Saturday, 26 March 2016

Dive of the Fossil Penguins

This is the post that will be current as of 1st April 2016. I will therefore be following my long-running annual tradition (that's "long-running" as in, "I've kept up this annual tradition for a whole year now") of posting about something that isn't a mammal, or even a synapsid.

When we think of 'birds' the ability to fly is surely one of the first things that come to mind. Indeed, the development of flight was surely one of - if not the - defining moments in early bird evolution. But, whereas bats have never given up that particular evolutionary innovation, flightless birds have evolved more than once. And, because we know that the very first birds could already fly, it follows that all flightless birds have ancestors that were perfectly capable of flight, but, for whatever reason, lost the ability.

Many, such as ostriches, did so to develop a fully land-based lifestyle, often becoming much larger in the process. But others swapped flight for diving ability. Of course, there are a great many diving birds that are still able to fly, but sometimes there is an advantage in really specialising in diving to the point that flight is no longer an option. This has, in fact, happened more than once in the course of avian evolution, but the only flightless diving birds we have now are the penguins, propelling themselves through the water with wings shaped like flippers.

(Perhaps the best known other example, since it only went extinct in 1852, is the great auk, which was related to puffins and guillemots).

Although the exact number is disputed, there are something like 20 species of penguin alive today. Furthermore, when it comes to their fossil history, penguins have the advantage of possessing solid, heavy, bones - at least in comparison to other birds. Bird bones are generally hollow and fragile, ideal if you need to be light enough to fly, but not great for leaving intact fossils behind. For penguins, at least, that's not an issue.

On the other hand, there's the small matter of where those fossils happen to be. It's not true to say that all penguins today live in Antarctica - one species even lives on the Galapagos Islands, which are literally on the equator. And we do, in fact, have a number of penguin fossils from South America, South Africa, and New Zealand, all places where penguins are still found today. Still, quite a lot of their evolutionary history does seem to have occurred around Antarctica (although there are competing theories), and that does present certain logistical issues when it comes to fossil hunting.

Nonetheless, Antarctic palaeontology is further advanced than you might think. One particularly rich fossil site is the La Meseta Formation, at the eastern tip of Seymour Island, off the coast of Graham Land on the Antarctic Peninsula. The rocks here are over 30 million years old, dating from the Eocene to the early Oligocene epochs, just before the ice sheets began to form. While daytime temperatures there today regularly reach 1°C (34°F) in the height of summer, things would have been rather more comfortable during the Eocene.

Indeed, the first fossil mammal ever found in Antarctica was uncovered on this island. It was described in 1982, since when a number of others, including some opossums, have been added to the list. Penguin remains are, however, rather more numerous, and at least seven species have been described from the site. While it should be stressed that the late Eocene is nowhere near as far back as the actual origin of the penguin group (which appears to date back to around 62 million years ago, not long after the demise of the dinosaurs), and these birds were certainly already flightless, the fossils can provide us some insights into how penguins evolved to fit their unusual lifestyle.

That's because almost all known fossils are of "stem-penguins". That is, they belong to early side-branches in penguin evolutionary history that left no living descendants, and that diverged before the appearance of the common ancestor of all living species. By some definitions they are "almost-but-not-quite-penguins", although that rather depends on how widely you want to cast the word "penguin" when referring to extinct species.

Most fossils found on Seymour Island have tended to be frustrating incomplete. The penguin species have been identified mostly on the basis of various wing bones, and, while the development of penguin flippers is undoubtedly an important part of their evolutionary story, it isn't the whole of it. If we want to understand how these early penguins interacted with the world, we also need their skulls, to get an idea of the shape and size of their brains, and of their sensory systems.

The first fossil penguin to include a skull wasn't described until 1946, and, while we do have a few fossil penguin skulls from Seymour Island, the scattered nature of the bones means that it isn't always to know which wings they go with - and therefore which species they are. But, even if we don't know exactly which species we're looking at, we can still perform CT scans on the skulls and see what we can learn. Last year, researchers published a study doing just that for three Seymour Island fossils and that 1946 specimen, Paraptenodytes antarcticus, which dates from around 20 million years ago and was discovered in Patagonia. To get some idea of how they fit into the overall picture, and what makes penguins so unique, they compared the results with those from similar examinations of six species of living penguin, and of other diving birds that have not lost the power of flight.

So what sort of differences might we expect to find? For one, it has been noted that, relative to their body size, flightless birds tend to have smaller brains than those that can fly. It is, however, unclear as to whether this is genuinely meaningful, and to what extent it is true of penguins as opposed to ostriches and their relatives. At any rate, there seems to be little difference in relative brain size between the fossil penguins and living ones, so if anything did happen there, it probably happened pretty early on.

But, if we can't learn much from the overall size of the brain, its shape and the relative size of different parts might be more informative. In birds, as in mammals, there is thought to be a correlation between the relative size of an animal's olfactory bulbs and the effectiveness of its sense of smell. Living penguins, like many birds, have a relatively poor sense of smell, relying more on vision to catch their prey, and likely finding smell of limited use when they spend so much of their time underwater. However, the same is not true of their closest living relatives, the petrels and albatrosses, which have, by avian standards, a particularly good sense of smell.

It turns out that one of the older, unnamed, fossils, had olfactory bulbs three times larger than those of equivalently sized living penguins, and not far off what we'd expect in a petrel. This implies that a good sense of smell may have arisen in the ancestor of both groups, being kept in the petrels, but lost in the later penguins. Quite what early penguins would have wanted an acute sense of smell for is less clear, but it seems plausible that they used it as modern petrels do, being able to identify patches of sea with high plankton growth that ought to also be rich in fish.

The shape of the semicircular canals, the part of the inner ear responsible for the sense of balance, was also unusual. The sense of balance, which relates to things like posture and positioning in 3-dimensional space is one that's clearly critical to any animal that has to fly and manoeuvre in the air. One would also expect that it's fairly important for animals that do a lot of swimming in the open ocean where there are few visible markers of one's orientation, and, indeed, the semicircular canals of penguins are broadly similar to those of other diving birds, from the closely related petrels to more distantly related animals such as grebes. But the fossil penguins resembled neither, with much broader canals than expected; it's unclear quite what this means, but it's interesting that it's something that early penguins seem to have evolved on their own, but then lost again when the modern sort arose.

The more obvious function of the inner ear is for hearing, and, again, the shape can give us some clues here. In this case, however, it turns out that the fossil penguins had a hearing range similar to that of their modern kin, which is fairly unremarkable. It would, however, be perfectly suitable for finding squawking young amidst a large and noisy colony, as modern penguins do today. Given the number of penguin bones at Seymour Island, it's entirely possible that even very early penguins lived in this manner.

Nobody has ever discovered a fossil penguin so primitive that it could still fly. Not only would that have lived a very long time ago indeed, given how old the oldest fossils we already have are, it probably had bones that were fragile enough not to preserve well. But we can see that whatever changes they needed to make to give up their command of their air in order to better catch fish, those changes were still ongoing 30 million years later, half-way to the present.

[Photo by Liam Quinn, from Wikimedia Commons. Cladogram adapted from Baker et al 2006.]


  1. But the fossil penguins resembled neither, with much broader canals than expected; it's unclear quite what this means, but it's interesting that it's something that early penguins seem to have evolved on their own, but then lost again when the modern sort arose.

    Are these weird semicircular canals known to bracket the crown group, or is it possible it's a synapomorphy of a wholly extinct branch?

    1. My understanding is that, while they probably bracket the crown group, since the fossil species concerned don't *appear* to belong to a single branch, the data is sufficiently incomplete that that can't be entirely ruled out.