VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 69

Ornithological Studies at Casey, Antarctica, 1977-1978 By ALAN N. COWAN, 6A Vancouver Street, Red Hill, A.C.T., 2603.

Summary This paper reports on ornithological work carried out by the author during his stay at Casey, Antarctica, as Medical Officer to the J 977 wintering party from February 1977 to February 1978. Preliminary notes on the history, geography and climate of Casey and on previous local studies are followed by an outline of the programme undertaken and of the problems encountered. The annual cycle of birds and other animals is briefly described. Systematic accounts are then given of each bird species, with status in the area including numbers where estimates have been attempted, and notes on behaviour. Introduction Casey, the most recently built of Australia's Stations on the Antarctic continent, lies on the Budd Coast just above the Antarctic Circle at 66° l7'S. Lat., 110° 32'E. Long. The station is on the Bailey Peninsula at the N.E. end of Vincennes Bay, which was named by Mawson after the Vincennes, a vessel of the U.S. Exploring Expedition of 1838 to 1842 (Commander, Lt. Charles Wilkes, U.S.N.), the first explorers in the area. Wilkes Station, just N. of Casey, was built by the U.S. Navy in 1957 for the International Geophysical Year. Subsequently taken over by Australia in 1959, it was abandoned in 1969 when Casey was completed and occupied. Geography Several glaciers flow into Vincennes Bay, whose coastline is mainly one of steep slopes of permanent ice and snow broken here and there by moraines and nunataks. In the N.E. part of the bay is the complex chain of the Windmill Islands, the main scenic feature of the a rea and also the mai n focus of ornithological interest. In summer, many of these rocky islands become extensively denuded of snow, providing nesting sites for the Antarctic seabirds which throng the area for the brief breeding season. They are the only available exposed rock area of any size in at least 800 km. of coastline. Two of these islands, Ardery and Odbert constitute a Specially Protected Area under the Antarctic Treaty 1967. They support breeding populations of the Antarctic P etrel (Fig. 1) . Climate Considerably to the N. of such places as McMurdo Sound, Casey's climate is relatively mild. The mean daily maximum temperature for January is + 1 deg. C. and the mean daily minimum for July is -20 deg. C. (The Antarctic Pilot 1974, data for Wilkes). The wind is characteristi­ cally N.E. to S.E.. Blizzards may occur, often with little warning, at any season but are of highest velocity and greatest frequency and duration during the winter months. Katabatic winds are much less a feature of the Casey climate than they are at some other stations, and there are many calm days. During the summer months daylight is prolonged and the radiant energy of the sun is great, so that even on subzero days there is tremendous melting of snow and ice due to the heating of the dark rock beneath; rushing streams with profuse growth of algae are a feature of summer on these islands. AUSTRALIAN 70 COWAN BIRD WATCHER

The sea ice in this region is notoriously unstable, and except for certain protected channels and bays can break up quite suddenly under the influence of tide and wind. Previous Ornithological Work in the Area Since the U.S. withdrawal the Wilkes-Casey area has been rather neglected in respect of bird studies and Australian biological work in Antarctica has been concentrated elswhere. The reasons are obvious; unless boats are available for visiting the bird breeding islands through the summer, observations terminate when the sea ice breaks up in spring. The last extensive use of (U.S.) boats was by Orton in 1961 (Orton, 1963). Two trips were made by Hicks in 1963, and since then there has been no summer programme on the islands, except for brief helicopter visits dur· ing the few hurried days of changeover. In winter one can travel to most of the isl ands over the sea ice, but the birds are then absent, and even in winter the vicissitudes of the weather and its effects on th e sea ice can suddenly bring about a dangerous situation. It is very largely due to the keen amateur interest of previous Australian wintering expeditioners that there is a body of information on the zoology of the area. Most of this is unpublished (AN ARE Station Logs). The 1977 Biological Programme T he author chose Casey for his year's stay just because there was so much scope for research. It was clear that boats were essential and accordingly a biological programme using boats was written to be con­ ducted there during the year. The Director, Antarctic Division, approved the programme and authorised the purchase of two French-made Hutchin­ son "Mapa 401 " inflatable boats with Volvo "200" outboard motors. It may be difficult for the reader without Antarctic experience to appreciate the practical problems involved in such a programme. The prevailing wi nd is off the land. T he water is always very cold and one could not survive prolonged immersion. Travellers must be prepared to camp for several clays on an island if overtaken by the weather. Tent, food, radio etc. must be carried. The two boats must never separate. Waters are largely uncharted and shallow reefs abound. Pack ice is always a menace and may appear capri­ ciously and unexpectedly, forcing extensive detours by small inflatable boats. A precise calendar of the complex seasonal movements of some bird species is hard to obtain. Dates vary with latitude and from year to year depending on, among other factors, snow accumulation at nest sites and the occurrence of blizzards. Observations are limited by weather condi­ tions and by restrictions on travel imposed by essential work at the station. Even a brief study period on the islands requires a major logistic exercise involving at least 3 men, 2 boats and much equipment being absent for a cl ay. Observations were therefore far from continuous. However we made 15 trips in the inflatables during December 1977 and January 1978. We had in fact attempted our first trip in March 1977 but this proved abortive due to early autumn sea ice formation. In addition a final trip was made in the ship's boat during changeover in January 1978. Most of the trips were to Ardery and Odbert, but the Roll-O'Connor group and Peterson VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 71

= CHARLTON!. <\::::) DEWART I. NELLY l.:c::;::. - FRAilER IS.

1(,1;;.15'5 ~~~~~- <;,G " \ 5 '5 I SHIRLEY \. I

RIDGE. ARDE.RY ;V ODB~

HOLL \ \?c?))JS:D Qo I

PENNEY BAY

0 ' eooew" KILOME.TRE.S4 " '~

Figure 1. The Windmill Islands. For clarity, only those places mentioned in the text are labelled. AUSTRALIAN 72 COWAN BIRD WATCHER were also visited. Earlier in the season, of course, visits to many of the islands were made on foot across the sea ice. The boats were forced to make a long detour across the open sea out­ side Hollin and Midgley and their associated reefs until late December when the Robertson Channel became clear; this allowed us a shorter and more sheltered route. However, O'Brien Bay remained frozen during the whole summer and we always had to sail outside Beall, while the channel inside Shirley opened just in time for our very last trip in the inflatables. It was stipulated that the boats were only to be used for the one year, at the end of which they were to be brought back to Australia. In fact they suffered damage and, though repaired, could not have been regarded as seaworthy for a second season. The Annual Cycle It will perhaps be helpful to describe the general pattern of the breeding cycles of the Antarctic animals and of the observations recorded throughout a typical Antarctic year. With the exception of the Emperor Penguin, a winter breeder which visits the Casey area after the breeding season, all the birds are summer breeders. Most begin to arrive in October although the Southern Giant­ Petrel is earlier and Wilson's Storm-petrel later. In general the breeding cycles of the "fulmarine petrels" are highly sychronous. Throughout this paper the term "fulmarine petrel" is used to include the tour specJes Southern Fulmar, Antarctic Petrel, Cape Petrel and Snow Petrel, but to exclude the Giant-Petrels. The last•- named should properly be included in the expression, but I have adopted this course for the sake of brevity. On the other hand the cycles of the Southern Giant-Petrel and Wilson's Storm-petrel are much less synchronous. The later and less synchronous breeding cycle of Wilson's Storm-petrel is lucidly discussed by Beck (1970) and seems to be well adapted to the bird's burrow-nesting habits; it is often late in the season before the burrows can be cleared of snow by this weak digger of delicate build. A severe blizzard at a critical time could destroy the young of an entire area if all were at the same stage of being newly hatched and having built up no fat stores. Most of the species depend heavily on zooplankton which is super­ abundant during the summer and early autumn. Fledging and departure must be completed before the winter. Most species leave the colonies by late March but the Southern Giant-Petrel, Wilson's Storm-petrel and South Polar Skua stay later, until April or May. Some individuals of the "fulmarine petrel" species, after leaving the colonies, remain in the general area for some weeks. However, winter on the Antarctic coast is not entirely devoid of bird life. A big "blow" can break up the sea ice at any time, bringing open water up to the shore and bringing also the birds that winter in the loose pack and open sea. Snow Petrels, Antarctic Petrels and less often Cape Petrels may thus be seen, a reminder that though absent they have not been far away. Southern Giant-Petrels may also appear during the winter months. The cycles of the "fulmarine petrels" are characterised by a pre-laying exodus or "honeymoon". Having occupied the colonies, established nest sites and mated, the birds leave again to feed at sea and return to lay some two weeks later. This exodus seems to be more synchronous in the VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies a t Casey 73

Adelie Penguins approach route to colony, O'Connor Island. P late 25 A.N.A.R.E. Photo by Alan N. Cowan

Adelie Penguin colony, O'Connor Island- Ardrey Island beyond . P late 26 A .N.A.R.E. Photo by Alan N. Cowan AUSTRALIAN 74 COWAN BIRD WATCHER

Antarctic Petrel and Snow Petrel. It is presumably related to the harsh conditions and the long periods of continuous incubation requiring heavy feeding before laying. The Adelie Penguins arrive in October, bringing with them the Leopard Seals whose staple diet they furnish and the South Polar Skuas which al so largely depend on them for food and breed in close association with them. The Weddell Seals, which frequent the coast throughout the year, have already started to pup, at the end of September. Emperor Penguins though not breeding locally, visit the coast in summer, as do occasional Crabeater Seals, numerous Elephant Seals at Peterson and, looking very formidable from a small boat, packs of Killer Whales. Systematic Accounts Emperor Penguin Aptenodytes forsteri Status in the Area: A regular summer visitor. Notes: The unique breeding behaviour of these birds, incubating in the Antarctic winter, is well known. Two adults showing no sign of moult appeared on the sea ice of Newcomb Bay on November 19. Birds in juvenile plumage (with white neck patches) started to be seen early in December especially on the sea ice of O'Brien Bay. Two birds, one adult and one juvenile, were seen high up on O'Connor on January 2. Three ad ults were seen on a small floe in Sparkes Bay on January 3. With the exception of the first two birds seen, all showed signs of moult. Pryor (1 968) at Haswell Island found that most chicks of the season had fledged fully when they left the breeding grounds in late December to early January, but the adults then departing were in various stages of moult. Presumably birds in adult plumage seen in November at Casey, at least 400 km. from the nearest colony (Watson, et al., 1971), would be non-breeders which had not yet started to moult. The birds typically remain on floes during the moult, which takes 6 weeks (Watson, 1975). At stations near colonies, e.g. Davis, Emperor Penguins are to be seen throughout the year (Johnstone, et al., 1973). Adelie Penguin Pygoscelis adeliae Status in the Area: By far the most numerous breeding species. There are large rookeries on many islands and on the Clark Peninsula. Eklund (1961) estimated a combined breeding population for the area at 155,000 individuals. The Shirley rookeries are readily accessible from Casey over the sea ice until late in the season and a census has been done several times in recent years. Orton (1963) estimated 3,000 nests on Shirley in 1961. Counts in 1974, 1976 and my own in 1977 all indicated somewhat more than 7,000 nests (ANARE Station Logs). Clearly the numbers on Shirley have swelled in recent years despite the increased attentions of human investigators from the new station nearby. However the individual colonies are constantly changing, old areas being abandoned and new taken over, and it may be that numbers on some of the other islands have fallen. Thus, a large colony at the N. tip of O'Connor is now empty although many nest depressions are still apparent. Notes: The Adelie Penguins colonise islands which provide extensive areas of rock and gravel, especialJy plateaus and basins, and reasonably accessible shorelines although in places they choose regular landings VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 75 requiring a leap of at least 2 metres onto an ice foot in preference to easily sloping shores nearby. Sometimes high steep ridges are occupied first; often these are the first areas to become snow-free. A long stiff climb up from the sea seems to present no problem to these birds. The arrival of the first Adelie Penguins at the rookeries takes place in mid-October and is an eagerly awaited event for the men who have passed the long winter vigil. The first breeding birds to appear, they symbolise the coming of spring and the return of life to this dead land; to be per­ mitted to witness this resurrection is a privilege. It is unusual for the sea ice at this time to extend for more than a few km. from the shore, although in other areas the birds may have to cross up to 100 km. of ice (Stonehouse, 1968) and will continue to travel even in a blizzard (Sapin-Jaloustre, 1960). An excellent summary of the Adelie Penguins' breeding cycle and behaviour is given by Penney who studied them intensively for two years at Wilkes. They provide an excellent subject for study, being easily approached and pursuing their affairs noisily and demonstratively. Stones are the only available nesting material and during the early days the colonies present busy scenes of the collecting, arranging, stealing and reclaiming of stones as the male birds, which tend to arrive earlier, establish their nest sites, frequently in the same spot as the previous year (Penney, 1968). The first birds were seen to arrive on October 15. Laying began a month later. The normal clutch is two eggs. The male incubates for the first two weeks and by the time he is relieved has been ashore continu­ ously without food for about fi ve weeks (Penney, 1968). It is during this first stint of incubation that the census is best done, since the colonies present an orderly appearance with one male on each nest, the females away and only a few non-breeders hanging about. Yearling birds, distinguished by their white chins, were not seen until late in the season, and then only very few. Clearly most yearlings do not visit the colonies, whereas we saw numbers in the pack ice on our home­ ward voyage. Hatching began on December 14 and two days later the first clutch of two chicks was found. Hatching continued until early January, although copulation between unsuccessful breeders was seen well after the end of the hatching period. At three weeks the chicks started to show independent behaviour, standing and moving about without the parent trying to cover. A t four weeks the chicks were starting to group into creches, apparently round the nucleus of a chick still on its nest site. At this stage the parents of this chick would try to drive off the strange chicks but without much conviction, and they were by now large and well-covered enough to tolerate the attacks; earlier in the season many lost young chicks are killed by the adults across whose territories they wander. By six weeks after the beginning of hatching, large creches of fifty or more had formed and fl edging had begun, the white chin of the immature being the first feathering to appear through the down. At this time some adults also had started to moult. The chicks departed for the sea in February. Some adults stayed about the rookeries until late March. A feeding study on Adelie Penguins was carried out and will be the AUSTRALIAN 76 COWAN BIRD WATCHER subject of a separate paper. A stomach tube manufactured on the station was successfully used to withdraw samples of stomach contents. It is hoped to be able to define the diet of these birds in the area. Previous similar studies have been done only in much higher latitudes at R oss Island (E mison, 1968; Paulin, 1975). One albino was was seen and photographed by Geoff N aughten. Southern Giant-Petrel Macronectes giganteus Status in the Area: These birds breed on the Fraziers, the outermost of the Windmill Islands. This is one of only four known colonies on the edge of the Antarctic continent itself (Johnstone, et al., 1973; Watson, et al., 1971). There are between 100 and 200 nests. A banding programme has been conducted here for some years, using helicopters at changeover. Notes: T he nests are completely exposed on rocky ridges which are proba bl y bare of snow for m ost of the year. Stones and colony debris are the only materials available. Since these birds are sometimes seen at Casey in winter it seems possible that some individua ls m ay frequent the islands throughout the year. Very few Southern Giant-Petrels were seen at the Adelie Penguin colonies, although one chick vomited an Adelie flipper while being banded. They were more in evidence at the Weddell Seal pupping grounds; and particularly at Peterson where several old E lephant Seal carcases attracted a flock of twenty. The breeding of this species is highly asynchronous. We found one chick much too small to band on January 25, 1978, while some others had very well-developed feathering, and there were almost certainly some young of the year fl ying. One adult was even sitting on an egg, which appeared new. Also, a wide range was found in bill lengths measured on 57 chicks in F ebruary 1977 (Table I). A s a comparison, Watson (1 975) gives exposed bill lengths for adults as over 97 mm. for m ales and u nder 97 mm. for females.

TABLE 1: Bill measurements* of Southern Giant-Petrel chicks on the Frazier Islands, February 13 and 17, 1977. Bill length mm Number Bill length mm. Number 63 1 82 2 65 1 83 1 68 2 84 2 69 2 85 2 70 1 86 4 71 1 89 3 72 3 91 1 73 1 92 2 74 2 93 2 75 4 76 6 Total 57 78 9 80 3 81 2 Mean 78.9mm.

*Measurements made of the exposed bill from the tip to the end of the feathering ( = culmen). VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 77

A delie Penguin feeding chick, O'Connor Island. Plate 27 A.N.A.R.E. Photo by Alan N. Cowan

Southern Fulmar and egg. Plate 28 A.N.A.R.E. P hoto by A lan N. Cowan AUSTRALIAN 78 COWAN BIRD WATCHER

The "Fulmarine Petrels" As noted above, four species are included. These are the Southern Fulmar Fulmarus glacialoides, the Antarctic Petrel Thalassoica antarctica, the Cape Petrel Daption capense and the Snow Petrel Pagodroma nivea. All but the first of these are monospecific genera, while the first genus Fulmarus contains only the Northern and Southern Fulmars which are both derived from a single stock (Watson, et al., 1971). All show strong similarities in appearance, flight, breeding behaviour and ecology. All have a pre-laying exodus and a synchronous cycle. The Southern Fulmar and Cape Petrel range much further North in winter than do the Ant­ arctic Petrel and Snow Petrel which are birds of the open pack ice and adjacent water. Each of the four selects a different type of nest site on the rocky cliffs and islands. Southern Fulmar Fulmarus glacialoides Status in the Area: By far the most numerous of the "fulmarine petrels", it breeds on five islands: Ardery, Odbert, Holl, Dewart and Nelly. The results of thorough counts made on Ardery and Odbert, and of perforce more rapid counts on Holl and Nelly are shown in Table II, which also shows previous counts. It is only proper to point out that the counts in 1972 and 1976 had to be made while access over the sea ice was still possible and hence before the start of laying, so that it would have been very difficult to exclude non-breeding birds; even after laying it is not always easy from a distance to distinguish between resting birds and incu­ bating birds, especially as many breeding areas are inaccessible and count­ ing has to be done from sea level. The main concentrations are at the W. end and along the N. cliffs of Ardery, along both N. and S. cliffs of Odbert and on the N.W. aspect of Holl. TABLE II: Southern Folmar: Estimates of numbers of nests. Summer of: ARDERY ODBERT HOLL NELLY 1961-1962(1) 1,500 500+ 160 A few hundred 1972-1973(2) 10 000* 5-8 000* 1976-1977(3) ' 2'315* 1977-1978 1,750 1:600 400 ?100 (1) Orton, 1963 (2) Luders, 1977 (3) ANARE Station Logs (Tyrrell) *See text for comments. Notes: Of the "fulmarine petrels" the Southern Fulmars usually choose the least stable nest sites, small ledges and corners in areas of precipitous unstable scree and often in the path of melt streams. Very many eggs were lost. An exception was the small colony at Mast Point at the W. end of Ardery where the nests are on wide, well protected terraces on tabular boulders. At this colony eight banded birds were recovered in January 1978. All had been banded in January 1962 or December 1963 as breeding birds and were thus in their 15th or 17th breeding season at least, a remarkable record in this harsh environment. Southern Fulmars appeared on October 22. By a month later they were present at the colonies in large numbers pairing, mating, displaying and VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 79

Antarctic Petrel and chick. P late 29 A .N .A .R . E . P hoto by A la n N . Cowan

Antarctic Petrel chick. P late 30 A.N.A.R.E. Photo by Alan N. Cowan AUSTRALIAN 80 COWAN BIRD WATCHER

active ly clearing snow to expose nest sites. The extraordinary maniacal cackling display of these birds is the dominant noise in the mixed petrel colonies. Their Jack of concern for man is total. The relatively asynchronous pre-laying exodus of this species, together with the large numbers, meant that they were present in good numbers continuously from the time of their arrival. Opposing views have been expressed by two previous authors (Orton, 1968; Luders, 1977) about competition for nest sites between this species and the Antarctic Petrel. My own observations indicated no obvious conflict; it appeared that the two species prefer different types of site. Laying started at the beginning of December and was completed by the end of the third week in December. The single white egg, relatively ve ry large as are those of all the " fulmarine petrels" and of Wilson's Storm-petrel, could be seen by gently lifting the flank of the sitting bird; with a quiet lateral approach one would receive no more than an enquir­ ing nibble. Egg losses were very high. Many rolled off the ledges or were carried away by landslides or chilled by melt streams and abandoned. Skuas snatched others. One Southern Fulmar was seen eating an egg of its own species which was cold and deserted, but after tentative tasting it desisted. The colony at Mast Point already mentioned had a higher success rate than was general; when hatching began, at least 85% of the "sitting" birds were on an egg, compared with a rate of 60% elsewhere.

A pattern emerged as the season progressed. A single bird on a nest site would sit tight when approached, and would almost always be found to be on an egg. A pair, however, would fly off the nest site and would almost never have been on an egg. Many other birds sat resting about the colonies, on rocks clearly not the site of nests. These too would take off if disturbed. Some of these must have been the mates of incubating birds but not all were, judging by the numbers present. Apparently large numbers of non-breeders attended the colonies. This applied also to the Cape Petrels and Snow Petrels, but not to the Antarctic Petrels (see below). This is the latest-hatching of the "fulmarine petrels" and the first chicks were seen on January 23; none had been present on January 17. The chicks of each of the four species have distinctive characteristics. It is a pity that the old name "Silver-grey Petrel" has been abolished since it is exactly descriptive, especially of the chick. One has to lift the parent to see the chick, which is completely covered and closely guarded until it has learnt to defend itself by projecting its stomach contents at the intruder. This guarding behaviour applies also to the Antarctic Petrel and Cape Petrel. The parents "shuttle" with food loads, and many find food nearby. In calm weather large mixed flocks of petrels were to be seen swimming and feeding within a few hundred metres of the colonies. The Southern Fulmar, like the Antarctic Petrel, is reluctant to part with its stomach contents and will do so only under great provocation. It is the weakest on its legs of the "fulmarine petrels" , belly-flopping on landing and taking off with like clumsiness. This may be connected with its choice, in general, of precipitous slopes for nesting. VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 81

Antarctic Petrel Thalassoica antarctica Status in the Area: The species breeds on Ardery and Odbert. This is one of only nine known rookeries (Watson, 1975) although, as with some other species, most known colonies are near stations, and doubtless others exist on unexplored coasts. For estimates of numbers see Table III:

TABLE Ill: Antarctic Petrel: Estimates of numbers of nests. Summer of: ARDERY ODBERT 1961-1962(1) 50 35 1972-1973(2) 420* 90* 1977-1978 200 50

(I) Orton, 1963, 1968 (2) Luders, 1977 ' The same comments apply as to Table II (see text). Notes: The opportunity to study this bird was one of the reasons for my electing to spend the year at Casey. The literature on the species is scanty, but does include papers by two previous expeditioners at W il kes and Casey (Orton, 1963, 1968; Luders, 1977). Antarctic Petrels were first seen on October 20. T hey were thus the earliest of the four " fulmarine petrels" to arrive. The nest sites are established on long horizontal rock ledges and terraces. Being relatively proficient on their legs, they are able to take off from, and land on such fi at areas, which would be less well suited to the Southern Fulmars. Of the nests on Ardery, the biggest concentration was in a spot previ­ ously undescribed so far as I can find near the E. end of the northern cliffs. A natural cutting breaks the cliff line and one can descend on mainly gentle gradients along what is in part almost a natural track. This

A.RDE.RY ISLA.ND FROM THE NORTH

PENNEY H~FWAY CAVE I>.ND MAST POINT RAVINE RAVINE CI>.VE RI>.VINE AND BOAT LANDING SOUCEK RAVINE

Figure 2. Northern face of Ardery Island. The largest number of nests of the Antarctic Petrel is on the " platform" . Others are in the area of Penney and Soucek Ravines. AUSTRALIAN 82 COWAN BIRD WATCHER access to the birds is not dangerous provided ordinary care is exercised on patches of loose scree. This group of nests numbers upwards of 150 (Fig. 2). The pre-laying exodus appeared complete; not a single bird was seen on Odbert between November 13 and 22, but they had returned by November 26. This beautiful species is a gentle creature, both in its pair displays and in its incubating behaviour; the quiet observer produces no disturbance. The earliest of the "fulmarine petrels" to hatch, the first chicks were found on January 11. Very few, if any non-breeders seemed to be present at the breeding areas, and as these are discrete and circum­ scribed a good estimate is possible. Every individual not incubating seemed paired with one that was, and each incoming bird seemed to land at a nest site with mutual recognition displays. Further, the success rate of layi ng and hatching was the highest of the four species; between 80 and 90% . T his could be associated with the level, stable sites used. By two weeks of age the chicks were being left unattended, and were able to regurgitate and to vocalise in self-defence. They are light grey with a distinctly long·necked appearance and the black bill is long and slender. A consistent observation throughout the summer programme was that no Antarctic Petrel was ever seen away from the breeding colonies, in contrast to the other "fulmarine petrels". It seems to be very sedentary at its breeding grounds, obtaining its food close by.

Cape Petrel Daption capense Status in the Area: Cape Petrels breed in fair numbers, on ten of the islands and one mainland area. The biggest concentrations are shown in Table IV. Notes: We first saw three birds on October 29 at Odbert, together with six Southern F ulmars. Weather conditions were severe; a blizzard was developing, and in better weather we might have seen more since in general birds retire to sea in a blizzard to seek milder conditions. However the date correlates well with that of Pryor (1968) in a similar latitude. The pre-laying exodus of the Cape Petrel is incomplete at any one time, presumably due to staggering. These birds choose a different type of nest site again. There is always a vertical or overhanging rock face behind the nest, which is on a horizontal rock or gravel patch. Several nests may be found together within a few metres. T he steepest cliffs are avoided; terraced areas are preferred. A number of nests on Ardery and Odbert were on escarpments running across the island and not overlooking the sea. Cape Petrels and Snow Petrels were often found as neighbours, the Snow Petrels beneath the boulders and the Cape Petrels on them.

TABLE IV: Cape Petrel: Estimates of numbers of nests.

Summer of: ARDERY ODBERT O'CONNOR HOLL 1961-1962(1) 50 40 1977-1978 300 150 500 - I ,000 100

(1) Orton, 1963. VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 83

Cape Petrels at nest site. P late 31 A.N.A .R.E. P hoto by Alan N. Cowan

Cape Petrel in flight over breaking-up sea ice. Plate 32 A .N .A.R.E. P hoto by Alan N. Cowan AUSTRALIAN 84 COWAN BIRD WATCHER

The Cape Petrel's behaviour may fairly be described as nervous. It is by far the most easily disturbed of the "fulmarine petrels", particularly before laying; and even when incubating is relatively liable, if disturbed, to fly off the egg so that one has to move circumspectly. It also vomits most readily, powerfully and accurately, with an effective range of two metres. After hatching, however, it becomes much quieter and guards the black chick tenaciously. Laying began at the beginning of December and hatching on January 17; hatching was apparently complete by the end of January, but this is difficult to define precisely since as in all the "fulmarine petrels" there were a few birds still sitting on eggs at the end of hatching; these failed eggs are abandoned and taken by Skuas. However, one observation was of particular interest in relation to breeding synchrony. At O'Connor on January 2, a month after the beginning of laying, Barry Seedsman saw an egg drop during take-off from the cloaca of a Cape Petrel we had inadvertently flushed from her nest. Probably such late laying would have proved unsuccessful. As with the Southern Fulmar, many non-breeders attend the colonies, pair and sit. About 40 per cent. of pairs were found to have no egg, having either failed to lay or lost the egg, but continued to spend much time sitting together at the nest, readily flying off when approached. In contrast, of those pairs with an egg one of the pair was almost always absent from the nest. The incubating bird stays on the egg unless sharply disturbed, e.g. by one's foot being planted near a hidden nest, but is nothing like so immovable as are the other "fulmarine petrels". Snow Petrel Pagodroma nivea Status in the Area: A common breeding bird both on the mainland peninsulas and on most of the larger islands, except for the Swains and Donovans, almost anywhere in fact where its nesting requirements can be met; steep rocky areas usually at least 50 metres above sea level, with deep fissures, or low spaces under boulders. Numbers are very hard to estimate because of its nesting habits: even a painstaking search may miss many nests, and the areas involved are very extensive and largly inacces­ sible. However, I believe it is at least as common in the area as the Southern Fulmar (see Table V). Notes: One of the particular aims of the ornithological programme for 1977-1978 was to obtain measurements of Snow Petrels in the Casey area. It has long been known that there is great variation in individual size in the species. It has been postulated that there are two subspecies: P.n. major and P.n. nivea. The only breeding colonies of the "large subspecies" P.n. major, known hitherto were in French Antarctic Territory and at the Balleny Islands (Isenmann, 1970; Isenmann, et al, 1969; Kinsky, 1970, 1978; Lacan, et al, 1969; Mougin, 1968; Prevost, 1969). It was therefore of very great interest and satisfaction to the author to find numbers of extremely large individuals breeding in the Casey area and, perhaps even more significant, to find several pairs of "large with small" individuals which seems to challenge the concept of discrete subspecies or races held by previous authors. This study is being prepared for publication. It is difficult to pinpoint arrival dates for this species since it is so commonly seen during the winter. It is certainly common for birds to pay repeated visits to the breeding areas in September, and complex VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 85

TABLE V: Snow Petrel: Breeding populations. Origin of Figures ARDERY ODBERT Summer 1961-62 estimates(!) 20 nests Summer 1977-1978 estimates 200 - 300 150 nests nests Summer 1977-1978 number of individuals measured 105 39 birds birds (1) Orton 1963. comings and goings doubtless influenced by weather conditions may thus begin very early in the spring, and merge into the mass arrival followed in turn by the mass pre-laying exodus of the breeding season proper. Thus, Pryor (1968) describes the birds arriving in early October and deserting and re·-occupying "at least twice", during late October and late November. Our experience at Casey was a clear-cut arrival on November 1 of birds at breeding areas and a steady build-up of numbers until the pre­ laying exodus in late November. Several visits during October to known breeding areas had failed to show any; however, records from some previous years (ANARE Station Logs) indicated a pattern more like that described by Pryor. Most of the nesting clefts are blocked with snow and ice during the winter and the bird's bill and legs are ill-adapted for excavation until periods of thaw have largely cleared the sites. While tentative efforts may be made during repeated visits in the spring, it is doubtless the weather conditions which encourage or discourage prolonged attendances at the colonies during September and October. It is at the beginning of November when the birds arrive in earnest that they are best observed. This is the period of prolonged and noisy displays, particularly at about 5 to 6 a.m. The birds sit about in the sunshine on boulders, engaging in territorial disputes during which they readily eject stomach contents at each other, producing bright orange splashes on the white plumage which are easily visible during the flying displays. The calls are harsh and metallic. At this stage, birds occupying hidden nest sites were very vocal, betraying their presence by loud calls as one passed. After return and laying they became progressively less inclined to do so, and locating the nests required more and more deliberate searching, although with practice one became able to predict with fair accuracy in which clefts one might find nesting Snow Petrels. As with the other "fulmarine petrels", the pairs spent much time together until laying, after which only the incubating bird would normally be found at the nest. It was found that 70 per cent. of pairs produced an egg. Eggs in most sites are not liable to predation but many are a ban· doned; one would see several old eggs from previous years surrounding the sitting bird. Many factors may be involved in this, one of which is certainly weather: in some years a summer blizzard refills the clefts with snow after laying. No doubt many chicks die from the same cause. AUSTRALIAN 86 COWAN BIRD WATCHER

In order to measure the Snow Petrels, a technique for extracting them had to be evolved. Most birds sat within a metre of an exit. Most nest sites chosen have two or more escape routes. By lying down, one could usually introduce an arm and, once provoked into setting the sharp hooked tips of its bill into the (gloved) finger, the bird could be extracted by a grip on the bill with the free fingers. On occasion one could only grasp some other part of the bird, wh ich was then withdrawn by steady traction. Birds not on an egg were more elusive and often an assistant had to block a second exit to prevent escape. Much noise and regurgitation accompanied the extraction and measure­ ment, but once pointed in the right direction and released the incubating bird wou ld return at once to the egg. The first chicks were seen on January 16. The chick is rendered dis­ tinctive by a white forehead and mid-grey down. The hidden nests protect against predation, and the chicks are left alone by the foraging parents from the age of a week or so under their boulders, whence they call harshly and regurgitate boldly at the passing observer. Hatching was completed by the end of January. Wilson's Storm-petrel Oceanites oceanicus Status in the Area: This elusive little bird is a numerous breeding species in the area, and on the rocky Antarctic coasts generally. It is very widespread both on the islands and on the peninsulas, usually in fiat areas strewn with boulders under which tunnels and cracks provide it with shelter. Its nesting habits make it difficult to study, except where it burrows under moss, e.g. at the Argentine Islands, Graham Land, when the nest can be observed from above by lifting the roof (Roberts, 1940). Notes: Wilson's Storm-petrels were first seen on November 15, and sightings gradually became more frequent. An exasperating bird to watch, its habits at the breeding grounds are crepuscular and its flight swift and erratic. On November evenings when the visibility was bad one's eyes would attempt to follow the small dark white-rumped birds flitting low over the dark snow-flecked rocks and calling loudly; then typically, after repeated circling, they would disappear behind a boulder but this time fail to reappear on the other side, having gone to ground. Undoubtedly, these movements were effective in concealing the nest entrances! Later, on sunny nights and at favourable areas such as Robinson Ridge, study was easier and photography more successful, but remarkably few nests were located. One would occasionally see the birds in the daytime. One nest occupied by two adults with an egg was found on January 2 on O'Connor. We were able to extract the egg temporarily for examination. The late departure of these birds obliges them to defer their moult until after their arrival in winter quarters (Watson, 1975). It should be mentioned that the yellow webs of the feet, so widely advertised, are quite impossible to see in flight even in excellent light and at very close quarters. South Polar Skua Stercorarius maccormicki Status in the Area: A widespread breeding species. Total numbers are small since a large territory is defended by each pair. In general the species breeds in intimate relationship with the Adelie Penguin, and most VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 87

Snow Petrels at nest site, November. Pla te 33 A.N.A.R.E. P hoto by Ala n N . Cowan

South Polar Skua, Ardery Island. Plate 34 A .N .A .R.E. Photo by A la n N . Cowan AUSTRALIAN 88 COWAN BIRD W A TGiER nests are among the penguin colonies. Nests were also present on Ardery which has no penguins but supports great numbers of breeding petrels. Notes: The first Skua was seen on October 15, the same day as the first Adelie Penguins. One or two were reported in September but these records must be considered very doubtful, and were probably due to con­ fusion with Giant-Petrels. At first, behaviour is quiet, but the birds soon become more aggressive in defence of the nest, and from the time of laying an approach within a given radius will induce the familiar "dive­ bombing", although there is great variability in the aggressiveness of individual pairs. A pair with a particularly ferocious male (see below) nested at the W. end of Ardery above the boat landing place, commanding all approaches to and from the petrel colonies. Incubating Skuas when approached would emit a repeated hoarse, squeaking cry, clearly an alarm summons to the mate, and the bird thus summoned was always the one which actively attacked, screeching harshly; meanwhile the bird which had been sitting would circle overhead, continuing to make the hoarse alarm call. It was difficult not to believe that it is the female which does most of the incubating and this was the conclusion arrived at by Eklund (1961) using electronic telemetry tech­ niques. One could sustain a painful blow on the head from the wing of these birds. Laying begins in late November. The eggs, usually two, are perfectly camouflaged and invisible from a few metres away. They are laid on a level gravel patch with surrounding rocks. The first chicks were found on December 26. The young are highly nidifugous. In no case were both chicks seen to survive for more than a few days, although in a minority of clutches both may survive to the flying stage (Eklund, 1961). Numbers of non-breeding birds frequented the area in summer. Sku as spent much time about the station, particularly beneath the kitchen windows and at the rubbish dump. The main Weddell Seal breeding colony on the sea ice by Herring Island was also attended by a group of Skuas, which consumed the placentae within minutes of delivery. At the penguin and petrel colonies, Skuas were constantly patrolling and pre­ dating eggs and chicks, and this is probably their main food supply. The South Polar Skuas at Casey are of the light phase, typical of the Antarctic continent, and they become lighter as the season proceeds; juveniles are much darker (Watson, 1975). Conclusions During my preliminary reading, I was struck by the paucity of infor­ mation in the literature on many basic aspects of Antarctic bird life, but having myself experienced in a small way some of the difficulties and hazards involved I now understand why there are still so many unfilled gaps in our knowledge, not only of Antarctic ornithology in general but also of local areas. Even in the better-known places around Casey we were able to add to previous records. For example, Dewart Island, though often visited by helicopter at changeover in recent years, and previously, so far as I can discover, known as the breeding place of only one species, the Southern Giant•-Petrel, was found to be occupied by five further breeding species. The Casey area is one of rich ornithological interest and offers almost unlimited scope for further work, given only the availability of suitable boats. VOL. 8 (3) SEPTEMBER, 1979 Ornithological Studies at Casey 89

Acknowledgements My thanks are due to Dr. R. I. Garrod, Director, Antarctic Divis ion, Department of Science, Melbourne, who authorised the programme. Dr. G. W. Johnstone, Biologist, and Dr. D. J. Lugg, Senior Medical Officer, both of the Antarctic Division, gave me unstinting help and support. Much helpful information was also given me by Mr. D. Purchase, Dr. M. D. Murray and Mr. D. Luders before I left Australia. At Casey the unfailing encouragement and helpful interest of Barry Seedsman, Officer in Charge, were beyond words. Egon Wehrle, to whose skilled craftsmanship and resourceful intelligence I could di scern no limits, kept our outboard motors repaired and running sweetly and reliably. Finally, no biology programme would have been poss ibl e at all but for the Casey Wintering Party 1977-78 expeditioners, for whose help, enthusiasm and companionship on some very rough trips I shall always be grateful. The maps in Figures 1 and 2 were prepared by Mr. P. Richardson.

References A.N.A.R.E. Station Logs. Antarctic Division, Dept. of Science, Mel­ bourne. (Unpublished records.) Hydrographer of the Navy, 1974. The Antarctic Pilot, 4th Ed., Hydro­ graphy Office, U.K. Beck, J. R., 1970. Breeding seasons and moult in some smaller Antarctic petrels. Antarctic Ecology, Vol. II. Academic Press. Eklund, C. R., 1961. Distribution and life history studies of the Southern Polar Skua. Bird Banding 32: 187-223. Emison, W. B., 1968. Food preferences of the Adelie Penguin at Cape Crozier, Ross Island. Antarctic Bird Studies. American Geophysical Union, Washington. Isenmann, P., 1970. Contribution ala biologie de reproduction du Petrel des neiges. L 'Oiseau et R .F.O. 40, no. special: 7-134. Isenmann, P., Lacan, F., et Prevost J., 1969. Notes sur une breve visite a Cape Hunter et Cape Denison. L'Oiseau et R .F.O. 39, no. special: 2-10. Johnstone, G. W., Lugg, D. J., and Brown, D. A., 1973. The Biology of the Vestfold Hills, Antarctica. ANARE Scientific Reports. Antarctic Division, Dept. of Science, Melbourne. Kinsky, F. C., 1970. Quoted in addendum to Isenmann, P., 1970. Kinsky, F . C., 1978. Personal communication. Lacan, F., Prevost J., et Van Beveren M., 1969. Etude des populations d'oiseaux de Pointe Geologie de 1965 a 1968. L 'Oiseau et R .F.O. 39, no. special: 11-32. Luders, D. J., 1977. Behaviour of Antarctic Petrels and Antarctic Fulmars before laying. Emu 77: 208-214. Mougin J-L., 1968. Etude ecologique de quatre especes de Petrels antarc­ tiques. L'Oiseau et R.F.O. 38, no. special: 1-52. Orton, M. N., 1963. A brief survey of the fauna of the Windmill Islands, Wilkes Land, Antarctica. Emu 63 : 14-22. Orton, M. N., 1968. Notes on Antarctic Petrels. Emu 67: 225-229. Paulin, C. D., 1975. Feeding of the Adelie Penguin. Mauri Ora 3: 27-30. Christchurch. AUSTRALIAN 90 COWAN BIRD WATCHER

Penney, R . L., 1968. Territorial and social behaviour in the Adelie Pen­ guin. Antarctic Bird Studies. American Geophysical Union, Wash­ ington. Prevost, J., 1969. A propos des Petrels des neiges de la Terre Adelie. L 'Oiseau et R.F.O. 39, no. special: 33-49. Pryor, M. E., 1968. The avifauna of Haswell Island, Antarctica. Antarctic Bird Studies. American Geophysical Union, Washington. Roberts, B. , 1940. The life cycle of Wilson's Petrel. British Graham L and Expedition 1934-1937 Scientific R eports, Vol. I, No: 2: 141-194. British Museum. Sapin-Jaloustre, J., 1960. £cologie du Manchot Adelie. Expeditions Polaires Francases, Paris. Stonehouse, B., 1968. Penguins. London. Watson, G. E., 1975. Birds of the Antarctic and Sub-Antarctic. American Geophysical Union, Washington. Watson, G. E ., Angle, J. P., Harper, P. C., Bridge, M. A., Schlatter, R. P., Tickell, W. L. N., Boyd, J. C., and Boyd, M . M., 1971 Birds of the Antarctic and Subantarctic. Folio 14, American Geographical Society, New York. ABW