Introduction

Cannibalism is defined as the consumption of eggs or significant parts of the body of conspecifics (Stanback and Koenig 1992; Fouilloux et al. 2019). Among birds, the most widespread and common form of cannibalism is heterocannibalism, in which non-kin eggs or chicks are eaten (Stanback and Koenig 1992). Cannibalistic bird species share common characteristics, such as carnivory, colonial breeding habits and a beak morphology that allows them to tear pieces of flesh rather than having to eat the whole prey (Stanback and Koenig 1992). Seabirds share many of these characteristics, and in fact, cannibalism has been reported for many species (Peter et al. 1990; Stanback and Koenig 1992; Daigre et al. 2012; Hayward et al. 2014; Neves et al. 2015).

The southern giant petrel Macronectes giganteus (SGP) is a seabird species with scavenger and predator habits. Its diet includes both marine and land-based prey and carrion, such as, crustaceans, cephalopods, fishes, penguins, flying seabirds, seals and even whales (Hunter and Brooke 1992; Petry et al. 2010; Corá et al. 2020; Mills et al. 2021) However, records of petrels preying on conspecifics are speculative or unconclusive. We report the first confirmed cases of inter-generational cannibalism for SGP breeding in the Antarctic and explore hypothesis that could explain the occurrence of this behavior.

Material and methods

Harmony Point (62°18′S; 59°10′W) at Nelson Island (South Shetland Islands, maritime Antarctic) is home to large breeding population of seabirds (Silva et al. 1998). SGPs breed on plateaus and on relatively flat areas close to the beach (ca. 480 breeding pairs; Krüger 2019). Two breeding groups are located at 60 m (east-southeast) and 160 m (south-southwest) from the Francisco de Gurruchaga Argentine shelter (62°14′3″S; 59°10′2″W). Both lie near the sea level. One lies within a rock outcrop with small blocks and the latter is surrounded by rocks of 10 m height. During the 78 days (between 20 November 2019 and 5 February 2020) we stayed in the area, we frequently observed southern giant petrels from the shelter’s window. We recorded and photographed any unusual activity using a 300 mm lens coupled to a Nikon D90 DLSR camera from inside the shelter. In one of the two occasions when cannibalism occurred, one researcher moved towards the nest and recorded/photographed the behavior from a 20 m distance.

Results

We recorded two cases of cannibalism for SGPs during the breeding season of 2019/2020, which are described below. In both cases, one adult, possibly male due to bill length and ‘jizz’, was preying upon a nestling of the same species. On January 17, we sighted a SGP carrying in its beak a living conspecific chick of ca. 10 days of age. The adult was about 20 m to the north of a petrel colony (the furthest from the shelter), where adults were brooding their chicks. We did not see if the bird himself removed the chick from the nest. The bird started to peck the chick’s rump (Fig. 1a), when two brown skuas (Catharacta antarctica) came flying towards him. As the skuas started to invest towards the chick, the adult carried it backwards and displayed an aggressive behavior characterized by a low intensity-standing attitude with an erected (but not fanned) tail and slightly lowered wings aside the body (Fig. 1a). Two minutes later, the petrel was able to tear flesh, but soon after the skuas stole the chick from him. While one skua took the chick a few meters away, another started to attack the petrel by ground and by air (Fig. 1b). The SGP reacted by uttering whinnying and growling calls and displaying and erected tale and ruffled neck feathers (Fig. 1c). A few seconds later, he was able to return to the chick, which was still alive, and eat a few pieces of the chick’s rump muscle (Fig. 1d). However, after a few seconds the adult left the chick for the skuas and flew away without further conflict. During this act, no other giant petrel approached the area trying to defend or feed from the chick.

Fig. 1
figure 1

A male southern giant petrel Macronectes giganteus feeding on a living conspecific chick at Nelson Island, Maritime Antarctic Peninsula while disputing it with skuas. a Adult pecking the chick’s rump in attempt to tear flesh. b Skuas stealing the chick. c Adult displays aggressive behavior towards the skuas. d Recover of the chick by the adult, which was finally able to tear the skin and feed from the chick’s muscle

Full size image

The second case was registered in January 27. We saw another adult male feeding on a nestling of about 20 days of age (Fig. 2, see also Online Resource 1 for a video footage). Before the incident, we noticed that the chick was alone in the nest and there was no apparent sign of its parents near the colony, which was the same colony where the first case occurred. Then we saw the adult pulling the chick out of the nest by its neck. We then moved towards the colony to better record the behavior at a 20 m distance. By the time we approached the scene the chick was already dead. The adult then started to tear the chick’s chest skin and eat its muscle (Fig. 2). Few adults were in the colony at this moment and they once again did not approached the one that was preying the nestling. There were no skuas nearby when this second predation event started.

Fig. 2
figure 2

Another male southern giant petrel Macronectes giganteus feeding from a conspecific chick after removing it from its nest at Nelson Island, maritime Antarctic (see Online Resource 1 for a video footage)

Full size image

Discussion

There are no previous records of cannibalism in procellariiformes birds. Literature, however, reports a possible case for SGP registered by Conroy (1972) in Signy Island, South Orkneys. In his report, he wrote, “one (chick) was possibly cannibalized—at one site, a chick showed a marked increase in weight following the disappearance of an adjacent chick, and a fresh giant petrel tarsus bone was found in the nest with the chick; and one chick died for no ‘apparent reason’”. Although Conroy gathered evidence to suggest that the missing chick was predated by its neighbor, he did not record the act itself.

SGP is mostly a colonial species (Conroy 1972) which usually breeds in loose aggregations (e.g. Croxall and Prince 1980; Krüger 2019). During late chick rearing when chicks acquire thermal independence, adults leave the chicks alone in the nest for hours or even days (Warham 1962), when they become exposed to potential predation by other giant petrels. Since SGPs are opportunistic when it comes to their feeding choices (Hunter and Brooke 1992), an unguarded chick of about 20 days could be easily preyed by an adult.

Cannibalism is recorded usually in association with food shortages or high population density (Hayward et al. 2014; Fouilloux et al. 2019), which is not the case for Harmony Point, with its large populations of avian prey (e.g. large colonies of chinstrap and gentoo penguin; Silva et al. 1998) and decreased SGP density (Krüger 2019). Studies on other groups of seabirds suggest cannibalism can be a specialized behavior displayed under certain environmental conditions, rather than being simply opportunistic (Hayward et al. 2014), therefore it would be an interesting aspect to be further evaluated on SGP populations. We also cannot rule out that the presence of researchers on the area and frequent approximation to colonies over 78 days caused physiological stress to the birds, which could have influenced the cannibalistic behavior. SGP have shown to increase their heart rates in 63% when approached to 20 m during incubation (Pfeiffer and Peter 2004). This increase might transcribe into increments in energy expenditure since heart rate and oxygen consumption are directly related (Nolet et al. 1992; Weimerskirch et al. 2002). However, giant petrels, such as wandering albatrosses (Diomedea exulans), have large body sizes and reserves that could buffer them from the impact of energy expenditure caused by frequent stress (Weimerskirch et al. 2002). We also stress out that, when both cannibalism events started, we were inside the shelter, which sits 160 m away from the colony, therefore, it was not a consequence of our direct interference—usually, non-breeding SGPs fly away with human proximity, so it is very unlikely the cannibalism events would have started with our presence nearby the colony. Furthermore, only a small fraction of breeding SGPs tend to leave the nest with the proximity of a single person in Harmony Point (around 3% of a sampled population, Krüger 2019).

Our record shows that heterocannibalism is part of the behavior repertoire of male SGPs. In fact, it might be more common than expected, since two cases were registered only ten days apart at one breeding group of Harmony Point. The lack of previous reports of cannibalism for the species could not only be related to variance in the frequency of this trait in different populations, but also to unreported cases.