges' lab says that frogs rafted, too (UPDATE: This link seems to be dead now, but may return; in the meantime, here's a news story). As usual, the press release precedes the actual appearance of the paper at PNAS, but it's a pretty good press release:
Nearly all of the 162 land-breeding frog species on Caribbean islands, including the coqui frogs of Puerto Rico, originated from a single frog species that rafted on a sea voyage from South America about 30-to-50-million years ago, according to DNA-sequence analyses led by a research group at Penn State, which will be published in the 12 June 2007 issue of the Proceedings of the National Academy of Sciences and posted in the journal's online early edition this week. Similarly, the scientists found that the Central American relatives of these Caribbean frogs also arose from a single species that arrived by raft from South America.
"This discovery is surprising because no previous theories of how the frogs arrived had predicted a single origin for Caribbean terrestrial frogs and because groups of close relatives rarely dominate the fauna of an entire continent or major geographic region," explained Penn State's Blair Hedges, the evolutionary biologist and professor of biology who directed the research. "Because land connections among continents have allowed land-dwelling animals to disperse freely over millions of years, the fauna of any one continent is usually a composite of many types of animals."
So rafts were heading north into the proto-Caribbean from South America at around the same time they were floating from Africa to South America.
The report talks quite a bit about how the Eocene-Oligocene date range rules out the theory that the frogs surfed on waves from the Chicxulub impact; I think that scenario seems sort of silly.
The original frogs that successfully colonized the Caribbean islands likely hitched a ride on floating mats of vegetation called flotsam, which is the method typically used by land animals to travel across salt water. "Some rafts of flotsam, if they are washed out of rivers during storms and caught in ocean currents, can be more than a mile across and could include plants that trap fresh water and insect food for frogs," Hedges said. It is not likely that the frog species dispersed simply by swimming because frogs dry easily and are not very tolerant of salt water.
The frogs are kind of a limiting case; more vulnerable to dehydration and heat than the monkeys or rodents would have been.
The thing is that these raftings need not have been rare or exceptional events; the only thing exceptional is when one happens to carry a species primed for rapid population growth and diversification. The species has to be small enough to survive for a long time on the sea with the available resources, and generalized enough to succeed on whatever resources it finds. It's the "bio" part of the biogeography that is the limiting factor -- for instance, any monkeys that crossed the Atlantic during the Miocene would have found a bunch of well-adapted platyrrhine competitors already in place. The same would be true for back-crossing to Africa. So there should be a bias toward relatively early events -- if a lineage was capable of establishing successfully by this rafting mechanism, it would have done so relatively early rather than late.
From that perspective, the interesting thing about the recurrent Late Eocene-Early Oligocene time frame is that the dispersals weren't happening much earlier. For the frogs, the answer may be habitat-related: maybe the Caribbean islands weren't great frog habitat until that time. For primates, the dispersal seems to have followed fairly soon after the appearance of anthropoids in Africa, so they dispersed to South America perhaps as early as they could have.
Poux C, Chevret P, Huchon D, de Jong WW, Douzery EJP. 2006. Arrival and diversification of caviomorph rodents and platyrrhine primates in South America. Syst Biol 55:228-244. doi:10.1080/10635150500481390
Schrago CG, Russo CAM. 2003. Timing the origin of New World monkeys. Mol Biol Evol 20:1620-1625. doi:10.1093/molbev/msg172
Takai M, Anaya F, Shigehara N, Setoguchi T. 2000. New fossil materials of the earliest New World monkey, Branisella boliviana. Am J Phys Anthropol 111:263-281.