It cannot all be climate

While momentum for this physical forcing/climate change paradigm accelerated, overtaking the species-interactions paradigm, Pauly et al. (Reference Pauly, Christiansen, Dalsgaard, Froeser and Torres1998) published an analysis that introduced the concept of “fishing down the food web.” This concept has gained acceptance for marine environments other than the SO as more and more researchers detail the loss of top-down trophic structuring owing to over-fishing of top- and middle-trophic level species (e.g. NSF 1998, Schindler et al. Reference Schindler, Essington, Kitchell, Boggs and Hilborn2002, Soulé et al. Reference Soulé, Estes, Miller and Honnold2005, Scheffer et al. Reference Scheffer, Carpenter and De Young2005, Frank et al. Reference Frank, Petrie, Choi and Leggett2005). Included in the Pauly et al. analysis, among examples from around the world, was a tally of the SO (non-cetacean) fishery landings that showed the now characteristic successively downward depletion of trophic levels in order to maintain economic viability of burgeoning fishing fleets (see also Myers & Worm Reference Myers and Worm2003, for independent, corroborating analysis for the Scotia Sea). According to this analysis, ostensible fishery-driven food-web shift in the SO (not to mention the large-scale whale extraction), also took place during the 1960s, much later than elsewhere on the globe. And, more critical to the old SO paradigm and perhaps providing reasons for inconsistent support of a krill surplus, it was not just cetaceans that had been depleted. Also removed were large predatory fish, a fact that is critical to consider in an analysis of food-web structure as fish are the most important predators in most marine ecosystems (Sheffer et al. 2005, and others). What is not surprising within the current confines of normal science is that we have yet to find even one paper in the SO literature that has cited Pauly et al., even though that paper provides crucial support for the importance of interspecific interactions required for the credibility of the between-species-interactions paradigm. Moreover, many Antarctic papers published over the past decade that have shown correlations between a long-term biological variable and a physical one, such as sea-surface temperature or sea ice extent (or their proxies), probably would find an equally compelling correlation if observations were compared instead against the declining trophic level/commercial fisheries data of Pauly et al. The latter are also characterized by a steep decline beginning in the 1960s (cf. De La Mare Reference De La Mare1997, Curran et al. Reference Curran, Van Ommen, Morgan, Phillips and Palmer2003, Atkinson et al. Reference Atkinson, Siegel, Pakhomov and Rothery2004; Fig. 1). The ocean around the colonies of polar penguin species (whose diet trophic level has also been declining - a metric that has been suggested as a proxy for decreased phytoplankton production (Hilton et al. Reference Hilton, Thompson, Sagar, Cuthbert, Cherel and Bury2006)) historically has been one of the most heavily fished, and depleted, waters on the globe (Pauly et al. Reference Pauly, Watson and Alder2005). It seems, therefore, that the Pauly et al. analyses are included in those that most researchers have relegated, under the climate change paradigm, to the category of anomaly, even though they certainly appear to be relevant to understanding the AME.

Fig. 1. Comparison of trends in 1) mean trophic level (×10^-1) of fishery landings from FAO areas 48, 58 (Scotia Sea) and northern 88 (does not include cetaceans already depleted in 1950s–early 1960s; from Pauly et al. Reference Pauly, Christiansen, Dalsgaard, Froeser and Torres1998), 2) krill density from net tows taken in Scotia Sea region, with salps having a mirror-image trend [no. m^-2 1000^-1; line shows predicted values from locally-weighted regression (lowess smoothing), data from Atkinson et al. Reference Atkinson, Siegel, Pakhomov and Rothery2004], and 3) 20 yr running mean of methanesulphonic acid (MSA, μM), a derivative of phytoplankton-produced dimethylsulphide (DMS), found in glacial ice-core annual layers at Law Dome, (western) East Antarctica (Curran et al. Reference Curran, Van Ommen, Morgan, Phillips and Palmer2003). Less DMS and MSA are produced if salps rather than krill are grazing the phytoplankton (Katamatsu et al. 2004).

Assuming that the rate of commercial take of resources can be a proxy for their availability (see references above regarding top-down forcing; also Myers & Worm Reference Myers and Worm2003), these data also indicate that a major shift has occurred in the food-web structure of the AME (Pauly et al. Reference Pauly, Christiansen, Dalsgaard, Froeser and Torres1998), and this shift has little to do with climate. Many feed-back loops and competitive and synergistic relationships are now probably gone (Pauly & Maclean Reference Pauly and Maclean2003) due to the loss of major portions of upper- and middle-trophic levels - especially around most sub-Antarctic and lower latitude Antarctic islands and northern Antarctic continental shelves, although food webs in some more southerly areas, e.g. the Ross Sea, remain to date intact. This loss occurred within the professional lifetimes of still-practicing SO ecologists and their students, and occurred just before the dramatic advances of modern technology that now define SO marine science (mid-1970s onward), as noted above. Thus, the vast majority of data supporting the physically-forced view of the AME have been collected well after the depletion of top predators in the system was well underway.

Unfortunately this depletion probably continued coincident with apparent or real shifts in physical factors (Zwally et al. Reference Zwally, Comiso, Parkinson, Cavalieri and Gloersen2002, Parkinson Reference Parkinson2002, Jacobs et al. Reference Jacobs, Guilivi and Mele2002, Jacobs Reference Jacobs2006). Because fishing down the food web can affect ecosystems just as dramatically as climate change (Pauly & Maclean Reference Pauly and Maclean2003), causes of patterns have merged and what is concluded about the AME may primarily be distorted by researcher bias and the lack of long-term data on food-web structure. In the Bering and North seas and Benguela Current, for instance, where researchers accept the idea of major climate change as well as heavy fishing pressure, the dramatic rise in the abundance of gelatinous organisms has been explained by the fishery pressure that has reduced grazing on their larvae (Brierly et al. 2001, Brodeur et al. Reference Brodeur, Sugisaki and Hunt2002, Heymans et al. Reference Heymans, Shannon and Jarre2004, Lynam et al. Reference Lynam, Hay and Brierley2005), with similar arguments being made for other, smaller systems (e.g. Mills Reference Mills2001, Xian et al. Reference Xian, Kang and Liu2005). On the other hand, in the mind-set of most scientists working under the climate change paradigm in the “relatively pristine” AME (e.g. a concept used in the summaries by Croxall et al. Reference Croxall, Trathan and Murphy2002, Smetacek & Nicol Reference Nicol2005), the current hypothesis is that it is simply changed sea ice that is responsible for a similar dramatic rise in gelatinous creatures (Atkinson et al. Reference Atkinson, Siegel, Pakhomov and Rothery2004, Kawaguchi et al. 2005). However, because both heavy fishing and salps are concentrated and/or dominant in waters of non-existent, reduced or declining ice cover in the AME (see Nicol et al. Reference Nicol, Pauly, Bindoff, Wright, Thiele, Hosie, Strutton and Woehler2000), a positive correlation between heavy fishing and increased salps could well be involved (see also Lynam et al. Reference Lynam, Gibbons, Axelsen, Sparks, Coetzee, Heywood and Brierley2006). In the case of the gelatinous salps, although notoriously difficult to detect in the highly acidic stomachs of fish, they have been found in the diet of certain SO piscine species (Casaux et al. Reference Casaux, Mazzotta and Barrera-Oro1990, Pakhomov Reference Pakhomov1997, Barrera-Oro Reference Barrera-Oro2003, Bushula et al. Reference Bushula, Pakhomov, Kaehler, Davis and Kalin2005), though not yet in the little-researched diets of pelagic juveniles of commercially important benthic fish (e.g. Pakhomov & Pankratov Reference Pakhomov and Pankratov1992, Koch & Eversen 1997, Barrera-Oro et al. Reference Barrera-Oro, Casaux and Marschoff2005). Finally, it is very likely that salps eat eggs and early larvae of krill (reviewed by Siegel Reference Siegel2005), a scenario consistent with the decline in krill as salps proliferate (cf. Atkinson et al. Reference Atkinson, Siegel, Pakhomov and Rothery2004).

Whale extraction may well have had a major effect

Some of us, exercising “normal science,” recently presented ideas about how changes in climate, weather and sea ice may have affected the long-term population trends of certain Antarctic penguin species breeding in the Ross Sea/western Pacific Ocean sector and elsewhere in the SO (Ainley et al. Reference Ainley, Clarke, Arrigo, Fraser, Kato, Barton and Wilson2005). By contrast, we compare the Ross Sea penguin trends to the rate of extraction of a major trophic competitor of Adélie penguins, the Antarctic minke whale Balaenoptera bonaerensis Burmeister, from the same region (Fig. 2). The population of this pagophilic cetacean, which completely overlaps the penguins' habitat and food preferences (an important coincidence; Murphy Reference Murphy1995), was aggressively targeted two decades later than the other, more open-water baleen whales (Bengtson & Laws Reference Bengtson, Laws, Siegfried, Condy and Laws1985). The effects to the population of the more than 116 600 taken (Clapham & Baker Reference Clapham and Baker2001), about 20% from the region of the Ross Sea (Brown & Brownell Reference Brown and Brownell2001) is unknown. Adélie penguins, seemingly released from trophic competition, show the appropriate demographic lag as increased penguin productivity and survival result in population growth for this slow-to-mature (up to seven years) upper-level predator. The fact that coastal polynyas were increasing in size during this time as well would facilitate the penguin increase (cf. Ainley Reference Ainley2002a, Parkinson Reference Parkinson2002, Ainley et al. Reference Ainley, Clarke, Arrigo, Fraser, Kato, Barton and Wilson2005). As the take of minke whales slackened beginning in 1987 (Brown & Brownell Reference Brown and Brownell2001), and best demonstrated in the Ross Sea, the penguin increase slackened.

Fig. 2. Numbers of minke whales (×100) removed from IWC areas V and VI (Adélie Land and Ross Sea sector) by decade in the industrial and scientific whaling eras, compared to percent differences relative to 1960s numbers of Adélie penguins breeding at capes Royds and Bird, Victoria Land, and emperor penguins at Pointe Géologie, Adélie Land. The Cape Royds colony, the longest monitored in the Antarctic, was the same size in 1959 as it was in 1909, with no evidence that it had been larger in historical times (Ainley Reference Ainley2002a). These penguins and whales spend the late summer, autumn and winter in the same habitat and region; minke whales and Adélie penguins have the same diet (Ainley Reference Ainley2002b). Data on penguins are from Wilson et al. (Reference Wilson, Ainley, Nur, Jacobs, Barton, Ballard and Comiso2001) and Weimerskirch et al. (Reference Weimerskirch, Inchausti, Guinet and Barbraud2003); those for whales from Brown & Brownell (Reference Brown and Brownell2001).

We then began to question the rapid decline attributed to a mysterious short-term episode of adult mortality in emperor penguins Aptenodytes forsteri Gray in the same region (Weimerskirch et al. Reference Weimerskirch, Inchausti, Guinet and Barbraud2003), a mortality event correlated to climate effects (Barbraud & Weimerskirch Reference Barbraud and Weimerskirch2001, Jenouvrier et al. Reference Jenouvrier, Barbraud and Weimerskirch2005). Also considered subsequently have been effects of changed climate on this species' inability to recover from that mortality event (Jenouvrier et al. Reference Jenouvrier, Weimerskirch, Barbraud, Park and Cazelles2004, Ainley et al. Reference Ainley, Clarke, Arrigo, Fraser, Kato, Barton and Wilson2005). In accord with massive adult mortality, the population decline in Adélie Land, a short distance west of the Ross Sea, was sharp (see Jenouvrier et al. Reference Jenouvrier, Barbraud and Weimerskirch2005) and it occurred in the early part of the Adélie penguin increase (Fig. 2). Not coincidentally we suggest, the mortality event happened at the time when minke-whale-eating killer whales Orcinus orca (L.) type A (Pitman & Ensor Reference Pitman and Ensor2003), inhabiting the waters frequented by these penguins (again, see Murphy Reference Murphy1995), were being challenged by a major, rapid large-scale removal of their main prey, i.e. the minke whale. Until that time it was thought that minke whale and killer whale populations, prey and predator, were in equilibrium (N.V. Doroshenko in Mikhalev et al. Reference Mikhalev, Ivahim, Savusin and Zelenaya1981). Did these top predators switch to alternate prey including emperor penguins, especially penguins whose vulnerability was increased by a possible short-term divergence in sea ice (as proposed by Jenouvrier et al. Reference Jenouvrier, Barbraud and Weimerskirch2005), and a habitat frequented only at its periphery by type A killer whales (Pitman & Ensor Reference Pitman and Ensor2003)? Emperor penguins have been found in killer whale stomachs (type unknown; Prévost Reference Prévost1961), and it is well known that these predators (type A) eat high numbers of the closely related king penguin A. patagonicus in the northern part of this SO region (e.g. Condy et al. Reference Condy, Van Aarde and Bester1978, Guinet Reference Guinet1992, Guinet & Bouvier Reference Guinet and Bouvier1995). Owing to confounding climate effects (warmer winter temperatures, stronger winds, thinner sea ice on which to breed), the emperor penguin population has not been able to recover from the initial massive mortality (cf. Barbraud & Weimerskirch Reference Barbraud and Weimerskirch2001, Jenouvrier et al. Reference Jenouvrier, Weimerskirch, Barbraud, Park and Cazelles2004, Ainley et al. Reference Ainley, Clarke, Arrigo, Fraser, Kato, Barton and Wilson2005). This scenario is reminiscent of a hypothesized killer whale prey switch, in this case from whales to sea otters Enhydra lutris L., an idea developed to explain changes in the fauna of the Aleutian Islands (Springer et al. Reference Springer, Estes, Van Vliet, Williams, Doak, Danner, Forney and Pfister2003), and a hypothesized switch to sea lions (Otaria flavescens (Shaw)) and elephant seals in the Southern Ocean (Branch & Williams Reference Branch, Williams, Estes, Demaster, Doak, Williams and Brownell2006). Moreover, an emperor penguin is far more of a meal for a killer whale than is the similar sized sea otter, both weighing 25–35 kg as the high fat content of the penguin would make it superior prey, probably equivalent to an energy-rich sea lion pup (Williams et al. Reference Williams, Estes, Doak and Springer2004).

Modelling Southern Ocean ecosystem structure and change

We will end here with some questions, hoping that the reader has at least given some credence to the hypothesis that the AME is vastly different from the pre-1970s due not just to climate effects but equally to biotic extractions from the ecosystem. We point out that in ecology, but not the physical sciences, it is important to keep in mind how paradigms have evolved or been replaced (Dunlap Reference Dunlap2006). Why did the paradigm involving the massive 1960s whale loss and its possible effect on the AME fall out of favour so quickly and with very little contest? Why have the possible ecological effects of the concurrent massive extraction of finfish from the SO not been acknowledged by ecologists? By keying on physical, bottom-up forcing, are most SO marine biologists, in effect, now acknowledging that the AME no longer has the resiliency it once had when numerous biotic feedback loops were in place to ameliorate climate effects? That is, lacking the biotic feed-backs, is the AME now more susceptible to physical forcing (see Balance et al. 2006), as is now being argued for depleted non-marine ecosystems (e.g. Post et al. Reference Post, Conners and Goldberg2000, Wilmers et al. Reference Wilmers, Sinha and Brede2002, Wilmers & Getz Reference Wilmers and Getz2005)? Indeed, the dramatic year-class cycling evident in the krill that dominate the Adélie penguin diet off the Antarctic Peninsula (Fraser & Hofmann Reference Fraser and Hofmann2003) - and where alternate prey (fish) are no longer available in quantity (see above, also Emslie & McDaniel Reference Emslie and McDaniel2002) - is reminiscent of the simple food webs of the terrestrial Arctic, where predator populations closely track their limited prey. No wonder that in this area strongly affected by both climate change and fishery depletion, unlike high-latitude portions of the Antarctic, the Adélie penguins are being replaced only in part by other, open water species (Fraser & Patterson Reference Fraser, Patterson, Bataglia, Valencia and Walton1997, Ainley et al. Reference Ainley, Clarke, Arrigo, Fraser, Kato, Barton and Wilson2005, Forcada et al. Reference Forcada, Trathan, Reid, Murphy and Croxall2006).

No doubt a strong mix of factors is involved in the re-structuring of the AME food web, with both climate change and depletion of top- and middle-trophic level species playing a role (as noted earlier, e.g. by Croxall Reference Croxall1992), but surely it is time for some serious re-thinking on their relative contributions and what it is we are actually measuring or managing in the observed trends of SO organisms? Finally, will we not be able to answer some of these questions better if portions of the SO, such as the Ross Sea, are kept free of intensive biotic extractions? The latter, following the progression experienced by other ocean systems (Pauly et al. Reference Pauly, Christiansen, Dalsgaard, Froeser and Torres1998), has now begun in that neritic system (Ainley Reference Ainley2002b).