Sea Otters, Hunters, and Steller's Sea Cows - Replaying a Recent Extinction


The nearly complete skeleton of a Steller's sea cow (Hydrodamalis gigas) - it is missing bones from the wrist and hand. From Woodward, 1885.

It did not take long for the last remaining population of Steller's sea cow to be driven into extinction. Discovered by the German naturalist Georg Steller around the Bering Sea's Commander Islands in 1741, this enormous and peculiar sirenian became an easy target for Russian hunters. By 1768, it was gone. (The marine mammal would not be scientifically described until 1780, and today it is formally known as Hydrodamalis gigas.) Yet, despite the clear role of hunting in the disappearance of this species, some scientists have suggested that an ecological cascade triggered by the hunting of another species also helped tip the massive dugongs into extinction, and so scientists S.T. Turvey and C.L. Risley have dug back into earlier research about the sea cows to find out which factors were most important in their demise.

Figuring out why a species became extinct is often a tricky business, especially in distinguishing direct and indirect causes. In the case of Steller's sea cow, it is hypothesized that they were easy targets for prehistoric hunters and extirpated from most of their range before Russian sailors began to lay into them. Hence hunting by humans obviously played an important, direct role in the extinction of Steller's sea cow, but decreases among sea otter populations may have played an indirect role. Sea otters have long been highly prized for their thick pelts, and as they were killed (by both aboriginal people and Russian fur hunters) populations of sea urchins - among their most regular prey - exploded. The inflated sea urchin populations then fed on shallow-water kelp unhindered, and since Steller's sea cow relied upon kelp for sustenance their decline may have been at least partly caused by these changes. The extinction of Steller's sea cow would have been caused by a one-two punch of ecological destabilization and overhunting, both of which were attributable to human activities.

It is extremely difficult to test whether it was hunting alone or hunting plus ecological disturbance which led to the extinction of Steller's sea cow. Most of what we know about the natural history of the animal comes from the notes of Georg Steller, and he did not keep exact counts of how many sea cows there were when he discovered them. Likewise, hunters did not keep precise details of how many sea cows they killed - we only have rough estimates of how much meat a sea cow yielded and how long that might last a crew. On the basis of these notes the 19th century naturalist Leonhard Stejneger tried to come up with rough estimates of how many sea cows were killed by each crew each year between 1743 and 1762, inflating the number to account for those which must have been killed but not eaten, and he presumed that the starting number of the population was less than 1,500 individual animals.

Although Stejneger's back-of-the-envelope calculations were made over a century ago and have not been verified, Turvey and Risley combined them with what is known about the life history of living dugongs in an attempt to find out if hunting alone could account for the extinction of the sea cows. They estimated how sea cow populations would have been affected by different hunting models (i.e. non-wasteful hunting or hunting for immediate consumption only) and tried to figure out what a sustainable take of Steller's sea cows would have been. As it turned out, the sea cows could have only sustained a very light level of hunting. Assuming Stejneger's hypothesized starting point of 1,500 individuals, a sustainable harvest of the slow-reproducing mammals would have been only 17 individuals a year. This is far, far below the estimated yearly average take - about 123 individuals - and so it is little wonder that Steller's sea cow was wiped out so quickly.

Additionally, the authors found that Stejneger's estimate for the starting sea cow population was probably too low. If hunters killed sea cows at the rate he proposed, they probably would have only lasted until 1756. Given that we know hunters stuffed their holds with sea cow meat and probably killed far more than they actually needed, it is more likely that the starting population was around 2,900 animals. This was still a small number - the remnants of a once more widespread species - but it attests to the extreme amount of animals killed but left to rot in the cold sea.


Estimated Steller's sea cow kills between 1743 and 1762 as proposed by Stejneger. In each bar, the white sections with slashes represent animals killed and consumed immediately, grey sections represent animals killed for immediately consumption but wasted, white sections represent animals killed for provisions, and the black sections represent animals killed for provisions but wasted. From Turvey and Risley, 2010.

Whether these estimates accurately represent what occurred and the broader implications that might have is another matter. The new estimates are directly based off of old estimates made by Stejneger which were based upon the limited information he was able to gather about the crews working around the Commander Islands. There is no doubt that there were more crews there than he could account for - he recognized himself that the records he had to work with were "very defective" in this regard - and this makes it all the more difficult to come up with an accurate understanding of the last days of the sea cows. Turvey and Risley's new estimates are certainly interesting, but as estimates based upon estimates without any way to verify that these numbers are correct, it is hard to know if they were right. Indeed, tracking the decline of the sea cows depends upon knowing how many there were to start with and the rate at which they were being killed, both of which are based upon assumptions which can't be verified.

Despite my reservations about the data used to come up with the new estimates, however, there is no reason to doubt that hunting played a major role in the extinction of Steller's sea cow. From studies of their living relatives and other large mammals we can safely hypothesize that they would have reproduced slowly, and if hunters were killing numerous individuals to provision their ships (with animals which were killed but never eaten added to this number) it is reasonable that hunting was the chief culprit behind the extinction. As the authors state, the proposed ecological effects of the sea otter decline/sea urchin rise may have still occurred, but its influence on populations of sea cows between 1741 and 1768 appears to have been minimal compared to hunting. It may be more profitable to explore this hypothesis for the time period before the 1740's, during the time when sea cow populations were disappearing, but such details are not considered in the new paper.

The rapid extermination of such large mammals by small groups of humans using unsophisticated weapons (i.e. harpoons) leads Turvey and Risley to suggest that the human-caused extinction of the sea cow throws credence to the controversial "overkill hypothesis" for the extinction of the Pleistocene megafauna. If, like the sea cows, large Pleistocene mammals became restricted to small populations then they could be quickly wiped out by small bands of people. This is certainly plausible, but the picture may be too narrow. Even if humans delivered the ultimate coup de grace to some megafaunal species, we must ask why those species became restricted to refugia or scattered in smaller populations in the first place. Ecological changes driven by climate change may have made some species more extinction prone, and the ecological consequences of the removal of large species - which often play a role in shaping local ecology - are just beginning to be understood. The caricature of "humans show up, large animals go extinct" does not work and masks more complex interactions which we do not yet fully understand (such as the extinction of species which show no sign of being hunted by humans and the changes in populations of small mammals). Arguing over single causes to explain the whole of the last extinction may be fertile fodder for pop sci articles and documentaries, but, in the case of the end-Pleistocene mass extinction, it seems that there were an intertwining array of factors which varied from continent to continent, from the dispersal of humans to the development of a warmer, wetter global climate. Perhaps, as some of the intense rhetoric surrounding the issue dies down, more nuanced approaches will help us better understand why some of the last great mammals disappeared..

Turvey, S., & Risley, C. (2006). Modelling the extinction of Steller's sea cow Biology Letters, 2 (1), 94-97 DOI: 10.1098/rsbl.2005.0415


More like this

The interconnectedness of ecosystems and their components is, today, a familiar concept. Top predators eat herbivores, herbivores eat plants, and top predators keep so-called meso-predators in check too. But perhaps it isn't appreciated enough just how interconnected things can be. Cristina…
A muskox (Ovibos moschatus), photographed in Alaska. From Flickr user drurydrama. Of all the mass extinctions that have occurred during earth's history, among the most hotly debated is the one which wiped out mammoths, saber-toothed cats, giant ground sloths, and the other peculiar members of the…
A golden-mantled ground squirrel (Spermophilus lateralis), photographed in Bryce Canyon National Park, Utah. Though abundant at the Samwell Cave Popcorn Dome, California site during the Late Pleistocene, its numbers in the area decline at the beginning of the present Holocene epoch."One of the…
One of Charles R. Knight's wonderful paintings of woolly mammoths walking through the snow of ancient Europe. On display at the Field Museum in Chicago. When did the last woolly mammoths die? There is no easy answer to the question. In its heyday the woolly mammoth (Mammuthus primigenius) was…

Humans again could be the answer for animals heading to the hills. I imagine megafauna would have large territorial demands. Megafauna that stay where the humans are get killed, so they learn to avoid humans, restricting their territory. Population adjusts to fit the new, limited territory. Said territory is invaded by humans, and ...

But I'm basing this idea on the way modern humans have caused extinctions (ecologists are always talking about species threatened by habitat loss; suburban sprawl continues unabated), which is bound to be different because our population is much larger. If early humans weren't as intimidating as we are now, there's no reason for retreat and back to square one.

By CS Shelton (not verified) on 24 Jun 2010 #permalink

Anyone look into what effect the evolution of commensal specis would have had on even the small animal populations?

Seriously, our non-human camp followers such as dogs, semi-domesticated goats, precursors to house mice & rats and our non-species specific parasites must have contributed to the cascading effects that humans caused.

A few years back there was news from the North pacific about behavioral plasticity in Orcas: in response to anthropogenic environmental change: a reduction in some other (I forget which) prey led to increased Orca predation on ... sea otters.

So, perhaps there's yet another, though probably minor, causal route between human hunting and the extinction of Hydrodamalis: fewer sea-otters (and whatever else the new human groups took) leading to increased willingness of Orcas to attack sea cows. (As in "Yeah, I know they're not very appetizing and have tough skin and live in dangerously shallow water, but I'm hungry.")

By Allen Hazen (not verified) on 24 Jun 2010 #permalink

Brian, do you have more info about megafauna that went extinct but show no signs of being hunted by humans? I have read some pretty convincing arguments in favor of the overkill hypothesis, but I'm always open to more info.

I agree that "humans show up, large animals go extinct" might be a bit of an oversimplification, but it still seems (to me, anyway) the most likely. Even the example you cite above is overkill extinction, it's just a more nuanced view.

Matt - You might want to check out the recent review by Koch and Baronsky "Late Quaternary Extinctions: State of the Debate" - - especially Table 3, and also "A requiem for North American overkill" by Grayson and Meltzer - To sum things up briefly here, most of the identified "kill sites" in North America contain mammoth remains (12), with two mastodon sites, two Equus sites, and one Camelops site. As far as large, extinct North American mammals are concerned, there have been no confirmed signs that humans hunted Arctodus ("short-faced" bear), Capromeryx (pronghorn), Glossotherium (ground sloth), Hemiauchenia (llama), Holmensina (armadillo), Megalonyx (ground sloth), Platygonus (peccary), Smilodon (saber-toothed cat), or Tapirus (tapir). Granted, there are still areas which have either not been sampled or are unlikely to contain archaeological assemblages bearing on this question, and there is the matter of how likely it is that signs of hunting are going to make it into the fossil/archaeological record.

As you can probably tell from my concluding paragraph - as well as previous posts - I have reservations about the overkill hypothesis, or at least the idea that humans killing off "naive" prey was the sole, primary factor behind the Pleistocene extinction. From what I have read, it appears that extinction triggers differed from continent to continent, and the question is not "Was it climate or humans?" but "How important was climate or human presence to the extinction of a given species?" I don't think either hypothesis alone can provide a blanket explanation, and I don't think the overkill hypothesis as originally proposed by Martin is very helpful at this point. (And all of this is to say nothing about how recognition of modern ecological destruction by our species may influence our ideas about what happened at the end of the Pleistocene.) To me, at least, it doesn't make sense to prefer overkill or climate change to explain the global pattern for all species - there is still a lot we don't know about individual species, how those species may have shaped local ecology, how humans interacted with those species, and the precise details of when and when species disappeared. To come to any conclusion now would be, I think, premature, and I feel that if we are going to understand this extinction we need to get past the climate change versus overkill rhetoric which has often dominated the conversation since the 1980's.

Rob - Check out the review by Koch and Baronsky (i.e. Table 1). I know the influence of "pest" species has been considered before, but not to the same degree as direct hunting.

With a population of 1500-3000, Hydrodamalis chances of long-term survival would presumably have been fairly slim even if the Russians hadn't turned up. Is it known when they were reduced to such small numbers? Is it plausible they were already dying out, and that Russian hunting merely sped the process?

By Andreas Johansson (not verified) on 25 Jun 2010 #permalink

Andreas - The authors conclude the paper by saying "However, sea cows had such a low maximum population growth rate and sustained yield that the species would unfortunately have become extinct even under a careful and conservative hunting regime." If the estimates of the authors were correct - which, as I mentioned above, I have my doubts about - then the population was pretty close to the tipping point around 1741. By that time the only existed in a last refuge around the Commander Islands and apparently had undergone a major range reduction due to hunting by aboriginal people. The details of this are still a bit fuzzy, but it appears that the Russian hunters finished off a small, remnant population rather than wiping out a species which was otherwise doing well.

If you look at the big picture, during the Pliocene and Pleistocene Hydrodamalis ranged down into California and Japan; the Pliocene range in the NE Pacific extended all the way down to Baja California, while during the Pleistocene (at least till ~300 Ka) Hydrodamalis is known from Monterey Bay (Central California). Between the Pliocene and the 18th century, Hydrodamalis had already suffered a major range reduction.

Based on records of Holocene Hydrodamalis bones, Daryl Domning and a couple others published a paper in Marine Mammal Science suggesting that by the Holocene, Hydrodamalis was already confined to the Arctic (i.e. Aleutians, Komandorskye Islands, etc.). I don't think that sea cow bones have yet turned up in any middens along the west coast of North America.

There are multiple reasons why human hunters present such attractive options as the primary or even exclusive drivers of the Pleistocene extinction event. These include the frequency through time of waves of extinctions following human colonization of new lands (esp. islands), the apparent inadequacy of climate change alone as a causal factor, and of course the pertinence of such arguments to human activities relative to large-scale extinctions today. But these reasons don't really hold up under much scrutiny. Analogies with island extinctions, for example, are probably overblown; island faunas can often consist of relatively "naive" animals with few defensive mechanisms, but it's a stretch to argue that since this is the case, continental faunas (including *numerous* large predators that islands lack) would have to have been similarly naive. And arguments framing the Pleistocene extinction as some sort of moral lesson for our actions today (humans are bad, look at all those Ice Age animals we killed, we have to be more environmentally sensitive and aware today in order to avoid similar results) are legion, but that doesn't mean the science underlying them is fully supported.

The key argument in favor of overkill seems not to be the strength of the hypothesis itself, but rather the perceived weakness of other scenarios, particularly the climate change hypothesis. Because the megafauna survived numerous previous Pleistocene glacial-interglacial transitions without any major extinction pulses, so the argument goes, climate cannot have been responsible for the terminal Pleistocene extinction. And because of this perception, many of the weaknesses inherent in the overkill concept are largely ignored. See the rebuttal to the Grayson and Meltzer "Requiem" paper by Fiedel and Haynes (doi:10.1016/j.jas.2003.06.004), for example, and you'll see that about as much time is spent by those authors in criticizing climate change as in arguing for overkill.

What's been missing is some consideration of the other side of the climate-megafauna equation: the animals themselves. In North America, for example, large mammal faunas were changing rather markedly towards the end of the Pleistocene, with bison entering the continent ~240 ka or thereabouts and then becoming abundant only towards the very end of the epoch. As I argued in Quaternary International back in April (doi:10.1016/j.quaint.2009.11.003), the addition and increasing abundance of these animals may have augmented competitive pressures on other megafauna during periods of climatic stress, which could have helped drive the extinction pulse, or at least contributed to it. But determining this will take a lot of "nuts and bolts" paleontology -- species distributions, relative MNIs of taxa across time and space, patterns of competition for resources, etc. -- yet such studies often seem relatively unsexy if not pointless where climate change is rejected as a causal factor out of hand and overkill "wins" the debate by default.

I agree that coming to any firm conclusions about this topic now is premature. I also agree that seeking single, blanket explanations for the extinction is not necessarily warranted. I'm hopeful that more detailed analyses of large mammal faunas -- more and better numbers, more and better dates, new data from isotopes, consideration of taphonomic factors, and so forth -- will better inform the debate. But we need to move away from the "either/or" mindset that has characterized much of this debate.

By Eric Scott (not verified) on 25 Jun 2010 #permalink

Thanks for the follow up. Actually, I guess the abstract for "Late Quaternary Extinctions: State of the Debate" sums up my position quite nicely. It does say "Results from recent studies suggest that humans precipitated extinction in many parts of the globe..."

That second link was broken, but I would like to read through it if you can find it.

I agree that humans show up and kill all the helpless naive animals is incorrect, but I certainly think that human impact on an ecosystem was the deciding factor in most or all of the pleistocene extinctions.

Have you read Tim Flannery's take on part of this debate - The Future Eaters? I thought it was a pretty good treatment of parts of the issue.

BTW, congrats on the book reviews! You have one of the few blogs where I enjoy and read pretty much every entry. Keep up the good work.

it is missing bones from the wrist and hand

If I recall correctly, in life Hydrodamalis gigas is though to have lacked phalanges entirely, its arm ending in just a pair of wrist bones.

Steller was really specific about the manual morphology of Hydrodamalis, which is discussed further in Domning (1978). Hydrodamalis should have had metacarpals, at the very least (IIRC). Also, most of the known skeletons are composites scavenged from bones that were sitting out on the island surface long after they had been hunted, so it isn't too surprising that some elements are missing.

Hydrodamalis cuestae, H. gigas as well as their presumed ancestor, Dusisiren dewana, all lacked phalanges (although I'm not sure about the metacarpals). Other morphological details of the forelimb of H. gigas agree with Steller's description of the way it used them for "walking", and probably with the lack of phalanges.