Polar bear gene flow blamed on melting ice is another model result that doesn’t make sense

Polar bear researchers just published a study that suggests polar bears have moved around the Arctic in direct response to recent sea ice changes — a conclusion I suggest you take with a grain of salt and a raised eyebrow.

That’s because they have also proposed, among other things, that the Svalbard Archipelago was a sea ice refugium during warm interglacial periods, and could be again if the Arctic warms as predicted. That they would accept and promote such a model-based conclusion, which has no relationship with reality, calls their scientific judgment into question.

Svalbard as a potential warm refugium_Jan 8 2015_PolarBearScience

Based on genetic model results, the Svalbard Archipelago (circled) has been proposed as a sea ice refugium for polar bears during previous warm Interglacial periods and during predicted sea ice declines in the future. Yet most years since 1979 (2014 was one exception), this region has been ice free during the summer, making Svalbard a decidedly poor candidate for retaining sea ice when it’s much warmer than today.

Late last year it was the population crowd gone wild (Hamilton et al. 2014), generating extinction predictions based on worst-case scenario climate change predictions1 for bears in the Canadian Arctic Archipelago; this week, it’s geneticists. I didn’t bother commenting on the over-the-top predictions of polar doom in November – I found it fear-mongering posing as science.

But this new paper deserves a bit of attention because it claims to reveal observed genetic impacts from a “rapidly warming Arctic” on polar bears. In fact, it does nothing of the kind.

The paper was published Monday at PLOS One, by a long list of co-authors led by IUCN Polar Bear Specialist Group member Lily Peacock.2 The paper is called “Implications of the Circumpolar Genetic Structure of Polar Bears for Their Conservation in a Rapidly Warming Arctic.

The main press release claim?

This circumpolar, multi-national effort provides a timely perspective on how a rapidly changing Arctic is influencing the gene flow and likely future distribution of a species of worldwide conservation concern.”

In fact, their “effort” is the output of a long list of statistical assumptions and/or computer models, not actual evidence that more than a few bears moved from one region to another in response to declining sea ice.

The paper’s primary claim is described in the graphic below: the authors suggest the model outputs they generated constitute “documentation” that polar bears have moved, in the directions suggested on the map, over the last 30 to 100 years (“3-10 generations”) in response to sea ice declines. The wider the arrow, the more bears assumed to be involved.

If these proposed movements really are driven by sea ice declines as the authors suggest, it implies that the Western Polar Basin (primarily the Southern Beaufort and the Chukchi Sea) has recently had better habitat for polar bears than the Russian Arctic and much better than the Canadian Archipelago. That doesn’t seem to fit with what we’ve been hearing about how bad the polar bear situation has been in the Southern Beaufort (it’s periodically bad, but due to heavy spring sea ice, not declining summer ice).

Figure 3 from Peacock et al. 2015.

Figure 3 from Peacock et al. 2015. Based on models that assume movement of bears (“gene flow”) accounts for the differences in gene frequencies found amongst bears sampled from different regions, at different times, since the 1980s.

The problem is, this “gene flow” result was generated by models — models based on particular assumptions of how and why genes might be different amongst individuals in different regions.The researchers used DNA samples collected over four decades: 1980s, 1990s, 2000s, 2010s (with only a few from the 1980s).

However, the “gene flow” pictured above represents assumed movements over 30-100 years (10 years per generation, defined elsewhere in the paper) — the caption says “recent directional gene flow (ca. 3-10 generations)“. I fail to see how such movements over 30-100 years could correlate with changes in sea ice over the last 20 years, as stated in the paper’s abstract:

We provide an expansive analysis of polar bear (Ursus maritimus) circumpolar genetic variation during the last two decades of decline in their sea-ice habitat.” [my bold]

Perhaps they simply explained it poorly. But no other paper that used the same method was discussed in the Peacock et al. paper, so that readers might judge the suitability and credibility of the method for polar bears. They also failed to offer any empirical evidence (i.e., proof of actual bears that moved) to support their conclusion that the model results of gene flow correlated with recent sea ice changes. In addition, they did not suggest how many bears might have been involved in these movements — it might be only a few but they imply it could be many.

Polar bears (and their ringed seal prey) move around the Arctic — that’s been demonstrated by direct observation. But I find such well defined movements in the particular directions shown on their map implausible, and implausible that they would be correlated with recent summer sea ice declines — especially the movement proposed from the Russian Arctic to the Western Arctic.

As for the Svalbard refugium nonsense, here’s what I had to say a couple of years ago when another group of geneticists proposed Svalbard as an interglacial refugium for polar bears (Miller et al. 2012):

Polyak et al (2010:1769) mention evidence of low sea ice or ice-free conditions during the last Interglacial (MIS 5e, ca. 130,000-115,000 years ago) in several western Arctic locations (e.g. Bering Strait, Chukchi Sea) but there is no mention of the region around Svalbard being a sea ice refugium (persistent multiyear sea ice) during interglacial periods. In fact, the only regions mentioned by these authors in regards to persistent multiyear sea ice during interglacials are northern Greenland and Ellesmere Island.” [my bold]

As far as I’m concerned, the ridiculous suggestion that the Svalbard region served as a refugium during warm interglacial periods calls into question the plausibility of their other conclusions. If they cannot be bothered to check a genetic model-suggested result against geographic and historical reality, why should we trust their judgment about anything?

Points of Interest in Peacock et al. 2015:
1. Generation time – the models used to support the listing of polar bears as ‘threatened’ in 2008 under the American ESA used a generation time of 15 years (thus 3 generations = 45 years). This paper used a generation time of 10 years, the estimate suggested by geneticist Matt Cronin and colleagues in 2009 (but whose work is not cited). I have not seen a 10 year generation time used before now in any published work. For example, Amstrup et al. (2010), published after Cronin et al. 2009, used 15 years for predicting future population declines. The difference between the two generation times estimates (15 years vs. 10 years) means that predicted changes in sea ice assumed to have a negative effect on future polar bear populations must be shown within 30 years rather than 45 years (3 generations).

2. Virtually every genetics paper published recently regarding the speciation event that generated the polar bear, and/or its genetic history, has given a different answer. These authors attempted another go, on top of the gene flow analysis. Surprise, surprise — they got another different answer.

3. The authors assumed that polar bears that once spent their summers on St. Matthew Island in the Bering Sea would be extinct now because of sea ice declines if they hadn’t already been slaughtered out of existence, and suggest that along with few bears reaching Newfoundland in spring, this is evidence of a recent contraction of the polar bears range due to sea ice changes. As I’ve pointed out previously, it appears that even in the late 1800s, St. Matthew Island bears spent about 5 months onshore during the summer, yet were depicted as fat and healthy.  If the bears that summered there had survived into modern times and found they had to move from St. Matthew Island because it was no longer a viable denning area (as they do today when local conditions change due to natural variations in sea ice), they would only have had to go as far north as Wrangel Island — not even into another subpopulation region — to find replacement habitat. Lack of polar bears summering on St. Matthew Island today (and few venturing to Newfoundland and Labrador) is hardly compelling evidence that the range of polar bears has contracted in any appreciable way due to sea ice changes because the change in range is geographically very small and biologically insignificant.

1. The Hamilton et al. (2014) prediction paper said this:

“Under business-as-usual climate projections, polar bears may face starvation and reproductive failure across the entire Archipelago by the year 2100.”

“Our model projection is based on the RCP8.5 scenario, which estimates the global average radiative forcing at 8.5 W/m 2 by 2100, and mean global temperature changes of ~3.5°C in 2071–2100 when compared with the historical period of 1961–1990 [35] , and represents a worst-case scenario.[my bold]

2. Other co-authors that are PBSG members (Obbard, Boltunov, Regehr, Ovsyanikov, Aars, Amstrup, Belikov, Derocher, Stirling, Wiig – and ex-member Taylor).  However, most of the listed “co-authors” (including Taylor) were acknowledged because they collected samples used in the analysis. USGS affiliated geneticists (including Peacock) plus one privately employed geneticist actually wrote the paper.

Amstrup, S.C., DeWeaver, E.T., Douglas, D.C., Marcot, B.G., Durner, G.M., Bitz, C.M., and Bailey, D.A. 2010. Greenhouse gas mitigation can reduce sea-ice loss and increase polar bear persistence. Nature 468:955-958.

Cronin, M.A., Amstrup, S.C., Talbot, S.L., Sage, G.K. and Amstrup, K.S. 2009. Genetic variation, relatedness, and effective population size of polar bears (Ursus maritimus) in the southern Beaufort Sea, Alaska. Journal of Heredity 100: 681-690. See NIPCC summary here.

Hamilton SG, Castro de la Guardia L, Derocher AE, Sahanatien V, Tremblay B, et al. 2014. Projected Polar Bear Sea Ice Habitat in the Canadian Arctic Archipelago. PLoS ONE 9(11): e113746. Open access doi:10.1371/journal.pone.0113746

Miller, W., Schuster, S.C., Welch, A.J., Ratan, A., Bedoya-Reina, O.C., Zhao, F., Kim, H.L., Burhans, R.C., Drautz, D.I., Wittekindt, N. E., Tomsho, L. P., Ibarra-Laclette, E., Herrera-Estrella, L., Peacock, E., Farley, S., Sage, G.K., Rode, K., Obbard, M., Montiel, R., Bachmann, L., Ingolfsson, O., Aars, J., Mailund, T., Wiig, O., Talbot, S.L., and Lindqvist, C. 2012. Polar and brown bear genomes reveal ancient admixture and demographic footprints of past climate change. Proceedings of the National Academy of Sciences 109:E2382-E2390. Open access: http://www.pnas.org/content/109/36/E2382.full
Text plus supplemental info: http://www.pnas.org/content/109/36/E2382.full.pdf+html?with-ds=yes

Peacock E, Sonsthagen SA, Obbard ME, Boltunov A, Regehr EV, et al. 2015. Implications of the Circumpolar Genetic Structure of Polar Bears for Their Conservation in a Rapidly Warming Arctic. PLoS ONE 10(1): e112021. doi:10.1371/journal.pone.0112021 Open access: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0112021

Polyak, L., Alley, R.B., Andrews, J.T., Brigham-Grette, J., Cronin, T.M., Darby, D.A., Dyke, A.S., Fitzpatrick, J.J., Funder, S., Holland, M., Jennings, A.E., Miller, G.H., O’Regan, M., Savelle, J., Serreze, M., St. John, K., White, J.W.C. and Wolff, E. 2010. History of sea ice in the Arctic. Quaternary Science Reviews 29:1757-1778. http://bprc.osu.edu/geo/publications/polyak_etal_seaice_QSR_10.pdf

Comments are closed.