Search

The diving behaviour of king penguins (Aptenodytes patagonicus) was studied on the Falkland Islands, where a small population (ca. 300 fledglings [year.sup.-1] is located at the geographical limit of their breeding range. King penguins rearing newly hatched chicks were equipped with time-depth recorders before leaving for sea. In total, 20,175 dives > 3 m were recorded from 12 birds during 15 foraging trips with a mean duration of 5.7 [+ or -] 2.3 days. The majority of the trips was directed up to 500 km to the northeast of the breeding colony in slope waters of, and oceanic waters beyond, the Patagonian shelf. Mean time spent underwater accounted for 42 [+ or -] 9% of the foraging trip. Mean dive depth achieved was 55 [+ or -] 16 m; maximum dive depth recorded was 343 m. Mean dive duration was 159 [+ or -] 25 s; maximum dive duration was 480 s. The mean vertical distance covered was 140 [+ or -] 65 km [trip.sup.-1]; and on average birds covered 25 km [day.sup.-1]. Synchronous diving behaviour was observed in two birds for a period of about 24 h after leaving the colony. Dive depth correlated positively with: (1) light intensity, (2) dive duration and (3) vertical velocities, thus confirming previous findings obtained from conspecifics at other breeding sites and indicating comparable diving behaviour. However, separation of dives according to their profile--V-, U-, or W-shaped--revealed significant differences between certain dive parameters. For a given depth range, bottom time was longer and vertical velocities higher in W-dives than in U-dives. This, together with a higher number of W-dives at dawn and dusk, suggests that foraging is more effective during W-dives than U-dives, and during twilight. These findings imply that king penguins have to make more complex decisions, individually and socially, on the performance of the subsequent dive than previously thought.

The present study addresses the ecology of two dominant copepod species in the Bay of Morbihan, Kerguelen Archipelago. The biomass of the herbivore Drepanopus pectinatus (from 2 mg [m.sup.-3] in winter up to 500 mg [m.sup.-3] in summer) is tightly coupled to seasonal changes in chlorophyll a concentration in the region, whereas the biomass of the predatory euchaetiid Paraeuchaeta antarctica increases during two distinct periods over the year: 250 mg [m.sup.-3] in early summer, with the recruitment of the annual generation, and 100 mg [m.sup.-3] in autumn, with the deposition of lipids as energy reserves in C5 stages and adults. The juvenile growth rates predicted by temperature-dependent models (0.09 [day.sup.-1]) closely approximate those observed in D. pectinatusin summer, but are much greater than those observed in P. antarctica (from 0.001 to 0.04 [day.sup.-1] depending on developmental stages). This difference can be explained by the reproductive strategies and trophic positions of the two species and may also result from the dependence of larval growth on energy reserves in P. antarctica. The production rates are five- and tenfold greater in juvenile stages than in adults, respectively, for D. pectinatus and P. antarctica. The secondary production by D. pectinatusis insufficient to support P. antarctica during winter, when the predatory species probably shifts to alternate prey. In summer the predation by P. antarctica accounts for only a minor part of the mortality estimated for D. pectinatus (from 20% to 60% depending on the examined station). At two of the three stations examined in the Bay of Morbihan, the production of P. antarctica could potentially support the dietary requirements of planktivorous seabirds in the region (~2,000 kg prey [day.sup.-1] for common diving petrels, Pelecanoides urinatrix, and ~90 kg prey [day.sup.-1] for rockhopper penguins, Eudyptes chysocome filholi).

The food of three closely related and sympatric species of mollymawks (Diomedea spp.), the black-browed (BBA, D. melanophrys), grey-headed (GHA, D. chrysostoma) and yellow-nosed (YNA, D. chlororhynchos) albatrosses, was compared at Iles Nuageuses, Kerguelen, during the chick-rearing period. BBA preyed almost equally upon cephalopods (39% by fresh mass), fish (31%) and penguins (31%), while GHA fed more on squids (52%, 16% and 28%, respectively) and YNA fed more on fish and not on penguins (13%, 84% and 0%, respectively). Crustaceans were always a minor component of the diet (< 3%). Patagonian toothfish was the main fish prey, and Todarodes sp. the main cephalopod prey for the community. Accumulated beaks emphasise the importance of juvenile ommastrephid squids in the diet of mollymawks, accounting for 81%, 71% and 55% of the total number of beaks in BBA, GHA and YNA samples, respectively. BBA preyed also upon a significant number of the octopod Ben thoctopus thielei (12%) and of the cranchiid squid Galiteuthis glacialis (4%), while GHA fed more on G. glacialis (18%) and on the onychoteuthid Kondakovia longimana (8%). When feeding on the same prey, prey size was similar for the albatross species. Comparison of overall prey biogeography together with the presence/absence of prey species indicators of water masses indicates segregation through different foraging areas among the three mollymawks. BBA forage almost exclusively over the shelf and upper slope waters surrounding the Kerguelen Archipelago. By contrast, GHA and YNA feed mainly in oceanic waters, YNA favouring the warm subtropical waters, and GHA the cold Antarctic waters. It is thus remarkable that birds from the same breeding grounds forage over such a wide latitudinal range, from about 35-40[degrees]S to 60-65[degrees]S, encompassing the Subtropical Zone for YNA, the Antarctic Zone for GHA and the Antarctic Polar Frontal Zone (where Kerguelen is located) for the three species.

Female northern rockhopper penguins of Amsterdam Island typically begin forage diving at dawn, returning to the colony late in the afternoon. Forage consists primarily of Thysanoessa gregaria, supplemented with squid and fishes. In studies of 14 females, mean dive depth was 18.4 m; mean dive duration was 57 s.

The stomach contents of the emperor penguin were collected at Cape Washington, Coulman Island and Cape Roget in the western Ross Sea. Samples were placed in a huge, flat-bottomed tray and classified. The emperor penguin's diet at Cape Washington was composed of 89% to 95% fish and 11% crustaceans, but did not include fresh squid. Fish also comprised majority of the diet of emperor penguins at Cape Roget and Coulman Island. However, squid was a small but significant part of the diet of the penguins at the last two sites.

Squid and fishes are main winter diet constituents of the chicks of the king penguins, Aptenodytes patagonicus, found on the sub-antarctic Crozet Islands. The percentages of wet and reconstituted masses shows that the nutritional intake of chicks changes quantitatively and qualitatively in the winter season as compared to the summer. The change in feeding behaviour is due to a change in myctophid species availability. The breeding adults forage in the outer shelf, upper slope and oceanic areas close to the Crozet Islands to feed their chicks.

Understanding the energy flux through food webs is important for estimating the capacity of marine ecosystems to support stocks of living resources. The energy density of species involved in trophic energy transfer has been measured in a large number of small studies, scattered over a 40-year publication record. Here, we reviewed energy density records of Southern Ocean zooplankton, nekton and several benthic taxa, including previously unpublished data. Comparing measured taxa, energy densities were highest in myctophid fishes (ranging from 17.1 to 39.3 kJ g -1 DW), intermediate in crustaceans (7.1 to 25.3 kJ g -1 DW), squid (16.2 to 24.0 kJ g -1 DW) and other fish families (14.8 to 29.9 kJ g -1 DW), and lowest in jelly fish (10.8 to 18.0 kJ g -1 DW), polychaetes (9.2 to 14.2 kJ g -1 DW) and chaetognaths (5.0-11.7 kJ g -1 DW). Data reveals differences in energy density within and between species related to size, age and other life cycle parameters. Important taxa in Antarctic food webs, such as copepods, squid and small euphausiids, remain under-sampled. The variability in energy density of Electrona antarctica was likely regional rather than seasonal, although for many species with limited data it remains difficult to disentangle regional and seasonal variability. Models are provided to estimate energy density more quickly using a species' physical parameters. It will become increasingly important to close knowledge gaps to improve the ability of bioenergetic and food web models to predict changes in the capacity of Antarctic ecosystems to support marine life., Competing Interests: Compliance with ethical standardsThe authors declare that they have no conflict of interests.All applicable international, national, and/or institutional guidelines for the care and use of animals were followed and all necessary approvals have been obtained.