Crabeater seal

The crabeater is the most abundant seal species in the world (Laws Reference Laws and Llano1977). It has evolved to be a specialist in its foraging habits and feeds almost exclusively on Antarctic krill (Euphausia superba Dana) in pelagic and slope areas, and crystal krill (E. crystallorophias Holt & Tattersall) in neritic areas. So important have krill become to its diet that its dentition has evolved to filter krill from sea water. The mating and pupping season begins in October and probably continues to early November, with the exact timing varying by region (Southwell et al. Reference Testa, Oehlert, Ainley, Bengston, Siniff, Laws and Rounsevell2003). During this period females occupy ice floes that are selected for their size and physical characteristics (Siniff et al. Reference Siniff, Stirling, Bengtson and Reichle1979). Here, a female gives birth to a pup and is soon joined by a male, who is apparently waiting for the pup to be weaned and the female to enter oestrus and become available for mating. The ice floe occupied must be large enough that it does not break up until the pup is weaned in about 4–6 weeks. Also, the selected floes usually have some structure that the female uses to advantage in order to thwart the advances of the male who constantly attempts to separate her from the pup (forced weaning), thus, to advance the onset of oestrus. The surface relief may also provide a visual shield against both leopard seals and killer whales (Orcinus orca (L.)), which are major predators (Siniff & Bengtson Reference Siniff and Bengtson1977). Killer whales have been known to create waves that displace seals from floes (Smith et al. Reference Smith, Siniff, Reichle and Stone1981, Pitman & Ensor Reference Pitman and Ensor2003, Visser et al. 2008). Of the four species of pack ice seals, the crabeater may be the most vulnerable to predation. Leopard seals focus particularly on pups, and killer whales are a constant threat to all age classes. In nearly every region of the Antarctic Ocean, a significant proportion (60–70%) of crabeater seals show scars that indicate escape from a leopard seal (Siniff & Stone Reference Siniff, Stone, Siegfried, Condy and Laws1985). Thus, the structure and size of ice floes are very important to protect this species from predation.

Any environmental changes that will affect the availability of Antarctic krill will probably have an impact on crabeater seals (see similar comments for certain seabirds in Croxall et al. Reference Croxall, Trathan and Murphy2002). It is well known that Antarctic krill are highly dependent on sea ice, particularly during winter, when juvenile krill feed on under-ice microbial organisms and escape predation; as sea ice disappears so has the local abundance of Antarctic krill (Fraser & Hofmann Reference Fraser and Hofmann2003, Atkinson et al. Reference Atkinson, Siegel, Pakhomov and Rothery2005). On the other hand, on more restricted continental shelf areas, more persistent and growing coastal polynyas should lead to increased prevalence of crystal krill (see Pakhomov & Perissinotto Reference Pakhomov and Perissinotto1996).

Areas where sea ice has and will continue to exhibit the above changes should exhibit a decline in crabeater abundance due to direct (pupping platform and protection from predators) and indirect (food supply) influences. On the other hand, as some changes may initially increase the availability of crystal krill, regional contradictions in population trends could become evident.

Leopard seal

The leopard seal has a diverse diet and is a major predator on many vertebrate species in the sea ice regions (Laws Reference Laws and Llano1977). Its diet choice varies seasonally and spatially, with krill sometimes dominating during winter, switching to crabeater seal pups during spring, and focusing on fledgling penguins as they go to sea in late summer. Fish are also a common prey item and can be taken throughout the annual cycle. The leopard seal is known to visit islands in the lower latitudes, and sometimes remains there throughout the year. Many of these are juveniles and often in poor condition. Whether more adults would join this trend as climate change proceeds is unknown, but such behaviour modification should be considered (Bester et al. Reference Bester, Hofmeyr, Kirkman, Chauke, De Bruyn, Ferreira, Makhado, Maswime, McIntyre, Mulaudzi, Munyai, Pistorius, Radzilani, Ramunasi, Tshithabane and Wilkinson2006). In spite of this general trophic pattern (Siniff & Stone Reference Siniff, Stone, Siegfried, Condy and Laws1985), some individuals become specialists and remain in one region, concentrating on a particularly vulnerable prey such as penguins (Kooyman Reference Kooyman1965). Such behaviour has been documented also in the case of predation on fur seal (A. gazella) pups (Hiruki et al. Reference Hiruki, Schwartz and Boveng1999).

The pupping season occurs in spring, but few observations have been made of pups with females (Siniff & Stone Reference Siniff, Stone, Siegfried, Condy and Laws1985, Southwell et al. Reference Testa, Oehlert, Ainley, Bengston, Siniff, Laws and Rounsevell2003). Based on appearance, it is hypothesized that the pup can enter the water fairly early in the lactation period and may accompany the female as she forages. Breeding may not follow weaning immediately, as is the case with the other pack ice species, but may be postponed, perhaps as much as a month after weaning has occurred (Siniff & Stone Reference Siniff, Stone, Siegfried, Condy and Laws1985). However, the exact timing of oestrus and breeding needs further data.

Leopard seals are known to use many different types and sizes of ice floes for resting, pupping and moulting. They are seen on bits of ice bergs as well. They are preyed upon by killer whales (see Ainley et al. Reference Ainley, Ballard, Karl and Dugger2005a), so larger, more protective ice floes are probably selected if available. Satellite telemetry shows that individuals may stay within one general area for extended periods (Rogers et al. Reference Rogers, Hogg and Irvine2005).

In considering the four factors concerned with ice in our sensitivity matrix, we predict that changes in any of them will not have a large influence on this species. Sea ice persistence and floe size and structure would seem to have little effect because of the ability of this species to use many different ice floe types and move location if ice floe availability becomes limited. Because of their diverse diet, food shortages may not be a problem and thus we have lowered the degree of negative impact (Table I). However, we offer a caveat to this assessment as there is a lack of information on the diet of juveniles, and juvenile survival may be a critical life-history stage.

Table I.

A matrix summarizing how environmental change may influence life-history patterns and, ultimately, the population dynamics of Southern Ocean seals. Arbitrary values between ± 4 were chosen to show the predicted degree of effect. Positive values indicate improved conditions because of climate change, while negative values indicate detrimental effects. “Food web changes” relate only to the effect of climate change on the pack ice ecosystem and ultimately krill availability.

Weddell seal

This species spends much of its time close to the Antarctic continent, where colonies form in early spring along predictable cracks in the persistent fast ice. In late winter and early spring, prior to pupping, these seals spend much of their time under this fast ice and use their teeth to keep breathing holes open where the ice is thin near perennial tide cracks. At high latitudes, e.g. McMurdo Sound (78°S), females haulout in late October and early November to give birth; at lower latitudes, e.g. the South Orkneys (60°S), pupping is in September (Testa et al. Reference Thomas, Demaster, Stone and Andriashek1990). The pups are weaned 6–7 weeks after birth, and copulation occurs under the ice. As the fast ice begins to deteriorate, adults and newly-weaned pups disperse. Satellite tracking of McMurdo Sound animals shows that some individuals move north into the Ross Sea (Burns et al. Reference Burns, Castellini and Testa1999, Ainley et al. Reference Ainley, Toniolo, Ballard, Barton, Eastman, Karl, Focardi, Kooyman, Lyver, Olmastroni, Stewart, Testa and Wilson2006). Weddell seals are known to feed principally on Antarctic silver fish (Pleuragramma antarcticum Boulenger), the Antarctic toothfish (Dissostichus mawsoni Norman) and several species of cephalopods (Burns et al. Reference Burns, Trumble, Castellini and Testa1998, La Mesa et al. Reference La Mesa, Eastman and Vacchi2004, Kim et al. Reference Kim, Conlan, Malone and Lewis2005, Ainley et al. Reference Ainley, Toniolo, Ballard, Barton, Eastman, Karl, Focardi, Kooyman, Lyver, Olmastroni, Stewart, Testa and Wilson2006, Pongannis & Stockard Reference Ponganis and Stockard2007).

In McMurdo Sound, where all pups born to this population have been marked since the 1970s and annual estimates of adults present have been made, relationships to several ice-related factors that affect population trends, behaviour and physiology have become obvious (Cameron & Siniff Reference Cameron and Siniff2004). Testa et al. (Reference Testa, Siniff, Croxall and Burton1991), linking climatological patterns to the dynamics of this population, found a relationship to El Niño–Southern Oscillation (ENSO), probably mediated through changes in ice extent and concentration (lower extent during ENSO; see Ledley Reference Ledley1997, Jacobs & Giulivi Reference Jacobs and Giulivi1998). More recently, elaborating on this relationship and with a longer time series, Hadley et al. (Reference Hadley, Rotella and Garrott2007a) reported that pup incidence among females tended to be lower when sea ice was more extensive in the previous year's post-weaning period and hypothesized that more sea ice in that period would lead to decreased foraging success for pregnant females. Possibly the decreased foraging success with more pack ice is related to the reduced amount of open water available for phytoplankton blooms and primary productivity (Arrigo & van Dijken Reference Arrigo and Van Dijken2004), which ultimately should affect fish availability. In support of this explanation, Proffitt et al. (Reference Proffitt, Garrott, Rotella, Siniff and Testa2007) demonstrated that foraging success of pregnant Weddell seals (reflected by the weaning mass of their eventual pups) decreased during summers characterized by extensive sea ice. Further, Hadley (Reference Hadley2006) found modest evidence that years of extensive winter sea ice were followed by years of higher survival among adult females. This is in accord with the increased productivity hypothesis, in that winters of extensive ice (which tend also to have more concentrated ice) are followed by summers of minimal ice (Jacobs & Comiso Reference Jacobs and Comiso1989). Finally, Hadley et al. (Reference Hadley, Rotella and Garrott2007b) reported that females were more likely to give birth to their first pup during a spring following a winter with more sea ice in the Ross Sea region, i.e. a winter often following a summer of minimal ice.

For Weddell seals, insight into the possible demographic consequences of long-term fluctuations in fast ice persistence was provided when a large portion of the Ross Ice Shelf broke off in late 2000. The resulting iceberg blocked the usual advection of sea ice from McMurdo Sound until winter 2006. As a result, the fast ice increased in extent, thickness, and seasonal persistence. This is indicated by the relative difficulty for an icebreaker to make a channel to McMurdo Station (Fig. 1a). Weddell seals were directly affected, presumably from the increased ice thickness, which in some cases closed off cracks that had been predictably present in previous years. As a result, fewer adults showed up for breeding and fewer pups were born (Fig. 1b). It is unknown whether McMurdo Sound females pupped elsewhere. This could certainly be the case, but, if so, based on normally high philopatry, it seems likely that pupping elsewhere would result in higher pup mortality. Over a 10-year period a strong negative relationship existed between the proxy of ice thickness and the number of pups born (r = -0.93, P < 0.01; Fig 1b). However, it is interesting to note how quickly the number of pups born to this population returned to previous levels when the effect of the iceberg was removed. This suggests that the seals were in the vicinity but, in the situation of reduced productivity owing to regionally more extensive, concentrated, and persistent pack ice (Arrigo & van Dijken Reference Arrigo and Van Dijken2004), were “choosing” not to breed. No doubt other species in the fast ice ecosystem, which have evolved to take advantage of the stable conditions that exist in fast-ice areas, were also affected (e.g. Dayton Reference Dayton1989).

In contrast, in the Antarctic Peninsula region over the past few decades, sea ice has diminished in extent, concentration, and especially persistence (Zwally et al. Reference Zwally, Comiso, Parkinson, Cavalieri and Gloersen2002, Parkinson Reference Parkinson2002, Stammerjohn et al. Reference Stammerjohn, Martinson, Smith, Yuan and Rind2008; Fig. 2a). Weekly counts of seals instituted in 1974 from various shore locations in Arthur Harbour (vicinity of Palmer Station) show that the number of Weddell seals has gradually declined (Fig. 2b; seals vs year, r_s = -0.896, P < 0.001). We do not know the exact reason for the decline but presume it has to do with the reduction in the region of the amount of fast ice necessary for breeding (Fig. 1a, ice extent, r_s = 0.457, p = 0.07). The general picture that emerges for Weddell seals with respect to the sensitivity matrix suggests that this species will be influenced either by increasing thickness of near-shore ice, or change in prevalence or persistence of fast ice areas. That is, any environmental event that holds potential to alter what historically have been seasonally predictable fast ice areas will greatly influence this species through the pupping and breeding seasons and thus we scored the maximum effect for sea ice persistence (Table I). We predict that both sea ice extent and ice floe size and structure would have less effect upon this species. In this case we felt that floe size and structure would be most important for protection from killer whale predation. In neritic areas, larger polynyas could mean enhanced productivity and availability of prey, mediated through crystal krill.

Fig. 2.

Ice area in the “Long Term Ecological Research (LTER) study area,” west coast of Antarctic Peninsula, 1979–2004, as determined by satellite imagery (Spearman correlation, ice vs year: r = -0.43; p = 0.03; data from Ducklow et al. Reference Ducklow, Baker, Martinson, Quetin, Ross, Smith, Stammerjohn, Vernet and Fraser2007). The maximum number of pinnipeds of three species counted among all censuses in a given year on the islands within that study area: Arthur Harbour, Anvers Island, 1974–2005 (see text for statistical relationship between ice and seal numbers).

Ross seal

This species is much less abundant than the other pack ice seals. It is highly pelagic, seldom seen, and large concentrations have never been noted. Few pups have been sighted, but the occasional records suggest pupping occurs on ice floes in spring, with exact timing dependent on the region of the Antarctic (Thomas et al. Reference Van Den Hoff and Morrice1980, Southwell et al. Reference Testa, Oehlert, Ainley, Bengston, Siniff, Laws and Rounsevell2003). Like the other species, the pupping season probably varies with location. During a cruise into the Amundsen Sea sector during late 1999 and early 2000 (Ackley et al. Reference Ackley, Bengtson, Boveng, Castellini, Daly, Jacobs, Kooyman, Laake, Quetin, Ross, Siniff, Stewart, Stirling, Torres and Yochem2003), Ross seals were found moulting on very large ice floes, often a kilometre across (Siniff, personal observation). Their appearance at this time was of sloughing skin similar to elephant seals. These large floes may be very critical at this time, as the seals may be at a disadvantage energetically if forced to enter the water. In regard to diet, analyses indicate that Ross seals feed predominantly on cephalopods. Recent satellite tracking indicates that this species may move north and beyond 60^oS, and remain in ice-free waters for some time (Blix & Nordoy Reference Blix and Nordoy2007).

With respect to our sensitivity matrix, only ice floe size during pupping and moulting is probably important to this species, and we have ranked this factor the most detrimental. Sea ice concentration co-varies with ice floe size and thus we considered this factor next most important (Table I). The effects of changes in sea ice on Ross seal foraging seems unlikely to be detrimental since they often forage at considerable distances from the continent, beyond the sea ice, and thus may be preying on species not particularly tied to the pack ice ecosystem.

Southern elephant seal

This species breeds on beaches, including pebble, cobble and boulder beaches, as well as on vegetated areas behind landing beaches, and exceptionally on ice (e.g. Laws Reference Laws1956, Condy Reference Condy1978a). As the females come ashore, groups coalesce, sometimes with several hundred on one beach. Meanwhile, males compete for supremacy to insure access to females. One “beachmaster” might control 40–50 females depending on structure of the beach. Females give birth to pups within a few days after arriving. They enter oestrus about 19 days after giving birth and are receptive for breeding for about four days. They often copulate with the dominant male of the area, although low ranking males may also participate in the breeding. Females remain with the pup until weaning (23 days after being born), which usually occurs in about 30 days after hauling out. At this time females leave the beach to feed and recover weight lost during lactation, after which they return to moult. Males return to moult later than females; younger animals return to moult before the adults (Condy Reference Condy1979, Kirkman et al. Reference Kirkman, Bester, Pistorius, Hofmeyr, Jonker, Owen and Stryydom2003a, Reference Kirkman, Bester, Hofmeyr, Jonker, Pistorius, Owen and Stryydom2004). The moulting process takes a few weeks during which they seldom go to sea; at other times of the year elephant seals are pelagic in mostly ice free waters (Bonner Reference Bonner1990, Borneman et al. Reference Bornemann, Kreyscher, Ramdohr, Martin, Carlini, Sellmann and Plötz2000). Elephant seals feed on squid and fish, including toothfish (summarized in Hindell et al. Reference Hindell, Bradshaw, Sumner, Michael and Burton2003). They are preyed upon by killer whales (Keith et al. Reference Keith, Bester, Bartlett and Baker2001), and in sub-Antarctic latitudes by white sharks (Carcharodon carcharias; van der Hoff & Morrice 2008).

We predict that southern elephant seals may benefit from regional climate change. Warming in the Antarctic Peninsula, with disappearance of sea ice as well as retreat of ice fields and glaciers, has already created additional beach areas for moulting, pupping and breeding and their populations are expanding into these areas, e.g. on Anvers Island (Fig. 2b; seals vs year, r_s = 0.454, P < 0.02). The relationship between seals and ice extent and ice area is inverse (r_s = -0.221, r_s = -0.257), but statistically weak, indicating that other factors besides large-scale sea ice are involved in trends. Establishment of breeding at Arthur Harbour, Anvers Island, as was the case in the re-invasion of the northern elephant seal's (M. angustirostris (Gill)) former colonies, began with several years of increasing numbers of moulting individuals (e.g. LeBoeuf et al. Reference LeBoeuf, Ainley and Lewis1974, for a particularly well documented case). Although the evidence is somewhat controversial, elephant seals may have bred on the coast of Victoria Land during a warm period in the mid-Holocene (Hall et al. Reference Hall, Hoelzel, Baroni, Denton, Le Boeuf, Overturf and Töpf2006). We have indicated (Table I) favourable reactions to decreases in the first five factors dealing with loss of sea ice and shelf ice, which will result in the exposure of more beach areas. The major uncertainty here is how changes in the Antarctic food webs will influence their populations.

Fur seal

The Antarctic fur seal (and closely related taxa) breeds on sand or cobbled beach as well as on rocky headlands at South Georgia, Macquarie, Heard, Kerguelen and the South Shetland islands (Condy Reference Condy1978b). Adult males hold territories, beginning October to December, and may have 10–120 females within the defended boundaries. Females give birth within these territories over a three-week period in November and December, suckling the pup intermittently for four months and making repeated trips to sea to feed (Robinson et al. Reference Robinson, Goldsworthy, Van Den Hoff and Hindell2002, Kirkman et al. Reference Kirkman, Bester, Makhado and Pistorius2003b). Females enter the period of oestrus 6–7 days after giving birth. Fur seals during the non-breeding season are pelagic for long periods; in higher latitudes males, especially, haulout regularly on beaches (e.g. Riedman Reference Riedman1990, Green et al. Reference Green, Williams and Burton1991, Daneri & Coria Reference Danieri and Coria1992). Antarctic fur seals feed on krill (at high latitudes), myctophid fish, and mackerel icefish (Champsocephalus gunnari Lönnberg) in near-surface waters (Green et al. Reference Green, Burton and Williams1989, Goldsworthy et al. Reference Goldsworthy, Hindell, Crowley, Hindell and Kemper1997, Klages & Bester Reference Klages and Bester1998, Reid et al. Reference Reid, Hill, Diniz and Collins2005, Murphy et al. Reference Murphy, Watkins, Trathan, Reid, Meredith, Thorpe, Johnston, Clarke, Tarling, Collins, Forcada, Shreeve, Atkinson, Korb, Whitehouse, Ward, Rodhouse, Enderlein, Hirst, Martin, Hill, Staniland, Pond, Briggs, Cunningham and Fleming2007). The lanternfish are caught during night when the fish rise to near the surface. During their pelagic phase, individuals in the high latitudes haulout on ice floes near the large-scale edge of the sea ice (Ribic et al. Reference Ribic, Fraser and Ainley1991).

Antarctic fur seals at high latitudes will probably be disadvantaged the most from the lack of krill offshore of their pupping and breeding beaches. At South Georgia foraging energetics and pup survival have been shown to depend on the availability of krill for foraging females during lactation (e.g. Costa et al. Reference Costa, Croxall and Duck1989, Forcada et al. Reference Forcada, Trathan, Reid and Murphy2005). Thus, we have predicted that changes in the food web will have the most effect on this species. At Anvers Island, the fur seals have been increasing (r_s = 0.719, P < 0.001). As with elephant seals there, the relationship to ice extent and area is inverse but weak (r_s = -0.390, r_s = -0.261) indicating other factors are also involved in the trends and probably could involve krill prevalence. At lower latitudes, fur seals prey heavily on fish (references above), which increases our uncertainty about the effect of lower krill availability over the long term. Loss of sea ice is predicted to have little effect (except as previously noted via its negative consequences on krill abundance), but increased beach availability could provide additional pupping and breeding areas (Table I).