Polar Biol (2010) 33:1149 1153
DOI 10.1007/s00300-010-0791-2
SHORT NOTE
Has early ice clearance increased predation on breeding birds
by polar bears?
" " "
Paul A. Smith Kyle H. Elliott Anthony J. Gaston
H. Grant Gilchrist
Received: 15 September 2009 / Revised: 27 February 2010 / Accepted: 2 March 2010 / Published online: 20 March 2010
Ó Springer-Verlag 2010
Abstract Past studies suggest that polar bears (Ursus Introduction
maritimus) consume terrestrial food only opportunistically
and derive little nutritional benefit from it. Here, we present The Arctic climate is changing at an unprecedented rate, and
observations of at least 6 bears consuming large numbers evidence of effects on wildlife is growing (Post et al. 2009).
of snow goose (Chen caerulescens) eggs at two locations in Many of the emerging examples are intuitive and direct.
the eastern low Arctic in 2004 and 2006. We also report Advancing schedules of ice break-up in western Hudson Bay
two records of a polar bear eating the eggs and chicks of force polar bears (Ursus maritimus) ashore earlier, leaving
cliff-nesting thick-billed murres (Uria lomvia) in 2000 and bears in poorer condition when sea-ice returns in fall (Stirling
2003. Climatic warming has resulted in progressively et al. 1999; Obbard et al. 2006; Stirling and Parkinson 2006).
earlier ice break-up in Hudson Bay, forcing bears ashore For arctic breeding birds such as the lesser snow goose (Chen
much earlier than historical records indicate. Advancement caerulescens caerulescens), changing climatic conditions on
in the nesting dates of birds has been more modest, and this the breeding grounds can affect timing of breeding, clutch
mismatch in timing could lead to an increasing overlap size and egg size (Skinner et al. 1998).
between the nesting period of birds and the period during In addition to these direct effects of climate change on
which bears are on land. At these sites in these years, bears wildlife, complex interactions between trophic levels are
were on land prior to the hatch of nests, and the predation predicted to occur (Gauthier et al. 2004; Berteauxet al. 2004).
that ensued was catastrophic for the birds at a local scale. However, the complexity of the interactions makes precise
Although anecdotal, our observations highlight the com- predictions difficult. Mismatches in the timing of phenolog-
plexity of trophic interactions that may occur in a changing ical events are considered a key mechanism through which a
Arctic. changing climate might adversely affect arctic wildlife
(Gaston et al. 2009; Post et al. 2009). Here, we document an
Keywords Polar bear Ursus maritimus Snow goose advance in the timing of appearance of polar bears on shore at
Chen caerulescens Climatic change Thick-billed murre one site and describe several recent instances of extensive
Uria lomvia Predation predation of snow goose and thick-billed murre (Uria lomvia)
eggs by polar bears in years when the bears came ashore
early. Although anecdotal, our observations suggest an indi-
rect link between marine ice conditions and the reproductive
success of birds, highlighting the complexity of interactions
that may occur in a changing Arctic.
P. A. Smith (&) A. J. Gaston H. G. Gilchrist
Environment Canada National Wildlife Research Centre,
Ottawa, ON K1A 0H3, Canada
e-mail: [e-mail]
Study species
K. H. Elliott
Polar bears are highly adapted to life on the frozen ocean
Department of Zoology, University of Manitoba,
Winnipeg, MB R3T 2N2, Canada and depend on ice for long-range movements, finding
123
1150 Polar Biol (2010) 33:1149 1153
mates and accessing prey (Stirling and Derocher 1993).
The most carnivorous of the Ursids, polar bears favour
ringed seals (Phoca hispida) but also prey upon a variety of
other marine mammals (e.g. Stirling and McEwan 1975;
Smith 1980; Lowry et al. 1987). They have difficulty
capturing prey in open water, and come ashore during ice-
free periods in regions with seasonal sea ice. They rely
primarily on stored fat throughout the ice-free period but
have been observed to feed opportunistically on terrestrial
foods including grasses (e.g. Elymus arenarius), marine
algae (e.g. Laminaria spp., Fucus spp.), berries (Vaccinium
uliginosum, Empetrum nigrum), carrion including conspe-
cifics and remains left from human hunting, and occa-
sionally caribou, fish, rodents and birds (Russell 1975;
Donaldson et al. 1995; Stempniewicz 2006; Schliebe et al.
2008; Rockwell and Gormezano 2009). Despite these
Fig. 1 The locations of our observations of bears consuming the eggs
numerous accounts of bears taking terrestrial foods, these
of snow geese and the eggs and chicks of thick-billed murres
behaviours have traditionally been considered opportunis-
tic, and the food considered to be of limited energetic
importance (e.g. Ramsay and Hobson 1991; Hobson et al.
2009; but see Derocher et al. 1993).
The lesser snow goose is widespread and abundant in
the eastern Arctic, and nests colonially in coastal tundra
(Mowbray et al. 2000). Laying dates are highly syn-
chronised and are strongly associated with local weather
and snow conditions (Skinner et al. 1998). Hatching dates
vary geographically and seasonally but typically occur
between mid-June and early July in Hudson Bay and the
Foxe Basin (e.g. Skinner et al. 1998; this study).
Thick-billed murres nest in dense colonies, typically
numbering in the tens to hundreds of thousands of indi-
viduals, on sea-cliffs throughout much of the circumpolar
Arctic. In the eastern Canadian Arctic, they lay their single
egg in late June or early July and incubate for approxi-
mately 33 days (Gaston and Hipfner 2000).
Fig. 2 The number of days upon which polar bears were seen (mean
among years Ä… SE) from a research camp at a colony of thick-billed
Observations
murres on Coats Island, Nunavut
Arrival of polar bears on shore
relatively stable but early ice break-up (2000 2007) as
observed at our study site (Gaston et al. 2005). The earliest
Thick-billed murre population studies have been conducted
sighting of a polar bear during 1985 1991 was 11 20 July,
at a research camp on northern Coats Island annually since
while during 2000 2007, the earliest sightings were during
1985 (N62° 570 W82° 000, Fig. 1). Researchers recorded
1 10 July. Acknowledging the limitations of the data, our
days upon which bears were sighted, and observations were
observations suggest that bears were on land in the vicinity
binned into 10- or 11-day periods. Observations were made
of the murre colony earlier in the season in 2000 2007 vs.
incidentally to our other work, and the period during which
1985 1991 (Fig. 2).
camp was occupied varied among years. We considered
only years in which camp was occupied early enough in the
Snow goose nest predation
season to capture first sightings of polar bears. We divided
the data into three periods, representing a period of rela-
Our first observation of predation on snow goose nests
tively late ice break-up (1985 1991), a period of rapid
was recorded in 2004 on southern Southampton Island,
advancement in ice break-up (1992 1999) and a period of
123
Polar Biol (2010) 33:1149 1153 1151
Table 1 Summary of bear observations, ice conditions and hatch of snow goose nests at Southampton and Coats Islands, 2004 and 2006
Site First bear Regional ice Local ice First gosling
sighted concentration conditions seen
Southampton island goose colony, 2004 23-Jun 63% on 25 Jun 1 km landfast, sparse pack [30-Jun
beyond on 30-Jun
Coats island goose colony, 2006 24-Jun 32% on 25 Jun Ice-free by 28-Jun 27-Jun
Regional ice concentration data (  North Hudson Bay Narrows  ) retrieved from the Canadian Ice Service (www.ice-glaces.ec.gc.ca)
approximately 9 km from the coast at N63° 440 W84° 010 June, we saw the first evidence of ice break-up several
(Fig. 1). At 2330 h, 23 June, a lone bear was spotted eating kilometres offshore, to the north of our study site. By 16 June,
snow goose eggs (Table 1). The bear was an adult female, open water was visible across much of the horizon. From
visibly fat and moved slowly from nest to nest, consuming 20 26 June, the ice was completely broken and moving
the entire contents of each nest. The bear remained in the rapidly in and out with the changing winds and tides. By 28
area, within the goose colony, and was sighted regularly June, the ice had completely cleared the area (Table 1).
until 10 July, when the crew left the area. It was observed Because ice-out was atypically early, bears had come
eating eggs directly on many occasions, and the effect of ashore before snow goose nests had begun to hatch. The
the predation on local nest success of snow geese was first snow goose gosling was seen on 27 June. By 2 July,
pronounced. Although we have no precise count of the the majority of the snow goose nests had been consumed,
number of nests consumed, approximately 400 nests were and the few nests remaining had hatched. Bears were
present in the regularly surveyed areas near camp. We sighted only irregularly after this date, primarily resting
observed 100% nest failure in these areas and attribute the along the coast or travelling overland to or from the south
vast majority of this failure to predation from this single coast of the island (a distance of 40 km).
bear. This bear was also regularly observed eating snow
goose eggs in areas where we did not survey; so the total Thick-billed murre predation
effect of this single bear on local reproductive success of
geese was substantial. We observed a single polar bear entering the thick-billed
At the time that the bear was first sighted, landfast ice murre colony at Coats Island, Nunavut, Canada during
was still present immediately south of the camp. On 30 20 25 July 2000 and 26 31 July 2003. The latter obser-
June, we estimated that at least 1 km of landfast ice vation is presumed to be a different individual, because it
remained, with sparse, drifting pack beyond it. This was a was a young bear and likely too young to have been
particularly late year for snow goose breeding across much independent 3 years prior. In both years, the bear climbed
of the eastern Arctic, and all geese were still incubating onto cliff ledges and consumed all murre chicks and eggs
(i.e., no nests had hatched) at this time. that were available. The bears consumed more than 100
The second observation comes from a research camp eggs and chicks in both years (Mallory et al. 2009). No
approximately 120 km to the southeast, on the northern predation by bears had been observed in 17 seasons of
coast of Coats Island (N62° 550 W82° 260), in 2006. A observations at the same site prior to 2000.
small colony of breeding snow geese is found near the
coast, containing approximately 350 nests with a density of
roughly 4 nests/ha. Discussion
Bears were first seen in the area of the snow goose colony
on 24 June (Table 1). We saw three bears feeding within the At the southern edge of their range, in Hudson Bay, the
colony and an additional two resting on blocks of ice at the retreat of sea ice forces bears ashore for periods exceeding
shore. Bears were sighted in the colony daily until 2 July, but it 4 months (Stirling et al. 1977; Stirling and Derocher 1993).
was difficult to determine the total number of individuals The timing of this retreat to land is influenced by the timing
consuming eggs. On 25 June, we counted at least five bears of ice break-up, and recent data suggest that in much of the
within the colony: a mother and cub, a large adult male and eastern low Arctic, the break-up of sea ice is up to three
two smaller individuals. While feeding, bears consumed the weeks earlier than in the 1970s (Gough et al. 2004; Stirling
entire contents of a nest and walked directly to the next; it was et al. 2004; Stirling and Parkinson 2006). Polar bears are
clear that bears recognised the scent or appearance of the nests curious and opportunistic animals, and as the ice-free
and were searching for them. season increases in length and bears spend more time on
Ice-out was early in 2006, and polar bears were first land, it is not surprising that individuals would search for
sighted on land when the ice had nearly left the area. On 13 novel opportunities to supplement their diet.
123
1152 Polar Biol (2010) 33:1149 1153
Although bears have been observed to eat eggs previ- 590 W81° 400), first sightings ranged from 25 June 5 July,
ously, accounts of this behaviour seem increasingly com- with a mean of 29 June. The goose nests that we observed
mon. Stempniewicz (2006) and Drent and Prop (2008) at Coats Island (100 km south of East Bay) began to hatch
reported that bears have become more numerous during the on 28 June 2005 and 27 June 2006 (for the small number
bird-breeding season in west Spitsbergen and have been that were not predated by bears and achieved hatch). Our
observed consuming the eggs of barnacle geese (Branta observations therefore suggest that bears come ashore
leucopsis), thick-billed murres and little auks (Alle alle). much earlier than they previously did and now overlap with
Similarly, we recorded raids by bears on a breeding colony the incubation period of both geese and thick-billed murres.
of thick-billed murres at Coats Island twice since 2000, Similarly, researchers on the Cape Churchill peninsula
although no bear had been seen consuming eggs or nes- have noted that the earliest bears to arrive on land now
tlings in 17 previous field seasons. Bears have entered a overlap with nesting snow geese and consume some eggs
snow goose colony at Churchill, Manitoba, 6 times in the (Rockwell and Gormezano 2009).
past 40 years, with 4 of these occasions occurring since Our observations suggest that some polar bears are
2000 (Rockwell and Gormezano 2009). We suggest that consuming substantial quantities of eggs with dramatic
this apparent increase in the frequency of egg consumption local impacts on the reproductive success of snow geese
among bears is a result of bears coming ashore earlier than (e.g. apparently complete breeding failure in the vicinity of
they previously did, and prior to the peak hatch of nesting our camp on Southampton Island, and nearly complete
birds. failure near our Coats Island camp) and potential nutri-
Information from radio-collared bears suggests that they tional benefits to individual bears. We suggest that this is
remain on the ice hunting seals as long as possible and primarily a result of earlier ice clearance forcing bears
come ashore only when ice conditions are no longer suit- ashore prior to snow goose hatch, but other factors might
able (Stirling and Derocher 1993; Stirling et al. 1999). Our also increase the spatial and temporal overlap between
first sightings of bears at Coats Island advanced by bears and geese.
approximately 20 days between 1985 and 2007. Over this Expanding populations of snow geese mean that existing
same period, a regression of ice conditions over year sug- colonies are increasing in size, and new colonies are being
gests an advancement of 19.7 days in the date of 50% ice formed. The first record of breeding by snow geese
coverage in Hudson Bay (see Gaston et al. 2009). Thus, the on Coats Island, for example, was in 2002 (AJG and
timing of arrival on land by polar bears appears to be P. Mineau, unpubl.), and surveys as recent as 1994 found
tracking the rapidly advancing schedule of ice break-up. no evidence of nesting geese (Gaston and Ouellet 1997).
The timing of birds breeding, in contrast, has advanced at Thus, while mismatched phenology may have played a role
a more modest rate. in the overlap of bears and geese here, this colony was
For snow geese in the Hudson Bay region, nest initiation unavailable to bears previously.
dates are correlated with spring temperatures and local Still, our observations demonstrate an earlier arrival of
snow cover (Cooke et al. 1995; Skinner et al. 1998). bears to land in this region, suggest an increase in their
Although we have no information on advancement in consumption of eggs and highlight the complexity of
goose nesting dates for the areas of our observations, geese ecological interactions that may occur in a changing arctic
at the Cape Churchill Peninsula in western Hudson Bay environment. The recent intrusions of polar bears onto
have advanced their dates of nesting by less than 2 days per near-vertical cliffs to consume eggs and chicks of thick-
decade since 1968 (Rockwell and Gormezano 2009), in billed murres, a potentially hazardous situation for bears,
contrast to the changes of a week or more per decade in the further demonstrate the lengths to which these opportu-
timing of ice clearance from Hudson Bay. Likewise, the nistic animals may go to supplement their diet during a
timing of breeding of thick-billed murres at the Coats longer ice-free season.
Island colony has advanced by only 5 days since 1988
Acknowledgments Field work at all sites was funded by Environ-
(2.5 days per decade), during which time the date of 50%
ment Canada, and logistical support provided by the Polar Continental
ice cover has advanced by 9 days per decade (Gaston et al.
Shelf Project. We appreciated the helpful comments provided by
2009).
A. Tarroux, D. Berteaux, G. Donaldson and one anonymous reviewer.
Traditionally, bears in the Hudson Bay region came
ashore in late July (e.g. 23 July in 1991; Stirling and
Derocher 1993), once the eggs of most birds had hatched.
References
The accounts above describe bears observed on land as
early as 23 June (in 2004) and 24 June (in 2006) at
Berteaux D, Réale D, McAdam AD, Boutin S (2004) Keeping pace
Southampton Island and Coats Island. From 2000 2007 at
with fast climate change: can arctic life count on evolution?
Integrat Comp Biol 44:140 151
a research camp at East Bay, Southampton Island (N63°
123
Polar Biol (2010) 33:1149 1153 1153
Cooke F, Rockwell RF, Lank DB (1995) The Snow Geese of La Post E, Forchhammer MC, Bret-Harte S, Callaghan TV, Christensen
Pérouse Bay: natural selection in the wild. Oxford University TR, Elberling B, Fox AD, Gilg O, Hik DS, HÅ‚ye TT, Ims RA,
Press, Oxford Jeppesen E, Klein DR, Madsen J, McGuire AD, Rysgaard S,
Derocher AE, Andriashek D, Stirling I (1993) Terrestrial foraging by Schindler DE, Stirling I, Tamstorf MP, Tyler NJC, van der Wal
polar bears during the ice-free period in Western Hudson Bay. R, Welker J, Wookey PA, Schmidt NM, Aastrup P (2009)
Arctic 46:251 254 Ecological dynamics across the arctic associated with recent
Donaldson GM, Chapdelaine G, Andrews JD (1995) Predation of climate change. Science 235:1355 1358
thick-billed murres, Uria lomvia, at two breeding colonies by Ramsay MA, Hobson KA (1991) Polar bears make little use of
polar bears, Ursus maritimus, and walruses, Odobenus rosmarus. terrestrial food webs: evidence from stable-carbon isotope
Can Field-Nat 109:112 114 analysis. Oecologia 86:598 600
Drent RH, Prop J (2008) Barnacle goose Branta luecopsis survey on Rockwell RF, Gormezano LJ (2009) The early bear gets the goose:
Nordenskiöldkysten, west Spitsbergen 1975 2007: breeding climate change, polar bears and lesser snow geese in western
in relation to carrying capacity and predator impact. In: Hacque- Hudson Bay. Polar Biol 32:539 547
bord L, Boschman N (eds) A passion for the pole: ethological Russell RH (1975) The food habits of polar bears of James Bay and
research in polar regions. Barkhuis Publishing, Groningen southwest Hudson Bay in summer and autumn. Arctic
Gaston AJ, Hipfner JM (2000) The thick-billed murre (Uria lomvia). 28:117 129
In: Poole A, Gill F (eds) The birds of North America No. 497. Schliebe S, Rode KD, Gleason JS, Wilder J, Proffitt K, Evans TJ,
The Birds of North America, Inc., Philadelphia Miller S (2008) Effects of sea ice extent and food availability on
Gaston AJ, Ouellet H (1997) Birds and mammals of Coats Island. spatial and temporal distribution of polar bears during the fall
NWT Arctic 50:101 118 open-water period in the Southern Beaufort Sea. Polar Biol
Gaston AJ, Gilchrist HG, Hipfner JM (2005) Climate change, ice 31:999 1010
conditions and reproduction in an Arctic nesting marine bird: the Skinner WR, Jeffries RL, Carleton TJ, Rockwell RF, Abraham KF
thick-billed murre (Uria lomvia L.). J Anim Ecol 74:832 841 (1998) Prediction of reproductive success and failure in lesser
Gaston AJ, Gilchrist HG, Mallory ML, Smith PA (2009) Changes in snow geese based on early season climatic variables. Global
seasonal events, peak food availability and consequent breeding Change Biol 4:3 16
adjustment in a marine bird: a case of progressive mis-matching. Smith TG (1980) Polar bear predation of ringed and bearded seals in
Condor 111:111 119 the land-fast sea ice habitat. Can J Zool 58:2201 2209
Gauthier G, Bęty J, Giroux JF, Rochefort L (2004) Trophic Stempniewicz L (2006) Polar bear predatory behaviour towards
interactions in a High Arctic Snow Goose colony. Integrat molting barnacle geese and nesting glaucous gulls on Spitsber-
Comp Biol 44:119 129 gen. Arctic 59:247 251
Gough WA, Cornwell AR, Tsuji LJS (2004) Trends in seasonal sea- Stirling I, Derocher AE (1993) Possible impacts of climatic warming
ice duration in southwestern Hudson Bay. Arctic 57:298 304 on polar bears. Arctic 46:240 245
Hobson KA, Stirling I, Andriashek DS (2009) Isotopic homogeneity Stirling I, McEwan EH (1975) The calorific value of whole ringed
of breath CO2 from fasting and berry-eating polar bears: seals (Phoca hispida) in relation to polar bear (Ursus maritimus)
implications for tracing reliance on terrestrial foods in a ecology and hunting behaviour. Can J Zool 53:1021 1027
changing Arctic. Can J Zool 87:50 55 Stirling I, Parkinson CL (2006) Possible effects of climate warming
Lowry LF, Burns JJ, Nelson RR (1987) Polar bear, Ursus maritimus, on selected populations of polar bears (Ursus maritimus) in the
predation on belugas, Delphinapterus leucas, in the Bering and Canadian Arctic. Arctic 59:261 275
Chukchi Seas. Can Field-Nat 101:141 146 Stirling I, Jonkel C, Smith P, Robertson R (1977) The ecology of the
Mallory ML, Gaston AJ, Gilchrist HG (2009) Sources of breeding polar bear (Ursus maritimus) along the western coast of Hudson
season mortality in Canadian Arctic seabirds. Arctic 62:333 341 Bay. Canadian Wildlife Service Occasional Paper 33
Mowbray TB, Cooke F, Ganter B (2000) Snow Goose (Chen Stirling I, Lunn NJ, Iacozza J (1999) Long-term trends in the
caerulescens). In: Poole A, Gill F (eds) The birds of North population ecology of polar bears in western Hudson Bay in
America No. 514. The Birds of North America, Inc., relation to climatic change. Arctic 52:294 306
Philadelphia Stirling I, Lunn NJ, Iacozza J, Elliott C, Obbard M (2004) Polar bear
Obbard ME, Cattet MRL, Moody T, Walton LR, Potter D, Inglis J, distribution and abundance on the southwestern Hudson Bay
Chenier C (2006) Temporal trends in the body condition of coast during the open water season, in relation to population
southern Hudson Bay polar bears. Climate Change Research trends and annual ice patterns. Arctic 57:15 26
Information Note 3, Ontario Ministry of Natural Resources
123