The Snow Petrel (Pagodroma nivea) is a fairly small petrel (30 to 40 cm in length with a wingspan of 75 to 95 cm and a mass of 240 to 460 g) with entirely white plumage. It has conspicuous dark eyes, a very small black bill, and bluish gray feet. In poor light the plumage appears gray, especially on the underparts. Females average slightly smaller than males and juveniles have more extensive grayish barring on the upperparts. (Watson 1975; Carboneras 1992)

Snow Petrels are year-round residents of Antarctica and are very closely associated with pack ice or icebergs and ice floes. Snow Petrel flight is very erratic, almost batlike, with short, rapid wingbeats and infrequent gliding. Snow Petrels tend to fly around 10 meters above the water and may fly very high over land. They are frequently seen hovering low over the water, but rarely observed swimming, preferring to rest on icebergs or ice floes.They eat mainly fish, along with cephalopods, other mollusks, and euphausiids (krill). They apparently also feed on seal placentas and carcasses of seals, whales, seabirds, and penguins, as well as excreta. (Watson 1975; Carboneras 1992)

Snow Petrels are endemic to Antarctica and the surrounding Southern Ocean, with a circumpolar breeding distribution (see Croxall et al. 1995 for detailed maps of all reported breeding localities).

The Snow Petrel is a fairly small petrel (30 to 40 cm in length with a wingspan of 75 to 95 cm and a mass of 240 to 460 g) with entirely white plumage. It has conspicuous dark eyes, a very small black bill, and bluish gray feet. In poor light the plumage appears gray, especially on the underparts. Females average slightly smaller than males and juveniles have more extensive grayish barring on the upperparts. Beneath the white plumage, the down is dark gray. The chick is lavender gray on the upperparts, head, and foreneck; ivory or grayish white on the abdomen; and pure white on the forehead. The second down is slightly lighter than the first. (Watson 1975; Carboneras 1992)

Molt occurs between November and April, during the breeding cycle. Body molt begins during incubation. Flight feathers molt as young near fledging. Unsuccessful and pre-breeders molt earliest. (Watson 1975)

• Terrestrial
• Marine

Snow Petrels are almost entirely restricted to the colder antarctic waters with pack ice or icebergs and ice floes (Watson 1975).

The Snow Petrel is closely associated with pack ice, occurring mainly in areas with 10 to 50% ice cover. Snow Petrels do not land on the water, preferring to rest on icebergs or ice floes. (Carboneras 1992)

Snow Petrels appear to be relatively sedentary in some locations, with most birds staying in the vicinity of the colony throughout the year, dispersing only to adjacent waters. Snow Petrels seldom stray far from pack ice except at their northernmost breeding grounds (Bouvetoya, South Georgia). (Carboneras 1992)

Snow Petrels eat mainly fish, along with cephalopods, other mollusks, and euphausiids (krill). They apparently also feed on seal placentas and carcasses of seals, whales, seabirds, and penguins, as well as excreta. (Watson 1975; Carboneras 1992)

Skuas (Catharacta spp.) are important predators of Snow Petrels. Ectoparasites include the feather mite Zachvatkinia hydrobatidii; the feather lice Ancistrona sp., Austromenopon (?) daptionis, Pseudonirmus charcoti, and Saemundssonia antarctica; and the flea Glaciopsyllus antarcticus. Glaciopsyllus antarcticus is endemic to the Antarctic continent, where it is known to parasitize a variety of seabird species. (Watson 1975; Pilgrim and Palma 1994; Steele et al. 1997).

Berman et al. (2009) studied senescence (decline in annual rate of survivorship or fecundity as a function of increasing age) in the Snow Petrel. Snow Petrels have low fecundity, high annual adult survivorship, and long lifespans. The authors' data and analysis indicated an annual increase in breeding success (probability of a chick fledging from a laid egg) for about the first 10 years of life (at which point breeding success rate was 50%). High breeding success was then maintained (and possibly even increased) until the oldest ages. Breeding probability (probability of a bird of a particular age breeding) increased from 0 at 5 years of age to 45% at 6 years, then increased at an annual rate of 0.02, but over an extended period (between 6 and 34 years of age); by the end of this period, approximately 80 per cent of birds were breeding. After this, breeding probability dropped abruptly. Thus, Snow Petrels did not show any sign of senescence in breeding success, and breeding probability did not decrease before 34 years of age (maximum longevity is more than 46 years).

Ageing is associated with a decline in basal metabolic rate (BMR) in many species, including humans. Moe et al. (2007) undertook a cross-sectional study to measure BMR of Snow Petrels aged between 8 and 39 years. They found that the BMR of the Snow Petrel does not decrease with increasing age. BMR seems to be sustained at a fixed level throughout the investigated age range, a finding the authors' discuss in the context of evolutionary theories of ageing.

Snow Petrel flight is very erratic, almost batlike, with short, rapid wingbeats and infrequent gliding. Snow Petrels tend to fly around 10 meters above the water and may fly very high over land. They are frequently seen hovering low over the water, but rarely observed swimming. Flocks are typically seen sitting on ridges of icebergs. (Watson 1975)

Snow Petrels feed on the wing by dipping and, occasionally, by diving and surface-seizing. They sometimes scavenge on land. (Carboneras 1992)

Adult annual mortality is 4 to 7%, yielding a mean adult life expectancy of 14 to 20 years (Carboneras 1992).

Snow Petrels breed colonially, nesting in small to large cliff colonies, usually near the sea but also several hundred km inland (as much as 440 km, Goldsworthy and Thomson 2000), at altitudes up to 2400 meters. Breeding starts in November in most areas. The nest is a simple pebble-lined scrape, usually in a deep rock crevice with overhanging protection. Nests often contain dehydrated eggs and mummified corpses of chicks from previous years. Just a single, white egg is laid. The incubation period is 41 to 49 days and young are brooded an additional 8 days. Fledging occurs at an age of around 41 to 54 days. The chick remains in the nest for about 7 weeks. (Watson 1975; Carboneras 1992)

Severe weather can be an important source of mortality, especially heavy snow that blocks nest entrances and causes adults to abandon eggs and allows chicks to starve. Egg mortality is around 50% and chick mortality is 10 to 15%, but mean annual adult survival is 93 to 96%. (Watson 1975; Carboneras 1992)

The extent of sea-ice during fall and winter influences percentage of birds breeding during the following summer (many Snow Petrels do not breed each year), breeding success of breeders, and fledgling body condition. Extensive sea-ice cover during the winter months appears to be associated with high population densities of krill (an important food for Snow Petrels) during summer, when petrels are feeding their young. Decisions to breed are affected by food availability in winter, whereas breeding success and fledgling quality are affected by food availability in summer, which in turn is affected by sea-ice extent in winter because of a time lag. Snow Petrels feed largely on krill (Euphausia superba and E. crystallorophias) and on fish (Antarctic Silverfish, Pleuragramma antarcticum) both during and outside the breeding season, although there are indications that cephalopods (e.g., the squids Psychroteuthis glacialis and Gonatus antarcticus) may play a more important role in some localities. Spring air temperature also influences breeding success. Increased air temperature during spring increases breeding success, presumably because low air temperature in spring results in some nest sites (crevices) being made inaccessible by snow and ice, a common cause of Snow Petrel breeding failure. (Barbraud and Weimerskirch 2001 and references therein; Jenouvrier et al. 2005)

Two Snow Petrel subspecies, differing markedly in size, are often recognized. These are reported to breed sympatrically yet remain distinct in some locations, but other reports indicate extensive hybridization. Some researchers have proposed treating the two forms as distinct species, but others have concluded that although they have evolved in partial isolation, they are not sufficiently distinct to warrant treatment as separate species. (Carboneras 1992)

Snow Petrels show a tendency to mate assortatively by size morph, but the occurrence of mixed pairs producing viable young reveals that reproductive isolation between the two size morphs is incomplete. The degree of isolation depends on breeding locality. A given area can harbor (1) only small birds, (2) only large birds, (3) colonies of small birds close to colonies of large ones, or (4) mixed pairs, the percentage of which varies from one locality to another. Data from a 34-year demographic study in Terre Adelie, Antarctica, showed that mate and nest fidelity were very high and that pairs involving mates of the same size morph and mixed pairs had similar fecundity. Despite its heterogeneity, the breeding habitat of snow petrels was relatively predictable. The primary requirement for breeding in Snow Petrels is a nest that is not filled with ice. Ice repeatedly made some nests unsuitable for breeding. When Antarctic ice and snow are prevalent in bad years for Snow Petrel breeding, ice may obstruct more than 40% of nests, causing many nesting failures. Obtaining a nest that is not frozen is therefore the primary requirement for breeding. Strong competition for nests may explain high fidelity rates and, combined with the absence of reproductive costs in mixed pairs, may have promoted decreased choosiness during mate choice, preventing total reproductive isolation between the two morphs. (Barbraud and Jouventin 1998; Jouventin and Bried 2001)

The Snow Petrel is not globally threatened. Despite habitat restriction, it is apparently abundant, with a total population of perhaps several million individuals. Overall numbers appear to be stable and possibly slightly increasing. There are an estimated 2,000,000 birds at colonies around the Ross Sea. (Carboneras 1992) However, given the strong dependence of Snow Petrels on sea-ice, Jenouvrier et al. (2005) predict that Snow Petrel populations will be negatively affected by reduced sea-ice in the context of regional warming, although the impact of changes in sea-ice extent may depends on its seasonality.