A 15-Year Quest to Protect Polar Bears

Finally, after 15 years of inaction on climate change to help save polar bears, a groundbreaking new paper led by our chief scientist emeritus, Dr. Steven Amstrup, clears the way to address greenhouse gas emissions under the U.S. Endangered Species Act.

The paper details, for the first time, the impact of a specific quantity of emissions on sea ice loss and polar bear survival, paving the way to remove the roadblock that has prevented action. We caught up with Steve to talk about the findings and what they mean for polar bears, the Endangered Species Act, and other wildlife.

Q: What inspired you to write this paper?

The genesis goes back to 2007, when I was Polar Bear Project Leader for the U.S. Geological Survey. I led the team that was tasked with gathering data on whether the polar bear should be listed as a threatened species.

Working under great pressure and a tight deadline, we produced a series of nine reports that told a dire story. Our findings projected that we could lose two-thirds of the world’s polar bear by the middle of this century, and possibly all of them by century’s end, without significant greenhouse gas reductions. The science in those reports was so clear that the U.S. decided to list the polar bear as a threatened species in 2008—the first species ever to be listed due to future threats from human-caused global warming.

Unfortunately, David Bernhardt, the solicitor of the Department of Interior at the time, issued a memo stating that unless scientists could separate the impact of emissions generated by specific projects from all emissions generated since the beginning of the Industrial Age, the impact of emissions on sea ice and polar bears from proposed projects could not be considered in Endangered Species Act reviews.

My first thought was: “Well, this is silly. Polar bears were listed because of global warming, but the Endangered Species Act can’t address it?”

Over time, my thoughts moved from “this is silly” to “how can we overcome this?”

Q: Tell us about the state of the science at the time with respect to emissions, sea ice loss, and polar bears. What gaps were flagged in the Bernhardt Memo?

At the time, we knew in a general sense that polar bears depended on sea ice for their survival. We knew that emissions were warming the planet and causing the sea ice to melt, and that projected sea ice losses would mean dramatically fewer polar bears as we moved through this century. The science on that was quite clear. But we didn’t have a way to quantify the impact of a specific amount of emissions on the vital demographic rates (survival or reproduction) that allow polar bear populations to persist.

There were two stumbling blocks. First, we didn’t have a link between sea ice trends and demographic response. Even though we understood that less sea ice would ultimately mean poorer polar bear reproduction and survival, we had not been able to define an attribute of the sea ice—like sea ice area, extent, or volume—that was quantitatively linked to demography. Second, we didn’t have a definitive link showing the impact of a specific amount of emissions on sea ice loss. This combination of gaps prevented us from overcoming the obstacles described in the Bernhardt Memo.

The memo was clearly designed to block action on climate change, making the Endangered Species Act powerless in terms of helping a climate-threatened species like the polar bear. Overcoming those gaps was a tall order but necessary to secure the protections needed to sustain the polar bears’ future.

Q: What was your first step in working to address these gaps?

In 2010, two years after the polar bear was listed, I was the lead author of a paper published in Nature that showed that there was no tipping point in Arctic sea ice, and that sea ice would stabilize if we halted emissions. Between the submission and publication of that paper, I decided to leave my job as a government scientist and become chief scientist at Polar Bears International. I wanted to focus my attention on communicating about the importance of addressing climate change, for polar bears and all of us, and to focus my research on addressing the gaps flagged in the Bernhardt Memo.

Q: What happened after that?

Through years of careful research, study by study, polar bear scientists gained a deeper understanding of the polar bears’ reliance on sea ice to hunt their seal prey and learned more about their energy needs. I was becoming increasingly frustrated at being unable to find a link between estimates of polar bear reproductive and survival rates and sea ice decline. I recalled early work done by Dr. Peter Molnár that explored whether energetic models could provide a different way to understand how changing ice dynamics can affect polar bears. I think it was 2014 when I called Peter, now at the University of Toronto Scarborough, and asked if he would be interested in refining his preliminary models to address polar bear reproductive and survival rates in relation to sea ice—in the hopes this could establish the link between demography and sea ice that we needed. I also called Dr. Cecilia Bitz, who was a coauthor on the 2010 paper in Nature, to see if she would be interested in providing the necessary sea ice/climate modeling expertise for a new project. We also engaged Dr. Jennifer Kay and Dr. Marika Holland for added support on the climate science side of the study’s team. Later, we brought in one of Peter’s PhD students, Stephanie Penk, to help with the final analyses.

The result of this collaboration was a landmark paper published in 2020 in Nature Climate Change. In it, Peter Molnár’s expanded modeling of the polar bear’s energetic balance linked polar bear survival rates with the annual number of ice-free days. (Ice-free days determine the fasting duration polar bears face when the ice melts every summer, forcing them to move on land or to follow the remaining sea ice over deep and unproductive waters.) The study detailed the fasting impact thresholds for different age and sex groups of polar bears—that is, how many days of fasting they can endure before their survival begins to decline rapidly. Based on those fasting impact thresholds and two different future greenhouse gas emissions scenarios that projected different possible annual fasting durations, we projected when each polar bear population was likely to disappear.

Identifying ice-free days as the sea ice attribute that provided a quantitative link between polar bear survival and declining ice availability was a critical outcome of that 2020 paper. Previous studies had attempted to link sea ice decline to the impact on polar bears but had not been successful. Molnár et al., however, showed that ice-free days was the critical sea ice attribute to focus on; it could be directly linked, through Peter’s energetic modeling, to polar bear survival and recruitment.

That 2020 paper inspired me to think, “Well, now that we know the number of ice-free days provides a quantitative link between polar bear demographics and sea ice, can we link ice-free days to emissions, overcoming the final hurdle in the Bernhardt Memo?”

Q: After that final piece fell into place, you began work on your current paper, which was published in the September 1, 2020 issue of Science. Can you tell us about that?

After the Molnár study came out, it was an obvious next step to pull the data together into one paper to overcome the challenges in the Bernhardt Memo. In fact, using regression analysis to determine the sensitivity of ice-free days to emissions, and then linking that to the regression relationship between ice-free days and survival, that we identified in the 2020 paper, seemed like such a simple idea I’ve been surprised no one else has thought of it.

I asked Cecilia Bitz, my coauthor on the 2020 paper as well as the 2010 paper in Nature, whether sufficient data on emissions were available that we could link them to ice-free days and fasting duration. She confirmed that this was possible and enthusiastically agreed to collaborate on a study that connected the dots from emissions, to ice-free days, to polar bear survival rates.

Building on the approach shown in the Molnár paper, I decided to focus on survival of cubs. Although our 2020 paper described the impact of more ice free days on all sex and age groups it made it clear that suvival of cubs was the first demographic parameter affected by longer fasting periods. Also, if cubs in a population don’t survive, the population can only decline regardless of whether adult bears can withstand longer annual fasting periods.. Meanwhile, Cecilia assessed available data on global emissions and performed the calculations to determine the number of ice-free days and the impact on polar bear cub survival caused by a specific amount of emissions.

Q: The past 15 years have been like a chess game for you, strategically moving piece by piece across the board to finally overcome the Bernhardt Memo restrictions. How does it feel to finally slide the last piece into place?

It is a relief to finally have the pieces of the puzzle assembled. And I am thrilled to see how well our estimates of the impact of emissions on recruitment explain the evidence of decline that we’ve seen in portions of the polar bears’ range. In the Western Hudson Bay and Southern Beaufort Sea populations, for example, the declines in cub recruitment—which we projected would occur when these bears crossed their fasting impact thresholds—match the outcomes from decades of study in those areas. Our outcomes therefore, not only address the critical policy stumbling block, but will provide an important tool for projecting the future of polar bears around the world.

Overall, the feeling is bittersweet. It’s satisfying, to be sure, to finally fill in the knowledge gaps flagged in the Bernhardt memo. We now have the scientific evidence to rescind that Memo, which is huge. But at the same time, I’m incredibly frustrated that the memo blocked 15 years of climate progress in the U.S.

When a species is listed under the Endangered Species Act, the government is required to use “the best available science” to inform a species recovery plan. We have now shown that the best available science, expressed in our new paper, overcomes the Bernhardt restriction. I’m hopeful that this paper will finally allow the government to ensure climate protections for polar bears under the Endangered Species Act—based on the best available science—and that the approach will serve as a model for other species and for countries weighing development projects. And that would be a win not just for polar bears, but for all of us.