There are seven emperor penguin (Aptenodytes forsteri) colonies distributed throughout the traditional boundaries of the Ross Sea from Cape Roget to Cape Colbeck. This coastline is c. 10% of the entire coast of Antarctica. From 2000 to 2012, there has been a nearly continuous record of population size of most, and sometimes all, of these colonies. Data were obtained by analysing aerial photographs. We found large annual variations in populations of individual colonies, and conclude that a trend from a single emperor penguin colony may not be a good environmental sentinel. There are at least four possibilities for census count fluctuations: i) this species is not bound to a nesting site like other penguins, and birds move within the colony and possibly to other colonies, ii) harsh environmental conditions cause a die-off of chicks in the colony or of adults elsewhere, iii) the adults skip a year of breeding if pre-breeding foraging is inadequate and iv) if sea ice conditions are unsatisfactory at autumn arrival of the adults, they skip breeding or go elsewhere. Such variability indicates that birds at all Ross Sea colonies should be counted annually if there is to be any possibility of understanding the causes of population changes.

The emperor penguin colony at Coulman Island is reputedly the largest known. This reputation is based on intermittent ground and aerial surveys performed since 1958. From an aerial survey obtained on 28 October 2010 we discovered that the total number of chicks was 56% of the lowest previous estimate of 2006 and only 41% of the most recent estimate in 2008. All of the counts tallied since 1983 were determined either by ground counts or from aerial film or digital photographs, or estimates from adult counts. We also determined the sea ice conditions in autumn, which is close to the time the adults arrive to breed. We present three hypotheses of what might have happened from 2008–10 to cause the step change in chick production, the small recovery of chick numbers in 2011, and the complete recovery of number of adults from 2010–11. We conclude that local circumstances may have strongly influenced the breeding behaviour of the emperor penguins in 2010 and to a lesser degree in 2011 when many adults elected not to breed.

The emperor penguin (Aptenodytes forsteri) is highly dependent on sea ice conditions, and future climate change may affect its distribution and numbers. Most studies on the demography and population dynamics of emperor penguins in relation to sea ice characteristics were conducted at a single colony (Pointe Géologie). Several non-exclusive hypotheses have been proposed to explain the dramatic decline of this colony, including changes in sea ice conditions, predation, flipper banding and human disturbance. Here, we report and analyse updated long-term trends in numbers of breeding pairs made at two colonies (Pointe Géologie and Haswell Island) where counts are comparable. Similar changes were observed for both colonies and paralleled changes in sea ice extent. At Pointe Géologie and Haswell Island, populations declined similarly and later growth rates were also similar since the early 1990s for Haswell and early 1980s for Pointe Géologie. The magnitude of the decline was similar between both colonies when numbers of breeding pairs were assessed. This study suggests that a common large-scale environmental factor has probably negatively affected both colonies.

The emperor penguin (Aptenodytes forsteri) is extremely dependent on the extent and stability of sea ice, which may make the species particularly susceptible to environmental change. In order to appraise the stability of the emperor penguin populations at six colonies in the western Ross Sea, we used linear regression analysis to evaluate chick abundance trends (1983–2005) and Pearson's r correlation to assess their relation to two local and two large-scale climate variables. We detected only one significant abundance trend; the Cape Roget colony increased from 1983 to 1996 (n = 6). Higher coefficients of variation in chick abundances at smaller colonies (Cape Crozier, Beaufort Island, Franklin Island) suggest that such colonies occupy marginal habitat, and are more susceptible to environmental change. We determined chick abundance to be most often correlated with local Ross Sea climate variables (sea ice extent and sea surface temperature), but not in consistent patterns across the colonies. We propose that chick abundance is most impacted by fine scale sea ice extent and local weather events, which are best evaluated by on-site assessments. We did not find sufficient evidence to reject the hypothesis that the overall emperor penguin population in the Ross Sea was stable during this period.

The arrival in January 2001 in the south-west Ross Sea of two giant icebergs, C16 and B15A, subsequently had dramatic affects on two emperor penguin colonies. B15A collided with the north-west tongue of the Ross Ice Shelf at Cape Crozier, Ross Island, in the following months and destroyed the penguins' nesting habitat. The colony totally failed in 2001, and years after, with the icebergs still in place, exhibited reduced production that ranged from 0 to 40% of the 1201 chicks produced in 2000. At Beaufort Island, 70 km NW of Crozier, chick production declined to 6% of the 2000 count by 2004. Collisions with the Ross Ice Shelf at Cape Crozier caused incubating adults to be crushed, trapped in ravines, or to abandon the colony and, since 2001, to occupy poorer habitat. The icebergs separated Beaufort Island from the Ross Sea Polynya, formerly an easy route to feeding and wintering areas. This episode has provided a glimpse of events which have probably occurred infrequently since the West Antarctic Ice Sheet began to retreat 12 000 years ago. The results allow assessment of recovery rates for one colony decimated by both adult and chick mortality, and the other colony by adult abandonment and chick mortality.

Within the Western Ross Sea, there are six emperor penguin colonies of widely different size that occur exclusively on sea ice. In 1990 a survey of all six sites, two by close overflights and four from the ground, showed that the breeding habitats were highly variable. The most important physical characteristics of these habitats appear to be stable fast ice, nearby open water, access to fresh snow, and shelter from the wind.

The data presented provides new information on the distribution of Antarctic squids and on the summer diet of the emperor penguins. The diet of 58 adult emperor penguins (Aptenodytes forsteri) on the fast ice of the Drescher Inlet, Vestkapp Ice Shelf (72°52′S, 19°25′W) in the eastern Weddell Sea was investigated. Prey consisted principally of squid, fish, krill, amphipods and isopods. Squids were identified by the lower beaks and allometric equations were used to estimate the squid biomass represented. Beaks occurred in 93% of the stomach samples. Each sample contained a mean of 27 beaks (range 1–206). Ninety-two percent of the squids could be identified by the lower beaks and belonged to four families (Onychoteuthidae, Psychroteuthidae, Neoteuthidae and Gonatidae). The most abundant squid was Psychroteuthis glacialis which occurred in 52 samples with lower rostral lengths (LRL) ranging from 1.4–7.2 mm. Forty-five samples contained Alluroteuthis antarcticus (LRL range 1.8–5.8 mm), 17 Kondakovia longimana (LRL range 4–12.1 mm), and four Gonatus antarcticus (LRL range 4.1–6.1 mm). In terms of biomass K. longimana was the most important species taken by the penguins comprising 50% of total estimated squid wet mass (245348 g) in 1990 and 48% in 1992 (154873 g). However, if only fresh beaks were considered for estimations of squid consumption, i.e. beaks that have been accumulated for not longer than 5–6 days in the stomachs, squid diet was of minor importance. Then total squid wet mass accounted for only 4809 g in 1990 and 5445 g in 1992 which implies that one penguin took c.30 g squid d−1 with P. glacialis and A. antarcticus being the most important by mass. The prey composition suggests that emperor penguins take squid at the steep slope regions of the eastern Weddell Sea.

The food and energy requirements of adult emperor penguins parenting chicks were examined at Auster (about 11 000 pairs) and Taylor Glacier (about 2900 pairs) colonies during the winter, spring and summer of 1988. Tritiated water-derived estimates of food consumption trebled during the five-month period of chick care, ranging from 2.3 kg d−1 in winter when chicks were <5% of adult mass, to 6.3 kg d−1 in summer, when chicks represented 40–50% of adult mass. These food consumption rates were equivalent to the acquisition of 11.4 MJ d−1 and 33.4 MJ d−1 metabolizable energy in winter and summer respectively. Chick provisioning was not accompanied by a major increase in food consumption by adults. Adults assimilated 84–92% of their daily food intake themselves and retained the remainder for the chick. The food ration of chick for the three seasons (42 kg) constituted only about 9.5% of adult maintenance requirements during the same period. Adults consumed an estimated 482 kg of food (including the ration for the chick), which amounted to about 10 700 t and 2800 t of fish and squid consumed by the breeding populations at Auster and Taylor Glacier, respectively.