Many W Hudson Bay polar bears still offshore at 7 August despite apparent low ice levels

Contrary to all expert expectations, five female polar bears (45%) out of eleven that had tracking collars attached last year were still out on the sea ice that’s lingering along the western shore of Hudson Bay as of 7 August. And if five collared bears are out there, then there are almost certainly many more without collars doing the same thing. This pattern of bears staying out on the ice long after the so-called ‘critical threshold’ of 50% concentration has passed has been going on since at least 2015 and many bears on tracking maps in July and August appear to be on ice that doesn’t exist.

Chukchi Sea polar bear Arctic_early August 2018_A Khan NSIDC small

There are two explanations for this pattern and both are likely true: 1) much more ice actually exists on Hudson Bay than satellites can detect and 2) polar bear experts are wrong that Western Hudson Bay polar bears head to land soon after sea ice concentration drops below 50%. Models that predict a catastrophic future for Western Hudson Bay polar bears (Castro de la Guardia et al. 2013; Molnar et al. 2020) assume that ice coverage of less than 50% in summer greatly reduces polar bear survival. However, if polar bears do not always head to land when sea ice drops below 50% then the models cannot possibly describe their future accurately. In other words, depending on the discredited ‘worst case’ RCP8.5 climate scenario for the most recent polar bear survival model that extrapolates from Western Hudson Bay bear data to many other subpopulations, as I discussed previously, may not be its only fault.

For at least the last five years now Western Hudson Bay polar bears have stayed out on melting summer ice when it was well below 50% concentration; the same phenomenon of bears tracking to what looks like open water has also shown up for Southern Beaufort bears.

Below is biologist Andrew Derocher’s tracking map at 7 August 2020 for the bears he and colleagues put collars on last year, which shows five out of eleven bears still out on the ice:

Derocher 2020 WHB tracking map 7 Aug_5 bears still on the ice

As I’ve pointed out previously, more ice is present on the bay than shows on the tracking map above because it only shows ice that’s >50% concentration. Here are the Canadian Ice Service charts for the northern and southern portions of the bay on 8 Aug:

Hudson Bay North daily stage of development 2020 Aug 8

Hudson Bay South daily stage of development 2020 Aug 8

All of the ice remaining is thick first year ice, which means there is likely to be more ice than satellites can visualize because of melt-water ponding on top of the ice floes at this time of year. Polar bears are excellent swimmers and deal easily with such melt ponds. These are a regular feature of the Arctic landscape at this time of year and part of the polar bears’ natural habitat. Below is a photo of melt ponds in the Beaufort Sea at mid-July 2016 (taken by NASA).

Melt ponds in the Beaufort 14 July 2016 NASA sm

According to the satellite charts shown above, only a bit of remnant thick first year ice is left on Hudson Bay. However, even polar bear researchers know that satellites are notoriously bad at getting the amount of ice correct at this time of year and can underestimate ice amounts by up to 50% in Hudson Bay. Polar bear biologists that work with Western Hudson Bay bears, like Andrew Derocher, are aware that this is the case, as the quote below from one of his student’s papers shows (Castro de la Guardia et al. 2017:227) [my bold]:

In general, passive microwave derived sea ice data are associated with an underestimation error of up to 30% during breakup and freeze-up throughout the marginal ice zone and seasonal ice regions in the Northern Hemisphere (e.g. Cavalieri et al. 1991, Comiso et al. 1997, Markus & Dokken 2002). In Hudson Bay, passive microwave sea ice concentration can underestimate sea ice concentration by up to 50% compared with CISDA (Agnew & Howell 2003). Underestimation biases of passive microwave data are associated with the presence of wet snow and melt ponds during breakup, and with areas covered by frazil ice and young ice during freeze-up (Agnew & Howell 2003).

However, there is another possible explanation for why many WH polar bears this year (and back to 2015 at least) have remained on the ice despite the apparent ice coverage being well below 50%. That’s the probability that low ice concentrations during sea ice breakup in the summer is not a death sentence for polar bears as researchers intend on modeling their future have long assumed. Below is Derocher’s tracking map for last year at the end of July, where we see the same pattern:

Derocher 2019 WHB 12 pbs ashore 12 still on ice with some unaccounted for at 27 July chart

In 2018, the pattern was similar: by late July ice cover in Western Hudson Bay was apparently almost non-existent but four bears with tracking collars were still offshore (and if four collared bears were on the ice, it’s almost certain there were many more without collars):

Derocher 2018 WHB collared females at 19 July

And in 2017, the story was the same. That summer was the first time I noticed Derocher suggesting that “bears may be shifting behaviour to stay out on less ice” to explain why the tagged bears were not coming ashore when the sea ice dropped below 50% as he and his colleagues expect them to do.

Derocher said something similar a few weeks ago (29 July 2020):

“4 W Hudson Bay polar bears ashore. Amazing how long they’re staying offshore as ice this low would usually have them all on land. So many still on the ice suggests a behaviour shift (i.e., behavioural plasticity). The @ualbertaScience polar bear crew is studying this.”

Only now, apparently, are Derocher and his colleagues giving any thought to the possibility that polar bears might not have rigid responses to changing sea ice conditions. No, polar bear experts looked at the sea ice was like in the 1980s and figured those conditions were what polar bears absolutely required (Amstrup et al. 2007; Castro de la Guardia et al. 2013, 2017; Durner et al. 2007; Stirling et al. 1999): they never considered the possibility that the bears could do perfectly well with less.

[I’ve been pointing out for years that after about late May to early June on Hudson Bay, there are few opportunities for polar bears to catch seals which means they would eat as little on the ice as they would do if they were on land. It’s cooler on the ice and there are no bugs, so some bears may choose to stay out as long as possible but garner no particular advantage by doing so.]

By all accounts, 2016 was a great summer for WH bears and many were out on the rapidly declining ice:

Derocher tweet_11 July 2016 WHB bear tracker

And in 2015, there were again bears on ice that shouldn’t exist:

In summary, polar bears have been doing very well in Western Hudson Bay since at least 2015 despite the fact that quite a few bears have stayed out on ice less than 50% concentration for weeks at the end of the melt season. Virtually all bears have come ashore fat and healthy over that period, whether they came ashore late or early. So the surprise about Western Hudson Bay is not just that for the last two years bears have been coming ashore fat and healthy and as late as they did in the 1980s but that their relationship with the melting ice contradicts a very basic assumption held by polar bear experts. It is apparent from what’s been happening in Hudson Bay since 2015 that polar bears do not require sea ice cover that is >50% concentration during the summer, which means that polar bear survival models developed to predict the future of the species based on this assumption are almost certainly getting the wrong answers.

PS. So far, there has not been a report from the Polar Bear Alert Program regarding problem bears in Churchill, Western Hudson Bay. As long as I’ve been collecting these published reports (2015), there has not been a first report of the season issued later than the second week in July. Even if it is issued this week, the second week in August is extraordinarily late for the first report of the season. However, it’s not clear if that’s because the bears are so late off the ice that there have been no problems yet or for some other reason. Last week I contacted the Town of Churchill, who usually post these reports on their website, and got this reply (12:46 PM 7 August 2020):

Good afternoon, thank you for your message. We have not yet received any reports from Manitoba Conservation and Climate regarding Polar Bears in the area. We have reached out, and will be posting them as they are received. Thank you.


Amstrup, S.C., Marcot, B.G. & Douglas, D.C. 2007. Forecasting the rangewide status of polar bears at selected times in the 21st century. US Geological Survey. Reston, VA. Pdf here

Castro de la Guardia, L., Derocher, A.E., Myers, P.G., Terwisscha van Scheltinga, A.D. and Lunn, N.J. 2013. Future sea ice conditions in Western Hudson Bay and consequences for polar bears in the 21st century. Global Change Biology 19:2675–2687. doi: 10.1111/gcb.12272

Castro de la Guardia, L., Myers, P.G., Derocher, A.E., Lunn, N.J., Terwisscha van Scheltinga, A.D. 2017. Sea ice cycle in western Hudson Bay, Canada, from a polar bear perspective. Marine Ecology Progress Series 564: 225–233.

Durner, G.M., Douglas, D.C., Nielson, R.M., Amstrup, S.C. and McDonald, T.L. 2007. Predicting 21st-century polar bear habitat distribution from global climate models. US Geological Survey. Reston, Virginia. Pdf here.

Molnár, P.K., Bitz, C.M., Holland, M.M., Kay, J.E., Penk, S.R. and Amstrup, S.C. 2020. Fasting season length sets temporal limits for global polar bear persistence. Nature Climate Change.

Stirling, I., Lunn, N.J. and Iacozza, J. 1999. Long-term trends in the population ecology of polar bears in Western Hudson Bay in relation to climate change. Arctic 52:294-306. [open access]

Comments are closed.