Flight of the Fossil Pelicans

Dalmatian pelican

Pelicans are distinctive birds with long necks, short legs, and a remarkably long beak the bottom half of which is attached to a large extensible pouch for holding captured fish. They are also large, with even the smallest species having a 180 cm (6 foot) wingspan and the largest being half that again. They are, of course, water birds, with many living along coasts or in brackish waters, but others found in lakes and rivers far inland.

They are also not mammals, which is a timely reminder that, yes, this is the post that will be live on 1st April, when I switch things about for one post a year.

It's probably not surprising to discover that there aren't a great many species of pelican in the world today - although there's possibly more than you'd think. There are, in fact, eight recognised species (compared with, say, the 53 species of gull) and they're found on every continent except Antarctica. These eight species are all physically similar to one another, to the extent that they are all placed in a single genus, Pelecanus. Significantly, though, they are sufficiently distinctive that that genus gets a family all to itself, much as all living members of the horse family belong to the single genus Equus.

Depending on the status of the domesticated forms, there are 7-9 living species of horse, which is comparable to the pelicans, but a much larger array of known extinct species. The same cannot be said for the pelicans, since, like most birds, the fragile skeletons necessary for flight do not preserve well. It's also possible that they were never very diverse anyway and, indeed most fossil species that we have found still belong to that one genus, Pelecanus. Come to that, we can't even tell any Ice Age pelican fossils apart from the living species, implying an almost total absence of change in the last two million years or so.

All eight living species are, however, at least moderately common so it's relatively easy to obtain samples that we can do genetic testing on to see how they are related to one another. It should be immediately obvious, looking at the resulting cladogram shown below, that, even if they are all so similar as to be placed into a single genus, the genetic evidence shows that pelicans can be neatly divided into two sub-groups.

Is there any obvious difference between the birds in those two groups? Well, yes there is, and it's the sort of thing that supports the fact that this sort of genetic testing works: the three species I have placed at the bottom are only found in the Americas, and the others only in the Old World. 

Clearly, at some point, pelicans crossed between continents and, if we want to know which one they started from, the fossil record gives us a clear answer. Most fossil pelicans from the New World are indistinguishable from the living species, including an apparent American white pelican that slightly predates the Ice Ages. The one clear exception is Schreiber's pelican (P. schreiberi) which lived in Florida and South Carolina around 5 million years ago. Older fossils cannot be reliably placed in any named species, but date back over 10 million years... and are still in the same genus we have today.

This suggests that pelicans entered the Americas in the Late Miocene, with all living New World species, and all the extinct ones we know of, having descended from them. If that's true, of course, we should find a different pattern in the Old World.

And we do.

One fossil species from Ukraine, two from India, and a possible fourth species from Tanzania all postdate the oldest New World species and thus prove little beyond the fact that Afro-Eurasia had a greater pre-Ice Age diversity of pelicans than the Americas. Fossils from France, Germany, and Australia, however, are much older, with one French specimen dating back 30 million years to the Early Oligocene - and it was still so similar to what we have today that it's placed in Pelecanus. 

It's got to be a successful body form if it's changed that little in such a long period of time.

But, if pelicans first appeared in the Old World, and then migrated to the Americas, what route did they take? The traditional assumption is that, given the age of the French specimen, the genus started off in Europe, and then spread east and south, eventually travelling so far east that they crossed the Bering Strait from Siberia into Alaska. 

That may, however, not be the only explanation. I stated above that the only undisputed named species of fossil pelican in the Americas was Schreiber's pelican. This was a large bird, with even the females around the size of the largest pelicans of today, probably with a wingspan of around 3 metres (10 feet). Since male pelicans are larger than females, those of this lost species would have been impressive indeed.

However, the claim that it's the only confirmed fossil species in the Americas is no longer true. Last year, a new species, the Paraná pelican (P. paranensis) was described from deposits in southern Argentina dated to around 10 million years ago, during the Late Miocene. The authors of that study argue that this is too far south, too early for the northern route to be the most likely explanation for pelican migration to the New World. Moreover, it more closely resembles the brown and Peruvian pelicans - and, crucially, Schreiber's pelican - than it does the American white pelican. 

This suggests, they argue, that the latter species is an offshoot of a clade of pelicans that originated in the south, and then travelled north. The scenario that they propose is that some, probably unknown, species of pelican flew from Africa to South America around 12 million years ago and took up residence there. At the time, large portions of Argentina and Brazil were covered in shallow seas and brackish coastal wetlands, a consequence of generally higher sea levels and continental shifting with the rise of the Andes.

The newly arrived pelicans adapted to this habitat, which explains why brown and Peruvian pelicans, uniquely among their living kin, live only in coastal, saltwater habitats, and not inland. Schreiber's pelican, given where its fossils have been found, may well have done the same. The American white pelican, found inland from Mexico to Canada, would have divided off from the others at a very early date. 

While the researchers argue that this is a simpler explanation than a northern route, even if that's right, it's not proof of anything. It is, however, a valid alternative hypothesis, with its main weakness, perhaps, being that the South Atlantic was, even then, much wider than the Bering Strait... but that's not proof of anything, either, since these things have happened to other groups of animals.

Either way, it leaves open the question of where Pelecanus came from in the first place. To answer that, we have to turn back to the Old World. For it is here, and only here, that we find pelicans belonging to anything other than the modern genus. Although a few others have been proposed over the years, only two genera seem to have stood the test of time: Miopelecanus from France, Germany, and possibly Kenya, and the Egyptian Eopelecanus. This latter is the oldest known member of the pelican family, dating back 36 million years to the end of the Eocene epoch. 

This, by some estimates, is close in time to the very origin of the pelicans, when they diverged from their closest living relatives, the shoebill and hamerkop, but long after their common ancestor had diverged from the herons, 52 million years ago, 80% of the way back to the time of the dinosaurs.

[Photo by Charles J. Sharp, from Wikimedia Commons. Cladogram adapted from Kennedy et al. 2013.]