Ornithological research at the sub-Antarctic Prince Edward Islands: a review of achievements

J Cooper Percy PitzPatrick Institute of African Ornithology, University of Cape Town, Rondebosch 7700, South Africa CR Brown Department of Zoology and Entomology, Rhodes University, PO Box 94, Grahamstown 6140, South Africa Ornithological research conducted at the Prince Edward aard met 'n yl, tundra-agtige plantegroei. Daar word Islands in the southern Indian Ocean since 1951 is geskat dat nege-en-twintig voelsoorte op die Prince reviewed. The two islands ofMarion and Prince Edward Edward-eilande broei en nog 44 soorte is opgemerk, maar are sub-Antarctic in nature, with a sparse tundra-like vege­ broei nie daar nie. Vir die grater seevoels wat op Marion­ tation. Twenty-nine species of birds are known or thought eiland oppervlakneste het, is daar onlangse opnamegege­ to breed at the Prince Edward Islands and a further 44 wens, maar nie vir die grawende stormvoels of vir die species have been recorded as non-breeders. Recent cen­ meeste voelsoorte wat op die onbewoonde Prince Edward­ sus data are available for the larger, surface-nesting sea­ eiland broei nie. Pikkewyne vorm die grootste persen­ birds ofMarion Island, but not for the burrowing petrels tasie van die voelfauna met jaarlikse oppervlakneste op or for most species breeding at uninhabited Prince Ed­ die Prince Edward-eilande. Tussen 1951 en 1987 is 12 424 ward Island. Penguins form the greatest percentage of individuele voels van 26 soorte gering, waarvan slegs 48 the annual surface-nesting avifauna of the Prince Edward daarna elders aangetref is. Aspekte van die broeigedrag Islands. Between 1951 and 1987, 12 424 individuals of van 21 voelsoorte is al bestudeer. Broeigeslaagdheid wis­ 26 species have been ringed,. of which only 48 individu­ sel grootliks tussen soorte en van jaar tot jaar. Relatief als have been reported away.from the islands. Aspects weinig navorsing is gedoen oor broeigedrag. Voels op of breeding biology have been studied in detail for 21 die Prince Edward-eilande belnvloed die terrestriele om­ species. Breeding success varies greatly between species gewing deur hul bydraes van energie en voedingstowwe, and shows year-to-year variations. Relatively little deur erosie en deur die verspreiding van saad. Die eet­ research has been undertaken on breeding behaviour. gewoontes van die meeste voelsoorte is al ondersoek, 20 Birds at the Prince Edward Islands affect the terrestrial daarvan omvangryk. Daar is hoofsaaklik twee dieet­ environment by contributing energy and nutrients, by ero­ groepe; soorte wat hulle voedse/ uit die see alJeen ver­ sion and by seed dispersal. Most species have had their kry en daardie wat, al is dit gedeeltelik, ook op land kos diets studied, 20 comprehensively. There are two broad soek. Eersgenoemde groep vreet in hoofsaak skaaldiere dietary groups: species obtaining their food exclusively (21 geidentifiseer), sefalapode (53) en vis (20). Onder­ from the sea, and those that forage, at least partially, ter­ soeke is uitgevoer na die verteringsfisiologie van seevoels restrially. The former group feeds primarily on crusta­ by die Prince Edward-eilande, insluitende die opnemings­ ceans (21 taxa identified) cephalopods (53) and fish (20). rendement en vertertingstempos. Die ekologie van vreet­ Studies have been undertaken ofdigestive physiology of gewoontes van pikkewyne op die Prince Edward-eilande seabirds at the Prince Edward Islands, including assimi­ is bestudeer met die hulp van spoed- en duikmeters. lation efficiency and digestion rates. The foraging ecol­ Fisiologiese ondersoeke oor seevoels op Marioneiland ogy of Prince Edward Island penguins has been studied het gekonsentreer op die meet van metaboliese tempos by the use ofspeed and dive meters. Physiological studies van nesmakende, ververende, broeiiende en kossoekende on seabirds at Marion Island have concentrated on meas­ individue van vier pikkewynsoorte, die albatros Di­ ,uring metabolic rates ofnesting, moulting, incubating and omedea exulans en verskillende stormvoelsoorte. Hier­ foraging individuals of four species ofpenguins , the wan­ die Jesings is aangewend om die daaglikse en totale be­ dering albatross Diomedea exulans and several species hoeftes vir oorlewing en groei van 'n hele aantal soorte of burrowing petrels. Such measurements have been used te bepaal. Hierdie energiebegrotings is gebruik om die to assess daily and total demands for maintenance and totale energie- en voedselbehoeftes van verskeie voel­ growth ofseveral species. These energy budgets have been gemeenskappe op die Prince Edward-eilande te bepaal. used to estimate the total energy and food requirements Beheer en uiteindelike uitroeiing van die huiskat Pelis of several avian communities of the Prince Edward Is­ catus bevolking word beskou as die belangrikste be­ lands. Control and eventual eradication of feral cats Pe­ waringsondememing om die voelbevolking van die Prince lis catus is considered to be the most important conser­ Edward-eilande te beskerm. vation effort required to protect the avifauna of the Prince Edward Islands. Introduction . 'n Oorsig oor omitologiese navorsing by die Prince Edward- eilande in die Suid-Indiese Oseaan word gegee. The Prince Edward Islands (the larger Marion Island and Marion- en Prince Edward-eiland is subantarkties van the smaller Prince Edward Island) are situated at approx-

40 S. Afr. T. Nav. Antarkt. , Vol 20 No 2, 1990 imately 47° S, 38° E in the southern Indian Ocean. They 37 40E 38E are sub-Antarctic in nature, a habitat not present on main­ PRINCE EDWARD ISLAND land South Africa.

Information on birds of the Prince Edward Islands be­ 0 5 10 km ~e~'' ':)' tory. Prior to annexation in 1947, short visits by scien­ 46 40S ~· So(JtJ, 1 tists and sealers resulted in some published information Ca/Je (eg Hutton 1865, Moseley 1879). After annexation, mem­ bers of early meteorological teams on Marion Island pub­ lished observations on birds (Bennets 1948, 1949, Craw­ ford 1952, la Grange 1962). The first professional biol­ ogist at the island was R W Rand, who made observa­ tions in 1951 - 1952 on numbers, distribution patterns and mensural and breeding parameters on nearly all bird species occurring at Marion Island (eg Rand 1954, 1955, 1956). A hiatus in ornithological research at Marion Island ·followed until the first "South African Biological and Ge­ ological Expedition" to the islands in 1965 - 1966 (van Zinderen Bakker Sr et al 1971) . An ornithologist, E M van Zinderen Bakker Jr, undertook studies on gentoo pen­ guins (see Table 1 for scientific names) and wandering Cape Hooker

albatrosses at this time. In addition, he also undertook 47S the first known ornithological visit, of five days, to unin­ habited Prince Edward Island in March - April 1966. Fig 1: The Prince Edward Islands Subsequently, roughly annual visits to Prince Edward Is­ land for surveys, counts and sample collecting of up to The islands are volcanic in origin, the most recent erup­ a week it a time have been made (Appendix 2 in Coop­ tion having occurred in 1980 (Verwoerd et al 1981). The er & Avery 1986). The 1965 - 1966 expedition marks islands are rugged with a sparse, tundra-like vegetation the initiation of sustained ornithological research at the and are, for the most part, surrounded by steep, coastal Prince Edward Islands. cliffs between 5 and 500 m high with few suitable land­ The only previous review of ornithological research ing beaches. This has implications for the distribution at the Prince Edward Islands is that of Siegfried (1978). of penguins which rely on such beaches for access to Much ornithological research has been undertaken since their breeding colonies (Williams 1980a). Coastal cliffs then, but a recent review of the environment and biota are utilized by scattered breeding colonies of imperial of Marion Island (Smith 1987a) gave relatively little at­ cormorants and kelp gulls. The steep, vegetated coastal tention to birds. A bibliography of scientific research, and adjacent slopes, dominated by Poa cookii tussock including ornithological literature, at the Prince Edward grassland, provide breeding habitat for several species Islands has been published (Siegfried et al 1979). This of albatrosses. These slopes are also occupied seasonal­ bibliography is kept up to date as part of a larger bib­ ly by burrowing petrels (Schramm 1986a). liography of Antarctic and sub-Antarctic biological Inland of the coastal cliffs is a relatively flat, discon­ research (Smith 1987b). This review covers ornithologi­ tinuous, poorly drained coastal plain dominated by Agros­ cal research conducted at the Prince Edward Islands, es­ tis magellanica mires and Acaena adscendens herbfield. pecially that undertaken since Siegfried (1978) . This plain is the favoured breeding habitat of wandering albatrosses and giant petrels. Between the coastal plain A brief description of the islands and the central highlands Blechnum penna-marina fern slopes and vegetated black-lava hummocks are used by burrowing petrels. Azorella selago fjaeldmark, overly­ Marion (290 km2) and Prince Edward Island (44 km2), ing older grey lavas, is used by Antarctic and Kerguelen 22 km apart (Fig 1), lie between 200 and 250 km north terns. of the Antarctic Polar Front (Lutjeharms & Valentine The central, mountainous region of Marion Island, 1984). Their climate is typically oceanic with little diel which rises to about 1 200 m, has a permanent ice-plateau and seasonal range. Annual temperatures average and is snow covered for about eight months of the year. 5,0 °C, with an absolute maximum and minimum of Prince Edward Island rises only to 670 m and has no 22,3 °C and -6,8 °C, respectively. Sea surface tempera­ ice-plateau. tures range from 2,1 °C to 8,0 °C, averaging 5,0 °C. Precipitation, mainly in the form of rain, is high, aver­ Species and populations aging > 2 500 mm yr- 1. Winds are predominantly westerly and frequently of gale force. Cloud cover is high and the sun shines on average under 4 h day-1. More Twenty-nine species of birds, 28 of them seabirds, are detailed accounts of the climate are provided by Schulze known or are thought to breed at the Prince Edward (1971), Gremmen (1981) and Smith (1987a) . Islands (Williams et al. 1979, Cooper & Brooke 1984,

S. Afr. J. Antarct. Res. , Vol 20 No 2, 1990 41 Table 1). However, at Marion Island, where all the in­ 1987). Because none of these non-breeding species oc­ tensive ornithological research has taken place to date, curs ashore in any numbers, they do not play significant only 27 species have been proven breeding. One of these, roles at the Prince Edward Islands as they do in con­ the common diving petrel, is thought no longer to breed tinental situations (Burger et al 1980). due to the depredations of feral cats. Storm petrels have Census data are reasonably good for the large surface­ only been proven to breed at Prince Edward Island, which nesting seabirds: penguins, albatrosses, giant petrels, cor­ also supports the whole population of yellownosed morants, skuas and gulls, some of which occur in the albatrosses. tens or even hundreds of thousands (Williams et al 1979, Twenty species of nonbreeding seabirds have been Watkins 1987, FitzPatrick Institute unpubl data, Table 1), sighted on or from the islands, along with 24 identified but are less so for the smaller burrowing petrels which species of vagrant terrestrial/fresh water birds (Gartshore are extremely difficult to count accurately (Schramm

Table 1 Most recent estimates of annual population sizes of breeding seabirds at the Prince Edward Islands. Species marked 'a' breed at intervals of more than one year. 'b' denotes breeding suspected but not proven

Species Marion Prince Source Island Edward Island

King penguina Aptenodytes pat.agonicus 215 230 5 000 1 Gentoo penguin Pygoscelis papua 888 655 2 Macaroni penguin Eudyptes chrysolophus 405 084 17 000 3,1 Rockhopper penguin Eudyptes chrysocome 137 652 35 000 4,1 Wandering albatrossa Diomedea exulans 1 533 1 277 4,3 Greyheaded albatross3 D. chrysostoma 5 037 1 500 4,5 Yellownosed albatross D. chlororhynchos 0 7 000 5 Sooty albatross Phoebetria fusca 2 055 700 4,1 Lightmantled sooty albatross P palpebrat.a 201 40 4,1 Northern giant petrel Macronectes halli 314 180 4 Southern giant petrel M. giganteus 2 891 410 4,1 Fairy prion Pachyptila turtur 100s 100s 4 Salvin's prion P vitt.at.a salvini 100s of 1 OOOs 10s of 1 OOOs 4 Blue petrel Halobaena caerulea 10s of 1 OOOs 100s of 1 000s 4 Greatwinged petrel Pterodroma macroptera 10s of 1 OOOs 1 OOOs 4 Kerguelen petrel P brevirostris 10s of 1 OOOs b 4,6 Softplumaged petrel P mollis 1 OOOs 1 OOOs 4 Grey petrel Procellaria cinerea 1 OOOs 1 OOOs 4 Whitechinned petrel P aequinoctialis 10s of 1 OOOs 1 OOOs 4 Blackbellied stormpetrel Fregett.a tropica b 1 OOOs 4 Greybacked stormpetrel Garrodia nereis b b 4,6 South Georgian diving petrel Pelecanoides georgicus 100s b 4,6 Common diving petrel P urinatrix 0 b 4,6 Imperial cormorant Phalacrocorax atriceps 589 120 4,7 Sub-Antarctic skua Catharact.a ant.arctica 900 60 8,1 Kelp gull Larus dominicanus 200 30 1 Antarctic tern Sterna vitt.ata <25 <25 4 Kerguelen tern S. virgat.a 10 20 ·9 Lesser sheathbill Chionis minor 980 420 1

1 Williams et al 1979; 2 Adams & Wilson 1987; 3 Watkins 1987; 4 1983 - 1990 censuses and estimates, FitzPatrick Institute unpubl data; 5 Berruti et al 1981; 6 Cooper & Brooke 1984; 7 Ryan & Hunter 1985; 8 Hunter 1990; 9 Ryan 1987a

42 S. Afr. T. Nav. Antarkt., Vol 20 No 2, 1990 1986a, Table 1). Of the nearly 800 000 pairs of surface­ nesting seabirds breeding annually, 91 % occur on the larg­ JUN JUL AUG SEP OCT NOV DEC JAN FE8 MASI APR MAY I er Marion Island. Macaroni and king penguins form 81 % by number of the annual surface-nesting avifauna of the Prince Edward Islands. The large numbers of burrow­ Macaroru Penguin ing petrels, especially on cat-free Prince Edward Island Aockhopper Penguin (see later), suggest there are as many, if not more, bur­ Wandering AlbatrossL--- -~,=«~,- ---C!::~Eim=••:::;:::=;::;l-~=-,~ rowing birds than there are surface-nesting birds.

Not all Southern Ocean seabirds breed annually. SOO!y AlbalfOSS

Knowledge of the species' breeding biologies is therefore L.M Sooty Albat,oss p i r:::c=ecm----:::;i..___ -'c="'=-'' - - ~ needed before realistic population estimates can be made. N Giant Petrel Eogs .._!__ 6"'~'-- -' Towards this end, which requires studies of individually

Salvln's Pnon marked birds, 12 424 individuals of 26 species were Fcimw: I ' Eogs? Oi::*5 :i I

CPresent~ ringed up to July 1987 (Gartshore et al 1988a). Nearly Blle Petrel ~ I ' Eggs j l l Chocks J j all of these were ringed at Marion Island. To date, only Greatwinged Petrelr "~"":::::::==~====:::J 58 individuals have been recaptured alive or recovered dead away from the islands (Gartshore et al 1988, un­ publ data). eggs d Grey Pet1el l'----,Chd<=c-,-- --, Ringing is now concentrated on a few surface-nesting Wh!techlmed Petrel CAeoco..ipat,on I , Egos .._)•-'-=en"'='--' species which are normally colour-banded as well (Coop­ Subaniarctic Skua I Azmrrm 1 er & Oatley 1985), so that individuals can be identified F Oiicks without subsequent handling. Resighting such birds, at marked nest sites from year to year, allows information to become available on the species' demographies (most species of Southern Ocean seabirds are long-lived; the two oldest known wandering albatrosses at the Prince Edward Islands were ringed off Australia and recaptured Fig 2: Breeding schedules of21 species ofbreeding sea­ over 24 years later (Gartshore et al 1988a). birds at Marion Island. Boxes span earliest and latest Species studied in this way currently include the wan­ known dates for the presence of eggs and chicks and dering albatross, both species of giant petrels and the moulting individuals. Arrows signify mean laying (A and king penguin. Resightings of flipper-banded individuals B eggs marked separately for macaroni and rockhopper of the last species should help unravel its unusual breeding penguins), hatching, fledging and moult commencement interval (Stonehouse 1960) and and allow its population and completion dates. Solid lines denote presence ashore size to be estimated. In addition, regular counts at select­ outside of the breeding season. Dotted lines and ques­ ed colonies of macaroni and rockhopper penguins and tion marks denote no or incomplete data . Sources ofin­ grey headed albatrosses allow assessment of year-to-year formation are cited in the text or are FitzPatrick Insti­ variations in numbers of breeding birds and breeding tute unpubl data success.

Breeding studies in September, the birds return, as pairs, to their burrows in April - May after having completed a post-breeding moult at sea (Fugler et al 1987). The reason(s) for this Aspects of the breeding biology have been studied for post-breeding visit may be related to pair bond and/or 21 of the 26 currently breeding seabirds of Marion Is­ burrow maintenance. It is not yet known whether the land. Most species are summer breeders, exceptions being other species of burrowing petrels breeding at the Prince two burrowing petrels, the greatwinged and the grey, and Edward Islands practice post-breeding visits (Schramm the gentoo penguin (Fig 2). The king penguin and wan­ 1983a, Berruti & Hunter 1986), although of these spe­ dering albatross, although laying their eggs in summer, cies, only the Kerguelen petrel is known to do so at the have laying to fledging periods of over 12 months, so that iles Crozet (Jouventin et al 1985). chicks are present through the winter months (Fig 2). Breeding success differs greatly between species and Gentoo penguins and imperial cormorants can relay af­ also shows year-to-year variations. For example, annual ter failure and so have extended breeding seasons in com­ breeding success of the wandering albatross varied be­ parison with most other species breeding at the Prince Ed­ tween 63 and 96% over a ten-year period (FitzPatrick ward Islands (Fig 2). Most species are absent at sea from Institute unpubl data). The breeding success of three gad­ the islands outside their breeding seasons. Some species fly petrels in 1979 - 1980 was 53 % for Kerguelen petrels, (eg gentoo penguin, kelp gull, imperial cormorant and 7% for softplumaged petrels and 0% for great­ the only non-seabird breeding on the islands, the lesser winged petrels (Schramm 1983a). The very low breed­ sheathbill, Fig 2) are resident all year round. However, ing successes of softplumaged and greatwinged petrels some species do visit their nest sites outside their breeding are attributed to predation by feral cats (Schramm 1983a, seasons. An example is the blue petrel whose chicks Newton & Fugler 1989). fledge in January and February. Although egg-laying does Most of the autecological breeding studies of seabirds not take place until October after a pre-breeding return at Marion Island have been undertaken as part of or at

S. Afr. J. Antarct. Res., Vol 20 No 2, 1990 43 the same time as more integrated studies on nutrient cy­ Table 2 cling, food and feeding ecology and energy and food re­ Quantities (tons dry mas.5) of avian products deposited quirements, aspects of which are discussed in detail be­ annually by 13 species of surface-nesting seabirds in low. It is therefore not intended to review details of in­ the coastal lowland of Marion Island (after Siegfried dividual species' breeding biologies (eg Berruti 1979, 1981) Burger 1979, Williams & Burger 1979, Williams 1980b,c, Schramm 1983, Berruti & Hunter 1986, Fugler et al 1987, Newton & Fugler 1989). Guano Feathers Eggs Carcasses Totals Most breeding biology studies at Marion Island have been ecological in nature. Exceptions are descriptions Penguins 3 552 428 43 126 4 149 of breeding displays of gentoo penguins, albatrosses and sheathbills (van Zinderen Bakker Jr 197la,b, Burger 1980, Other species 41 0,5 1 43,5 Berruti 1981) and the vocal systems of whitechinned and grey petrels (Brooke 1986). In the last study, it was found 429 43,5 by use of playback experiments that birds were able to Totals 3 593 127 4 192,5 recognize calls of their mates in the absence of visual cues. A J Williams studied aspects of the eggs and incuba­ in the form of guano, feathers, eggs and carcasses. Pen­ tion of Marion Island seabirds during the mid 1970s, plac­ guins contribute fully 99 % by mass of the over 4 000 ing them in taxonomic context with a wider suite of spe­ tonnes of avian products deposited annually on the is­ cies, finding for example, that penguins lay eggs of ex­ land by surface-nesting birds (Table 2, Siegfried 1982). pected size for heavy birds producing semi-altricial young Of this total, 86% is guano. However, since most pen­ (Williams 1980d) and that the long laying intervals of guins breed close to the sea on bare rock (Williams penguins may be related to their relatively thick egg shells 1980a), much of this guano is lost from the terrestrial and the time taken for mobilization and deposition of environment as run-off to the sea. Birds are considered mineral elements, such as calcium, in the shell (Williams to be the major source of nitrogen and phosphorus to 1981a). Egg composition (relative proportions of yolk, Marion Island (Lindeboom 1979, 1984, Smith 1976, 1978, albumen and shell) for 16 species of Marion Island sea­ 1979). birds and nine species of seabirds breeding in South Afri­ The roles of burrow-nesting birds in contributing ener­ I ca showed that variation in the proportionate mass of gy and minerals to the island's terrestrial ecosystem are ... eggshell is not related to hatchling precocity, but is prob­ not so well known as those of the surface-nesters, large­ ably due to adaptive differences in nest substrata and agil­ ly owing to inadequate information on the numbers of ity of incubating birds (Williams et al 1982). It was found birds present. Guano analyses and obvious vegetational that there was no simple relationship between egg com­ changes in areas where burrowing petrels, especially blue position and hatchling precocity. petrels, breed in high densities suggest that the manur­ Rockhopper and macaroni penguins lay two eggs of ing roles of burrowing birds are as important, if not more dissimilar size, the larger egg being the second laid but so, than are those of the surface-nesting species (Lin- the first to hatch (Williams 1980e, 198lb,c). Burger & deboom 1979, 1984, Smith 1976, Fugler 1985, Fugler et Williams (1979) suggested this was due to a delay in the al 1987). The contributions of burrowing birds, especially onset of incubation until the second egg was laid and, guano produced by their chicks, enter directly into the because the smaller egg is incubated in the less favoura­ soil and are not washed away by rainfall as in the case ble anterior position in the brood pouch, to the slightly at most penguin colonies. lower temperature which it generally experienced. Sub­ The manuring effects of guano and other avian products sequently, physiological work on embryonic metabolism entering the surrounding seas and resulting in phytoplank­ has shown that the slight temperature differences ex­ ton blooms at the Prince Edward Islands have been re­ perienced by the eggs and the delay in incubating the cently studied (Perissonotto & Duncombe Rae 1990, smaller egg are insufficient to explain why the larger last Perissonotto et al 1990), with effects being found broad­ laid egg always hatches first. The smaller eggs have an ly similar to those adjacent to Antarctic penguin colo­ inherently lower embryonic metabolism than do the larger nies (Delille 1987). Thus the energy and mineral contri­ eggs and consequently take longer to reach the metabol­ butions of birds at the Prince Edward Islands may be of ic level required for hatching (Brown 1988a). equal (or more) importance to the marine as they are · to the terrestrial environment. Avian effects on the terrestrial environment Burrowing petrels and, to a lesser extent, cliff-nesting Birds at the Prince Edward Islands affect the terrestrial albatrosses, remove soil and vegetation and affect slope environment in three ways : by contributing energy and stability, leading to "peat slips" which can cause exposure nutrients in the form of avian products; by erosion; and of bedrock. It has been estimated that 2 900 m3 of peat by dispersal of propagules. have been removed by the erosive effect of king and Only the first aspect has been studied in detail at Mari­ macaroni penguins at the two largest breeding colonies on Island (Burger et al 1978, Siegfried et al 1978, Wil­ (Hall & Williams 1981). A less important erosive effect liams 1978, Williams & Berruti 1978, Williams et al 1978) is grooving and polishing of rocks near landing beaches who report on the mineral and energy contributions of caused by penguin claws (illustrated in Hall & Williams the surface-nesting birds to the terrestrial environment 1981).

44 S. Afr. T Nav. Antarkt. , Vol 20 No 2, 1990 At least six species of birds have been reported as act­ sub-Antarctic skua, kelp gull, terns and lesser sheathbill). ing as potential seed dispersers at Marion Island (Kok Members of the former group feed on crustaceans, 1975). However, the importance of birds as dispersal cephalopods (primarily squid) and fish, the relative agents for propagules is unknown. proportions of which differ between seabird species and may further differ seasonally and annually (see below). Diets of birds at the Prince Edward Islands The general composition of the diet for those species of exclusively marine-feeding seabirds for which informa­ To date, information on the diets of most of the 29 spe­ tion is currently available is summarized in Table 3. cies breeding or thought to be breeding at the Prince Ed­ Of the four species of penguins which breed at the ward Islands is available, but only 20 species have been Prince Edward Islands, the king penguin feeds studied comprehensively. predominantly on fish (Adams & Klages 1987}. In con­ Early information on the diets of seabirds at the Prince trast, the diets of macaroni and rockhopper penguins are Edward Islands was based on analysis of relatively in­ broadly similar with both species consuming more crusta­ digestible hard parts of prey such as the chitinous man­ ceans than cephalopods and fish (Brown & Klages 1987, dibles (beaks) of squid (Furness et al 1984), which are Table 3). The inshore-feeding gentoo penguins feed about often regurgitated. Analysis of regurgitated casts has its equally on crustaceans and fish (La Cock et al 1984, limitations, over-emphasizing certain components of the Adams & Klages 1989). Diets of burrowing petrels have diet. Preferable are whole samples collected from spe­ been relatively poorly studied and quantitative informa­ cies which readily regurgitate their stomach contents upon tion is limited to five species (Table 3). capture (eg albatrosses and petrels) or by a non-lethal, Within the broad classes of marine prey consumed by stomach-flushing technique, used especially on penguins seabirds at the Prince Edward Islands, 21 crustacean, 53 (Wilson 1984, Ryan & Jackson 1986). Whole regurgita­ cephalopod and 20 fish taxa have been identified, along tions from greyheaded albatrosses have demonstrated that, with eight taxa of other marine organisms and two _of despite the wide range of squid prey consumed by these insects (Table 4). These totals do not include seabirds birds at the Prince Edward Islands, squid comprises only and marine mammal remains which are taken by sever­ 34 % of the diet by mass, most of the rest being fish (Ta­ al species, notably Phoebetria albatrosses and giant ble 3, Hunter & Klages 1989). petrels. Detailed breakdowns of crustacean, cephalopod Birds at the Prince Edward Islands fall into two broad and fish prey are presented in Tables 5, 6 and 7. groups; those that obtain their food exclusively from the Several species of crustaceans, notably the euphausiid sea (penguins, albatrosses, petrels and imperial cor­ Euphausia vallentini and the amphipod · Themisto morants) and those that obtain their food, or at least a gaudichaudii, are important in the diets of gentoo , proportion of it, from terrestrial sources (giant petrels, macaroni and rockhopper penguins, Salvin's prions and

Table 3 Relative proportions of crustaceans, cephalopods and fish in the diets of 14 species of exclusively marine­ feeding seabirds at Marion Island

Species Period of No Age class% Relative proportions by mass Source sampling Samples sampleda crustaceans cephalopods fish otherb

King penguin Mar 1984 - Mar 1985 120 A+C <0,1 13,5 86,5 0,0 Adams & Klages (1987) Gentoo penguin Sept 1982 64 A 30,0 < 0,1 70,0 0,0 La Cock et al (1984) Mar 1984 - Mar 1985 144 A 44,4 2,1 53,5 0,0 Adams & Klages (1989) Macaroni penguin Dec 1983 - Feb 1984 & Dec 1984 - Feb 1985 75 A 69,8 11,6 18,7 0,0 Brown & Klages (1987) Rockhopper penguin Jan - Mar 1984 & Jan - Mar 1985 85 A 84,4 5,0 10,6 0,0 Brown & Klages (1987) Wandering albatross Sept - Dec 1988 & 50 C <0,1 56,7 35,4 7,9 FitzPatrick Institute Jun - Aug 1989 (unpubl data) Greyheaded albatross Feb 1985 - Apr 1987c 88 A 3,0 34,2 58,0 4,9 Hunter & Klages (1989) Sooty albatross Feb - Apr 1990 37 C 0,5 44,4 31,6 23,6 FitzPatrick Institute (unpubl data) Lightmantled sooty Mar - May 1990 32 C 3,9 33,9 45,7 16,5 FitzPatrick Institute albatross (unpubl data) Salvin's prion Jan - Feb 1987 39 A 44,4 13,9 41,9 0,0 Gartshore et al (1988) Blue petrel Nov 1984 - Jan 1985 49 A+C 59,5 15,7 21,2 3,6 Steele & Klages (1986) Greatwinged petrel Aug - Sep 1979 21 C 6,0 90,0 4,0 0,0 Schramm (1986b) Kerguelen petrel Dec 1979 22 C 24,0 70,0 6,0 0,0 Schramm (1986b) Softplumaged petrel Feb 1980 9 C 10,0 89,0 1,0 0,0 Schramm (1986b) Imperial cormorant Apr 1984 - Mar 1985 47 A 18,8 7,1 71,4 2,7 Espitalier-Noel et al (1988) a - A = adult, C = chick b - includes remains of seabird and marine mammals c - includes 42 samples from Prince Edward Island

S. Afr. J. Antarct. Res. , Vol 20 No 2, 1990 45 blue petrels. The decapod crustacean Nauticaris mario­ Table 4 nis is also of importance to these three species of pen­ guins and to imperial cormorants (Table 5). Amphipods, Marine prey (excluding remains of seabirds and ma­ mysids and decapods comprise the main crustacean prey rine mammals) of 19 speciesa of seabirds at the Prince of greatwinged, Kerguelen and softplumaged petrels. An­ Edward Islands: number of taxa in major taxonomic tarctic krill E. superba has been recorded only in the diet groupings of the two sooty albatrosses. Cephalopod prey comprise 53 % of all marine taxa Groupings No of taxa % taken by marine-feeding seabirds at the Prince Edward Islands (Table 4). Onychoteuthid squid, especially Kon­ dakovia Jongimana, are of particular importance in the Cephalopods 53 51,0 diets of penguins and albatrosses (Table 6), which take Crustaceans 21 20,2 specimens between 0,5 g or less in the case of rockhop­ Fish 20 19,2 per and macaroni penguins and up to 480 g in the king Gastropods and bivalves 3 2,9 penguin. Albatrosses take by far the widest range of squid Echinoderms (sea stars prey (especially onychoteuthid, cranchiid and histioteuthid and brittle stars) 2 1,9 squid) presumably reflecting their wide foraging ranges Insectsc 2 1,9 compared to those of other seabirds, especially penguins. Holothurian 1 1,0 However, 26 squid taxa have also been identified from Salp 1 1,0 greatwinged petrels and the fact that relatively few taxa Polychaete 1 1,0 have been identified from the other two gadfly petrels (Table 6) is probably due to the few identifiable beaks so far obtained rather than to differences in foraging Total 104 patterns. Identification of fish prey from sub-Antarctic seabirds a - Eighteen species listed in Table 6, with the addition is generally difficult because of their highly digested state of the kelp gull (Blankley 1981, Branch 1985). in most samples. Knowledge of the fish prey of seabirds b - Includes a single specimen of Alepisauris breviros­ at the Prince Edward Islands is based mainly on the iden­ tris (Alepisauridae) taken by a giant petrel Macro­ tification of otoliths. Mesopelagic myctophids (lantern nectes sp. (Gon & Klages 1988) fish), in particular Krefftichthys anderssoni, Protomyc­ c - Taken at sea by blue petrels and salvin's prions (Steele tophum tenisoni, P. normani and Electrona carlsbergi, & Crafford 1987, Gartshore et al 1988) although not are the most important fish prey of several of the pen­ strictly marine prey guin species and are also among the species found in the petrels, although data in the latter group are gener­ marionis was the predominant crustacean in both spe­ ally limiting owing to small numbers of fish identified cies in one year but was replaced by Euphausia vallenti­ (Table 7). Several nototheniid fish, especially Lepidono­ ni and Thyssanoessa vicina the following year. Similar­ tothen squamifrons, are important in the diets of gentoo ly, Electrona carlsbergi comprised an appreciable propor­ penguins and imperial cormorants, which forage closer tion of fish consumed by macaroni penguins in one year inshore. but was not present the next. In addition to sub-Antarctic Species for which detailed quantitative information for species, the diets of both macaroni and rockhopper pen­ one year or longer is available show both short-term and guins include species regarded as sub-tropical and Ant­ longer- term changes in their diets (Adams & Brown 1989, arctic, probably introduced into the region through peri­ Brown et al 1990). For example, the proportion of fish odic incursions of foreign water masses from north and in the diet of king penguins increased from 70 % in winter south of the islands (Miller 1985, Boden & Parker 1986). to 100 % in summer, suggesting a change in availability. Some species of seabirds at the Prince Edward Islands The relative proportions of each prey type eaten by derive only a portion of their food from the sea. Kelp macaroni and rockhopper penguins also change seasonal­ gulls and lesser sheathbills obtain a proportion of their ly, with crustaceans comprising almost the entire diet prey from the intertidal zone. Kelp gulls feed mainly on when chicks are small and pelagic myctophid fish and the limpet Nacella delesserti which is caught in shallow squid increasing to form almost the whole diet when water in calm conditions (preferentially selecting large, chicks are older and adults forage farther offshore (see pale individuals), and the bivalve Gaimardia trapesina below). Consequently, despite the very similar diets which is obtained from the fronds of kelp Macrocystis (Brown & Klages 1987, Cooper et al 1990), food resources sp. offshore (Blankley 1981, Branch 1985). Lesser sheath­ are nevertheless partitioned between the two species since bills forage intertidally on the membranous alga Porphyra macaroni penguins begin breeding three to four weeks sp. which they pull off the rocks at low tide and also before rockhopper penguins (Fig 2). By the time rock­ on several other intertidal and beach organisms (Burger hopper penguin chicks hatch and are fed crustaceans, 1981a). Sheathbills also scavenge eggs and carcasses and macaroni penguins are foraging farther offshore (Brown kleptoparasitize adult penguins feeding chicks. In winter, 1987a) for fish and cephalopods. both kelp gulls and sheathbills feed mainly on terrestri­ In audition to seasonal changes, large year-to-year al invertebrates found on the vegetated coastal plains changes were evident in the diets of macaroni and rock­ (Burger 1978, 1981a, 1982, 1984). Antarctic and Kerguelen hopper penguins (Brown et al 1990). For example, N. terns at the Prince Edward Islands feed close inshore on

46 S. Afr. T. Nav. Antarkt., Vol 20 No 2, 1990 unidentified fish but also forage over stranded kelp wrack er studies found the diet to consist predominantly of bur­ and coastal marshes (Berruti & Harris 1976, Ryan 1987a). rowing petrels, particularly Salvin's prions, blue petrels, Although penguins form the most important prey type Kerguelen petrels and softplumaged petrels (Adams 1982, of the two closely related species of giant petrels at the Schramm 1983b) which are attacked in the air and fre­ Prince Edward Islands, they are of greater importance quently soaked in water at ponds or seal wallows before to the males and, in particular, to males of the southern being eaten (Sinclair 1980). In addition, skuas also giant petrel which prey extensively on king penguin chicks scavenge from carcasses of seals and penguins and prey in winter (Hunter ms, Hunter & Brooke ms). Northern on penguin chicks and eggs; the latter being selected on giant petrel males rely to some extent on seal carrion the basis of location (eggs inside penguin colonies are and female northern giant petrels feed predominantly on selected in preference to those outside colonies) and qual­ squid, fish and burrowing petrels taken at sea (Hunter ity (intact or apparently intact eggs are selected in prefer­ & Brooke ms). ence to holed eggs) (Brooke 1985). Sub-Antarctic skuas feed entirely terrestrially at Marion A more recent study on skua diet has shown that 70 % Island during the breeding season (Hunter 1990). Earli- now consists of penguin carrion, mainly from macaroni

Table 5 Crustacean prey of U species of seabirds at Marion Island

Euphausiacea + + Euphausia vallentini +++ +++ +++ +++ +++ E. lucens + + Thysanoessa vicina +++ T. macrura + + Amphipoda + Themisto gaudichaudii + +++ + + +++ ++ Primno sp. + ++ + Vibilia sp. + + + ++ ++ ++ Cyllopus sp. ++ + ++ Hyperiella sp. + + ++ ++ Eurythenes obesus +++ +++ +++ Eurythenes sp. + ++ Lysiannassidae + Mysidacea Gnathophausia gigas +++ +++ ++ Decapoda + + Nauticaris marionis + +++ +++ +++ +++ Nematocarcinas + + + Jongirostris Parapasiphae sp. ++ +++ +++ Sergestes sp. ++ ++ ++ Acanthephyra sp. ++ Lithodes sp. + Cladocera + Ostracoda + +

No of taxa 3 7 10 7 2 3 7 6 5 7 6 2

+, + + and + + + signify increasing relative abundance.

Sources: Adams & Klages 1987, Adams & Klages 1989, Brown & Klages 1987, Espitalier-Noel et al 1988, Fitz­ Patrick Institute unpubl. data, Hunter & Klages 1989, Gartshore et al 1988, La Cock: et al 1984, Schramm 1986b, Steele & Klages 1986.

S. Afr. J. Antarct. Res., Vol 20 No 2, 1990 47 Tobie 6 Cephalopod prey of 18 species of seabirds at Marion Island

~ -~ ~ ~ ] ] ] ] .5 ~ e ] ·t ! ~ ~ l ! ...... -e= ~ .,Q i 0 .5 = .,Q ] ! ! ! ~ ";J "' ! E3 .5 1:)1) ";J ~ = ... ~ -= - -= .9= -=~ -=~ ... .Sb ~ -! 0 1:)1) ... ., -= t -=~ -= ~ ~ ... 1:)1) ~ = ! Q., ... Q., .§ ! ·a .5 ~ ~ = 0 ... ~ ~ E E .5 = E3 .c ";J Prey taxa ! e ~ 1.c .§ ~ ~ ~"' ~ al ·c: 1:)1) .c .5 l = ~ ~ ' ~ :!ll -= j ~ .c ~ -a .~ ~ t' ~ ! ~ - ~ c.,= c.,... .!:fl j- -; = ~ .§ ~ ~ ~ I ~ ~ ..l ~ 00 = c., ~ i ~ Onychoteuthidae +++ + ++ ++ +++ +++ ~ Moroteuthis knipovitchi + +++ ++ ++ ++ ++ + I M. robsoni + ++ ++ + + + M. ingens ++ + + + Moroteuthis sp. + :, Kondakovia longimana + + + + + +++ +++ +++ +++ +++ ++ +++ ++ ++ Octopoteuthidae Octopoteuthis rugosa + Octopoteuthis sp. + + ++ + Taningia danae + + + Histioteuthidae Histioteuthis adantica ++ + + +++ H. dolfleini + + ++ H. eltaninae ++ +++ +++ +++ + ++ H. macrohista + + ++ + H. meleagroteuthis ++ ++ + + H. miranda + + + ++ Histioteuthis sp. + + ++ Gonatidae Gonatus antarcticus + ++ ++ ++ ++ + + +++ + + ""' G. phoebetria + Gonatus sp. ++ Enoploteuthidae Ancistrocheirus Jesueuri + + ++ Orrunastrephidae + Notodarus sp. + Martialia hyadesi + + Chiroteuthidae Chiroteuthis capensis ++ C. picteti + + ++ +++ + C. veranyi + + C. macrosoma + + ++ Chiroteuthis sp. E + + Chiroteuthis sp. ++ ++ ++ + ++ + + + Cycloteuthidae Discoteuthis Jaciniosa ++ D. discus + Discoteuthis sp. C + Discoteuthis sp. + + ++ + Cycloteuthis akimushkini + C. sirventi ++ + Brachioteuthidae

Brachioteuthis picta ++ '4 Brachioteuthis sp. + Mastigoteuthidae Mastigoteuthis sp. A + + Mastigoteuthis sp. C + + Mastigoteuthis agassizi + + Psychroteuthidae I Psychroteuthis glacialis + + ++ ~ Psychroteuthis sp. B + + Cranchiidae + + Taonius belone + + ++ T. cymoctypus ++ + + ++ T. pavo ++ + + + ++ + Taonius sp. ++ Teuthowenia antarctica ++ +-l +++ +++ + ++ ++ + T. megalops impennis + + ++

48 S. Afr. T. Nav. Antarkt. , Vol 20 No 2, 1990 Tobie 6 continued

.5 t j

T pellucida + Phasmatopsis sp. + Galiteuthis armata + + + + G. glacialis + ++ ++ ++ +++ ++ Me.sonychoteuthis hamiltoni + + + Megalochranchia maxima + Bathothauma lyromma + Neoteuthidae Alluroteuthis antaICticus + ++ + + Alluroteuthis sp. + Bathyteuthidae Bathyteuthis abyssicola + Lepidoteuthidae Lepidoteuthis grimaldi + Pholidototeuthidae Pholidoteuthis boschmae + + Architeuthidae Architeuthis sp. + Oegopsid A + Octopodidae ++ Alloposus mollis + + ++ Octopus sp. +++ ++ +++ ++ + +++

No of identified 6122 865 1171 484 1288 901 68 3442 426 34 165 13 22 124 12 11 31 37 lower beaks

No of taxa 9 2 4 2 37 11 4 33 19 9 9 2 26 6 7 7 1

+, + + and + + + signify increasing relative abundance.

Sources: as for Table 5, with the addition of Berruti & Harcus (1978), Brooke & Klages (1986), Imber & Berruti (1981), Lipinski & Jackson 1989, Hunter & Brooke ms and NJ Adams (pers comm). and rockhopper penguins (Hunter 1990). This change Jackson & Ryan (1986) have shown that whitechinned in diet is regarded as a response to increased predation petrels digest fish faster than they do squid and crusta­ on burrowing petrels by feral cats (van Rensburg 1985, ceans, although their assimilation efficiency of each prey 1986, van Rensburg & Bester 1988a, Bloomer & Bester type is similar (Jackson 1986). Digestion rates, gut pas­ 1990) and is reflected in a high breeding success of skua sage rates and assimilation efficiencies of penguins, al­ pairs which nest close to penguin colonies and a lower batrosses and petrels at Marion Island have recently been success in pairs breeding farther away in areas of tradi­ studied in detail (Jackson & Cooper 1988, Jackson 1990, tionally high burrowing petrel concentrations (Hunter Jackson & Place 1989, 1990) No clear trends were found 1990). The relative abundance of burrowing petrel re­ in assimilation efficiencies of several Marion Island sea­ mains in the vicinity of skua nests have also previously birds in relation to diet and intraspecific variability was been used as a means of rapidly assessing their relative high (Jackson 1990). Seabirds were found to be able to abundance at Marion Island (Schramm 1983b) and at digest wax esters and, for those species naturally feed­ Prince Edward Island (Adams 1982). ing on crustaceans, chitin. The latter ability is presumed Less well studied than the diets is the efficiency with to enhance the efficiency of digestion of the soft tissues which seabirds at Marion Island assimilate their food. of crustaceans (Jackson 1990). Adams (1984, 1990) has shown that king penguins have an 81 % assimilation efficiency for squid and also that Foraging ecology the time taken to digest a fish meal was correlated to the size of the meal, larger meals taking longer to digest. Feeding methods of sub-Antarctic seabirds and their re-

S. Afr. J. Antarct. Res. , Vol 20 No 2, 1990 49 Tobie 7 Fish prey of 13 species of seabirds at Marion Island

C: l ·~ "O C: ~ .r Prey taxa j j

Myctophidae Krefftichthys anderssoni +++ ++ +++ +++ Protomyctophum tenisoni +++ ++ +++ +++ ++ P. normani ++ ++ +++ +++ P. bolini + ++ + + Protomyctophum sp. + + + Gymnoscopelus nicholsi ++ + Gymnoscopelus sp. + + + + Electrona carlsbergi + + + + + + + + + ++ + E. subaspera + + + Diaphus sp. + + Nototheniidae +++ + Lepidonotothen squamifrons + + + + + +++ Notothenia coriiceps + Gobionotothen acuta + + Paranotothenia magellanica + ++ + Dissostichus eleginoides + + Paralepididae Magnisudis prionosa + + Harpagiferidae + Harpagifer georgianus +? + Channichthyidae Channichthys rhinoceratus + + + Chiasmodontidae Dysalotus cf. alcocki + Muraenolepidae Muraenolepis sp. + ++ Trichiuridae Benthodesmus elongatus +

No of identified fish 13 399 2 257+ 5 167 893 2 11 3 5 376 2

No of taxa 10 13 12 7 2 3 3 2 3 2

+, ++ and +++ signify increasing relative abundance. Sources: as for Table 5 with the addition of Blankley (1981) and NJ Adams (pers comm). Taxonomic treatment after Gon & Klages (1988).

lation to morphology have been discussed in detail else­ mean maximum foraging ranges of 300 km. Macaroni where (eg Croxall & Prince 1980, Prince 1980, Harper and rockhopper penguins probably forage up to 200 km et al 1985, Harper 1987). Research at the Prince Edward from Marion Island and gentoo penguins from 15 - 40 Islands has focused on determining foraging ranges of km. There is considerable variation in mean maximum penguins, which account for the greater avian biomass foraging range, related to the age of the chicks at the time at the islands. Although estimates of foraging ranges of of measurement (Table 8). penguins may be calculated from the frequency with The diving capabilities of penguins at Marion Island which they feed their chicks (eg Williams & Siegfried have been investigated using simple, capillary-type depth 1980), such estimates are subject to assumptions on swim­ gauges. Gentoo penguins at the islands are capable of ming speed and the amount of time the birds spend swim­ 70-m dives, although most dives seldom exceed 20 m, ming. The deployment of direct recording autoradiograph­ which is consistent with what is known of their foraging ic speed meters (Wilson & Bain 1984, Adams et al 1988) range and the distribution of their prey (Adams & Brown on penguins at Marion Island has shown that they nor­ 1983). Information based on only two measurements from mally swim at speeds of 7,4 - 8,7 km h 1 (Table 8). rockhopper penguins at Marion Island suggests that these King penguins forage farthest from their colonies, with birds are also capable of dives ofup to 100 m (CR Brown

50 S. Afr. T. Nav. Antarkt. , Vol 20 No 2, 1990 unpubl data) and measurements from king penguins else­ Energy and food requirements where are of depths in excess of 240 m (Kooyman et i1 1982). Apart from an endocrine study of sheathbills (Burger & Despite estimates of the penguins' potential foraging Millar 1980), physiological studies on seabirds at Marion ranges, we do not know exactly where penguins feed. Island began in 1981. Primary objectives of these studies Although macaroni, rockhopper and king penguins have were to understand the ecological adaptations of the sea­ the potential to forage south of the Antarctic Polar Front birds to a sub-Antarctic environment and to assess their (between 200 and 250 km south of the islands), there potential impact on marine resources in the Southern is no dietary evidence to suggest that they do so because Ocean by measuring their energy requirements. all prey species present in their diets are known to oc­ Breeding albatrosses, petrels and penguins, in particu­ cur in the vicinity of the Prince Edward Islands (Allan­ lar, spend long periods ashore, especially during incu­ son et al 1985, Boden & Parker 1986). Furthermore, we bation, chick-brooding and, in penguins, during moult. do not know where king, macaroni and rockhopper pen­ It is during these periods that the birds are available for guins go outside their breeding seasons, although vagrants physiological studies. may wander long distances (Enticott 1986, Gartshore et Initially, studies concentrated on measuring resting al 1988). metabolic rates, using oxygen consumption as a meas­ To date, we have little idea of where albatrosses and ure of metabolism, to establish baseline measurements petrels breeding at the Prince Edward Islands forage, against which energy expenditures of other activities although in some cases inferences may be made from could be compared. Resting metabolic rates of the larg­ their prey. For example, the relative proportions of Ant­ er albatrosses and giant petrels are similar to those of arctic versus sub-Antarctic squid in the diets of sooty birds of similar size (Adams & Brown 1984), but the rest­ ing metabolism of the smaller burrowing petrels was and lightmantled sooty albatrosses suggest that lightman­ higher than expected for birds of equivalent body mass, tled sooty albatrosses forage mainly to the south of the as were the metabolic rates of five species measured at Antarctic Polar Front and sooty albatrosses mainly to the Gough Island (C R Brown unpubl data). A high rate north (Berruti 1979, Frost 1979, FitzPatrick Institute un­ of resting metabolism may represent an adaptation in the publ data). Elucidation of these aspectss of foraging ecol­ smaller : petrels to cold sub-Antarctic and temperate en­ ogy requires the use of satellite-based telemetry (eg Par­ vironments. In contrast, small tropical petrels have meta­ melee et al 1985, Strikwerda et al 1986, Nowak & Ber­ bolic rates lower than expected, presumably an adapta­ thold 1987, Jouventin & Weimerskirch 1990) or dead­ tion to reduce heat loading (Pettit et al 1985). reckoning devices which accumulate data on vectors and With the exception of the king penguin, resting meta­ distances travelled (Wilson & Wilson 1988). bolic rates of the penguins at Marion Island were also In contrast to marine foraging species, the foraging be­ higher than expected (Brown 1984). However, it is nota­ haviour of giant petrels, skuas and sheathbills on land ble that penguins held in captivity for long periods of are better documented (Sinclair 1980, Burger 1981a, 1982, time have metabolic rates close to expected values (Drent 1984, Hunter 1990, ms, Hunter & Brooke ms) as are & Stonehouse 1971, Erasmus & Wessels 1985). those of kelp gulls inter- and subtidally (Branch 1985) During incubation, albatrosses, petrels and penguins and imperial cormorants and Kerguelen terns subtidally undertake long shifts during which they fast. In addi­ (Cooper 1985, Ryan 1987a). tion, egg temperatures must be maintained at levels well

Tobie 8 Mean foraging parameters of breeding penguins at Marion Island. Values for time spent swimming and foraging range are separated into adults feeding large and small chicks

Species N Travelling speed N Time Foraging range Source swimming Km-1h % Km small large small large range

King penguin 15 8,7 6 19 36 255 301 75 - 902 Adams (1987)

Gentoo penguin 53 7,9 38 16 14 1 - 103 Adams & Wilson (1987)

Macaroni penguin 5 7,5 3 - 38 178 59 - 303 Brown (1987a)

Rockhopper penguin 7 7,4 7 30 33 4 - 157 Brown (1987a)

S. Afr. J. Antarct. Res., Vol 20 No 2, 1990 51 above ambient temperatures. Oxygen consumptions of Studies of embryonic metabolism of Procellariiformes penguins, wandering albatrosses and burrowing petrels, have shown that in most respects sub-Antarctic species measured during incubation both in the field and the are similar to tropical species (Whittow 1983). However, laboratory at Marion Island, suggest that energy expen­ total oxygen requirements for development are higher in diture is seldom elevated above resting levels (Brown sub-Antarctic species as a result of higher hatchling lev­ 1984, Brown & Adams 1984, Adams 1990, CR Brown els of metabolism, a feature which may pre-adapt chicks unpubl data). This feature has been observed in other to the higher thermoregulatory demands experienced in petrels (Grant & Whittow 1983, Ricklefs et al 1986) and a cold environment (Brown & Adams 1988). may represent an adaptation to reduce energy require­ Although energy, and hence food, demands of seabirds ments during long incubation shifts. Measurement of oxy­ at Marion Island are highest during summer when most gen consumption during incubation is in contrast to most species breed, they are especially high during chick­ studies on these species to date, which have estimated rearing. Albatrosses and petrels have extended nestling energy requirements from rates of loss of body mass of periods, a characteristic that has been attributed to a individuals during incubation fasts (Croxall 1982). The patchy, sparse distribution of prey, subject to unpredic­ latter estimates are sensitive to assumptions on how much table fluctuations (Lack 1968). More recently, constraints of the mass loss comprises fat, protein and water (Grosco­ on the ability of the adult to transport food between the las 1988). foraging area and the nest site have been used to explain long nestling periods (Ricklefs 1983). Measurements of the energy requirements of chicks have been made at Marion Island for king, gentoo, macaroni and rockhop­ per penguins (Brown 1987c, Adams 1990), wandering al­

1500 Macaroni batross and southern giant petrel (N J Adams unpubl penguins it Lipid data) and whitechinned, greatwinged, blue and Salvin's ~ Protein • Biosynthesis prion (Berruti et al 1985, Brown 1988b). Examples of 0 Maintenance energy budgets, calculated from measurements of oxy­ gen consumption throughout the fledging period and from ';' 1000 body composition analysis, of macaroni and rockhop­ 'O -, per penguin chicks are illustrated in Fig 3. Overall, energy :. requirements for growth and maintenance of chicks ranged from 9 250 kJ for Salvin's prions to about 747 000 kJ for wandering albatrosses, equivalent to food re­ 500 quirements of 0,93 and 220 kg of food , respectively. The cost for procellariiform seabirds of producing a chick (70 kJ per g of chick produced) is high compared with that of penguins ( 40 kJ per g of chick) and is a result of the long nestling periods characteristic of this order. Ricklefs (1983) suggested that in some seabirds maxi­ mum energy demands occur during brooding when adult foraging is constrained by the need for one parent to re­ main at the nest site. He further suggested that the high demands of brooding might explain the early develop­ ment of homeothermy in the chicks of some species. In­ Rockhopper penguins vestigations at Marion Island have shown that burrow­ 1000 ing petrel chicks can maintain, by metabolic adjustment, near-adult body temperatures (about 38° C) at ambient .. temperatures of 5 °C from within 1 - 5 days of hatch- I ing (Brown & Prys-Jones 1988). Analysis of cooling 'O -, curves of chicks at different wind speeds further sug­ :!. 500 gest that thermoregulation is facilitated by the favoura­ ble burrow rnicroclimate and the chicks' thick down at hatching. In most seabirds moult extends over a period of time and the daily cost of moult is subsequently negligible

7 14 21 28 35 42 49 56 63 70 in relation to normal existence energy requirements. In contrast, penguins undergo a post-breeding moult while Age(days) fasting ashore in which the entire plumage is replaced in 4-5 weeks. New feather synthesis and reduced ther­ mal insulation makes moult in penguins, as indicated by a high rate of mass loss, an energetically expensive Fig 3: Energy budgets of macaroni and rockhopper pen­ process. The actual energy cost can be estimated from guin chicks at Marion Island from hatching to fledging rates of mass loss and changes in body composition. (after Brown 1987, 1989) These changes have been measured for macaroni and

52 S. Afr. T. Nav. Antarkt. , Vol 20 No 2, 1990 rockhopper penguins at Marion Island (Williams et al Conservation research on the avifauna of 1977) and have been used to calculate the energetic cost of moult in other species of penguins (Croxall 1982). the Prince Edward Islands Measurement of changes in mass-specific oxygen con­ sumption in the same two species has also provided in­ Conservation problems of birds at Marion and Prince formation on the energy requirements of the different Edward Islands are all the result of man's activities (Wil­ processes of moult (Brown 1985, Adams & Brown 1990), liams et al 1979, Williams 1984). The conservation which overall exceed the energy requirements of rest­ management of the two islands themselves has recently ing, non-moulting birds by 20 - 30%. In addition, a been discussed (Bonner & Lewis-Smith 1985, Clark & study of the development of the new feathers has indi­ Dingwall 1985, Walton 1986, Smith 1987a, Smith & Lewis cated that these begin developing 3 - 4 days before the Smith 1987, Cooper & Condy 1988, Cooper & Berruti birds first return ashore, thus increasing reported moult 1989). The greatest threat to the birdlife is undoubtedly periods by an equivalent period of time (Brown 1986). that of feral cats at Marion Island which have reduced Although spending relatively long periods ashore dur­ population sizes and lowered breeding success of sever­ ing parts of the breeding season, most albatrosses and al species of burrowing petrels (eg van Aarde 1980, petrels and some species of penguins are pelagic (in­ Schramm 1983, van Rensburg 1985, Fugler et al 1987, dividual wandering albatrosses, for example, spend 80% van Rensburg & Bester 1988a, Newton & Fugler 1989, of their breeding season at sea foraging). Measurements Bloomer & Bester 1990, FitzPatrick Institute unpubl data). of activity budgets and energy requirements during this One species, the common diving petrel, is thought to be period require the use of remote-sensing devices ( eg extinct as a breeding species at Marion Island and great­ Prince & Francis 1984, Cairns et al 1987) and doubly­ winged, softplumaged and grey petrels are regarded by . labelled water (eg Davis et al 1983, Nagy et al 1984, Costa Brooke (1984) as severely threatened as a result of the et al 1986). Doubly-labelled water at Marion Island has activities of cats, as is probably also the Kerguelen petrel. been used to estimate the cost of foraging in wandering In 1977 the introduction of the host-specific contagious albatrosses, demonstrating that energy expenditures dur­ feline panleucopaenia virus (FPL) led to a reduction in ing a foraging trip are less than twice that of resting birds cat numbers, now being further reduced by shooting and · (Adams et al 1986). This low value is attributed to the trapping (van Aarde 1984, van Rensburg 1986, van Rens­ highly efficient soaring mode of locomotion of these birds burg et al 1987, van Rensburg & Bester 1988b, Bloomer and has subsequently aiso been demonstrated for grey­ & Bester 1990, M N Bester pers comm). Cats are not headed albatrosses at South Georgia (Costa & Prince present at Prince Edward Island, which therefore serves 1987) and Laysan albatrosses at Hawaii (Pettit et al 1988). as a "sanctuary" for burrowing petrels. The ultimate objective of energetic studies of seabirds Potential or known threats to bird life are from chlori­ at Marion Island is to construct models of the timing and nated hydrocarbons (Gardner et al 1985), underwater magnitude of the impact of the different species on ma­ blasting (Brown & Adams 1983), human disturbance, fuel rine resources in the vicinity by integrating information spills and entanglement (Cooper & Condy 1988) and in­ on energy requirements, population numbers, breeding gestion of plastic particles (Ryan 1987b). Plastic parti­ seasons and success, diet and assimilation efficiency. cles were found in varying amounts in all of 19 species Although preliminary, such models exist for lesser sheath­ of seabirds sampled at the Prince Edward Islands (Ryan bills (Burger 1981b) and wandering albatrosses (Adams 1987b, 1988). However, no significant differences were et al 1986). The most detailed models are for the four found in assimilation efficiencies of whitechinned petrels species of penguins which together consume 880 000 artificially fed large quantities of plastic particles (Ryan tonnes of crustaceans, fish and cephalopods during their & Jackson 1987). Also, ingested particles had no appar­ seven-month breeding and moulting cycles (Brown 1987b, ent effect on blue petrels, using bird mass and fat reserves Brown 1989, Adams 1990). Of this figure, king penguins as measures of condition (Ryan 1987c). account for 74 % , macaroni penguins 21 % , rockhopper A recent proposal to construct an emergency landing penguins 5 % and gentoo penguins less than 1% (Adams facility for fixed-wing aircraft at Marion Island was not 1990). Pelagic fish account for 70 % of total prey biomass, approved, after an environmental impact assessment. followed by crustaceans (18 %) and cephalopods (11 %) Such a development would have had both short- and long­ (Adams 1990). term deleterious effects on the island's birds (Heymann Several such models (eg Wiens & Scott 1975, Furness et al 1987). 1978, Furness & Cooper 1982) rely on estimates of energy requirements for different activities based on multiples of basal or existence energy requirements calculated from Acknowledgements allometric equations or, in some cases, from few em­ pirical data (Pettit et al 1984, Croxall et al 1984). Although This review commemorates the 25th anniversary of the many aspects of seabird ecology relating to model in­ First South African Biological Expedition to Marion and puts still require further study at Marion Island, the in­ Prince Edward Islands, 1965 - 1966. We thank N J formation on energy expenditures represents a substan­ Adams, SR Henley, S Hunter, S Jackson, NT W 'KJ.ages, tial advancement in estimating food consumption of sea­ P G Ryan, W R Siegfried, W K Steele and B P Watkins bird populations at the Prince Edward Islands and cur­ and three anonymous referees for allowing us the use of rently outweighs information on the available prey their unpublished information and/or for commenting on populations. the text. Ornithological research at the Prince Edward

S. Afr. J. Antarct. Res., Vol 20 No 2, 1990 53 Islands forms part of the South African National Antarctic bills and Imperial Cormorants at Marion Island (subantarctic). Cor­ Research Programme and receives the logistic and finan­ morant 9: 77-84 cial support of the South African Department of Environ­ BLOOMER JP & BESTER MN 1990. Diet of a declining feral cat ment Affairs. Fe/is catus population on Marion Island. S. Ali: J. Wild/. Res. 20: 1-4 BODEN BP & PARKER LD 1986. The plankton of the Prince Ed­ ward Islands. Polar Biol. 5: 81-93 BONNER WN & LEWIS SMITH RI 1985. Conservation areas in References the Antarctic. Cambridge: Scientific Committee on Antarctic Research BRANCH GM 1985. The impact of predation by Kelp Gulls Larus dominicanus on the sub-Antarctic limpet Nacella delesserti. Polar Biol. ADAMS NJ 1982. Subantarctic skua prey remains as an aid for rapidly 4: 171-177 assessing the status of burrowing petrels at Prince Edward Island. BROOKE M de L 1985. Skua predation on penguin eggs: the in­ Cormorant 10: 97-102 fluence of egg quality and location. Wilson Bull. 97: 366-368 ADAMS NJ 1984. Utilization efficiency of a squid diet by adult King BROOKE M de L 1986. The vocal systems of two nocturnal burrow­ Penguins (Aptenodytes patagonicus). Auk 101: 884-886 ing petrels, the White-chinned Procellaria aequinoctialis and the Grey ADAMS NJ 1987. Foraging range of King Penguins Aptenodytes P. cinerea. Ibis 128: 502-512 patagonicus during summer at Marion Island. J. Zool., Lond. 212: BROOKE M de L & KLAGES N 1986. Squid beaks regurgitated by 475-482 Greyheaded and Yellownosed Albatrosses, Diomedea chrysostoma and ADAMS NJ 1990. Feeding biology and energetics of King Apteno­ D. chlororhynchos at the Prince Edward Islands. Ostrich 57: 203-206 dytes patagonicus and Gentoo Pygoscelis papua Penguins at sub­ BROOKE RK 1984. South African Red Data Book-birds. S. Afr. Antarctic Marion Island. PhD thesis, University of Cape Town. 246 pp Natn. Sci. Prag. Rpt. 97: 1-213 ADAMS NJ & BROWN CR 1983. Diving depths of the Gentoo Pen­ BROWN CR 1984. Resting metabolic rate and energetic cost of in­ guin (Pygoscelis papua). Condor 85: 503-504 cubation in Macaroni Penguins (Eudyptes chrysolophus) and Rock­ ADAMS NJ & BROWN CR 1984. Metabolic rates of sub-Antarctic hopper Penguins (E. chrysocome). Comp. Biochem. Physiol. 77A: Procellariiformes: a comparative study. Comp. Biochem. Physiol. 345-350 77A: 169-173 BROWN CR 1985. Energetic cost of moult in Macaroni Penguins ADAMS NJ & BROWN CR 1989. Dietary differentiation and troph­ (Eudyptes chrysolophus) and Rockhopper Penguins (E. chrysocome). ic relationships in a sub-Antarctic penguin community at Marion Is­ J. Comp. Physiol. 155: 515-520 land. Mar. Ecol. Prag. Ser. 57: 249-258 BROWN CR 1986. Feather growth, mass loss and duration of moult ADAMS NJ & BROWN CR 1990. Energetics of molt in penguins. in Macaroni and Rockhopper Penguins. Ostrich 57: 180-184 In: Penguin biology, eds LS David & JT Darby. Academic Press, BROWN CR 1987a. Traveling speed and foraging range of macaroni San Diego. pp 297-315 and rockhopper penguins at Marion Island. J. Field Om. 58: 118-125 ADAMS NJ, BROWN CR & NAGY KA 1986. Energy expenditure BROWN CR 1987b. Ecological energetics of Eudyptes penguins at of free-ranging Wandering albatrosses Diomedea exulans. Physiol. Marion Island. PhD thesis, University of Cape Town. 205 pp Zoo] . 59: 583-591 BROWN CR 1987c. Energy requirements for growth and maintenance ADAMS NJ & KLAGES NT 1987. Seasonal variation in the diet in Macaroni and Rockhopper Penguins. Polar Biol. 8: 95-102 of the King Penguin (Aptenodytes patagonicus) at sub-Antarctic Mari­ BROWN CR 1988a. Egg temperature and embryonic metabolism of on Island. J. Zool., Lond. 212: 303-324 A- and B-eggs of macaroni and rockhopper penguins. S. Ali: J. Zool. ADAMS NJ & KLAGES NT 1989. Temporal variation in the diet 23: 166-172 of the Gentoo Penguin Pygoscelis papua at sub-Antarctic Marion Is­ BROWN CR 1988b. Energy requirements for growth in Salvin's Pri­ land. Colonial Waterbirds 12: 30-36 ons Pachyptila vittata salvini, Blue Petrels Halobaena caerulea and ADAMS NJ & WILSON M-P 1987. Foraging parameters of Gentoo Great-winged Petrels Pterodroma macroptera. Ibis 130: 527-534 Penguins Pygoscelis papua at Marion Island. Polar Biol. 7: 51-56 BROWN CR 1988c. Energy expenditure during incubation in four ADAMS NJ, WILSON RP & BAIN CAR 1988. The use of thallium species of sub-Antarctic burrowing petrels. Ostrich 59: 67-70 as a radioactive source in autoradiographic devices for penguins at BROWN CR 1989. Energy requirements and food consumption of sea. J. Field Om. 59: 89-94 Eudyptes penguins at the Prince Edward Islands. Antarct. Sci. 1: 15-21 ALLANSON BR, BODEN B, PARKER L & DUNCOMBE RAE BROWN CR & ADAMS NJ 1983. The effect of underwater explo­ C 1985. A contribution to the oceanology of the Prince Edward Is­ sions on Rockhopper Penguins Eudyptes chrysocome. Cormorant 11: lands. In: Antarctic nutrient cycles and food webs, eds WR Siegfried, 68-69 PR Condy & RM Laws. Springer, Berlin. pp 38-45 BROWN CR & ADAMS NJ 1984. Basal metabolic rate and energy BENNETS JA 1948. King Penguins own Marion Island. Afr. Wild/. expenditure during incubation in the Wandering Albatross (Diomedea 2(3): 57-63 exulans). Condor 86: 182-186 BENNETS JA 1949. Island home of "Wandering Albatross". Ali: BROWN CR & ADAMS NJ 1988. Egg temperature, embryonic Wild/. 2(4) 22-29,31 metabolism and water loss from the eggs of sub-Antarctic Procella­ BERRUTI A 1979. The breeding biologies of the sooty albatrosses riiformes. Physiol. Zool. 61: 126-136 Phoebetria fusca and P. palpebrata. Emu 79: 161-175 BROWN CR & KLAGES NT 1987. Seasonal and annual variation BERRUTI A 1981. Displays of the sooty albatrosses Phoebetria fus­ in diets of Macaroni (Eudyptes chrysolophus chrysolophus) and ca and P. palpebrata. Ostrich 52: 98-103 Southern Rockhopper (E. chrysocome chrysocome) Penguins at sub­ BERRUTI A, ADAMS NJ & BROWN CR 1985. Chick energy Antarctic Marion Island. J. Zoo]., Lond. 212: 7-28 balance in the Whitechinned Petrel Procellaria aequinoctialis. In: Ant­ BROWN CR, KLAGES NT & ADAMS NJ 1990. Short and medium­ arctic nutrient cycles and food webs, eds WR Siegfried, PR Condy term variation in the diets of penguins at Marion Island. S. Afr. J. & RM Laws. Springer, Berlin. pp 460-465 Antarct. Res. 20: 13-20 BERRUTI A, GRIFFITHS AM, IMBER MJ, SCHRAMM M & BROWN CR & PRYS-JONES RP 1988. Development of homeother­ SINCLAIR JC 1981. The status of seabirds at Prince Edward Island. my in chicks of sub-Antarctic burrowing petrels. S. Afr. J. Zool. 23: S. Afr. J. Antarct. Res. 10/11: 31-32 288-294 BERRUTI A & HARCUS T 1978. Cephalopod prey of the sooty al­ BURGER AE 1978. Terrestrial invertebrates: a food resource for birds batrosses Phoebetria fusca and P. palpebrata at Marion Island. S. Ali: at Marion Island. S. Ali: J. Antarct. Res. 8: 87-99 J. Antarct. Res. 8: 99-103 BURGER AE 1979. Breeding biology, moult and survival of Lesser BERRUTI A & HARRIS A 1976. Breeding schedules of Antarctic Sheathbills Chionis minor at Marion Island. Ardea 67: 1-14 and Kerguelen Terns at Marion Island. Notomis 23: 243-245 BURGER AE 1980. An analysis of the displays of Lesser Sheathbills BERRUTI & HUNTER S 1986. Some aspects of the breeding bio­ Chionis minor. Z. Tierpsychol. 52: 381-396 logy of Salvin's Prion Pachyptila vittata salvini at Marion Island. Cor­ BURGER AE 1981a. Food and foraging behaviour of Lesser Sheath­ morant 13: 98-106 bills at Marion Island. Ardea 69: 167-180 BLANKLEY WO 1981. Marine food of Kelp Gulls, Lesser Sheath- BURGER AE 1981b. TlIIle budgets, energy needs and kleptoparasitism

54 S. Afr. T. Nav. Antarkt., Vol 20 No 2, 1990 in breeding Lesser Sheathbills (Chionis minor). Condor 83: 106-112 southeastern Atlantic and southwestern Indian Oceans. Cormorant BURGER AE 1982. Foraging behaviour of Lesser Sheathbills Chio­ 13: 118-142 nis minor exploiting invertebrates on a sub-Antarctic island. Oecolo­ ERASMUS T & WESSELS ED 1985. Heat production studies on gia 52: 236-245 normal and oil covered Jackass penguins (Spheniscus demersus) in BURGER AE 1984. Winter territoriality in Lesser Sheathbills on air and water. S. Afr. J. Zool. 20: 209-212 breeding · grounds at Marion Island. Wilson Bull. 96: 20-33 ESPITALIER-NOEL G, ADAMS NJ & KLAGES NT 1988. Diet BURGER AE, LINDEBOOM HJ & WILLIAMS AJ 1978. The of the Imperial Cormorant Phalacrocorax atriceps at sub-Antarctic mineral and energy contributions of guano of selected species of birds Marion Island. Emu 88: 43-46 to the Marion Island terrestrial ecosystem. S. Afr. J. Antarct. Res. FROST PGH 1979. Seabird distribution and the transport of nutrients 8: 59-70 from marine to terrestrial ecosystems. S. Afr. J. Antarct. Res. 9: 20-27 BURGER AE & MILLAR RP 1980. Seasonal changes of sexual and FUGLER S 1985. Chemical composition of guano of burrowing petrel territorial behaviour and plasma testosterone levels in male Lesser chicks (Procellariidae) at Marion Island. In: Antarctic nutrient cy­ Sheathbills Chionis minor. Z. Tierpsychol. 52: 397-406 cles and food webs, eds WR Siegfried, PR Condy & RM Laws. BURGER AE & WILLIAMS AJ 1979. Egg temperatures of the rock­ Springer, Berlin. pp 169-172 hopper penguin and some other penguins. Auk 96: 100-105 FUGLER SR, HUNTER S, NEW10N IP & STEELE WK 1987. BURGER AE, WILLIAMS AJ & SINCLAIR JC 1980. Vagrants and Breeding biology of Blue Petrels Halobaena caerulea at the Prince the paucity of land bird species at the Prince Edward Islands. J. Bio­ Edward Islands. Emu 87: 103-110 geogr. 7: 305-310 FURNESS BL, LAUGKSCH R & DUFFY DC 1984. Cephalopod CAIRNS DK, BREDIN KA & MONTEVECCHI WH 1987. Activi­ beaks and studies of seabird diets. Auk 101: 619-620 ty budgets and foraging ranges of breeding Common Murres. Auk FURNESS RW 1978. Energy requirements of seabird communities: 104: 218-224 a bioenergetics model. J. Anim. Ecol. 47: 39-53 CLARK MR & DINGWALL PR 1985. Conservation of islands in FURNESS RW & COOPER J 1982. Interactions between breeding the Subantarctic: a review of the protected areas of Insulantarctica. seabirds and pelagic fish populations in the southern Benguela region. International Union for the Conservation of Nature and Natural Mar. Ecol. Prog. Ser. 8: 243-250 Resources, Gland, Switzerland GARDNER BD, SIEGFRIED WR & CONNEL AD 1985. Chlori­ COOPER J 1985. Foraging behaviour of non-breeding Imperial Cor­ nated hydrocarbons in seabird eggs from the southern Atlantic and morants at the Prince Edward Islands. Ostrich 56: 96-100 Indian Oceans. In: Antarctic nutrient cycles and food webs, eds WR COOPER J & AVERY G 1986. Historical sites at the Prince Edward Siegfried, PR Condy & RM Laws. Springer, Berlin. pp 647-651 Islands. S. Afr. Natn. Sci. Prag. Rpt. 128: 1-82 GARfSHORE NA 1987. Rare bird sightings at the Prince Edward COOPER J & BERRUTI A 1989. The conservation status of South Islands, December 1983-May 1987. Cormorant 15: 48-58 Africa's continental and oceanic islands. In: Biotic diversity ofsouthern GARI'SHORE NA, COOPER J & HUNTER S 1988a. Bird ringing Africa: conservation and management, ed BJ Huntley. Oxford Univer­ at Marion and Prince Edward Islands, 1982-1987; with an analysis sity Press, Cape Town. PD 239-253 of movements since 1951. S. Afr. J. Antarct. Res. 18: 23-29 COOPER J & BROOKE M de L 1984. Breeding status of burrow­ GARfSHORE NA, STEELE WK & KLAGES NT 1988b. Summer ing petrels at Prince Edward Island. S. Afr. J. Antarct. Res. 14: 34-35 diet of the Salvin's prion at sub-Antarctic Marion Island. S. Afr. J. COOPijR J, BROWN CR, GALES RP, HINDELL. MA, KLAGES Zoo]. 23: 309-313 NTW, MOORS PJ, PEMBER10N D, RIDOUX V, THOMPSON GON O & KLAGES NT 1988. The marine fish fauna of the sub­ KR & VAN HEEZIK YM 1990. The diets and dietary segregation Antarctic Prince Edward Islands. S. Afr. J. Antarct. Res. 18: 32-54 of crested penguins (Eudyptes) . In: Penguin biology, eds LS David GRANT GS & WHITfOW GC 1983. Metabolic cost of incubation & JT Darby. Academic Press, San Diego. pp 131-156 in the Laysan Albatross and Bonin Petrel. Comp. Biochem. Physiol. COOPER J & CONDY PR 1988. Environmental conservation at the 74A: 77-82 sub-Antarctic Prince Edward Islands: a review and recommendations. GREMMEN NJM 1981. The vegetation of the subantarctic islands Environ. Conserv. 15: 317-326 Marion and Prince Edward. Geobotany 3: 149 pp. W Junk, The Hague COOPER J & OATLEY TB 1988. A first inventory of colour-banding GROSCOLAS R 1988. The use of mass loss to estimate metabolic projects in the Subantarctic and Antarctic 1965-1984. Cormorant 13: rate in fasting seabirds: a critical examination based on Emperor Pen­ 43-45 guins (Aptenodytes forsten) . Comp. Biochem. Physiol. 90A: 361-366 COSTA DP, DANN P & DISHER W 1986. Energy requirements of HALL AJ 1987. The breeding biology of the White-chilllled Petrel free-ranging Little Penguins, Eudyptula minor. Comp. Biochem. Phys­ Procellaria aequinoctialis at South Georgia. J. Zoo]., Lond. 212: iol. 85A: 135-138 605-618 COSTA DP & PRINCE PA 1987. Foraging energetics Qf-Orey-headed HALL KJ & WILLIAMS AJ 1981. Animals as agents of erosion at Albatrosses Diomedea chrysostoma at Bird Island, South· Georgia. sub-Antarctic Marion Island. S. Afr. J. Antarct. Res. 10/11: 18-24 Ibis 129: 149-158 HARPER PC 1987. Feeding behaviour and other notes on 20 spe­ CRAWFORD AB 1952. The birds of Marion Island, south Indian cies of Procellariiformes at sea. Notomis 34: 169-192 Ocean. Emu 52: 73-85 HARPER PC, CROXALL JP & COOPER J 1985. A guide to forag­ CROXALL JP 1982. Energy cost of incubation and moult in petrels ing methods used by marine birds in Antarctic and sub-Antarctic seas. and penguins. J. Anim. Ecol. 51: 177-194 BIOMASS Handbook 24: 1-22 CROXALL JP 1984. Seabirds. In: Antarctic ecology, Vol 2, ed RM HEYMANN G, ERASMUS T, HUNTLEY BJ, LIEBENBERG AC, Laws. Academic Press, London. pp 533-619 RETIEF G de F, CONOY PR & VAN DER WESTHUYZEN OA CROXALL JP & PRINCE PA 1980. Food, feeding ecology and eco­ 1987. Report to the Minister of Environment Affairs on an environ­ logical segregation of seabirds at South Georgia. Biol. J. Linn. Soc. mental impact assessment of a proposed emergency landing facility 14: 103-131 on Marion Island - 1987. S. Afr. Natn. Sci. Prog. Rpt. 140: 1-209 CROXALL JP, RICKETfS C & PRINCE, PA 1984. Impact of sea­ HUNTER S 1990. The impact of introduced cats on the predator­ birds on marine resources, chiefly krill, of South Georgian waters. prey interactions of a sub-Antarctic avian community. In: Antarctic In: Seabird energetics, eds GC Whittow & H Rahn. Plenum Pub­ ecosystems. Ecological changes and conservation, eds KR Kerry & lishing Corp, New York. pp 285-317 G Hempel. Springer, Berlin. pp 3 65-371 · . DAVIS RW, KOOYMAN GL & CROXALL JP 1983. Water flux and HUNTER S ms. The impact q( ayian predator-scavengers on King estimated metabolism of free-ranging Gentoo and Macaroni Penguins Penguins Aptenodytes patagonicus at Marion Island at South Georgia. Polar Biol. 2: 41-46 · HUNTER S & BROOKE M de L ms. The diet of giant petrels Ma­ DELILLE D 1987. Spatial distribution of coastal Antarctic seawater cronectes spp. at Marion Island, southern Indian Ocean bacteria: relationship with avifauna. Polar Biol. 7: 55-60 HUNTERS & KLAGES NT 1989. The diet of grey-headed albatross­ DRENT RH & SWNEHOUSE B 1971. Thermoregulatory responses es Diomedea chlororhynchos at the Prince Edward Islands. S. Afr. of the Peruvian Penguin Spheniscus humboldti. Comp. Biochem. l. Antarct. Res. 19: 31-33 Physiol. 40A: 689-710 HUT10N FW 1865. Notes on some of the birds inhabiting the ENTICOIT JW 1986. Distribution of penguins at sea in the Southern Ocean. Ibis ser. 2, Vol 1: 276-298

S. Afr. J. Antarct. Res., Vol 20 No 2, 1990 55 IMBER MJ & BERRUTI A 1981. Procellariiform seabirds as squid PETTIT TN, ELLIS HI & WHITIOW GC 1985. Basal metabolic predators. In: Proceedings ofthe Symposium on Birds of the Sea and rates in tropical seabirds. Auk 102: 172-174 Shore, 1979, ed J. Cooper. African Seabird Group, Cape Town. pp PETTIT TN, NAGY KA, ELLIS HI & WHITIOW CG 1988. Incu­ 43-61 bation energetics of the Laysan Albatross. Oecologia 74: 546-550 JACKSON S 1986. Assimilation efficiencies of Whitechinned Petrels PETTIT TN, WHITIOW GC & ELLIS HI 1984. Food and ener­ (Procellaria aequinoctialis) fed different prey. Comp. Biochem. Phys­ getic requirements of seabirds at French Frigate Shoals, Hawaii . In: iol. 85A: 301-303 Proceedings of the 2nd Symposium on Resource Investigation in the JACKSON S 1990. Seabird digestive physiology in relation to forag­ Northwestern Haw.uian Islands, Vol 2, eds RW Grigg & KY Tanou. ing ecology. PhD thesis, University of Cape Town. 210 pp University of Hawaii Sea Grant College Programme, Hawaii. pp JACKSONS & COOPER J 1988. Use of fibre-optic endoscopes in 265-282 studies of gastric digestion in carnivorous vertebrates. Comp. Biochem. PRINCE PA 1980. The food and feeding ecology of Blue Petrel Physiol. 91A: 305-308 (Halobaena caerulea) and Dove Prion (Pdchyptila desolata). J. Zoo/., JACKSON S & PLACE AR 1989. Seabird digestion: living off the Lond. 190: 59-76 fat of the sea. Nuclear Active 40: 38-41 PRINCE PA & FRANCIS MD 1984. Activity budgets of foraging JACKSON S & PLACE AR 1990. Gastrointestinal transit and lipid gray-headed albatrosses. Condor 86: 297-300 assimilation efficiencies in three species of sub-Antarctic seabird. J. RAND RW 1954. Notes on the birds of Marion Island. Ibis 96: Exp. Zool. 255: 141-154 173-206 JACKSON S & RYAN PG 1986. Differential digestion rates of prey RAND RW 1955. The penguins of Marion Island. Ostrich 26: 57-69 by Whitechinned Petrels (Procellaria aequinoctialis). Auk 103: 617-619 RAND RW 1956. Cormorants on Marion Island. Ostrich 27: 127-133 JOUVENTIN P, MOUGIN J-L, STAHL J-C & WEIMERSKIRCH RICKLEFS RE 1983. Some considerations on the reproductive ener­ H 1985. Comparative biology of the burrowing petrels at the Crozet getics of seabirds. Stud. Avian Biol. 8: 84-94 Islands. Notornis 32: 157-220 RICKLEFS RE, ROBY DD & WILLIAMS. IB 1986. Daily energy JOUVENTIN P & WEIMERSKIRCH H 1990. Satellite tracking of expenditure by adult Leach's Storm-petrels during the nesting cycle. Wandering Albatrosses. Nature 343: 746-748 Physiol. Zool. 59: 649-660 KOK OB 1975. Verspreiding van veervasgehegte sade deur voels. Os­ RYAN PG 1987a. The distribution, population size ancj foraging be­ trich 46: 261-263 haviour of Kergeulen terns at the Prince Edward Islands. S. Afr. J. KOOYMAN GL, DAVIS RW, CROXALL JP & COSTA DP 1982. Antarct. Res. 17: 163-166 Diving depths and energy requirements of king penguins. Science RYAN PG 1987b. The incidence and characteristics of plastic parti­ 217: 726-728 cles ingested by seabirds. Mar. Environ. Res. 23: 175-206 LACK D 1968. Ecological adaptations for breeding in birds. Methuen, RYAN PG 1987c. The effects of ingested plastic on seabirds: correla­ London tions between plastic load and body condition. Environ. Poll. 46: LA COCK GD, HECHT T & KLAGES N 1984. The winter diet 119-125 of gentoo penguins at Marion Island. Ostrich 55: 188-191 RYAN PG 1988. Intraspecific variation in plastic ingestion by sea­ LA GRANGE JJ 1962. Notes on the birds and mammals on Marion birds and the flux of plastic through seabird populations. Condor 90: Island and Antarctica (SANAE). 1 S. Ali: Biol. Soc. 3: 27-84 446-452 LINDEBOOM HJ 1979. Chemical and microbiological apsects ofthe RYAN PG & HUNTER S 1985. Early breeding of Imperial Cor­ nitrogen cycle on Marion Island (sub-antarctic). PhD thesis, Rijks­ morants Phalacrocorax atriceps at Prince Edward Island. Connorant universiteit, Groningen 138 pp 13: 31-34 LINDEBOOM JU 1984. The nitrogen pathway in a penguin rook­ RYAN PG & JACKSONS 1986. Stomach pumping: is killing sea­ ery. Ecology 65: 269-277 birds necessary? Auk 103: 427-428 LIPINSKI MR & JACKSON S 1989. Surface-feeding on cephalo­ RYAN PG & JACKSON S 1987. The lifespan of ingested plastic par­ pods by procellariiform seabirds in the southern Benguela region, ticles in seabirds and their effect on digestive efficiency. Mar. Poll. South Africa. J. Zool., Lond. 218: 549-563 Bull. 5: 217-219 LUTJEHARMS JRE & VALENTINE HR 1984. Southern Ocean SCHRAMM M 1983a. The breeding biologies of t'1e petrels Prerodrcr thermal fronts south of Africa. Deep-Sea Res. 31: 1461-1475 ma macroptera, P. brevirostris and P. mollis at Marion Island. Emu MILLER DGM 1985. Marine macro-plankton of two sub-Antarctic 83: 75-81 islands. In: Antarctic nutrient cycles and food webs, eds WR Sieg­ SCHRAMM M 1983b. Predation by Subantarctic Skuas Catharacta fried, PR Condy & RM Laws, Springer, Berlin. pp 355-361 antarctita on burrowing petrels at Marion Island. S. Ali: J. Antarct. MOSELEY NH 1879. Notes by a naturalist. An account of observa­ Res. 13: 41-44 tions made during the voyage of the HMS "Challenger" round the SCHRAMM M 1986a. Distribution and abundance of burrowing world in the years 1872-1876. John Murray, London petrels (Procellariidae) at Marion Island. Polar Biol. 6: 63-70 NAGY KA 1987. Food metabolic rate and food requirement scaling SCHRAMM M 1986b. The diet of chicks of Greatwinged, Kergue­ in mammals and birds. Ecol. Monogr. 57: lll-128 len and Softplumaged Petrels at the Prince Edward Islands. Ostrich NAGY KA, SIEGFRIED WR & WILSON RP 1984. Energy utiliza­ 57: 9-15 tion by free-ranging Jackass Penguins Spheniscus demersus. Ecolo­ SCHULZE BR 1971. The climate of Marion Island. In: Marion and gy 65: 1648-1655 Prince Edward Islands. Report of the Biological and Geological Ex­ NEW10N IP & FUGLER SR 1989. Notes on the winter-breeding pedition/1965-1966, eds EM van Zinderen Bakker Sr, JM Winterbot­ Greatwinged Petrel Prerodroma macroptera and Grey Petrel Procel­ tom & RA Dyer. AA Balkema, Cape Town. pp 16-31 laria cinerea at Marion Island. Connorant 17: 27-34 SIEGFRIED WR 1978. Ornithological research at the Prince Edward NOWAK E & BERTHOLD P 1987. Die Satelliten-Telemetrie in der Islands: a review of progress. S. Afr. J. Antarct. Res. 8: 30-34 Erforschung von Tierwanderungen: eine Ubersicht. J. Om 128: SIEGFRIED WR 1982. The roles of birds in ecological processes 405-422 affecting the functioning of the terrestrial ecosystem at sub-Antarctic PARMELEE DF, PARMELEE JM & FULLER M 1985. Ornitho­ Marion Island. Com. Nat. Franc. Rech. Antarct. 51: 493-499 logical investigations at Palmer Station: the first long-distance track­ SIEGFRIED WR, FORBES PF & CONOY PR 1979. Scientific ing of seabirds by satellites. Antarct. 1 U.S. 20: 162-163 research at the Prince Edward Islands, 1847-1979: a bibliography. S. PERISSONOTIO R, ALLANSON BR & BODEN BP 1990. Trophic Ali: J. Antarct. Res. 9: 35-41 relations within the island seas of the Prince Edward Archipelago, SIEGFRIED WR, WILLIAMS AJ, BURGER AE & BERRUTI A Southern Ocean. In: Trophic relationships in the marine environment, 1978. Mineral and energy contributions of eggs of selected species eds M Barnes & RN Gibson. Aberdeen University Press, Aberdeen. of seabirds to the Marion Island terrestrial ecosystem. S. Afr. J. Ant­ pp 296-314 arct. Res. 8: 75-87 PERISSONOITO R & DUNCOMBE RAE CM 1990. Occurrence SINCLAIR JC 1980. Subantarctic Skua Catharacta antarctica preda­ of anticyclonic eddies on the Prince Edward Plateau (Southern Ocean): tion techniques on land and at sea. Cormorant 8: 3-6 effects on phytoplankton biomass and production. Deep-Sea Res. 37: SMITH YR 1976. The effect of burrowing species of Procellariidae 777-793 on the nutrient status of inland tussock grasslands on Marion Island.

56 S. Afr. T. Nav. Antarkt., Vol 20 No 2, 1990 S. Ali: 1 Bot. 42: 265-272 WILLIAMS AJ 1980d. Penguin proportionate egg weights. Notor­ SMITH VR 1978. Animal-plant-soil nutrient relationships at Marion nis 27: 125-128 Island (Subantarctic). Oecologia 41: 239-253 WILLIAMS AJ 1980e. The breeding biology of Eudyptes penguins, SMITH ·VR 1979. The influence of seabird manuring on the phos­ with particular reference to egg-sae dimorphism. PhD thesis, Univer­ phorous status of Marion Island (Subantarctic) soils. Oecologia 41: sity of Cape Town 123-126 WILLIAMS AJ 1981a. Why do penguins have long laying intervals? SMITH VR 1987a. The environment and biota of Marion Island. S. Ibis 123: 202-204 Afr. I. Sci. 83: 211-220 WILLIAMS AJ 1981b. The clutch size of Macaroni and Rockhopper SMITH VR 1987b. A computerized bibliography of Antarctic and Penguins. Emu 81: 87-90 sub-Antarctic biological research. S. Afr. 1 Sci. 83: 190 WILLIAMS AJ 1981c. The laying interval and incubation period of SMITH VR & LEWIS SMITH RI 1987. The biota and conservation Rockhopper and Macaroni Penguins. Ostrich 52: 226-229 status of sub-Antarctic islands. Environ. Intemat. 13: 95-104 ' WILLIAMS AJ 1984. The status and conservation of seabirds on some STEELE WK & CRAFFORD JE 1987. Insects in the diet of the Blue islands in the African sector of the Southern Ocean. Intemat. Coun­ Petrel Halobaena caerulea. Cormorant 15: 93-94 cil Bird. PreseIY. Tech. Pub]. 2: 627-635 STEELE WK & KLAGES NT 1986. Diet of the Blue Petrel at sub­ WILLIAMS AJ & BERRUTI A 1978. Mineral and energy contribu­ Antarctic Marion Island. S. Afr. J. Zool. 21: 253-256 tions of feathers moulted by penguins, gulls and cormorants to the S10NEHOUSE B 1960. The King Penguin (Aptenodytes patagoni­ Marion Island terrestrial ecosystem. S. Afr. 1 Antarct. Res. 8: 71-74 ca) of South Georgia: 1. Breeding behaviour and development. Falk­ WILLIAMS AJ & BURGER AE 1979. Aspects of the breeding biol­ land Isl. Dep. SuIY. Sci. Rep. 23: 1-81 ogy of the Imperial Cormorant, Phalacrocorax atriceps, at Marion STRIKWERDA TE, FULLER MR, SEEGAR WS, HOWEY PW Island. Gerfaut 69: 407-423 & BLACK HD 1986. Bird-borne satellite transmitter and location pro­ WILLIAMS AJ, BURGER AE & BERRUTI A 1978. Mineral and gram. Johns Hopkins APL Tech . Digest 7: 203-208 energy contributions of carcasses of selected species of seabirds to VAN AARDE RJ 1980. The diet and feeding behaviour of feral cats, the Marion Island terrestrial ecosystem. S. Afr. 1 Antarct. Res. 8: Felis catus at Marion Island. S. Afr. 1 Wildl. Res. 10: 123-128 53-59 VAN AARDE RJ 1984. Population biology and the control of feral WILLIAMS AJ & SIEGFRIED WR 1980. Foraging ranges of krill­ cats on Marion Island. Acta Zool. Fenn. 172: 107-110 eating penguins. Polar Rec. 125: 159-161 VAN RENSBURG PJJ 1985. The feeding ecology of a decreasing WILLIAMS AJ, SIEGFRIED WR, BURGER AE & BERRUTI A feral house cat, Felis catus, population at Marion Island. In: Antarc­ 1977. Body composition and energy metabolism of moulting eudyp­ tic Nutrient Cycles and food Webs, eds WR Siegfried, PR Condy tid penguins. Comp. Biochem. Physiol. 56A: 27-30 & RM Laws. Springer, Berlin. pp 620-624 WILLIAMS AJ, SIEGFRIED WR, BURGER AE & BERRUTI A VAN RENSBURG PJJ 1986. Control of the Marion Island cat (Felis 1979. The Prince Edward Islands: a sanctuary for seabirds in the catus) population: why and how. S. Afr. 1 Antarct. Res. 16: 110-112 Southern Ocean. Biol. ConseIYa. 15:59-71 VAN RENSBURG PJJ & BESTER MN 1988a. The effect of cat Fe­ WILLIAMS AJ, SIEGFRIED WR & COOPER J 1982. Egg com­ lis catus predation on three breeding Procellariidae species on Ma­ position and hatchling precocity in seabirds. Ibis 124: 456-470 rion Island. S. Afr. J. Zoo/. 23: 301-305 WILSON RP 1984. An improved stomach pump for penguins and VAN RENSBURG PJJ & BESTER MN 1988b. Experiments in feral other seabirds. 1 Field Orn. 55: 109-112 cat population reduction by hunting on Marion Island. S. Afr.1 Wildl. WILSON RP & BAIN CAR 1984. An inexpensive speed meter for Res. 18: 47-50 penguins at sea. 1 Wildl. Manage. 48: 1360-1364 VAN RENSBURG PJJ, SKINNER JD & VAN AARDE RJ 1987. Ef­ WILSON RP & WILSON M-P 1988. Dead reckoning: a new tech­ fects of feline panleucopaenia on the population characteristics of feral nique for determining penguin movements at sea. Meeresforsch. 32: cats on Marion Island. J. Appl. Ecol. 24: 63-73 155-158 VAN ZINDEREN BAKKER EM Jr 1971a. A behaviour analysis of the Gentoo Penguin (Pygoscelis papua Forster). In: Marion and Prince Edward Islands. Report on the Biological and Geological Expedi­ tion/1965-1966, eds EM van Zinderen Bakker Sr, JM Winterbottom, & RA Dyer. AA Balkema, Cape Town. pp 251-272 VAN ZINDEREN BAKKER EM Sr, WINTERBOTIOM JM & DYER RA eds 1971. Marion and Prince Edward Islands. Report on the South African Biological and Geolgoical Expedition/1965-1966. AA Balkema, Cape Town VERWOERD WJ, RUSSELL S & BERRUTI A 1981. 1980 volcanic eruption reported on Marion Island. Earth & Planetary Sci. Letters 54: 153-156 WAL10N DWH 1986. The biological basis for conservation ofSubant ­ arctic islands. Report of the Joint SCAR/IUCN Workshop at P-aim­ pont, France 12-14 September 1986. Scientific Committee on Antarctic Research/International Union for the Conservation of Nature and Natural Resources, Cambridge. 32 pp WATKINS BP 1987. Population siz.es of king, rockhopper and macaro­ ni penguins and wandering albatrosses at the Prince Edward Islands and Gough Island, 1951-1986. S. Afr. 1 Antarct. Res. 17: 155-162 WHI'ITOW GC 1983. Physiological ecology of incubation in tropi­ cal seabirds. Stud. Avian Biol. 8: 47-72 WIENS JA & SCOIT JM 1975. Model estimation of energy flow in Oregon coastal seabird populations. Condor 77: 439-452 WILLIAMS AJ 1978. Mineral and energy contributions of petrels (Procellariiformes) killed by cats, to the Marion Island terrestrial ecosystem. S. Afr. 1 Antarct. Res. 8: 49-54 WILLIAMS AJ 1980a. Geology and the distribution of Macaroni Pen­ guin colonies at Marion Island. Polar Res. 19: 279-281 WILLIAMS AJ 1980b. Aspects of the breeding biology of the Gen­ too Penguin, Pygoscelis papua. Gerfaut 70: 283-295 WILLIAMS AJ 1980c. Aspects of the breeding biology of the Sub­ Antarctic Skua at Marion Island. Ostrich 51: 160-167

S. Afr. J. Antarct. Res., Vol 20 No 2, 1990 57