25
Chapter 4
Biological Responses to Climate Change
An exhaustive consideration of the biological conse-
contaminants. Terrestrial vegetation also has an indirect
quences of the kinds of physical change that are pre-
impact on contaminants by altering snow accumulation
dicted for the Arctic is not feasible at this time, nor is it
and soil temperature (Sturm et al., 2001).
warranted for identifying how biological changes might
Arctic terrestrial animals have provided some of the
effect major change in contaminant pathways. Here,
clearest examples of large cycles in their populations
ecosystem changes are highlighted that appear to have a
(Krebs et al., 2001; Predavec et al., 2001) and it is
strong potential to alter the exposure of Arctic biota to
against this natural background variability that the ef-
contaminants or to alter their resilience to that expo-
fects of global change will have to be evaluated. Warmer
sure. There is general agreement that the kinds of changes
winter temperatures promote the growth of woody
discussed below have, or will, take place in the Arctic
shrubs and encourage the northward migration of the
but much less agreement concerning their probable
tree line (MacDonald et al., 1993; Serreze et al., 2000;
scope and timing. The primary intent, therefore, is to
Vörösmarty et al., 2001). Although the advance of the
provide examples of processes that ought to be included
tree line (estimated at 100 km per °C warming (IPCC,
explicitly in models and to help focus future attention on
2002)) might be expected to occur slowly over time scales
biological connections of significance to contaminants.
measured in centuries, the particular sensitivity of tun-
Whereas aquatic food webs in the Arctic exhibit en-
dra to water-table fluctuations and permafrost melt could
demic contamination from biomagnifying chemicals, ter-
produce widespread alteration in ground cover more rap-
restrial food webs are among the world's cleanest (AMAP
idly with, for example, the replacement of tundra by
2003b; de March et al., 1998). Therefore, apex feeders
vascular plants (Gorham, 1991; Rouse et al., 1997;
that adapt to change by switching between land-based
Weller and Lange, 1999). Gradual climate change can
and aquatic food webs have a particularly large potential
affect species distribution, abundance, morphology, be-
to change their exposure to contaminants such as organ-
havior, population diet and community structure (East-
ochlorine compounds (OCs) and mercury. Humans prob-
erling et al., 2000; Predavec et al., 2001). Although there
ably provide the best example of such flexibility but other
appears to be no compelling evidence of recent large
animals (e.g., Arctic foxes (Alopex lagopus) and grizzly
change in the Arctic tundra ecosystem, models suggest
bears (Ursus arctos)) can also adjust diet to opportunity.
that tundra may decrease to one third of its present size
(Everett and Fitzharris, 1998).
Warmer summer temperatures are likely to promote
4.1. Terrestrial systems
forest fires which will be accompanied by direct emis-
In this report terrestrial systems are defined as including
sions of polyaromatic hydrocarbons (PAHs), polychlori-
forests, grasslands, tundra, agricultural crops, and soils.
nated dibenzo-p-dioxins (PCDDs), polychlorinated di-
Surface-air exchange between airborne contaminants
benzofurans (PCDFs), and other POPs produced by
and terrestrial systems is important in the overall fate
combustion (see for example, Gribble, 1994; Yunker et
and long-range transport of chemicals, especially for the
al., 2002). Forest fires will also damage terrestrial soils
semi-volatile chemicals which are split between the gas-
leading to erosion and an increased release of organic
eous and condensed states. As a result of their high or-
carbon, which in turn affect aquatic systems.
ganic content, terrestrial phases (e.g., soils, forests, grass-
lands) act as reservoirs for many persistent organic pol-
4.2. Aquatic systems
lutants (POPs) (Simonich and Hites, 1994), particularly
4.2.1. Lakes, rivers and estuaries
polychlorinated biphenyls (PCBs), DDT, hexachlorohex-
ane (HCH) and chlorobenzenes (AMAP, 2003b). Air-
The changes in snow and ice cover and in the hydrologi-
surface exchange of POPs into terrestrial phases is a dy-
cal cycle will alter the light and nutrient climate of fresh-
namic process that controls air burdens of chemicals.
water systems. These changes together with loss of per-
Thus any change in the extent of vegetation cover asso-
mafrost, which will enhance the supply of nutrients and
ciated with global warming will have implications for
particulates to lakes, will increase aquatic productivity
contaminant fate and transport. Wania and McLachlan
and particle flux (Douglas et al., 1994; McDonald et al.,
(2001) have shown that forests have a unique ability to
1996; Schindler, 1997). Although the spring bloom will
mitigate atmospheric concentrations of OCs by `pump-
probably advance with early loss of ice cover, hydrologi-
ing' chemicals from the atmosphere into foliage and
cal processes in a lake's drainage basin will probably
thence to a long-term reservoir in forest soil. This pro-
also advance. Increased summer temperatures will dis-
cess is likely to be most important for OC compounds
advantage fish such as trout (Salmo spp.) and grayling
with log KOA of ~9 to 10 and log KAW ~ 2 to 3 (where
(Thymallus arcticus) whereas winter temperature in-
KOA and KAW are octanolair and airwater partition co-
crease may enhance microbial decomposition. Shifts in
efficients see Wania, 2001). Because these key proper-
the seasonal light/temperature cycle may also advantage
ties are strongly temperature-dependent (see section 6.3.4.
or disadvantage species lower in the food web including
for greater detail) even a small change in climate may
phytoplankton, zooplankton and insects. Change in
alter the dynamics of this process and thus the cycling of
water level will have obvious effects on important fish
26
AMAP Assessment 2002: The Influence of Global Change on Contaminant Pathways
Ice algae
Kelp
Phyto-
plankton
Other
Ice
Copepods
Mya
Other
zooplankton
amphipods
truncata
bivalves
Other
Parathemisto
Chaetognaths
Arctic cod
Gelatinous
Walrus
Other
amphipods
spp.
zooplankton
benthos
Guillemot/
Fulmar
Murre
Ringed
Narwhal
Beluga
Harp
Bearded
Kittiwake
seal
seal
seal
Figure 4·1. A simplified
Fish, seals, whales
schematic diagram show-
ing the marine food web
(based on Welch et al.,
Gulls
Polar bear
Humans
Killer whale
1992).
stocks, especially species dependent on small refugia for
There are too many examples of how ice climate
over-wintering (Hammar, 1989) or species dependant on
variation can affect ecosystem structure to list them all
freshwater coastal corridors for their life cycle; the Arc-
(see for example Sakshaug and Slagstad, 1992; Tynan
tic cisco (Coregonus autumnalis) provides a relevant ex-
and DeMaster, 1997) and it is not likely that all changes
ample of the latter (see Gallaway et al., 1983). Warming
that have occurred in Arctic systems have been ob-
and loss of nearshore or estuarine ice in the Beaufort Sea
served. The thickness and distribution of ice can influ-
may eliminate indigenous fish which are then replaced
ence the amount of organic carbon produced, the types
by anadromous fish from the Pacific Ocean (see Babaluk
of algae that produce it, and connections between the
et al., 2000). Although warming is likely to result in
algal production and communities in the water column
widely-distributed shifts in zoogeographic distributions
or sediments (Niebauer and Alexander, 1985). Ice con-
that have the potential to affect every step in the fresh-
trols wind mixing and light penetration especially when
water food chain, prediction will probably founder on
covered with snow, and it may also support upwelling
`counter-intuitive' surprises (Schindler, 1997).
at the ice edge but suppress upwelling beneath the ice.
Through its annual cycle, ice formation decreases strati-
fication in winter but increases stratification when the
4.2.2. The ocean
ice melts in spring. These physical factors impact upon
The effect of ice on Arctic marine ecosystems has long
the nutrient supply to surface water, the light climate,
been understood by those who harvest the sea (Bock-
and the water stability which together control primary
stoce, 1986; McGhee, 1996; Scoresby, 1969; Vibe,
production. Furthermore, mats of algae that grow on
1967). Change in ice climate, therefore, has a large po-
the bottom of the ice support an epontic food web that
tential to modify marine ecosystems, either through a
ultimately feeds Arctic cod (Boreogadus saida), ringed
bottom-up reorganization of the food web by altering
seals (Phoca hispida) and polar bears (Ursus maritimus)
the nutrient or light cycle, or a top-down reorganization
or, alternatively, by being shed from melting ice in spring,
by altering critical habitat for higher trophic levels (see,
support a benthic food web that feeds molluscs, walrus
for example, Parsons, 1992). Any reorganization that
(Odobenus rosmarus), bearded seals (Erignathus barba-
changes the number of trophic levels in the food web or
tus) and king eiders (Somateria spectabilis). Similarly, pri-
alters the flow of carbon between, for example, pelagic
mary production within the water column may be par-
and benthic food webs would have particular signifi-
tially grazed to support a pelagic food web, or descend
cance for contaminants that biomagnify, such as mer-
ungrazed and, together with fecal pellets and zooplank-
cury and the OCs; the complexity of the interaction be-
ton carcasses, feed the benthos (Grebmeier and Dunton,
tween ice and aquatic ecosystems provides much scope
2000). The bifurcation between pelagic and benthic food
for alterations in contaminant pathways (Figure 4·1).
webs is strongly influenced by the distribution of ice and
Arctic and sub-arctic marine ecosystems are also altered
its impact on nutrient and light climates. These proces-
by ocean climate changes such as regime shifts involving
ses, which have great potential to alter the timing and
the displacement of water masses and associated popula-
amounts of vertical particle flux in the ocean, are likely
tions or temperature change (Figure 4·2; Dippner and
to have a greater impact on the sequestering of POPs
Ottersen, 2001; Hare and Mantua, 2000; Helland-
into the Arctic Ocean than air-sea exchange or the so-
Hansen and Nansen, 1909; Hunt et al., 1999; Loeng,
called `cold-condensation' effect. Dachs et al. (2002) show
2001; Saar, 2000; Sakshaug et al., 1991, 1994).
that in mid-latitudes, sinking particulate matter, which is
Biological Responses to Climate Change
27
the dominant export pathway for POPs from the ocean
production and less new production. The loss of rela-
surface layer, drives deposition at the ocean surface.
tively large diatoms could reduce the size of herbivores,
Shifts in benthic species distribution due to tempera-
potentially inserting an extra `small-carnivore' step at
ture, carbon flux or other climate-related change have
the bottom of the food web which would increase the
the potential to alter completely the coupling between
number of trophic levels. Because biomagnification of
sediments and bottom water. In one well-documented
OCs is often exponential (Fisk et al., 2001a), slightly
example from a temperate region, the invasion of Echi-
higher concentrations at low trophic levels (e.g., cope-
ura (Listriolobus pelodes) into coastal benthic commu-
pods) can have a large impact on apex feeders. Stratifi-
nities off California, for as yet unknown reasons, re-
cation, which is altered at the basin scale under AO/
sulted in aerated and biomixed sediments that reduced
NAO shifts, affects plankton composition and vertical
the evidence of wastewater impacts regionally (Stull et
flux dramatically as evident from studies in the Barents
al., 1986).
Sea (Wassmann, 2001). For example, Wassmann et al.
(1990) showed that algal blooms by Phaeocystis sp.
along the Greenland coast and in the Barents Sea tend
4.2.2.1 Bottom-up trophic change
not to get grazed resulting in a large transfer of organic
The projected loss of ice for the Arctic Ocean, particu-
carbon to the benthos. Climate change in the form of ei-
larly over the shelves, intuitively should increase primary
ther loss of ice cover or increase in stratification has the
production in the marginal seas through enhanced mix-
potential to alter the quantity of available food and to
ing, light penetration and upwelling. In other words,
redistribute its flow between epontic, pelagic and ben-
Arctic shelves would begin to look more `temperate'.
thic habitats.
Greater new production implies greater particle flux and
The Bering Sea provides another outstanding exam-
greater secondary production, but the complexity of ma-
ple of recent change from the bottom-up permeating an
rine ecosystems should forewarn of possible surprises.
entire ecosystem. In view of the Bering Sea's vulnerabil-
Massive blooms of jellyfish were observed in the Bering
ity to airborne contaminants from Asia (Bailey et al.,
Sea during in the 1990s (Brodeur et al., 1999; Hunt et
2000; Li et al., 2002), it is particularly regrettable that
al., 1999) and their emergence was ascribed to sea-sur-
the observations of ecosystem change since the 1970s
face temperature increase and loss of ice cover the
were not matched by contaminant pathway studies. Evi-
same two key changes poised over the Arctic Ocean.
dence from stable isotope records in bowhead whale
Parsons (1979) has drawn attention to the funda-
(Balaena mysticetus) baleen suggests that the carrying
mental ecological differences between western seaboards
capacity of the Bering Sea ecosystem began to decline in
in the Northern Hemisphere, where coastal water ex-
the mid 1970s (Figure 4·2a, Schell, 2000). This change
hibits divergence and upwelling, and eastern seaboards
may relate to a larger picture of change throughout the
which are convergent. The former have been of greater
North Pacific (Hare and Mantua, 2000) and, in particu-
commercial interest but are also characterized by jelly-
13C
fish (Parsons, 1979). The Arctic Oscillation (AO) does
14
not cause reversal of large-scale wind circulation but
Old partial baleen plate from Panuk Island
Old partial baleen plate from St. Lawrence Island
does produce more divergent Arctic Ocean margins un-
15
der AO/anticyclonic conditions and less divergent mar-
Baleen from 37 whales taken over past four decades
gins under AO+/cyclonic conditions. The inherently noisy
16
events of coastal upwelling and downwelling could then
17
act together with the AO in a form of `stochastic reso-
nance' (Rahmstorf and Alley, 2002) to enhance upwel-
18
ling during AO conditions. This enhanced upwelling
might then have the capacity to produce large-scale mo-
19
dal shifts in shelf ecosystems and their commercial po-
20
tential. Changes in ocean climate, such as those associ-
1945
1955
1965
1975
1985
1995
ated with the AO/NAO, have long been known to affect
Percent
fisheries in sub-polar seas either directly through water
100
Shrimp
property changes (T, S) or indirectly through changes in
community structure (Hare and Mantua, 2000; Klyash-
80
Flatfish
torin, 1998; Marteinsdottir and Thorarinsson, 1998).
A dramatic example of large-scale, bottom-up bio-
60
logical change was witnessed during the SHEBA drift
across the Beaufort and Chukchi Seas in 1997 to 1998
40
(Melnikov et al., 2002). Compared to Soviet observa-
Gadoid
tions from drifting stations that passed over the same re-
20
gion 20 years earlier, there was a marked decrease in
large diatoms in the water column and microfauna
0
Others
within the ice. The freshening and strong stratification
1975
1980
1985
1990
of the surface water, due to river discharge diverted into
Figure 4·2. Examples of significant change within Arctic and sub-
the basin under the strong AO+ conditions of the early
arctic ecosystems. This figure illustrates a) a change in the food web
structure probably commencing during the 1970s as reflected by a
1990s, reduced the supply of nutrients from below, and
decrease in the 13C of baleen from bowhead whales (Schell, 2000)
promoted species more typical of freshwater ecosystems.
and b) significant fish population changes in the Gulf of Alaska
Consequentially, there was a high proportion of recycled
31
since 1970 (modified from Anderson and Piatt, 1999).
28
AMAP Assessment 2002: The Influence of Global Change on Contaminant Pathways
lar, to the switch in the Pacific Decadal Oscillation
the period of pelagic drift from the main spawning
(PDO) from cold to warm phase in the mid 1970s. The
grounds in the Lofoten area in northern Norway to the
change in regime rapidly permeated the entire ecosystem
nursery grounds in the Barents Sea leads to above aver-
of the Bering Sea altering fish community structure
age year-class strength. This was attributed in part to
(shrimp and crab populations declined while pollock,
temperature and in part to added supply of zooplank-
cod and flatfish populations increased significantly Fig-
ton-rich water from the Norwegian Sea into the feeding
ure 4·2 b), and seabird and mammal populations (Sprin-
areas of the Barents Sea.
ger, 1998). More recently, blooms of small phytoplank-
For older fish, other factors may contribute to inter-
ton (Emiliania huxleyi) were observed in 1997 and 1998
annual variation. During periods of high abundance,
(Saar, 2000). Because these phytoplankton are smaller
fish density may cause the geographic range of Arcto-
than the diatoms that typically bloom in the Bering Sea,
Norwegian cod to expand or shift. Temperature has
they were grazed by smaller copepods instead of larger
been reported to cause displacement of Arcto-Norwe-
euphausids which in turn probably led to die-offs of the
gian cod toward the east and north during warm peri-
short-tailed shearwaters (Puffinus tenuirostris) that feed
ods and toward the south-western part of the Barents
on the latter (Stockwell et al., 1999). Similarly, the alter-
Sea in cold periods (Ottersen et al., 1998). These shifts
ation of primary production both in quantity and distri-
may not be caused by temperature itself but, rather, by
bution probably decreased food availability for fish,
temperature-induced changes in the distribution of prey
whale, seal and walrus populations forcing die-offs, mi-
organisms (Ottersen, 1996).
gration or redistribution throughout the food web (Bots-
Loeng (2001) has discussed the types of change that
ford et al., 1997; Grebmeier and Cooper, 1995; Greb-
may well occur in the Nordic seas should ocean temper-
meier and Dunton, 2000; Hare and Mantua, 2000; Rugh
ature rise, as projected, by 1 to 2°C (Figure 4·3). In the
et al., 1999; Stabeno and Overland, 2001). Large as these
Barents Sea the feeding area of capelin will be displaced
ecosystem changes appear to have been, they may pale
to the northeast and the spawning ground may move
in comparison to the natural fluctuations that have oc-
eastward along the northern coast of Russia. Cod will
curred during the past two millennia (Finney et al.,
distribute more toward the northeast partly because ac-
2002). Furthermore, these long-term proxy data provide
ceptable ocean temperature will be found there and
a strong warning that relationships between biological
partly because their main food item, capelin, will move
populations and physical forcing established from short
in that direction. These displacements will put such
observational records may not hold up over a longer pe-
stocks closer to contaminant sources in the eastern Bar-
riod when other non-linear factors may have a chance to
ents Sea. In the Norwegian Sea, the Norwegian spring
operate (initial conditions, or other cyclical forcing, for
spawning herring (Clupea harengus) may return to the
example). Clearly, the dramatic changes in the Bering
migration route they used prior to the mid 1960s, when
Sea system could spill over into the Chukchi Sea, and the
ocean temperature around Iceland was over 1°C higher
decline of Bering inflow by ~15% since the 1940s (Fig-
than today. Presently, adult herring over-winter in a
ure 3·20 a) suggests a matched decline in new and ad-
Norwegian fjord before commencing their spawning mi-
vected production in the Chukchi Sea simply due to re-
gration along the Norwegian coast followed by a feed-
duced nutrient and organic carbon supply.
ing migration into the Norwegian Sea. With temperature
In the Barents and Nordic Seas it has long been rec-
increase, the herring may over-winter after the feeding
ognized that fish populations respond to climate vari-
migration just east of Iceland as they did before 1965,
ability (Helland-Hansen and Nansen, 1909). Indeed, the
distribution of capelin (Mallotus villosus), the single
most important food species for Arcto-Norwegian cod,
is known to vary from year to year dependent on the in-
flow of Atlantic water (Sakshaug et al., 1994). Fluctua-
Capelin
tions in large- and regional-scale atmospheric pressure
conditions affect winds and upper ocean currents (Fig-
Atlantic cod
ures 3·2 and 3·17), modify water temperature, alter drift
patterns of fish larvae, and change availability of prey
items. Mixing during summer alters the nutrient cycle
N O R W E G I A N S E A
and the coupling between primary production and ben-
thos (Peinert et al., 2001; Wassmann, 2001). Details are
important. For example, while long and unrestricted lar-
Herring
val drift is crucial for the Arcto-Norwegian and Ice-
landic component of cod stocks at West Greenland, lar-
val retention on favourable banks is the key for recruit-
Mackerel
Mackerel,
ment to stocks residing in small and open systems (Ot-
bluefin tuna
tersen, 1996). The 600 to 1200 km drift of Arcto-Nor-
North Sea
wegian cod larvae from spawning grounds to nursery
herring
grounds where they settle on the bottom provides much
opportunity for interannual variation; pelagic juveniles
Anchovy, sardine
in the Barents Sea exhibit a typical westerly distribution
in some years while, in other years they distribute to the
Figure 4·3. Possible changes in the distribution of selected fish spe-
east (Ådlandsvik and Sundby, 1994). Ottersen and Sund-
cies in the Nordic and Barents Seas resulting from an increase in sea
by (1995) showed that southerly wind anomalies during
temperature of 1 to 2°C (modified from Loeng, 2001).
Biological Responses to Climate Change
29
and not return to the Norwegian fjord, completely alter-
change. Polar bears rely on ringed seals for food, and
ing their exposure to contaminants. New species may in-
ringed seals prefer landfast or stable first-year ice for
vade. Presently, mackerel (Scomber scombrus) is scarce
pupping (Finley et al., 1983; Stirling, 2002; Wiig et al.,
along the coast of northern Norway, but with ocean
1999). The loss of landfast ice in spring, the loss of food
warming might migrate as far as the Barents Sea. Blue
supply for seals, or the inability of bears to access seals
whiting (Micromesistius poutassou), bluefin tuna (Thun-
during the few critical weeks in spring when pupping oc-
nus thynnus) and sharks (Elasmobranchia) may also be-
curs, means life or death and can produce large popula-
come more frequent visitors to this area.
tion shifts (Harwood et al., 2000; Smith and Harwood,
2001; Stirling et al., 1999). In Hudson Bay, bears proba-
bly accumulate most of their annual energy require-
4.2.2.2. Top-down trophic change
ments during the few months of late spring prior to
Ice-covered seas have a unique capacity for top-down
breakup when they can access older pre-weaning ringed-
trophic change. To understand and predict how the par-
seal pups or naïve post-weaning pups exactly the pe-
tial or complete loss of ice will impact upon the trophic
riod of time that has seen recent dramatic change (Figure
structure requires a detailed understanding of how top
3·23). Furthermore, permafrost is a critical habitat for
predators take advantage of ice (Carmack and Macdon-
bears because they dig maternity dens in frozen peat,
ald, 2002; Lowry, 2000; Vibe, 1967). In an incisive re-
and this habitat is threatened by warming or increased
view, Tynan and deMaster (1997) discuss how whales,
incidence of fire initiated by more frequent lightning
walrus, seals, bears and cod, are likely to be affected by
strikes. In Hudson Bay, at the southern limit of their
change in ice climate and show that their response to
population, polar bears presently appear to be in a very
change depends on how `plastic' their dependence on ice
precarious position (Stirling and Derocher, 1993; Stir-
might be.
ling et al., 1999).
Change in the landfast ice may give the advantage to
Arctic cod is the most important forage fish in the
either seals or to bears (Carmack and Macdonald, 2002)
Arctic Ocean food web (Figure 4·1; Bradstreet et al.,
with the result that Arctic cod would be subject to more,
1986; Tynan and DeMaster, 1997; Welch, 1995). The
or less, predation, respectively. Walrus use drifting ice to
loss of ice, in either the marginal seas or, as projected by
haulout in winter because it provides better access to
models, for the entire ocean (Figure 2·2 b, Flato and
benthos, but they also use terrestrial haulouts in ice-free
Boer, 2001), would have a massive impact on the distri-
periods, perhaps with detrimental energy costs (Lowry,
bution and life history of Arctic cod and, therefore, on
2000; Tynan and DeMaster, 1997). In contrast, eiders
seals, beluga (Delphinapterus leucas) and birds who de-
(Somateria spp.) and other benthic-feeding birds prefer
pend heavily on them. One thing is clear: the ice edge is
open water with a relatively shallow bottom (< 50 m)
an especially critical habitat for cod and marine mam-
(Dickson and Gilchrist, 2002; Grebmeier et al., 1988;
mals and it is this region that is most vulnerable to
Suydam et al., 2000). Loss of ice (landfast or drifting) in
change.
critical regions or at critical times of the year, or move-
Finally, climate change can alter the routes and desti-
ment of the ice edge to deeper water where benthos can
nations of migratory species. For example, under the
no longer be accessed, therefore, can mean a substantial
AO+ conditions of the early 1990s, Pacific salmon (On-
rearrangement of the top of the food web advantaging
corhynchus spp.) began to enter Arctic rivers (Babaluk
some animals, disadvantaging others and possibly caus-
et al., 2000). Similarly, bowhead whales and belugas
ing wholesale migration (Dyke et al., 1996b, 1999;
range widely in search of food and their range varies
Dyke and Savelle, 2001; Fay, 1982; Lowry, 2000; Moore
enormously in time and space with changes in ice
and Clarke, 1986; Tynan and DeMaster, 1997; Woollett
climate (Dyke et al., 1996b; Dyke and Savelle, 2001;
et al., 2000). With benthos not readily available, walrus
McGhee, 1996). Nor are long migrations limited to
might turn to predation on seals thereby raising their
whales. Harp seals (Phoca groenlandica) of the North-
trophic position considerably (Muir et al., 1999), or
west Atlantic undergo 8000 km round trips to feed on
with absence of ice, killer whale (Orcinus orca) preda-
Arctic cod in Baffin Bay (Finley et al., 1990) and bird
tion on bowhead whales might decimate their popula-
species migrate inordinately long distances often de-
tion leaving their prey (zooplankton) as food for some-
pending on critical areas along their migration pathways
thing else. Early breakup in the Bering and Beaufort Seas
where they may ingest contaminants (see, for example,
during 1995 to 1998 probably led to the observed aban-
Braune et al., 1999; Savinova et al., 1995; Springer,
donment of seal pups in 1998 and the decline or starva-
1998). The extent to which migratory species are able to
tion of walrus.
adapt to potentially rapid changes in key staging areas
The Hudson Bay polar bear population provides per-
may be of critical importance to their future (Carmack
haps one of the clearest warnings of the consequence of
and Macdonald, 2002).