Skip to main content
U.S. flag

An official website of the United States government

New model gives insight to the potential future of the Pacific walrus

Right-click and save to download

Detailed Description

Walruses are important to human communities bordering the Chukchi and Bering seas in the United States and Russia, and the status of walrus provides information about the health of these highly productive marine ecosystems. Projecting the future population status of the Pacific walrus was investigated with a new model developed by scientists at the USGS Alaska Science Center.

The Bayesian network model integrates the potential effects of changing environmental conditions and human stressors to help identify the reasons associated with declines in projected walrus populations. Sea ice habitat, particularly in summer/fall, and harvest levels had the greatest influence on future population outcomes. The Bayesian network model for walrus provides the framework for an increased research effort on the Pacific walrus and its marine ecosystem, as part of the Changing Arctic Ecosystems initiative.

The purpose of this initiative is to understand how changes in the ice-dominated ecosystems of the Arctic affect biological communities. A report detailing this model and its findings are available in the journal Polar Biology.


Public Domain.


[Intro Music]

Paul Laustsen: Hello. This is Paul Laustsen with the U.S. Geological Survey, speaking with Dr. Chad Jay, a research ecologist specializing in walrus population ecology with the USGS Alaska Science Center.

Thanks for joining me today, Dr. Jay.

Chadwick Jay: Thanks for having me.

Paul Laustsen: You have a paper that is being published in the journal of Polar Biology about a new scientific model that projects the population of the Pacific walrus throughout the rest of the century. Why is it important to know what the walrus population might be in the year 2095?

Chadwick Jay: Well, there's a lot of concern with decreasing sea ice habitat up in the Arctic. Walruses use the sea ice for a number of life functions. One of them is give birth on the sea ice and nurse their young on the sea ice. They use the sea ice to rest, so with the reduction of sea ice, it really is forcing walruses into behaviors that they have not used in the past.

The other thing, sea ice acts as a barrier as well to potential anthropogenic stressors to species like walruses. But sea ice in the winter time, a lot of the industry is not able to go up into some of the areas that are important for walruses.

So with the loss of sea ice, it opens up the opportunity for other activities to occur in walrus habitat which can also impact the walruses, so it's a pretty complex system. And managers in society would like to know is how the reduction of sea ice has projected for the future to the end of the century, at least how walruses really respond to those reductions of sea ice.

Paul Laustsen: So how does this model actually project the walrus populations?

Chadwick Jay: The model is called the Bayesian network model and some of the inputs into the model have to do with sea ice projections, the extent of sea ice and the amount of time sea ice is present in areas that range to the walrus which is the Bering and the Chukchi seas.

It also has for input in the model is how the degree in which sea ice may cause a change in the arctic ecosystem by, for example, an earlier sea ice melt could result in a decrease organic matter into the bottom and that's important because walruses are bottom feeders, benthic feeders.

They feed on clams and snails and worms and a variety of other things at the bottom and so those prey species decrease in abundance that could also impact walruses. And then also as we mentioned before, with decreased sea ice there could be a decrease in other anthropogenic activities within their range and those could also act to modify the bottom substrate and affect the abundance of prey for walruses.

In addition, those activities could also have an impact on walruses through increase in waste levels or the potential for oil spills and things like that, which could disturb or have direct or indirect impacts on the walruses.

And then finally, walruses are harvested in Alaska. They're important food source for Alaskan natives and Russian natives as well and so the harvest of walrus has also impact the population obviously.

Paul Laustsen: Interesting. So what are the most likely factors in the model that will affect the outlook of the walrus population?

Chadwick Jay: Well, what we found through this model is that the reduction of sea ice, particularly in the summer/fall season in the Chukchi sea and potentially high level of harvest in the future could have the greatest impact on walrus outcome.

Paul Laustsen: Who will use this model and how will the information generated from it be used?

Chadwick Jay: I think the model is useful for a couple of things.

One is that different organizations and managers can look at the model and the outcome and how the outcomes are related to some of these specific stressors in walrus outcome to the future and to be able to help, plan and mitigate some of those stressors.

We also like the important use of the model is for us in planning our research and folks in our research – the future research – because it helps to identify what's link to what, you know, what things have the greatest impact on the population and things we may need more information on.

Paul Laustsen: How does this model fit into the larger context of your walrus studies?

Chadwick Jay: Right now, I think it is providing and then have a good framework for us such you to be able to pull in these various variables together and see how they're related and talk about how they're linked, and they just tell us to better perceive how these things together can impact the walrus population.

And then also, again, as a framework for us for future research, it allows us to look at what things in population might be well sensitive to, where are and the gaps are and where we maybe to increase our level of research effort.

And then also, again, as a framework for us for future research, it allows us to look at what things in population might be well sensitive to, where are and the gaps are and where we maybe to increase our level of research effort.

Paul Laustsen: Is there any other information about this new model that you'd like to share?

Chadwick Jay: This is kind of a first cut of the model. It's something that we're looking as not something as maybe static but a dynamic model so we'll revisit it in another few years. I think the model as we think we need to, sort of, gives us what we think is a better portrayal of the system and how walrus will respond to that system.

And also as we get new information through research monitoring, that would also help our end of the value system model in the future.

Paul Laustsen: That's great. It's really interesting work that you're doing Dr. Jay.

Chadwick Jay: Thank you.

Paul Laustsen: Thank you very much for taking the time to speak with me today and good luck in the future.

Chadwick Jay: Thanks.

Paul Laustsen: This is a production of the U.S. Geological Survey, U.S. Department of Interior.

[End Music]

Show Transcript