Introduction

Arctic seabirds nesting colonially have adopted certain strategies to minimize breeding losses during fledging. They are species-specific and can depend on many factors, such as the mode of chick development, fledgling mobility, the distance separating a colony from the sea, the way of nesting (open nesting on cliff ledges or in hidden scree burrows), the type of predator and the level of predatory pressure.

In this review, I focus on the fledging behaviour of Little Auk Alle alle and Brünnich's Guillemot Uria lomvia belonging to the family Alcidae, subfamily Alcinae and tribe Alcini. The latter group includes Common Guillemot Uria aalge, Brünnich's Guillemot, Razorbill Alca torda, the extinct Great Auk Pinguinus impennis and Little Auk. These species are therefore the most closely related to each other. The two guillemot species and Razorbill are morphologically and ecologically similar to each other, including their fledging behaviour, but Little Auk differs significantly from all three in every respect. I shall use the term “auks” whenever I make reference to both Little Auks and Brünnich’s Guillemots.

The mass departure of young auks from large breeding colonies is one of the most spectacular phenomena to be seen in the Arctic. During a few calm days, or actually “nights”, thousands of young auks in the company of their fathers leave the colony by flying or gliding, to get to the relatively safe sea as soon as possible. Those unable to reach the sea by air walk to the shore alone. The abundance of easy prey attracts many predators from the area, mainly large gulls Larus sp. and Arctic foxes Vulpes lagopus. The fledging of Brünnich’s Guillemots was first described by Pennycuick (1956) on Spitsbergen, Uspenski (1956) on Novaya Zemlya, and Williams (1975) on Bear Island. Later, several studies addressed various aspects of the biology, ecology, and behaviour of Brünnich's Guillemots during their departure from the colony (Daan & Tinbergen 1979; Gaston and Nettleship 1981; Gaston et al. 1983; Hatch 1983; Gilchrist and Gaston 1997a, b; Gilchrist et al. 1998; Elliott et al. 2009, 2014, 2017). Stempniewicz (1981, 1983, 1995) and Wojczulanis et al. (2005) described the departure of young Little Auks from the colony, including the behaviour of fledglings and parent birds, and also mortality due to gull predation. All these papers have been consulted for this review, but there are still many unexplored aspects of this phenomenon (Harris & Birkhead 1985; Gilchrist & Gaston 1997a,b; Elliott et al. 2017).

Material and methods

In addition to the available literature mentioned above, I also used my own unpublished observations in this review. The observations of the fledging of Brünnich's Guillemots, which I refer to in this review, were made at the Gnållberget colony in Hornsund (SW Spitsbergen) in 1983, 2005 and 2014–16, and also at the Fuglehuken colony on Prins Karls Forland (W Spitsbergen) in 2017. Continuous observations were carried out for a total of 180 h at Gnållberget and for 36 h at Fuglehuken, during which the behaviour of young guillemots leaving the colony in the company of adult birds was recorded, and attacks by gulls and foxes during the consecutive stages of the departure counted. The frequency of particular types of behaviour in young and adult guillemots and predators was also assessed. Observations were opportunistic and carried out over a vast area with a complicated topography, which made it difficult to count and track the exact sequence of events related to the departure. The quality of the material gathered was also influenced by the very variable dynamics of the fledging, which depended on weather conditions affecting visibility. For the above reasons, the materials presented in this study is mainly qualitative not quantitative. Hence, my own observations of guillemots should be treated as descriptive, only to a limited extent permitting inferences to be drawn regarding the frequency of particular types of behaviour. Our previously published observations on the fledging of Little Auks, made in Hornsund (Stempniewicz 1981, 1983, 1995; Wojczulanis et al. 2005), were used for comparative purposes.

The departure of young auks from their breeding colonies

The largest breeding colonies of alcids in the Arctic are of Little Auks and Brünnich’s Guillemots, the latter usually mixed with Black-legged Kittiwakes Rissa tridactyla. Colonies usually consist of several to several dozen or even several hundred thousand breeding pairs, although in the case of the Little Auks it is often difficult to separate the individual patches of colonies that make up multi-million aggregations (Gaston & Nettleship 1981; Birkhead & Harris 1985; Harris & Birkhead 1985; Stempniewicz 2001).

Among the Alcini, both parents rear the chick but it is the father that takes care of the young birds leaving the colony (Bradstreet 1979, 1982; Birkhead & Harris 1985). In the period immediately preceding the departure from the colony, the chicks become agitated, spend a long time in the close company of their father and make deliberate movements suggesting impending departure. They interact on the nest site by head bowing and vocalizing, which may stimulate the chick to fledge. At this stage, acts of violence of adults against young birds have been observed mainly in guillemots (Gilchrist & Gaston 1997a). These attacks by neighbours on the chick and its parent, which are moving towards the edge of the ledge and getting ready to leave the colony, are especially dangerous. Quite often, they result in the chick or the parent or both being thrown off the ledge. For a chick, even if it is not killed by the fall and hitting protruding rocks, it is a lonely journey to the sea on foot without the father’s protection, and is rarely successful (Gilchrist & Gaston 1997a; Ashbrook et al. 2008). In both studied auk species, the young bird always takes off first, followed by its parent. The already airborne pair of parent and young bird may be joined by other adult birds (Gaston and Nettleship 1981; Stempniewicz 1995). From that moment on, the chick calls rhythmically repeating the same single syllables. This voice contact forms the bond between the birds during departure and enables a young bird to be found if it lands prematurely on land or on water and is lost to its father. Unfortunately, this acoustic signal is also recognized by predators which, being physically more agile than auks, find and kill lost fledglings before their fathers can rescue them. In both guillemots and Little Auks, the most important cause of chick mortality when leaving the colony is the separation of the chick from the parent before reaching the sea (Stempniewicz 1995; Gilchrist & Gaston 1997a).

Once the young bird is flying/gliding, the parent follows, usually slightly behind and above it. If all goes well, both birds land on the sea and soon set off in tandem towards the open ocean. For various reasons, however, some young birds (especially guillemots) land before reaching the sea, so they have to try to get there on foot (Daan & Tinbergen 1979; Stempniewicz 1995; pers. obs.). Although many guillemot colonies are situated on coastal cliffs and young birds leave the breeding ledges by jumping directly into the sea, some colonies are located several hundred metres from the shore. Such premature landings may result from the fact that the colony is too far from the sea, the nest site is not high enough above the sea level, or the young bird, on leaving the colony, is not sufficiently developed to fly/glide the entire distance between the colony and the sea. Further reasons may be that the guillemot chick is knocked off the ledge by neighbours (Gilchrist & Gaston 1997a), or a strong gust of wind forces it to land on the tundra. Finally, it is common for a flying/gliding Little Auk or guillemot chick to be attacked in the air by a gull; this will deflect the young bird from its aerial path even if the gull fails to capture its intended prey. Young auks that fall on the tundra try to get to water as quickly as possible. The time it takes to reach the sea depends on the distance and the nature of the ground on which they are walking. Rock rubble en route significantly slows down walking birds, but does, if necessary, provide concealment from predators. Sometimes, they come across small ponds and swamps in the tundra, which they treat as destinations, but these often become death traps for the young auks, usually being too small and shallow for avoiding predator attacks (Williams 1975; Hatch 1983; Stempniewicz 1995; pers. obs.).

The importance of the height and location of the colony in relation to the sea

The farther from the seashore the colony is located, the longer the distance that must be covered by the young auks leaving the colony and the more opportunities to kill them the gulls and foxes hunting them will have. Little auk colonies are usually located a few kilometres from the sea, although small breeding groups are also observed on coastal rock cliffs and several kilometres from the shore (Stempniewicz 2001; Keslinka et al. 2019). Young Little Auks cover this distance by actively flying. Much more important is the location of the colony in relation to the sea for young guillemots that cover this route by gliding (Williams 1975; Hatch 1983; pers. obs.). This limits the maximum distance of the guillemot colony from the sea and thus facilitates observation, because the area over which the spectacle of these birds leaving the colony takes place is much smaller than in the case of Little Auks.

The distance between a guillemot colony and the shoreline and the height of the nesting ledge on the cliff above sea level largely determine the gliding range of young birds. Simply put, it is described by a right-angled triangle, of which (a) the vertical side corresponds to the height of the rock shelf in the colony above sea level, (b) the horizontal side corresponds to the horizontal distance separating the cliff wall from the sea shore, and (c) the hypotenuse reflects the path along which the young guillemot glides at an angle β from the colony to the sea (Fig. 1).

Fig. 1
figure 1

A young Brünnich’s Guillemot Uria lomvia accompanied by its father leaving the Gnållberget colony (a—the vertical side of the triangle corresponds to the height of the rock shelf in the colony above sea level, b—the horizontal side corresponds to the horizontal distance separating the cliff wall from the sea shore, c—the hypotenuse reflects the path along which the young guillemot glides at an angle β from the colony to the sea). Photo: L. Stempniewicz

Full size image

The characteristics of a breeding colony, including its height and position in relation to the seashore, most likely influence the breeding success of auks. This seems to be of particular importance in the case of guillemots, whose young, when leaving the colony, are not yet able to fly actively. The height of the rock ledge off which the young birds jump and the distance of the colony from the sea shore determine the proportions of young guillemots that reach the sea, whether by gliding or walking or both, and thus the proportions in which they fall prey to foxes and gulls. Almost all the young guillemots leaving the relatively low (450 m above sea level) colony of Fuglehuken, located c. 500 m from the shore, covered at least part of the way to the sea on foot, whereas the vast majority of fledglings from the Gnållberget colony, which is almost twice as high (790 m) above sea level and lies close to the shore (c. 250 m), reached the sea by gliding (pers. obs.). In a particular colony, young guillemots from the higher ledges have a much greater chance of reaching the sea than those from the lower ones. The latter probably make up the largest proportion of fledglings that travel at least part of the way to the sea on foot, and are much more frequently predated on by foxes. The high mortality rate of fledglings leaving the colony on foot results from the fact that they are subjected to predation pressure from both foxes and gulls on a dangerous stretch of land that is difficult for a small bird to negotiate. Having reached the sea, they still have to overcome the most treacherous section of the shallows with numerous rocks on which seagulls are lurking. These dangers are avoided by those young guillemots that leave the colony by gliding, landing on the sea some distance away from the shore (pers. obs.).

Therefore, below a certain height threshold and above a certain distance of the colony from the sea, the overall mortality of young guillemots when leaving the colony may be so high that such places are unsuitable for breeding. Of course, guillemot colonies on taller cliffs may be situated further from the shoreline than those on low ones. But this does not apply when guillemots breed on small, predator-free, coastal islands. On such islets, a colony may be located on low cliffs well away from the shoreline, or even on a flat surface, and chicks travel the entire distance from the colony to the sea on foot, accompanied by the parent birds. They often do not complete the journey to sea on the same evening they leave the cliffs, but spend one or more of the following days in ponds under adult supervision (Hatch 1983).

Paternal care

One peculiarity of the behaviour of alcine auks is that only the father cares for the chick during its departure and for many weeks afterwards at sea (Bradstreet 1979, 1982; Brown 1985). In Little Auks, the male takes over most of the care for his chick during the last week of its stay in the colony. He remains with it much longer within and outside the nest and feeds it much more often than the female does (Harding et al. 2004). In Brünnich's Guillemots, the female feeds the chick more often than the male throughout its stay in the colony, but the male spends much more time with the chick on the nesting ledge, defending it from predators (Paredes et al. 2006).

Various possible causes of this unusual strategy in a monogamous bird species have been considered, but an unequivocal explanation has yet to be proposed. There are differences in the investments of females and males in the successive stages of reproduction, with an initially significant investment of females in producing the large egg, which might be compensated for by the subsequent paternal-only care of the young bird. However, the situation is similar in all monogamous bird species. Numerous studies of Little Auks addressing this issue have taken into account the level of stress in birds of both sexes at different stages of the reproductive period, their involvement in egg incubation and their participation in feeding the chick (Wojczulanis-Jakubas et al. 2009, 2012, 2013, 2014, 2020; Wojczulanis-Jakubas & Jakubas 2012). But so far, they have failed to discover the causes. These authors emphasize the larger size of males and their higher level of aggressiveness, which may predispose them to defend the chick during its stay in the colony, when leaving it, and later at sea. A phylogenetic determinant, i.e. a tendency towards paternal care, quite often observed among Charadriiformes, cannot be ruled out.

Predation

On leaving their colony, young auks are hunted by Arctic foxes and large gulls, mainly Glaucous Gulls Larus hyperboreus, less often Great Black-backed Gulls L. marinus and, to a lesser extent still, skuas (Great Skua Stercorarius skua and Arctic Skua S. parasiticus) (Williams 1975; Daan & Tinbergen 1979; Stempniewicz 1995). Occasionally, walruses Odobenus rosmarus and polar bears Ursus maritimus also kill them (Donaldson et al. 1995; Mallory et al. 2004; Obbard et al. 2022). The two most important predators hunting young auks leaving the colony, i.e. Arctic fox and Glaucous Gull, differ in the place and manner of hunting. Foxes hunt young auks going to the sea over land, whereas gulls hunt them in the air, on land and at sea. During mass fledging, foxes do not consume the prey immediately but store it, which greatly enhances the overall effects of their hunting. Glaucous Gulls sometimes carry prey to their young in the nest and then return to hunting, but these are marginal incidents, because by this time the vast majority of young gulls will already have fledged (Stempniewicz 1995; pers. obs.).

Foxes are territorial and the hunting monopoly is held by local fox families. They are very skilled at finding and killing prey with one snap of their mouths (Fig. 2). They carefully observe young auks flying and walking to the sea, as well as gulls that had earlier managed to catch their prey. They are also guided by hearing, flawlessly locating the place where the characteristic rhythmic voice emitted by young auks comes from, even when they cannot see the bird itself (Stempniewicz 1995, pers. obs.). They kill primarily young birds, but on occasion the accompanying parents as well. During the large-scale departure of guillemots from the colony, when many young birds that have not reached the sea by gliding walk on land, foxes kill one bird after another, leaving them on the tundra and going after more. Later, they return to collect their victims, taking them to holes they have excavated, e.g. in a nearby marsh, where they store and cover them up (Stempniewicz, pers. obs.). Holes between or under boulders offer alternative storage sites. Gulls sometimes find such a cache and steal the foxes’ prey (Stempniewicz & Iliszko 2010).

Fig. 2
figure 2

A young Little Auk Alle alle being eaten by an Arctic fox Vulpes lagopus. Photo: C. Nelo

Full size image

Young auks landing on the tundra and continuing their journey to the sea on foot are also watched for and killed by gulls. A gull is faster than a fox and can more quickly catch a prey item that the fox is also chasing. However, a gull takes much longer to dispatch its prey, so that a fox can often take over the auk the gull has caught before the latter has time to kill and swallow it (Stempniewicz, pers. obs.). Gulls often try to take captured prey items from each other (Figs. 3, 4). Such antagonistic interactions between predators lead to much confusion, thus giving a young auk that has been captured but is still alive the opportunity to escape or hide (Stempniewicz 1995).

Fig. 3
figure 3

A Glaucous Gull Larus hyperboreus hunting a Little Auk Alle alle fledgling on the sea (a–c), carrying it to the nearest rock (d) and swallowing it (e). Note the Great Black-backed Gull Larus marinus among the competitors (c). Photo: C. Nelo

Full size image
Fig. 4
figure 4

A Glaucous Gull Larus hyperboreus hunting a Little Auk Alle alle fledgling in the air (a–f). Photo: C. Nelo

Full size image

Gulls are universal predators, hunting young auks in the colony, in the air, on the tundra and at sea. They patrol the entire flight route of the young auks, circling in the air, or keeping watch from elevated sites. The animated young auks preparing to leave the colony are already in the gulls’ sights, as it were, and some are taken by surprise before they even have time to hide or fly off. During this period, the colony is intensively patrolled by gulls. The second stage is the flight/glide from the colony to the sea, during which young auks are attacked over both land and water. The gulls’ superior manoeuvrability, especially in windy weather, gives them an advantage over the auks and a significant number of them are killed in the air (Stempniewicz 1995; Gilchrist et al. 1998). Smaller prey items (Little Auks) are usually swallowed immediately, whereas larger ones (guillemots) are carried to a secluded spot and swallowed there (Figs. 3, 4). Some of the birds are forced to land on the tundra, where they meet their end (Stempniewicz 1995, pers. obs.).

Fig. 5
figure 5

A Brünnich’s Guillemot Uria lomvia father's victorious battle in defence of a fledgling attacked by a gull (a–h). Photo: L. Stempniewicz

Full size image

Those young auks that make it to the sea undisturbed with their fathers are in principle safe. They can dive very well to escape under water and are protected and escorted by their parent birds. As a rule, gulls do not attack them then. But their interest is aroused by solitary fledglings which have either reached the water on foot or lost contact with their parent on the way. These, too, can dive and run underwater, but being completely disoriented, they cannot establish the direction in which they should go and keep going round in circles. Exhausted, they sooner or later succumb to predators (Stempniewicz 1995, pers. obs.). Gulls adopt certain hunting techniques for solitary fledglings at sea, after becoming separated from the escorting father. They approach swimming chicks in a fast, low-level glide, presumably to surprise the prey, and attempt to snatch them from the water surface. Gulls also swim rapidly towards them, zigzagging among small ice floes, presumably to confuse the victims and catch them before they can dive (Jakubas & Wojczulanis-Jakubas 2010).

Anti-predator behaviour of auks

Because of their small body size, neither young nor adult Little Auks can withstand a gull attack. When directly threatened in the air, young birds perform a characteristic nose-diving manoeuvre (Stempniewicz 1983, 1995; Fig. 4). This involves folding the wings and dropping almost vertically. The onrushing gull is unable to immediately follow the falling prey and some time elapses before it can loop around and return to the chase. Meanwhile, the young Little Auk escapes at a lower altitude and if close to the sea, can reach it in time. A juvenile Little Auk can repeat this manoeuvre if it is at a sufficiently high altitude, but if not, it lands on the tundra and seeks cover in the nearest hole or pond. The higher it flies, the greater its chances of escaping or finding a hideout after landing on the tundra. This is probably why young Little Auks fly high above land, but very low over the sea, so they can dive in at any moment (Stempniewicz 1995).

Unlike Little Auks, young guillemots do not nose-dive when attacked by a gull in the air, but when attacked on land or water, they actively defend themselves even though they too stand no chance against a predator. When threatened directly, they do not run away, but instead turn towards the attacker and peck at it ferociously. When defending its young, a parent guillemot launches a violent counterattack on the predator, running towards it with its beak stretched horizontally, supporting itself with its wings and pecking the attacker. They are often effective in these counterattacks against gulls (Fig. 5). Guillemots use a similar defensive behaviour on land against an attacking Arctic fox, although in this case it is rarely effective unless the attackers are this year's naive foxes (Stempniewicz, pers. obs.).

Fig. 6
figure 6

Adult Brünnich’s Guillemots Uria lomvia on a ledge and their young jumping off it (a); a young guillemot running to the sea after landing on the tundra (b). Photo: L. Stempniewicz

Full size image

Adult auks are generally reluctant to land on flat ground because of problems with taking off again. Many parent birds, when the young bird they are escorting falls to the ground, lower their flight, slow down, circle, but do not land and eventually fly out to sea. Although Little Auks fly much more efficiently, they land on the ground only exceptionally (Stempniewicz 1995). The less efficient guillemots land more often, but then many more young guillemots in any case travel from the colony to the sea on foot (Fig. 6). Some of the adult guillemots that land on the ground jump up after a while and fly to the water, only a few of them accompanying the young ones on foot to the sea (Stempniewicz, pers. obs.)

Fig. 7
figure 7

A Brünnich’s Guillemot Uria lomvia fledgling departing to the open sea in the care of its father, unmolested by Glaucous Gulls Larus hyperboreus (a); an adult guillemot rushing with flapping wings to the rescue of a fledgling attacked by a gull (b). Photo: L. Stempniewicz

Full size image

The parental bond and care behaviour appear to be essential for the survival of young auks leaving the colony. The most frequent prey items of predators are fledglings unattended by their parents (Stempniewicz 1995). Noteworthy is the wide spread of care behaviour among parent birds, especially in guillemots. Some parents take no action when a young one lands prematurely or is attacked by a predator. Other parents restrict themselves to looking for a fledgling lost in the water or on the tundra and escorting it to safety, but leaving it to its fate when threatened by a predator. Nevertheless, on seeing or hearing the call of a young bird being attacked by gulls, many adult guillemots counterattack, running across the water surface and flapping the water surface coot-like with their wings as they run towards the attacker (Fig. 5, 7b Finally, some of the parent birds exhibit an almost suicidal tenacity in defending their fledglings. Such fierce battles occur both on land against an attacking gull or fox (less frequently) and on water and coastal rocks against attacking gulls (more often) (Stempniewicz, pers. obs.; Fig. 5).

Fig. 8
figure 8

A Little Auk Alle alle hatchling and an adult bird with food (mainly Calanus copepods) in the gular pouch. Photo: L. Stempniewicz

Full size image

It seems that the offspring of parents that are heavily involved in the protection and defence of their young have a better chance of survival. This is the case regarding antagonistic interactions with Glaucous Gulls which, while risky, do usually not end in the death of the adult guillemots. But too much involvement, as in the case of guillemots attacking foxes, would be suicidal behaviour. This variability in response may be related to the age of the birds. It has been observed that young parent guillemots readily abandon their offspring when in danger, whereas old birds desperately defend them to the last until taken by a predator. Such a difference was found in response to polar bears plundering rock ledges in a guillemot colony (Elliott et al. 2014).

Biological and ecological conditions of fledging behaviour in Little Auks and Brünnich’s Guillemots

In both species, the single eggs are large (10–20% of the female body weight) and incubated by both parent birds forming a monogamous, often long-term pair. The chicks are also fed by both parents. At the end of the nesting season, the male Little Auk takes over the care of the chick, and in both species, only the male looks after the young bird when it leaves the colony and later at sea (Harris & Birkhead 1985; Stempniewicz & Jezierski 1987; Harding et al. 2004; Paredes 2006).

The most important similarities and differences in the breeding biology, development of chicks, way and timing of leaving the colony, anti-predatory behaviour of adult birds and chicks of Little Auks and Brünnich’s Guillemots are presented in Table 1. Weighing c. 160 g, the Little Auk is the smallest Atlantic alcid species (Fig. 8). Because of its small size, it has one of the lowest wing loads among auks (0.98 g/cm2), which ensures efficient flight, but at the same time makes wing-propelled diving difficult (Stempniewicz 1982). It nests in sheltered burrows among the rock debris lying on the gentle, eroded slopes of the mountains, usually at a considerable distance (1–5 km) from the sea shore. It lays one relatively large egg (31.3 g; 19.2% of the female's body weight), from which a semi-precocial chick hatches weighing 21 g (13% of the female's body weight) (Stempniewicz 1981; Gaston 1985).

Table 1 Similarities and differences in the biology, ecology and behaviour of Little Auks Alle alle and Brünnich's Guillemots Uria lomvia, influencing their fledging performance
Full size table

Little Auk chicks stay in the nest for 25–30 days, a duration similar to the incubation period (29 days). They have the fastest growth rate among auks, reaching a maximum weight of c. 123.5 g (77.3% of adult weight) by c. 20 days of age. This is followed by a significant pre-fledging weight recession, caused by intensive wing training and the reduced feeding rate owing to the female’s deserting the brood about one week before the young leave the colony. At the time of departure, young Little Auks weigh on average 105.5 g (67.3% of the adult body weight). Their wings are shorter but relatively wider than those of adults and have a very well-developed alula. Their wing load (0.79 g/cm2) is much lower (80.6%) than in adult birds. By the time they depart the nest, their legs are almost as long as those of adults (Fig. 9). In addition, their heart weight to body weight ratio is almost half (141.9%) as high again as that of adults. The pre-fledging fall in body weight improves the locomotory capabilities and physiological efficiency of juvenile birds leaving the breeding colony (Stempniewicz 1980, 1981, 1982).

Weighing approximately 1 kg, Brünnich's and Common Guillemots are the largest living auks (Fig. 6). They therefore have the largest wing load of flying seabirds (c. 2 g/cm2), which results in high flight costs, but at the same time permits very deep and effective wing-propelled diving (Livezey 1988). They nest on steep coastal cliffs and lay a single egg (100 g; 10–12% of the female's body weight) on open rocky shelves. After 32 days of incubation, a chick weighing 65–70 g hatches from the egg (7% of the female body weight), which then develops according to the intermediate model. During the first 10 days of life, it cannot thermoregulate. It remains in the colony only for about three weeks and is fed by both parents until departure, by which time it has reached a weight of 204–212 g, which is barely 22–24% of the body weight of an adult bird. The most striking feature is their then undeveloped primaries and secondaries that preclude active flight (Gaston 1985; Harris & Birkhead 1985). On leaving the colony, the chick throws itself off the ledge, flapping its short wings vigorously and, using its primary coverts and widely spaced legs with webbed feet as supporting surfaces, glides down to the sea (Fig. 1, 6). Most of them still have fluff on their heads and rumps). After landing on the water, the chick calls loudly and the father swims quickly towards it and greets it with a growl. They touch bills and frequently circle one another before starting to swim away from the colony, the father leading and the chick by his side (Williams 1975; Gaston & Nettleship 1981; Gilchrist & Gaston 1997a; Stempniewicz, pers. obs.).

Fig. 9
figure 9

Young Little Auk Alle alle behaviour preceding fledging: leaving the nest chamber and emerging on the surface of the colony (a), feeding outside the nest (b), wing exercising (c) and starting to fledge (d). Photo: C. Nelo

Full size image

Young guillemots leaving the colony are extremely resistant to collisions with hard ground. This is due to the well-developed feathers on the lower body that function like an airbag in a car and well-developed legs that act as a buffer during landing (Gaston 1985). After hitting the ground, water or rock, they shake themselves down quickly and are ready to continue their journey. Their gliding flight, which precludes any manoeuvring, means that they hit the ground with full force. Their incredible resistance to injury is also supported by our observations from the Gnållberget colony of fledglings, which were severely battered by gulls. In two cases, young guillemots, swallowed but immediately regurgitated, were able to pick themselves up and survive (Stempniewicz, pers. obs.).

Social factors

During the period when the young guillemots are leaving the colony, the adult birds, gathering at that time in a rafting aggregation on the sea near the colony and calling loudly, play a significant part (Williams 1975; Daan & Tinbergen 1979; Gilchrist and Gaston 1997a). Below the Gnållberget colony they remain at a compromise distance from the shore: close enough for average young birds to fly to, but beyond the zone of skerries and shallows where gulls are lurking. A chick landing in the latter zone also risks crashing into a rock or disappearing from the parent's sight and has difficulties in diving successfully. This floating life raft of adult birds plays an important role in guiding young birds to a safe landing. When a lone fledgling lands on the water away from this raft or reaches the water on foot, it is completely disoriented. It swims backwards and forwards, goes around in circles, and if spotted by a gull, its fate is sealed. But if the parent bird is nearby and hears the chick's call, it swims over to it and the chick is saved (Stempniewicz, pers. obs.).

The loud choral vocalizations of a group of adult guillemots at sea may stimulate young birds to leave their native rock shelf. This decision is probably not an easy one, given the abyss opening up against the wall of the colony, the young bird’s inability to fly actively, and its strong innate behaviour preventing it from falling off the ledge. Undoubtedly, the vocal activity of guillemots from the raft correlates with the timing of departure from the colony. This applies to both the intensity of departure on the following days and at specific times of the day. When there are no guillemots calling at sea below the colony, no fledging takes place (Stempniewicz, pers. obs.). The status of the birds making up such a raft is not known. Daan & Tinbergen (1979) suggest that these are birds that have lost their broods, or they may be non-breeders in a given season. The mothers of the departing young could also be present in a raft; unlike Little Auk females, they remain in the colony for several weeks after the departure of their young (Gaston and Nettleship 1981; Harris & Wanless 1990). Young guillemots leaving the Gnållberget colony and landing at sea are usually accompanied by several adult birds at first. Some time later, they go to the open sea with their fathers only, while the rest of the birds remain where they are (Stempniewicz, pers. obs.). The significance of adult guillemots aggregating on the sea near the colony during the fledging period is not clear. Besides the positive aspects of their presence, such as fledging stimulation and indicating relatively safe landing spots (gulls avoid these), there are also negative aspects. Birds in such rafts sometimes attack and even drown chicks, especially those that are swimming alone (Gilchrist and Gaston 1997a).

A young Little Auk is usually accompanied only by its father when leaving the colony. Only occasionally do other birds join them along the way but they do not engage in protecting the fledgling. This is probably the reason why many more young Little Auks are attacked and killed by gulls during their flight from the colony (Stempniewicz 1995). In the case of guillemots, the young bird glides from the colony in the company of its father and often several other adult birds. On landing on the water, the young guillemot swims out to sea with its father, while the neighbours that accompanied it in the air remain where they are. In guillemots, the parent bird can to some extent correct the gliding trajectory of the young. If it falls too early and is heading straight towards dangerous coastal rocks, parents can extend the gliding distance of the young bird by up to several dozen metres. This they do by flying under it, thereby creating a draught of air. This alters the angle of glide, so extending the distance travelled by the chick. In the opposite situation, when the fledgling is gliding at an angle that would take it too far, i.e. beyond the group of guillemots swimming on the sea below the colony, the parent “knocks” it down by flying over it, thus forcing it to land earlier (Stempniewicz, pers. obs.).

Brünnich’s Guillemots are much more social than Little Auks (Birkhead 1985). Although Little Auks form huge, compact colonies, each pair occupies a separate nest in the rock rubble. During incubation and in the first half of the nestling period, both adult birds stay with their chick in a sheltered burrow and are not in direct contact with their neighbours. The chicks only come to the surface of the colony when they are about two weeks old, where they exercise their wings, but do not interact with other birds. During the departure from the colony and after landing at sea, father and fledgling rely solely on each other (Stempniewicz 1995; Fig. 9). Guillemots lay their eggs, incubate them and feed their chicks on an open rock ledge that is usually occupied by many neighbours. The density of birds on the shelf is further increased by non-breeders, which remain on it throughout the breeding period. Living in a compact group in a very limited space, the distance between individuals is very small; moreover, all the birds in this group are involved in defending the colony from attacks by gulls trying to snatch eggs or chicks from the ledge (Birkhead 1985; Harris & Birkhead 1985). This creates a social bond between all the birds in the micro-community into which the chicks hatch. However, interactions between these birds manifest themselves in different forms. On the one hand, conflicts between birds from the same ledge sometimes make it difficult for chicks and their escorting fathers to leave it (Gilchrist and Gaston 1997a,b; Irvine et al. 2021). On the other hand, a few birds from the nesting ledge usually accompany the fledgling and, together with its father, protect the chick from gull attacks during the gliding flight and landing at sea (Gaston & Nettleship 1981; Stempniewicz, pers. obs.).

Timing of fledging and mortality rate

Despite the increased activity of predators, the overall mortality among young auks leaving the colony is not high in relation to the size of the local auk population. In Little Auks, it is estimated at 2.8–7.6% (Stempniewicz 1995; Wojczulanis et al. 2005). In guillemots, it is even less (0.6–2.2%), but this only applies to colonies or sub-colonies situated by the sea, where the young can jump directly into the water. However, guillemot mortality grows significantly to about 17.5–30.0% when the colony is located farther from the shore and the young birds have to travel part of the way on foot, especially over difficult terrain (Williams 1975; Gilchrist & Gaston 1997a). In such colonies, fledging synchronization, resulting in a swamping effect limiting losses, becomes of fundamental importance (Williams 1975; Daan & Tinbergen 1979; Gaston & Nettleship 1981).

The young auks’ departure from the colony is clearly condensed in time. This applies both to the duration of the departure period and to its diurnal schedule. Departure usually begins in the late afternoon and evening hours, reaches its greatest intensity at night (22:00–02:00 h), when the maximum number of adult birds is present in the colony, and abates in the early morning. Given the 24 h of the polar day, this activity rhythm is probably little related to the daily changes in light intensity. Colonies with a more or less southerly exposure are in the shade at night, and this may reduce the detection of young birds by predators Williams 1975; Stempniewicz 1986, 1995).

The timing of departure from the colony rarely takes the form of a Gaussian curve, with few fledglings in the early and late stages and one marked peak in between. Most often, several fledging waves are observed, during which all the young birds, more or less ready for departure, leave the colony. These waves may be separated by breaks of even a few days if the weather is bad (Daan & Tinbergen 1979; Stempniewicz 1995, pers. obs.). The intensity of fledging seems to be hampered by strong winds. In the 2016 season in Hornsund, a strong foehn wind blowing for a week prevented the departure of young guillemots from the Gnållberget colony. The next day, after the wind had died down, departure was intense, remaining so throughout that whole windless day. Rainfall or cloudiness does not appear to be of any importance (Stempniewicz, pers. obs.).

According to many authors (Williams 1975; Daan & Tinbergen 1979; Gaston & Nettleship 1981—in the case of guillemots, and Stempniewicz 1995—in the case of Little Auks), the synchronization and condensation of the young auks’ departure from the colony minimizes losses due to predation. The key to the fledging success of a young bird is the decision when (day and time) to fly out, because the pressure exerted by gulls varies greatly in time. The birds flying out first in each wave are disproportionately vulnerable to predation. The gulls are then hungry and attack any fledglings they spot, often still in the air. At the beginning of the daily departure, between 16:00 and 20:00 h, airborne Glaucous Gulls killed 18% of young guillemots leaving the colony at Ingeborgfjellet (W Spitsbergen). But large-scale departures at night (20:00–24:00 h) are safer, with hardly any attacks by gulls (Daan & Tinbergen 1979). During peak fledging, most gulls are busy processing and digesting previously caught and ingested prey. They now take much longer to handle their prey and, being unable to swallow another large prey item whole, they tear it apart and consume small pieces of the flesh (Stempniewicz, pers. obs.). Gulls overfed at peak fledging temporarily lose their motivation to hunt (Fig. 7a). If they do decide to attack, they go for easy victims, such as lone fledglings on water or on land. Young auks departing during this period, even those walking on the tundra, have a much greater chance of reaching the sea unmolested and surviving. This reduces losses among young birds leaving the colony during the fledging peak, but increases losses among those leaving the colony in the first wave (live shield effect) (Daan & Tinbergen 1979; Stempniewicz 1995).

The heightened mortality among the first and last young auks to leave the colony is a manifestation of the stabilizing effect of selection for the timing of fledging. This promotes birds that tend to fly at peak times rather than off peak, irrespective of the hour of the day. Usually, the rush-hour occurs only at "night". But after a long break due to unfavourable weather, the next day, in fine weather, young guillemots flew throughout the day with great intensity, as if making up for the shortage caused by the break (Stempniewicz, pers. obs.). The most important factor influencing mortality in fledglings is their degree of advancement, which will affect their fitness and potential flight range. The less advanced ones are unable to reach the sea and land on tundra or in a dangerous coastal zone with numerous shallows and skerries, with gulls lying in wait, as it were. The farther the young auks can fly, the greater their chances of survival. For Little Auks, it is best to fly as far as possible beyond the dangerous coastal zone, whereas for guillemots, the aim is to land as close as possible to a raft of adult birds (Daan & Tinbergen 1979; Stempniewicz 1995, pers. obs.).

Ydenberg et al. (1995) presented a model defining a colony departure strategy for young auks in relation to growth rate, timing and mortality in the colony and at sea. Two predictions follow from it. One is that faster-growing chicks can fly out at a younger age because they achieve a larger body weight faster than those growing more slowly. The other is that if later-hatched chicks are to leave the colony during the fledging peak, they will necessarily be younger, less well developed and lighter in weight. This model, however, does not take into account the differentiation in the mortality of young auks in the different phases of departure from the colony. It takes many days, does not resemble a Gaussian distribution, and usually has several peaks. It appears to be more important to fit into any of the fledging peaks, both during the day and throughout the fledging period (Daan & Tinbergen 1979; Stempniewicz 1995, pers. obs.).

A large number of young birds leaving the colony at the same time with a more or less constant number of predators reduces mortality (swamping effect) and to some extent reduces the importance of the age and body weight of the young birds leaving the colony. Many authors (Greenwood 1964; Williams 1975; Daan & Tinbergen 1979; Gaston & Nettleship 1981; Stempniewicz 1995) have assumed that the swamping effect reduces the magnitude of the overall losses caused by predation among young auks leaving the colony. The evident pre-fledging weight recession observed in Little Auks leaving the colony by active flight reduces their wing load and improves flight efficiency (Stempniewicz 1980, 1995). One would expect it to be greater in the faster-growing, earlier hatched nestlings than in the less advanced, later chicks, which cannot afford a significant weight loss.

The length of the auks’ breeding season depends mainly on latitude and the ice conditions near the colony. It is more concentrated in colonies located further north in more severe oceanographic and climatic conditions (Harris & Birkhead 1985). However, the degree of predation pressure may also affect breeding duration. In colonies where predatory pressure is low, e.g. on Horse Head Island (W Greenland; Evans 1981), the hatching period, and in consequence the fledging period of Little Auks, was around twice as long as in Hornsund, where the level of gull predation is high. The pressure to shorten the breeding period results from the fact that the magnitude of brood losses is correlated with the duration of exposure to predation, especially of such sensitive stages as young birds leaving the colony (Stempniewicz 1995).

The most important factor determining the magnitude of losses among young auks leaving the colony is the proportion of predators to prey. Owing to the disproportion in body size and development pattern, the duration of the reproductive period in auks and gulls differs significantly. As Glaucous Gulls are large and have a semi-altricial development pattern, their nesting period (incubation + chick feeding) is very long (c. 75 days). This period is much shorter in the semi-precocial Little Auks (c. 56 days) and intermediate guillemots (c. 53 days; Harris & Birkhead 1985). Thus, the duration of their stay in the colony and exposure to predatory pressure is also limited. When auks leave the colony, a significant source of food for gulls disappears. As a result, the size of the gull population trophically associated with colonial auks is maintained at a much lower level than the hypothetical level that could be maintained, if the gulls were able to rely on such an abundant source of food throughout their stay in the Arctic. This disproportion in the number of prey and predators most effectively reduces the magnitude of losses among fledging auks (Stempniewicz 1995).

Premature departure from the colony by young guillemots could be also explained by safety reasons. The predatory pressure of gulls on chicks in the colony accumulates with each successive day that they remain there. Early departure from the colony would reduce the total mortality of young birds. But this explanation does not seem convincing, because the earlier they leave the colony, the less advanced they are in development, which must result in higher mortality at this stage (Williams 1975; Daan & Tinbergen 1979). As demonstrated by Elliott et al. (2017), the mortality of young guillemots in the period after leaving the colony, when they are on the open sea in the care of their fathers, is similar to that experienced during their stay in the colony, so departure from the colony does not change this situation. According to these authors, the impact on the length of the stay of young guillemots in the colony and the age at which they leave the colony is not related to safety, but is determined by the amount of energy that a young guillemot can receive in the colony and at sea.

Ashmole’s halo

One of the hypotheses explaining the premature leaving of the colony by young auks is the Halo effect (Ashmole 1963). The Central Foraging Hypothesis predicts that during the chick feeding period, the range of foraging flights in colonial seabirds is limited by the need to return to the nest and feed the nestlings. Foraging in the area adjacent to the colony can therefore, over time, lead to the depletion of available local resources, limiting reproductive success and delaying the breeding age. This is one of the proposed reasons why young auks leave the colony prematurely and forage with their parents in an unrestricted sea area. In fact, Elliott et al. (2017) showed that, having left the colony, young guillemots receive more than twice as much energy per day as during their stay in the colony, even though at that time they were being fed by both parents, and now only by their father. At sea, however, the parent bird no longer needs to expend time and energy on flying to the feeding grounds and returning to the colony. The parent bird can also choose any feeding place, because it is not limited by the range of foraging flights from the colony.

Overfishing of stocks near the colony during the breeding season, though likely, is very difficult to test owing to the great mobility of shoal fish and planktonic crustaceans, which are the main food resources of seabirds. There is also the strong effect of sea currents carrying fish and crustaceans from other regions. The Ashmole hypothesis is supported by several pieces of indirect evidence, such as energy models, showing a close relationship between colony size and foraging range, and determining maximum colony size. This also suggests a negative correlation between colony size, growth rate and mean body weight of young guillemots (Furness 1982; Gaston et al. 1983; Furness & Birkhead 1984; Hunt et al. 1986; Jovani et al. 2016; Patterson et al. 2022). This hypothesis was directly tested with positive effect in the Shag Phalacrocorax auritus (Birt et al. 1987). However, this does not apply to auks, which feed on highly mobile prey, whereas Shags prey on demersal stationary fish. Also, Weber et al. (2021) reported the depletion of flying fish abundance within the range of foraging flights of boobies Sula dactylatra and S. leucogaster and Ascension Frigatebirds Fregata aquila breeding on Ascension Island. But this is also an example that does not apply to auks as it comes from very different ecological system.

A very interesting study by Elliott et al. (2009) on Brünnich's Guillemots equipped with GPS loggers showed that the length of their foraging flights increased over the course of the season. At the beginning of the breeding season, guillemots caught large fish, later medium-sized, and towards the end, small fish and invertebrates. Those authors believe that pressure from a large colony leads to a gradual thinning of low-mobility benthic prey, while pelagic prey, especially larger fish, escape from the zone adjacent to the colony, resulting in a Halo effect. They conclude that it is a mechanism that partially regulates the size of local populations of guillemots. However, there is still no direct evidence of a decline in pelagic stocks of fish and crustaceans in the vicinity of large auk colonies as a result of their foraging during the breeding season. In the case of the planktivorous Little Auks, this seems unlikely. But if this hypothesis were confirmed, it would lead to the prediction that the larger the colony, the earlier (at a younger age) the young Little Auks should leave it. Our studies on Little Auks from colonies that differ with respect to size and foraging flight distance did not reveal such a relationship. Both in Hornsund, where c. 500,000 pairs of Little Auks nest and parent birds feed close to the colony (up to 60 km), and in Magdalenefjorden (c. 20,000 breeding pairs; foraging flights up to 150 km), the fledging age was similar (Jakubas et al. 2012, 2013; Keslinka et al. 2019).

In conclusion, the size of an auk colony has both benefits and limitations for breeding birds. Large colonies are safer because of the better functioning warning system against predators (Little Auks; Stempniewicz 1995) and the more effective direct defence against them (guillemots; Gilchrist & Gaston 1997a, b; Gilchrist et al. 1998; Irvine et al. 2021). Owing to the seasonal nature of the colony as a food source for predators, which is limited to the breeding season, the larger the colony, the more favourable the ratio of the number of prey (auks) to predators (gulls and foxes). This results in fewer losses of eggs and chicks in relation to the entire local population of auks (Stempniewicz 1995). The abundance of feeding grounds available within the range of foraging flights of birds appears to be of fundamental importance for the size of the colony. An example illustrating the relationship between the size of the colony and the available food resources are the multi-million Little Auk colonies in the Thule region on the trophically rich North Water Polynya and the several hundred thousand guillemot colonies in the Canadian Arctic and on Bear Island. Certain negative factors observed in large, densely populated guillemot colonies, such as the increasing frequency of conflicts between birds nesting on the same rock ledge, which may result in increased losses among chicks leaving the colony, do not seem essential for colony recruitment.