DISTRIBUTION and LIFE HISTORY STUDIES of the SOUTH-POLAR SKUA by CARLR

Home , Brown skua, Chilean skua, East Base, Great skua, Holl Island, Skua

BIRD-BANDING A JOURNAL OF ORNITHOLOGICAL INVESTIGATION

VOL. XXXII OCTOBER,1961 No. 4

DISTRIBUTION AND LIFE HISTORY STUDIES OF THE SOUTH-POLAR SKUA By CARLR. EKLUND INTRODUCTION Skuasare in the Order Charadriiformes,and togetherwith jaegers of the genusStercorarius constitute the Family Stercorariidae.Taxon- omists disagree as to whether the skua comprisesa single bipolar speciesor whether it consistsof several speciesand sub-species.In this study four speciesare recognizedand all are classifiedin the bipolar genus Catharacta,with one speciesnormally inhabitifig the Northern Hemisphereand the other three the Southern. Theseinclude Catharactaskua, the great skuaof the NorthernHemisphere; 'C. chilen- sis t Saunders), the Chilean skua found generally in the temperate regions of South America; C. lonnbergi t Mathews), the brown skua which primarily inhabitsthe subantarcticareas; and C. maccormicki f Saunders},the south-polarskua which is confinedto Antarctica. This latter species,which is the subjectfor this study, has been commonly calledthe Antarcticskua as well as McCormick'sskua, but south-polar skua is the more common name and this term is followed.

ACKNOWLEDGMENTS: This study was used as basis for a doctoral disserta- tion at the University of Maryland. Funds were initially provided by •he United Sta,tes National Committee--IGY of the National Academy of Sciences, and sub- sequent banding work was done through a grant provided by the National Science Foundati,on. Without this aid the study could never have been carried ont. Banding records are being kept by the United States ,Fish and Wildlife Service and this is gratefully acknowledged.Special thanks are due many personsfrom the United States as well as ot,her countries who so ably cooperatedin this interna- tional program with their banding and observationwork. Whenever possible, an attempt has been made to give proper credit in the text for this work. The south-polarskua was first collectedduring Captain James Clark Ross'sexpedition in the Erebusand Terror to Victoria Land in 1839-43 (McCormick, 1884, p. 154). Saunders(1893) proposedthe name Stercorariusmaccormicki, and first describedit, over 50 years after it had beencollected, as a distinctspecies, on the basisof five specimens forwardedto the British Museumby Dr. McCormick. Thesehad been procuredat PossessionIsland, latitude 71 ø 14t S., longitude171 ø 15t W. The south-polarskua nests chiefly on lands within the Antarctic Circle, and on the basis of presentrecords the breeding range is generally confinedto the area southof the 65th parallel exceptin Palmer- land where it may extendnorthward to betweenlatitudes 62-64 ø S. This correspondsroughly to the 32øF isotherm ivan Rooy, 1957). Hamilton 11934,pp. 173-174) indicatesthat the speciesdoes not occur

187 ] 88 ] EKLUND,The South-Polar Skua Bird-BandingOctober

,40' 150' •ZO' •0' •OC' •O' SO' • ?O' 60' 50' 4•' PLATE Z

ANTARCTICA SOUTHPOLAR SKUA'. RANGE AND RECORDS

ß •_• ,•.

Pack- zce only part of the year and probable hmH of norma/winter range of SouthPolar Skuo ß t•ecord$ of South Polar Skua recoveries

north of latitude65 ø S. in Pahnerland.Gain (1914, p. 109) recorded it at the SouthShetland Islands, however, and foundit nestingas far north as Admiralty Bay, latitude 62ø 06' S. This record is further supportedby collectionsfrom the DancoCoast of Palmerland•Graham- land), latitude64 ø 30' S., longitude62 ø W. tMurphy,1936, p. 1010). Bagshawe(1939, p. 280) describesthe brown skua at Water Boat Point on the DancoCoast, latitude 64ø 48' S., longitude62 ø 43' W., but this appearsto be misidentified.He describesthe young as slate- colored. This being the case it must be the south-polarrather than the brown skua,since young of the latter are buff-colored. Four recordsof the occurrenceof the south-polarskua have been made in New Zealand at locations shown on Plate I. The first C. mac- cormickiresulted from a misidentifiedspecimen of C. lonnbergifrom Stewart Island •Falla, 1940). Falla recorded another occurrenceof C. maccormicki from the west coast of North Island, New Zealand, in 1940, and a third was recorded on the New Zealand coast in 1950. All these records were based on identification of dismembered birds. Fleming (1953) describesa fourth record of an exhaustedbird that came ashore on North Island. It was surmised that it had been blown into the Tasman Sea by a storm that originatedin the subantarctic winter range of the skua. Vol. XXXII 1961 EKLUND,The South-Polar Skua [ 189

Argentine naturalistsprocured a specimenof C. rnaccormickiat Laurie Island in the easternmostSouth Orkneys, latitude 60 ø 45' S., longitude44 ø 35t W., in November1904. Up to that time this was the most northern record of the species(Clark, 1906). A specimen collectedin the Sagami Sea off the coast of Japan was identified as C. maccormickiby Murphy (1936, p. 1007). Description Plumage. There is considerablevariation in the color plumageof south-polarskuas. The bird is muchpaler brownthan the brownskua, and it lacks the cinnamon or rufous-colored feathers so characteristic of most other forms. The ventral and dorsal surfacesare generally more light brown, with a somewhatstreaked appearance, becoming darker posteriorly. The head and neck have a light chamoisor straw- color tone. The darker brown of the wing feathers helps accentuate the large outstandingwhite patch on the primaries,which is so readily recognizedin flight. The feature that most clearly distinguishesC. maccormickifrom all othersis the downystage of the young. It is the only oneof the genus with young of a bluish or slate-graycolor. The nestlingsof all the other skuasare buff or rufous. Gain 11914, p. 115) maintainsthat this characteristicdistinguishes the south-polarfrom the brown skua of the subantarctic.In the courseof this studythe youngof the species have been observedat five differentregions of Antarcticaand all have been slate-colored.Buff-colored young have never been seenin these areas. Most of the fledgling down is shed by the seventhor eighth week, and the plumage at this stage is a uniform mouse-gray,which later becomesvery dark gray. Flight generallytakes place after about 9 weeksof age, during the latter half of February and the first half of March in the Windmill Islands. At this time the darker coloration of the youngbird makesit quite distinguishableas it flies with the adults. Body Measurements.Measurements taken by Friedmann 1,1945, p. 311) of south-polarskuas collected in Palmerlandshow an average wing dimensionfor 8 males of 373.6 mm., comparedwith an average of 380.1 for 10 females. Tail measurementsaveraged 151.9 for males as against150.1 for females. Tarsus measurementsaveraged 62.0 for males and 62.6 for females. The culmen, measured from the base, averaged 59.6 for males and 60.4 for females. The average of the middle toe without the claw was 54.2 for males and 56.8 for females. Sex differentiationthrough measurementsmay be indicated after a larger collectionof specimenshas been examined. Falla {1937, p. 248) examined 15 specimensof C. maccormicki and in the measurements found indicationsof a greatersize in the female,although this was not so markedas in C. lonnbergi. He felt that in any carefullysexed series of breedingadults the length of the wing might be a dimensionin which females would have a decided advantage. Friedmann's exam- inationsof C. maccormickiindicated an averagedifference of 6.5 mm. in wing measurementsbetween the sexes. Wilson (1907, p. 75) says there seemsto be a fairly constantthough slight advantagein point of size in the female over that of the male. Weight. Weights taken of 20 adults in January-February1941 at 190] EKLUND,The South-Polar Skua Bird-BandingOctober

the East Base in Palmerland (latitude 68 ø 12' S., longitude 67 ø 03' W.) averaged1,295 grams (Eklund, 1945). The average weight of 10 birds at the Hallett Station in January 1957 was 1,500; of 30 adults weighed at McMurdo Sound in the same month, 1,285; and of 50 adults taken at the Wilkes Station in February-March 1957, 1,390. The averageweight of these110 specimens,taken in thesefour different regionsin Antarctica,was 1,355 grams (2 lbs. 153/• oz.); maximum, 1,940; and minimum, 985. Body measurementsgiven by previouslycited authoritiesdo not in- clude weight. It would be interestingto obtain a truly representative sampleto determinewhether differencesexist here betweensexes, as in the caseof other polar-inhabitingspecies such as the giant petrel (Macronectesgiganteus) and the southernblack-backed or kelp gull (Larus dominicanus). Unfortunately,weights by sex of the above 110 specimenscould not be obtainedbecause all the birds, except5 of those taken at East Base, were banded and released for migration studies. Body Temperature. The south-polarskua seemsmore active in everythingit doesthan any other polar speciesobserved. The rapid physiologicor metabolicprocesses involved in such energy appear borne out by its high body temperatures.Temperature studies were carriedout in Palmerlandin 1939-41,and of 6 speciesthe skuashowed the highestaverage, as well as the maximum, of all recordingstaken (Eklund, 1942, p. 545). In the above cited study. 28 skuasshowed an average temperature of 106.14ø F., with a maximum of 108.40ø for a juvenile as ;veil as for an adult, and a minimum of 107.40ø for a juvenile and 103.40ø for an adult. This averagewas 2.84ø higher than that of the snow petrel(Pagodroma nivea); 1.11ø higherthan the southernblack-backed gull; 2.34ø higherthan the Adeliepenguin; and 0.51ø higherthan the antarctic tern (Sterna vittata). Eggs. A clutchusually consists of 2 eggs. The color varies slightly, but generallyis greenishin variousshades, with light-brownsplotches. Some eggs collectedin Palmerland in 1939-41 were more brownish than green,with dull, rather cloudedsplotches of brown. Many seen in the Windmill Islandsappeared much lighter greenthan others. From Cape Adare 26 collectionsmeasured between 65.5 and 78.7 mm. in length,and 47.7 and 52.0 in width (Oates,1901, p. 2261: and 8 eggscollected in AdelieLand (Etchecoparand Prevost,1054, p. 243) showedsomewhat greater uniformity, the maximum length, 73.7: the minimum, 70; and the mean, 71.9, respectively. The egg shellof the skuais quite fragile, comparedwith that of the Adelie penguin. Thicknesseswill undoubtedlyvary, but one measure- ment showed the skua shell to be 0.0232 inches, while that of the penguinwas 0.0325. Age. Little is yet knownof the longevityof the south-polarskua, but by bandingrecoveries over the years further data on this shouldbe provided. Recordsof the United StatesFish and Wildlife Serviceshow that an adult skua banded in March 1941 at our Palmerland East Base on StoningtonIsland in Marguerite Bay (USFWS 40-605604) was recoveredin the same area (Base E) in November 1949 (Sladen and Vol. XXXII •961 EKLt,ND, The South-Polar Skua [ 191

Tickell, 1958, p. 20• by Stonehouseof the British FalklandIslands DependenciesSurvey. This indicatesan age of at least 10 years. Banding A primary objectivein studyingthe movementsand migrationsof the specieswas to band or otherwisemark a representativenumber. and in suchmanner as to make the bandseasily recognizablein the fieldto avoidthe necessityof alwayshaving to captureor kill the birds. The ultimatesuccess of the studywas dependent upon continually ob- servingthe skuasand reportingon theseobservations. The bandingphase of the studywas begunin September1956 by forwardingletters to the participatingnations asking their cooperation. Sevencountries agreed to take part. Theseincluded Argentina, Aus- tralia, Japan, New Zealand, Norway, the United Kingdom, and the U.S.S.R. Belgiumparticipated starting in 1957-58, and banding was doneat a Chileanstation in 1960-61. Bandingwas carried out at the 16 areas listed in Table I. Banding kits were provided the cooperatingcountries, and each includeda map {Plate II), unnumberedcolored bands, colored vinyl- plastic adhesivetape, fishline, and record and instructionsheets. In addition. some kits contained United States Fish and Wildlife Service numberedaluminum bands, with pliers and instructionsfor use. The numberedbands were usedonly if the countrydoing the bandingdid not have a bird-bandingprogram or bandsof its own. As many personsas possiblewere alertedto the study,particularly thoseon the pan-antarctic islands and in countriesadjacent to Antarc- tica, so that observationsmight be made. Maps showinglocations of banding stationsand the color of bands used at each, togetherwith instructions and observation record sheets. were forwarded to the United Kingdomfor useat Tristan da Cunhaand GoughIslands. The French,in additionto forwardingthe informationto their Adelie Land Station. sent it also to the Amsterdam and Kerguelen Islands. New Zealand sent the informationto CampbellIsland, Australia to Mac- quarie Island. and South Africa to its Marian Island. To acquaintornithologists with the study,a brief descriptivearticle was publishedin variousornithological journals. Theseincluded The Bokmakierie, in South Africa (Eklund, 1957c•; the South Australian Ornithologist {1958b} and the Emu (1958c) in Australia; Notornis, in New Zealand i1958a); The Ring, in Great Britain (1957b} and, in addition, in a New Zealand news bulletin, •4ntarctic i1957a•.

Type o/ Bands Numbered aluminum-alloylock-on bands, a•2-inchinside diameter (size 7A}, suppliedby the United StatesFish and Wildlife Service. were usedat the United Statesstations. These bands were alsosupplied to Argentina,Norway, and Japan, althoughthe latter two countries were unableto band any skuas. Besidesa number,each band bore the inscription "Write-Avise Fish & Wildlife Service, Washington,D.C. USA." Stationsof the United Kingdom,Australia, New Zealand,and the U.S.S.R. used numberedbands employedin their own national bird-banding,or ringing, programs. 192] EKLUND,The South-Polar Skua Bircl-•Bana•ngOctoDer

o o o o o o o o •o o

• oooo Vol. XXXII 1961 E•rt UND, The South-Polar Skua [193

INDœXO__F $TAT[Ot•

PLATE 'IT ANTARCTIC AREA STATIONS

• s•c•tTO•

Besidesthe numberedmetal band, a coloredthermoplastic band made of rubberand styrenewas used. This wasthe first time sucha leg band had beentried, and it was developedfrom a suggestionby Rudersdorf and Van Deusen (1955). The rubber-styrenematerial is manufacturedin various thicknesses and colors,under the trade nameof Boltaron,by the Bolta Company, Lawrence, Massachusetts.The bands, 0.45 (1/22") thick, were made by cuttingpieces 1-5/8 by 3 inches. Thesewere then rolled insidea a/•-inch brass tube and heated at 315øF for 15 minutes. This molded the material in suchmanner that both endsoverlapped about halfway aroundthe band,to form an insidediameter of approximatelyY2 inch. The material shrinksslightly after heating,so the final band is about 1•2 incheslong. Three-quarterinch long bandswere usedfor immature birds. The band is appliedby springingthe endsapart, placingit around the tarsus,and letting it springback together. Once in place,there is 194] EKLUN.,The South-Polar Skua Bird-BandingOctober little chanceof its comingoff. As further insuranceagainst loss, however, a 3 to 4-inch strip of pressure-sensitive,adhesive, vinyl-plastic tape of the same color as the band was placed around it. This also helpedmaintain a constantdiameter. Subsequentobservations of bandedbirds over four breedingseasons failed to showthat any of the thermoplasticbands had comeoff. This could be checkedin many casesbecause birds seenwith metal bands only would indicate loss of the colored band, since all banding up through 1958-59 exceptfor someat McMurdo Soundand the Hallett Stationduring 1955-56,and that by the U.S.S.R. at their Mirny Station in 1956-57and the United KingdomAdmiralty Bay Stationin 1958-59 employedboth types. The manner in which the bandsstayed on the leg exceededthe most optimistichopes. The material, manufactured in many thicknessesand colors, holds much promise as an effective means of marking birds for field identification,either through use as a leg, neck,or wing marker. The colorsdo not fade, and in many casesthe thermoplasticmaterials outwore the metal band, as evidenced by thinnessof the metal or the metal band being missingfrom the leg. Six basic solid colors were used--yellow, red, green, blue, black, and white. These were chosen because there was less likelihood of human error in interpretation. Seventeendifferent colors or combinations were used, as listed in Table I. In addition, a plastic, orange-coloredring had been used previouslyat Australia's Mawson Station, and this was continued. Combinationswere obtainedby paintinghalf the band with a special vinyl-basepaint manufacturedby the InterchemicalCorporation of Newark, New Jersey. Only blue and yellow colorswere availablethe first year, so otherred and greenplastic paints were used.These were not so satisfactoryas the vinyl-basepaint. A gray thermoplasticband also was used in combinationwith red, yellow, and green paint, but the gray was employedat only three stations,since it might be too easily confusedwith white. A similar- colored vinyl-plasticadhesive tape was applied with the two-color combinationbands, although in somecases it was possibleto apply over only one of the colors. New Zealand'sScott Station and our country'sNaval Air Facility at McMurdo Sound used bands of the same color, becauseof their 2-mile proximity,as did Hallett Station,which is maintainedjointly by New Zealand and the United States. To provideeasier identification in subsequentyears of birds banded as juveniles,a 3/•-inchrather than the lX/•-inchcolored band was used. In the 1956-57 nestingseason at Hallett Station the a/•-inchband was solid red, and during the 1958-59 seasonit was red and white. No juvenileswere bandedhere in 1957-58. At Wilkes Station a green ,•/•-inchband was used in 1956-57, black and white in 1957-58, and greenand whitein 1959-60. Thermoplasticbands were not usedat any of the stations after the 1959-60 season. In the 1959-60 and 1960-61 seasonsa lX/•-inch green colored metal band was used at Wilkes for adults,and a similar red band was usedat Hallett Station. A yellow metalband was assigned the McMurdoSound Station at the beginning of the 1960-61 season. Vol. XXXII 1961 EKLUND,The South-Polar Skua [ 195

Methods o/ Capturing The south-polarskua is relativelyeasy to catch,and severalmethods were used. During the breedingseason, when the birds are ever active in defenseof their eggsand young,they can be readily taken with a hand net, and 156 were thus capturedat Wilkes Station in 1956-57 and 1957-58. The net hoop shouldbe 30 inchesor more across,with a handleup to 6 feet long. The frame shouldbe made of aluminumto make it light to swing. Snaringwith a fishlineis anothereasy method of capture,and over 200 were taken in this manner at Wilkes Station, and at McMurdo Soundand Hallerr Stationwhile our shipswere en route to Wilkes in 1956-57. Snaring was usually done at a dog-foc-dseal cache where skuascongregated to feed. The easiestmethod was to attach a 30 to 40-foot line to a long bamboopole, which was laid on the ground with the line run to the bait. The snare loop at the end was large enoughto permit the bird to enter well into it beforereaching the bait. The loop was most successfulwhen placed along the side of a seal with a knife slit throughthe skin so that the bird had to walk into the loop from one directionin order to get at the exposedportion of the meat or blubber. After a considerable number of birds were thus snared in the same area, the skuas became extremely cautious in steppinginto the loop. Snaringwas always most effective during snowy, stormy days. Another device, tried for the first time in Antarctica, was a cannon- projectilenet, a methoddeveloped for use on waterfowlby Dill and Thornsberry(1950•. Modificationsand improvementshave been made by the United StatesFish and Wildlife Service,as discussedby Salyer (1955) and by Crockford and Cannon (1956) of the Georgia State Game and Fish Commission. The net, measuring60 by 30 feet, is shot over the birds in a seal- baited area, with three cannon projectiles. Black powder, electric squibs,and a battery are usedfor firing. More recentdevelopments of the projectileemploy prepared cartridges. With this device,258 skuas and 11 giant petrelswere taken in 31 shots. The maximumper shot was 29 birds. Occasionally one or more of the cannons failed to detonatebecause of poor connections,or a projectile would break loose,making part of the net fall short. This resultedin the escape of over 75 birds. Juvenileswere easily taken at the nest by hand or with a net.

Marking Birds/or Field Identification To determinepossible reaction to color and to be better able to follow the daily feedingand other movements,some birds were dyed or sprayed with enamel. DuPont acid-color"milling yellow G. N. concentrate,"and "crocein scarlet N extra" dyes were used. These were preparedwith methyl alcoholand aceticacid in accordancewith the methodused by Winston (1954). At Hallett Station, 7 skuas were dyed yellow and a total of 182 adultswere either dyed or sprayedwith enamelduring the two seasons at WilkesStation. At thelatter station, 37 had beencolored scarlet by 196] EKLUND,The South-Polar Skua Bird-BandingOctober the end of the 1956-57 nesting season. In the following seasonat 6 different areas within the Windmill Islands, 66 were dyed various combinationsof yellow and scarlet,and 79 were sprayedwith enamel of various colorsto provide distinctiverecognition marks for each individual. Somebirds were dyed on the wings only, while otherswere colored on the entire body as well as on the wings. In doing the body coloring it was found that two birds died, possiblyfrom exposure,as do ducks when becoming soaked with oil. The acid-alcohol solution might have destroyedthe oil in the feathers,which in turn may have affected the bird's insulation. Thereafter only the white wing patcheswere dyed,but this still madethe birds easilydistinguishable in flight. Of the total birds dyed, 7 were recapturedthe same seasonwith somecolor still remaining,and 8 were retakenthe following year at Wilkes Station, but with no trace of the color. A dyed bird was the first recordedreturn away from the place of banding (USFWS 617- 005671. Banded and dyed scarlet on March 26, 1957, it was taken by scientistsaboard the Russianicebreaker OB on February 19, 1958, in the pack ice 120 miles westwardfrom Wilkes Station. Another (USFWS 617-00566), bandedand dyed scarletat the sametime and placeas the Russianrecovery, was retakenat McMurdoSound on January 3, 1959. One disadvantageof dyeingis that skuasoften bathe and thuswash off the dyes. One bird, dyed on November16, had faint traces of scarlet recognizablein flight on December26. During the secondnesting season, 79 adultswere sprayed with quick- drying Ducoenamel on one or both of the white wing patches.Yellow, red, blue, or green colorswere used accordingto the areas in which the birds were taken. This was by far the best method for quick identificationin flight, sincethe enameldid not affect flight and the color did not wash off, but lasted until the feathers were lost in the molt. It was also extremelyeasy to apply in the field from the spray can.

Birds Banded In March 1941 15 adult skuas (C. maccormicki) had been banded at the East Base in Palmerland,during the United StatesAntarctic Expedition. During the six nesting seasonsfrom 1955-61 there were 2,205 south-polar skuas reported banded at 5 United States and 11 foreign stations. Of this number 1,650 were adults and 555 were immature. Of these,19 adultsand 64 juvenilesbanded by Austin 11957,p. 13) at Hallett Stationand in the McMurdoSound area during the 1955-56 season were without colored bands. The 37 skuas banded by Syroechovski(Lapina, 1958) at the U.S.S.R. Mirny Station in 1956-57 were also without colored bands, as were the 16 banded by FIDS at Admiralty Bay. An additional 248 adults and 76 immature brown skuas(C. lonnbergi) were bandedat the United KingdomSigny Island Station. The reportedtotal of all skuasbanded to date is 2,544. The numbers, area banded, country, age, and bander or reporter are shown in Table II. Vol. XXXII •96• EKLU,D,The South-Polar Skua [ 197 198] EKLUND,The South-Polar Skua Bird-:BandingOctober

•o' •o no?o FPL•TE'nT WINDMILLISLANDS I WILKES LAND ANTARCTICA

ADELIEPENGUIN ANDSKUA NES'T1 N•GAREAS

BROWNING •-•

,•. 'SLAND

,',0

Li/e History Most of the detailed life history studiesof the south-polarskua were carried out in the Windmill Islands. These islands, within which Wilkes Station is located, extend 19 to 20 miles in a north-south line alongthe Knox-Buddcoasts in VincennesBay (Plate Ill). The group comprisesapproximately 65 to 70 rock outcroppings,of which five couldbe classedas peninsulasbut are now mappedas islands.These Vol.1961XXXII EKLUND,TheSouth-Polar Skua [199 includeClark, Bailey,Mitchell, and Browning"Islands," and Robinson Ridge. Someof the exposedrock areasextend as far as 3 to 4 miles westwardinto VincennesBay from the edgeof the polar icecap,while the north-southdimensions are up to 3 miles in length,as in the case of Mitchell. Including Frazier and ChappelIslets there are 60 to 65 remaining islands,islets, or smalleroutcroppings that have a long dimensionof at least500 feet. None are more than 2 mileslong. Ardery Island,the highest,rises to 371 feet abovesea level. The Windmillsare the only major exposedland area for at leasta 500-milesector of the coast. Skuas inhabit the islands within VincennesBay for a period of approximately6 months. The last seenduring the 1956-57 seasonwas on April 20, 19 daysafter the disappearanceof all Adeliepenguins. The first observednear Wilkes Station in the followingseason was on October19, 6 daysafter the appearanceof the first penguin,and 24 days beforethe first penguineggs were laid. Richard Penney,who was at Wilkes Stationas biologistfor a two-yearperiod, sightedthe first skua in the 1958-59 breedingseason on October17, and during the 1960-61 season he saw the first one on October 8. The last seen at the end of the 1958-59 and 1959-60 breeding seasonswere on April 26. There were 16 Adeliepenguin rookeries on the Windmill Islands during the 1957-58nesting season, in which approximately155,000 penguinswere breeding. South-polar skuas were foundnesting at all these,but the detailed studieswere carried out mainly at penguin rookeriesA and B on Clark Peninsula,•/• and 2 miles, respectively, from Wilkes Station.

Breeding-SiteTenacity The south-polarskua returns to the samearea on whichit nested the yearbefore. A skua (USFWS566-43502) banded during the 1955- 56 nestingseason by Austin (1957,p. 13) at CapeRoyds, McMurdo Sound,was recapturedas a nesterat the samearea on December24 of the followingseason by Honkalaand Long,Wilkes Station scientists. The day followinganother nester was capturedhere (USFWS 566- 43501) that had also been banded by Austin the previousseason. Thesewere the first known recordsof C. maccormickirecaptured the followingseason at the samenesting ground, and this indicatedthe speciesto be site tenacious. At CapeHallett, Austin banded 10 nestingadults and 52 juveniles during the 1955-56season. At the samearea the next season4 of these10 adultswere recapturedat their nests (USFWS 577-83065, -83064, -83023, and -83015). The last two birds were again retaken at the samearea for the third successivebreeding season in 1957-58 by RichardChappell, the Boy Scoutstationed at Little Americawho hadstopped over at HallettStation en routehome. He alsorecaptured three other nestingskuas that had been bandedthe previousseason at Hallett (USFWS 617-00132, -00123, and -00124). The last two werea matedpair that alsohad beencaptured as a nestingpair the yearprevious. Reid (1961) observed126 previously-bandedbreeders here during the 1960-61 season. 200] EKLUND,The South-Polar Stcua Bird-BandingOctober

At four United Kingdom Falkland Islands DependenciesSurvey basesthere were 639 skuas(C. maccormickiand C. lonnbergij marked with thermoplasticbands. Of this numberat least 107 were observed or recapturedduring the following summerat the baseswhere they had beenbanded. Twenty-nine were similarly observed two yearsafter- ward at the original place of banding (Scotland,19591. At Wilkes Station,277 adult skuaswere bandedduring the 1956-57 season,and the following year 69 t24.9 percent) of these were recaptured. These were then further distinguishedby placing a white strip of vinyl-plasticadhesive tape around the middle of the green band, to indicatethat they had been retaken. In addition to the recapturedbirds, at least 75 green-bandedskuas of the previousseason were seenat 16 separatenesting areas studied within the Windmill Islands,but could not be retaken. These,together with thoseactually recaptured, would indicatea known return of more than 50 percent of 1956-57 bandedbirds to their breedingareas the following season. On the basis of these returns, site-tenacitycan be considereda breedingtrait of the south-polarskua. No recordshave yet been obtainedof birds using different nestingareas in successive seasons. Site-tenacityis importantin the problemof speciation,since a species may consistof many populationswhich, thoughintermingling during the off-breeding season, seldom interbreed. Such isolation during breedingwould tend to lessenchances of geneticchanges spreading to other populations(Tinbergen, 1953, pp. 86-87).

,4ge at First Nesting In an attemptto determinethe age at which south-polarskuas first breed 172 immature birds were marked with a/•-inch colored thermoplasticbands at Wilkes Stationduring the 1956-57and 1957-58seasons. During thesesame two seasons,79 were similarly bandedat the U. S.- New Zealand Hallett Station and Austin banded 52 birds during the 1955-56 season. During the 1960-61 summer,Reid (1961} recaptureda 5-year-old nesting bird (577-83060) which Austin had banded as an immature bird in 1955-56. This is the first such record of C. maccormicki to indicate the age at which it first nests. It also indicatesthe breeding site tenacityfor birds hatchedin that area.

Territorialism South-polarskua nesting areas are governedby the presenceof Adelie penguinrookeries. Selectionof the territory within this area also appearsto be influencedby the availability of nestingpenguins, a primary sourceof food for both adultsand youngduring that period of the year. Nestswere found at 17 areas within the Windmills,and 16 (95 percent) of thesewere adjacent to the penguins.The seventeenth area wherenesting was observed was adjacent to nestsof the capepigeon (Daption capensis)and silver-grayfulmar (Priocella antarctica) on Ardery Island, and only one skua nest was seen •Plate III). Skuas were seennesting at all of 22 Adelie penguinrookeries visited in five differentregions of Antarcticain 1939-41and 1956-58. It is concluded Vol. XXXII 1961 EKLUNO,The South-Polar S•'ua [201

that the specieswill consistentlybe found nestingadjacent to Adelie penguin rookeries. Mating Three known nestingpairs of skuaswere banded at Wilkes Station during the 1956-57 season. In the 1957-58 seasontwo of these three pairswere foundnesting together again (USFWS 617-002139and -00290, and 617-00345 and -00347), and at the same areas where they had been banded. During the 19'58-59 nesting seasonTressler (1959), stationleader at Wilkes, observedone of thesepairs (617-00345 and -00347} at the exactterritory occupiedin 1957-58. This wasthe third successiveyear the pair had mated. In no instanceswere any birds of the other two pairs recapturedor observedwith a different mate. During the 1957-58 seasonat Wilkes Station, 17 additionalbreeding pairs were capturedwhile at the nest. Two hand or dip netswere used to catch them as they dived in defenseof the nest. All were banded with the green band and a numbered metal band. In addition, to keepthe pairs separate,each l•/•2-inch-wide green band wasdistinctively marked with a •/•-inch-widewhite, red, yellow, or blue vinyl-plastic adhesivetape, in variouscombinations, at the top, middle, or bottom of the band. All the above nests were marked with numbered stakes that showed also the color combination used on the band. Numbers were likewise sprayedin paint on nearbyrocks to simplifyrelocation of the territories in subsequentyears. Thesenest siteswere .checked.during the 19513-59 seasonby Tressler(19591 and by Penney (1961) during the 1959-60 and 1960-61 seasonsto determinewhether the pair-bondrelationship was maintainedas well as whetherthe pairs returnedto the samenest sites or territories. Tresslerobserved that 13 of the 19 pairs of recordin 1957-58were matedin 1958-59,and occupiedthe sameterritory as previously.Of the remainingmarked pairs 1 of each of the 4 pairs was back at the sameterritory occupiedthe previousyear, but with a differentmate. In no instancewas the otherof the originalpair observedon the nesting ground. This wouldindicate that the pair bondmight have been broken becauseof mortality to one of the birds, rather than to the pair split- ting up to matewith anotherbird. A skuarecovered dead in Australia (USFWS 617-00501), as discussedlater, had nested at Rookery B during 1957-58,but it wasnot one of the markednesting pairs. Penney found sevendead Wilkes bandedadult skuasduring the 1959-60 season. During the 1959-60 seasonPenney observed 6 of the 19 pairs of record back at their original nestingterritory, and in 1960-61 4 of thesepairs were again back at the sitesfor the fourthconsecutive year. Of the remaining15 pairs at the end of the 1960-61season, one of sevenpairs nested with a differentmate on the .sameterritory, one pair waskilled at the end of the fourth nestingseason, and two otherswere found dead. Recordsof all markedpairs are shownin Table III. Continuedobservation of thesemarked pairs will providefurther informationon how long skuasremain mated. Pairs seemto be to- getherearly as well as late in the season,and it is possiblethey remain 202] E•,,-,-D,The So•th-Polar S/,•ua Bird-BandingOctober Vol. XXXII 1901 EKLUND,The South-Polar Skua [203 togetherthroughout the winter. Fortunatelyobservation on the pair bondwill be possibleat WilkesStation over a longperiod because the AustralianGovernment took over operationof the stationfrom the UnitedStates at the IGY end in January1959. At Hallett Station,Reid (1961) recaptureda nestingpair (617- 00125 and -00126) duringthe 1960-61season which I had bandedas matesat this area in 1956-57. This couldrepresent at leastthe fifth nesting.season for this pair. No informationwas obtained as to whichof thepair selectsthe nesting territory. In determiningincubating egg temperatureswith an electronictelemeter, as discussedlater, one of the pair, the female, waskilled at the end of the studyto determinethe sex. The following nestingseason the matereoccupied the sametrritory with anotherbird. On the basisof thesebanding records it can be concludedthat the south-polarskua has a pair-bondrelationship and will reinatethe followingyear. Stonehouse(1956, p. 14) foundin Palmerlandthat some pairs of the brown skua. (C. lonnbergi)remated the followingyear and others did not.

Copulation Copulationmay take placeaway from the nestingterritory as well as on it. Two skuaswere seen copulatingon November17 while groupedwith otherskuas on the seaice near a penguinrookery where the pair might eventuallyhave nested. Anotherpair was observed copulatingwithin its nestingterritory. The actionis not unlike that of the Adeliepenguin, wherein the male standsposteriorly on the back of the female. On one occasion the male made contact for 10 to 15 seconds,covering the femalethree separate times in this manner,with all actiontaking place within 6 minutes.

Nesting On the basisof a surveyin which all land areaswere visited in the Windmillsduring the nestingseason, it is believedthat 95 percentor more of all skua nestswill be confinedto exposedlands on which Adeliepenguins are nesting(Plate III). A primaryreason for selecting theseareas is that penguinrookeries are a ready sourceof food. Only one skua nestwas found wherethere were no penguins.This was on Ardery Island, on which 1,500 to 2,000 silver-grayfulmers and cape pigeonswere nestingand from which the skuasobtained their food, basedon observationof empty eggsof thesespecies around the nest. The skua-nestingareas within the Windmill Islands are shown on Plate III. Four pairs of skuaswere found nestingnear a smallAdelie penguin rookery on one of the six BalaenaIslets, 25 miles northeastof Wilkes Station. Theseislets are not consideredin the Windmill Group. A detailedstudy was made of forty skua nestsat Adelie penguin rookeriesA and B on Clark Peninsula. Observationswere begun early in November, and each nest was visited at least once every 24 hours from the time it was found until the recordwas completed,either on destructionof the nestor terminationof the study. As discussedunder "Mating,"nesting pairs werecaptured and eachmarked conspicuously Bird-Banding 204] EKLUND,The South-Polar Skua October onthe green band with different colored tape. All nestingbirds captured were sprayedon the white wing patcheswith yellow,red, or green enamel.The studyterminated on January26, whena changeoverof personnelwas effected.

Nest Building Nestsare usuallymerely scraped or scoopedout of fine pebblesor gravelon a rockyknoll or mound.Fifteen of thesescoop-outs, hollows, or cupsmeasured 8 to 10 inchesin diameter.and up to 11/2to 2 inches in depth.Sometimes a rim of gravelraises the nestslightly above ground,but usuallythe hollow is belowground-surface level. Thereis no liningof any sort,and in thisrespect it differsfrom the brownskua in SouthGeorgia, which lines its nestwith grass(Stonehouse, 1956, p. 15). Oftentwo or morescoop-outs are frequent within each territory. The habit of makingthese is somewhatsimilar to the selectionof nestsites by herring gulls,as describedby Tinbergen(1953, pp. 127-129), whereinboth matesscrape several cups in the soil of the territory selectedprior to final selectionby the female.It wasnot possibleto determinewhich bird, or whetherboth, took the initiative in preparing scoop-outsand whichmade the final selection.

Egg Laying The first clutch of eggsduring the 1957-58 seasonwas found on November21 at Midgley Island "West." Skuasnear Wilkes Station, where the detailednesting studies were conducted,started occupying territoriesduring the secondweek in November,and the first egg was found on November23. A total of 72 eggs was laid in the 40 nests understudy. Of thesenests, 80 percenthad 2 eggs,while 8 (20 percent) containedonly 1. During the 1960-61season Penney found the first egg near Wilkes Stationon November26.

Incubation and Hatch Murphy •1936, p. 1019) statesthat the skuaincubates its eggsabout 4 weeks,while Ponting (1921, p. 217) saysthe period is 3 weeks. Wilson (1907, p. 71) saysthe period of incubationis 4 weeksfor C. maccormicki.Stonehouse (1956, p. 17) in three instancesfound the incubationperiod of C. lonnbergito be 32 days. In the Windmill Islandsthe first hatchingnoted during the 1957-58 seasonwas on December22, and the last on January 19. The peak period, when68 percentof the eggshatched, was from December26 to 31. Penneyobserved the first skuachick at Wilkes on December28 during the 1959-60 season. It was impossibleto determinethe precisehatching period of eggs found after the bird had begun incubation. However, 26 eggs were locatedwithin the first day of laying, and observationswere made on all theseonce every 24 hoursuntil hatching. It took 30 days of incubation for 15 (58 percent)of theseto hatch; 9 (35 percent)hatched in 29 days; I in 31 days; and I in 32 days. Pipping observation.s made on 40 eggsdisclosed that 27 pippedfor Vol. XXXII 1961 EKLUND,The South-Polar Skua [205

2 days before hatching, 12 required only 1 day, while 1 took 3 days from the first cracking of the shell until the chick emerged. Of the total 72 eggswithin the 40 nestsstudied, 28 (39 percent) were either destroyedor unhatched,while 44 (61 percent) were successfully hatched. More than 90 percent of the unhatchedeggs contained embryos,and in many of mere pippinghad begun. Both parents take part in incubation. During a 9-day continuous temperaturerecording of incubatingeggs, using an electronictelemeter, as discussedlater, one of the pair under studywas sprayedwith green enamelto provide easy differentiationof both birds on the nest. A continuous24-hour daily watch was made of the pair for 132 hours and45 minutes(51/._, days). After the temperaturestudy was concluded and the secondegg in the nest had hatched,one of the parent birds was collected to determine sex. During the 5•Jz days of continuousobservations, one or the other of the pair relievedthe mate on the nest 13 times, an averageof 2.36 times every 24 hours. The female (the bird collected) averaged7 hours and 33 minutessitting on the nestbefore being relieved, while the mate averaged2 hours and 9 minutes. The longesttime the femalewas on the nest was 15 hours and 22 minutes and the shortest 1 hour and 55 minutes;for the other, the long period was 7 hours and 16 minutes and the short, 9 minutes. There is little ceremonyduring the incubationchange. Usuallythe settingbird will fly from the nest to a nearby rock where the mate is roosting. The pair will then stand close together for about 30 secondsduring which both will raise the wings upward and call. hnmediatelyfollowing this the relief bird goesto the nestand resumes the incubating.

Temperatureof IncubatingEggs

Temperaturesof incubating eggs were obtained through use of an electronictemperature telemeter placed inside a skua egg. Another was placed in a penguin egg and simultaneousrecordings from both nestswere taken over a 9-day period from December11 to 19, 1957. Thisstudy has been previously reported by Eklundand Charlton (1958). In operationthe telemeteregg radiatesa modulatedelectromagnetic signalor pulse,the frequencyof modulationbeing determinedby the temperatureinside the egg. The signal is interceptedby the loop antenna,relayed by the transmissionline to the low-frequencyradio receiverin a nearby shelter,where it is amplified,detected, and sent to a loudspeakerand to a decadecounter, which recordsthe number of pulses. The pulseswere calibrated against a standardthermometer beforeand after the study. Althoughsignals were continuousa count wastaken for eachbird only onceevery 15 minuteson a 24-hourbasis, alternatelyon the skuaand the penguin. Resultsof the studyshow that the averagetemperature of the incubat. ing skuaegg over a 58-hourperiod of continuousrecord was 96.6øF., with a maximumof 103.5ø and a minimum of 87 ø. The averagebody temperatureof this speciesis 106.1øF., ranging from a maximum of 108.4ø to a lninimumof 103.4ø (Eklund, 1942). The averagetempera- 206] EKLUND,The South-Polar Skua Bird-BandingOctober ture of the incubatingegg was thus 9.5øF. lessthan the averagebody temperature.The averagetemperature of the incubatingAdelie penguin egg is 92.7øF., with a maximum of 98.2ø and a minimum of 84.5 ø. The greaterrange in the incubatingtemperature of the skua egg is reflectedby the numberof timesthe parentbird will changeincubation duties with its mate (2.36 times every 24 hours as discussedunder Incubationand Hatch), as comparedwith the long periodsthe penguin may sit withoutbeing relievedby its mate (usuallyat least 2 weeks continuouslyby the male at the beginningof incubation). Care o/ Eggs and Young To learn moreabout care of a nest,one of two skuaeggs was removed and placeda foot away. The bird returnedwithin two minutesand resumedincubation on the remainingegg. Both eggs,however, were in the nest 1 hour and 25 minutes later. On anotheroccasion a third egg was addedto a nest of two. The setting bird returned to the nest, looked at the eggs, and turned around and walkedto its mate. After 4 minutesit again returnedand startedincubating the threeeggs. Only oneof theseeventually hatched, but incubationof the unhatchedeggs took placefor 32 days. In anotherinstance an Adelie penguinegg was substitutedfor one of two skuaeggs. The parentbird on returningto the nestset on both eggswithout hesitation. After the skuaegg hatchedthe foster parents continuedincubation of the penguinegg for 2 more days,after which it was abandoned.Three dayslater it was eaten. The skuashad been on the penguinegg a total of 31 days. The parent skua will cover its chickswith the wings at times, but usuallyone or both fledglingsare left exposed.One parentis generally very near the youngwhile the other, if not off feeding,is within the territory and readyto drive off any maraudingbirds. A strangeskua has only to fly over the territory to be pursuedby a fast-flyingparent. To determinethe reaction of parentgirds to chicksother than their own, a 29-day-oldnatural-colored bird was taken from a nest, which alsocontained a red-dyedchick. This was placedwith a nearbypair, whichhad one 35-day-oldnatural-colored fledgling. The parentscared for both. On the next day both chickswere movedand placedin the nestfrom which the one had been taken originally. At the sametime the dyed chick was removedfrom the first nest and subsitutedfor the two which were taken from the second nest. The red chickwas acceptedby the fosterparents and was beingwell caredfor at the time the studywas concludedlate in January. Of the two chickstogether, one was dead the day followingtheir placement in the nestfroin whichthe dyedone had beenremoved. It had not been eaten. This experimentwould suggestthat skuaswill care for chicks other than their own, and that color made no difference. Dejenseo/Eggs and Young While incubating,the parentbird will usuallystay on the nestuntil approachedto within 20 to 50 feet. Then with wingsoften upraised it will walk stifflyaway while setting up a loud ha ha ha ha call. At rare timesa bird on the nestcan be nudgedwith the foot beforeit will leave. After flushingfrom the nestthe parentbird, togetherwith the mate, Vol. XXXII 1961 EKLUND,The South-Polar Sl,'ua [207 start "dive bombing"the intruder. This generallyinvolves a wide circle followedby a straight-awayapproach from about 100 feet. The start may be 10 to 50 feet abovethe ground. The parentsusually take turns, andthey may come from any angle.Often they approach within 1 to 5 feet of the headbefore turning off, and at this point they can be taken with a wide hand net, with certain agility on the part of the netter. In this manner 156 adults were caught at McMurdo Sound and the Hallettand Wilkes Stations. It is oneof thebest of Antarcticsports. Skuasdo not makecommunal attacks, and diving is carried out only by the parentswhen an intrudercomes within their territory. The birds have dived at me severalthousand times, and while the majority of timesthey will turn away without striking,there have beenmany occa- sionswhen they hit the head. Onepair, on the eighthday of incubating their eggs,struck 7 timeswithin 4 minuteswhile an attemptwas being made to count and record penguins.Three of the blows were with suchforce as to result in a splittingheadache. Another time my eye washit so hard it wasimpossible to seefor severalminutes. The skua has few equalsamong birds as a defenderof the home. Holding a stick over the head will ward off the birds, but this is difficultto do while climbingover rocks requiringuse of the hands, or whiletrying to recordpenguin counts, as was done every 48 hoursat one of the rookeries. Finally in "self defense"a long giant petrel primary wing featherwas sewnon my cap. The skuasthen contented themselveswith hitting the feather. The striking is usuallydone with the front of the feet as the bird "lowersits flaps"in brakingitself just beforereaching the person. This is generallya soft blow, thoughless frequently,they may strike with the front of the wing, and this is a much harder blow. Survival o/ Young Of the 44 chickshatched from the 40 nestsstudied, 34 (77 percent) survivedto the last week in January,when the studywas concluded. This representsa 47 percentproduction of chicksfrom the 72 eggs understudy. If one or both juvenilessurvived to the end of the study the nest was considered successful. There were 28 such successfulcases, which representsa nesting successof 70 percent, or mortality of 30 percent. It is interestingthat of 23 successfulnests from which eggswere not removedby man, there were only 6, or about one-fourth,in which both juveniles survivedto near flying stage. This number would probablyhave beengreater, however, if coloringhad not resultedin someincreased mortality. Although a rather heavy mortality among secondjuveniles is indicatedby this study,it doesnot appearso serious as that claimedby Wilson (1907, pp. 71-72). He saysthat becauseof thetendency of theyoung to wanderand get separated, it is a noticeable fact,and was "proved conclusively," that thoughtwo youngare almost alwayshatched, one of the two is invariablymissing after a few days or a week. While it was impossibleto followthe youngthrough to flight stage duringthe 1957-58season, this latter phasewas observedduring the 1956-57nesting season in the Windmills. Two knownpairs of young that had almostreached flight stagewere banded on February7 and 11. 208] EKLUNq),TheSouth-Polar S•ua Bird-BandingOctober

Othernestling pairs were probably also banded, but unfortunatelythese were not alwaysrecorded as pairs. Of the total of 40 nests,14 had either one or both eggsor chicks preyedupon by their own species.In conductingthis study 10 addi- tionalnests were interfered with in wholeor in part by man,or by both manand skuas, either through permanent removal or the shiftingof one or both eggs. Efject o/Color on Survival Wilson (1907, p. 70), in speakingof the skua'seggs and young says,"the bird makesno attemptat concealment,though its color as well as the colorof its eggsmight be consideredto be speciallyadapted to the nature of the ground, but with such habits as the skua's there is obviouslynothing in it." Data obtainedduring the presentstudies indicatethat protectivecoloration may be a factor in survival of the young. In an effort to learn more of the effect of color on survival, 12 of the 44 chickshatched were dyedeither red or purple. Thesewere dyed in the shellafter the methoddevised by Evans •1951), with the excep- tion that dye wasnot injectedinto the egg until it first beganpipping. This providedabout 2 daysfor the dye to permeatethroughout the egg. No loss of chicksin the eggsresulted from the coloring. and it lasted until the down molted after 7 to 8 weeks. No chickswere dyed in a nestthat containedonly one egg, and only one of the two eggsin a nestwas ever colored. Only six 150 percent) of the total dyed surviveduntil the end of the study. Of the 32 natural or slate-coloredchicks 28 (87 percent) survived. While this is hardly a sufficientsample, it does suggestthat the slate-graycolor. which generallyblends well with the rocky habitat, is a factor in protecting chicksfrom predationby adult skuas.

Renesting Skuas will renest if incubation has not started. On November 28 a pair wasobserved on a territoryhaving several scoop-outs. The nextday one egg had been laid, and on the following day it was gone. Since both birds were recognizableby the banding,a watch was kept on the territory. On December2 anotheregg was laid within 35 feet of the original nest. This was followedby a secondegg on the fifth day. Both eggswere successfullyhatched, although neither chick survived. This was the only record obtainedof renesting. No renestingoccurred with pairs that had startedincubation. only to have one or both eggsstolen. There were sevensuch records. Nests containing two eggs, one of which was removed, did not cause abandonment. One pair {USFWS 617-00345and -00347), which was the first re- matedpair ever recorded,incubated an egg for 12 days.starting on November28, 1957. On the twelfth day the egg was cracked,and the next day it had beeneaten. Renestingdid not take place.but the pair was still remainingon the territory at the time checkingwas stopped on January20. Thispair alsoremated on the sameterritory in 1958-59 and 19.59-60. vo•.1961XXXlI E}cLt-xu,The Sottth-Polar S/,'tta [209

ClitnaticFactors A•ecting Nesting There was no evidence that weather conditions, such as snow or wind,affected nesting of the skuain the WindmillIslands. Heavy snows duringthe nestingseason cause mortality among adult Addie penguins throughthe smothering of theincubating bird by snow,and the freezing of eggsand chicks. The samecould happen to the skua,but no serious snowstormsoccurred and there was no opportunityto observepossible nestloss from this causeduring the two nestingseasons studied. The winds were strongduring this time.

Feeding The first seasonalappearance of the skuain the WindmillIslands is aroundmid-October, about a week after arrival of the first penguins. This coincideswith puppingtime for the Weddellseals which appears to be a factorin theearly attraction of skuasto thearea, since the birds feedon sealplacenta. At this time they alsoutilize marine life, such as fish and red krill (Euphasiaj. Three to four weeksafter the skua arrives, or about the secondweek in November, the Adelie penguins beginlaying, and from this time until mid-December, when their chicks starthatching, the skuas will feedin largepart on the eggs. In stealingeggs a skuawill wait at the edgeof a penguingroup until a sittingbird can be drawn off its nest. Then it, or anotherskua, quickly dartsin and makesoff with an eggin its bill. Oddlyenough, penguins neverseem too disturbedat the nearbypresence of skuas.When one getstoo closeto the nest,the sittingbird, or others,merely peck at it to drive it away. Often the skuais actuallychased by the penguin. Sometimesa bird will fly amonga group and standthere waiting a chanceto takean egg. Muchof thetime they are unsuccessfulin their thieverv. Duringhigh windsa skuapair will hovera foot or two overa group of nestingpenguins, with wings outstretched against the wind. Mean- whilethe penguins stand up andpeck at theskuas above them. Finally onewill be drawnoff its nestlong enoughfor one of the pair to seize an eggand fly off. It seemslike a goodexample of teamworkon the part of the predators. During the temperature-telemeterexperiment one of the pair of skuasnot incubatingtook an abandonedegg from a nearbypenguin group.It flewwith the eggin its beakto within30 feetof its mateat the nest. Then for 11 minutesit tried to break the egg with over 30 hammerlikepecks, in whichthe head was raised high and brought down withgreat force on the egg. Finally it gaveup andflew away. Inspec- tion showedthe egg ratherheavily coated with mud, and therefore probablyharder to break.After 10 minutesthe bird returned to peck at the eggagain but finallygave up oncemore after 4 minutesof trying.An eggbeing incubated in a penguinnest was then placed on a nearbyrock. The same skua immediately flew to thisegg, picked it upin thebeak, and flew to within75 feetof its nest.Its matethen leftthe nest, flew to theegg, broke it, andthey took turns eating from it. Youngpenguins are a primaryfood source of the skuaand this supplyis doubtlessthe motivatingfactor in locationof skuanests. Sladen11958, p. 66'i presentsan excellentdiscussion of skuas feeding 210] EKLUN•,The South-Polar Skua Bird-BandingOctober on youngpenguins, and it appearsthere is selectivityin that the skuas tend to weedout the weaklingsor thosethat do not defendthemselves. Penguinchicks and eggs,however. are not the solesource of skua food at this time. Food disgorgedby adultsfor feedingtheir youngrevealed considerablequantities of fish and red krill obtainedfrom the seas. Thoughadults will eat youngskuas, ! have never seena parent kill and eat its own live chick. During the nestingstudy there were several instanceswhen the parent bird apparentlyate its own deadchick. This however,is no differentfrom that of many other animalsthat devour their own dead young. The parent bird doesnot hear the chick call or seeit move,and this probablymakes the chick entirelyforeign to them and thereforeits only significanceis as food. The skua'spropensity for eatingits own kind, while recognized,may be slightlyexaggerated. On April 1 a deadimmature bird, whichhad almostreached flight stage, was found with a brokenwing. It had lain exposedat the edgeof a penguinrookery for severalweeks, judging from its condition,and manyskuas could have seen it. In spiteof this the adult birds had not touchedit for food, whereasthey had eaten manynearby penguin chicks. Skuas are able to swallow large chunks of meat. To test this, hamburgerwas rolled into meat balls and frozen. These were then fed to a captivebird. The bird was unableto swallowa 2-inch-diameter ball. Neithercould it quite eat a 13•-inchpiece until after it had torn off a few small bits. In its efforts to swallow this it would take the ball in the bill, raisethe headstraight upward, and by a seriesof up and down jerks of the head and neck it eventuallydowned the meat. A 15/s-inch-diameterball was swallowed whole after 6 minutes of effort. Usuallya skuawill allowonly its mateto feednear it. When baiting the 60 by 30-footcannon net, it wasseldom possible to draw morethan 30 skuaswithin the baitedarea of the drop,because each bird seemed to require its own feedingspace. A pair of skuasonce kept all other birds from feedingin the baited area for more than an hour. This is usuallydone with upraisedwings and loud calls,often followedby an attack. In fall, followingthe disappearanceof young penguinsas a source of food, the skuasfeed on marine life. No records are available on their food habitsduring the winter, whenthey apparentlyare in the pack ice to the northward. Probablythey then feed on marine life exclusively. Skuasfeed their youngby disgorgingfood on the ground,where chickspick it up. When juvenilesare handledduring bandingthey occasionallyregurgitate their food. Of 16 suchrecords, 6 disgorged food consistingof penguinmeat, 9 of fish or red krill, and ! of the headof a skuanestling. This wouldindicate that parentsobtain a good part of the food for their youngfrom the sea. Judgingfrom the condition of the food, whereinsome of the fish and red krill appeared almostwhole, it doesnot seemlikely it was obtainedby the parent skuasfrom the stomachsof youngpenguins that theyhad eaten. Such food wouldbe considerablyground up, sinceit wouldfirst have been eatenby the penguinparent, following which it wouldha•'e been re- gurgitatedinto the nestlingwhere it wouldundergo further digestion. Vol.1901XXXII EKLUND,TheSmith-Polar Skua [211

Then it would had to have been swallowedby the parent skuas, regurgitatedon the ground,and finally eatenand again regurgitated by the young. While at Hallett Station an attempt was made to determinethe distanceat whichnesting skuas would range daily to find food. Seven birdscaptured near the shipwere dyed yellow, and two of thesewere later seenon their nests41/2 miles away. An adult capturedat Shirley Islet, in the Windmills, on December4, was retaken on January 11 at its nest on Holl Island, 81/:2 miles distant. Since it was defendingits 2 to 3-week-oldyoung at the time, it musthave been nesting when first captured.This wasthe longestdistance recorded; probably the daily range during the nestingseason is even greater.

Drinking Skuasare often seento eat snowor drink at freshwaterponds, and not uncommonlyto swallowsmall pieces of ice. Onceduring a bad windstorm,drifting snowfilled a large box in which two dyed skuas werebeing kept for feedingand molting studies. Both birds were buried alive; one died and the other was removedto a building where it re- mainedfor 36 hours. During this day and a half it ate chickenlivers and a half poundof partlyfrozen seal meat. No waterwas provided, althoughthe meatcontained some moisture. On returningthe skuato the box outdoorsit immediatelyate three chunks of broiledsteak. After this it startedeating powdered snow by dippingthe bill into it, each time raisingthe head to swallowthe snow. This was done 174 times within 25 minutes.

Senseof Smell Gain (1914) was of the opinionthat a skua,seen by Scottand his partyon the polaricecap at latitude80 ø 20' S. duringtheir ill-fated trek to the SouthPole in 1911-12,had been attractedthere by wind- borne odors from slain sledgedogs. Wilson (1907, pp. 67-68) says that the skua's sense of smell "must be a little short of marvellous." He citestheir surpriseat seeingone at their camp 170 milesfrom open water and 150 milesfrom the nearestspot known to be frequentedby skuas.Then they recalledthat they had killed a sealthe night before, that the wind was southerly,and that therefore"nothing but the scent of bloodcould have broughtthe bird thosemany miles." With all due respectto Gain and Wilson,I believetheir viewsare erroneous.That skuasare observedby trail parties at long distances from any naturalfood source is a fact, but I do not believethat they are guidedthere by smell. My experiencewas somewhatsimilar to Wilson'sregarding skuas travelinglong distancesfor food. On a 1,264-miledog sledgingtrip madewith CaptainFinn Ronneduring the 1939-41United States AntarcticExpedition (Ronne, 1945), we observedtwo skuasat a point about100 miles from open water at thesouthern end of KingGeorge VI Sound(Eklund, 1945, pp. 302-303). They werethe first and last seen after leavingthe edgeof the opensea ice. Like Wilson,the night previouslywe had killed one of ourdogs and the meat was exposed. Therewas very little wind at the time, however,and in my opinion 212 ] EKLUND,The South-Polar Skua Bird-BandingOctober

the birds locatedthe meat by their uncannysight. Our trail could easily have been followeda long distanceby the many visible dog droppings. Studies on the sense of smell in birds have been limited and incon- clusive. Romer (1955, p. 493) saysthe senseof smellin birds is rather feeble and appearsto play but a small role in sensoryreception. Tinbergen (1953. p. 21) found no indication whatsoeverthat the herring gull is guided by or has a reactionto smell. Walter presents a good reviewof the literatureon the senseof smellin birds, starting with the early experimentsof Audubonand Darwin. He concluded •1943, p. 69t that althoughbirds have a fairly well developedolfactory organthey still did not possessthe faculty of smelling. From a blind usedwhile bandingat Wilkes Station! often saw skuas locatesmall pieces of frozenseal meat in windsup to 35 knots. A bird would hover over the meat at heightsup to 30 feet and either drop down to it or circle aroundand comein low and upwind. With such wind velocitiesblowing meat odorslow and parallel with the ground it wouldseem impossible that smellscould reach up to the heightsat whichthe bird wasflying when it first locatedthe food. On five occasions,fresh sealmeat was sealedin a clear, plasticbag from whichno odor couldescape, and placedon the ground. In all casesthe skuasripped the plastic open within a minute after first sightingit. This, however,merely demonstrated that the birds located the food by sightrather than by smell,and doesnot provethey could not have found the meat by scent. On six occasions,however, when fresh meat was coveredwith a •/•2-inchlayer of powderedsnow, it was neverlocated. On the basisof studiesby Ishida and Shimizu (1958), it was determinedthat light snow,such as was usedto cover the meat, has a low densityas well as a low airflowresistance. This shouldthen permitodors to pass.This is furthersubstantiated by diffusiontests conductedbv Bender •1959) in which, using H.2Sgas as a tracer, he pickedup tracesof gasin 2 hoursthrough 6 feet of snow. Fresh chickenlivers were buried in a bottomlessbox containinga captiveskua. Thesewere sprinkled with light snowsufficiently to con- ceal them but through which odors could pass. The bird paid no attentionto the buried meat. In fact, it sat down on the slight mound covering the food. When liver was thrown there, the bird ate it immediately. Homing Instinct A skua taken at McMurdo Sound was flown the 780 miles to the South Pole Station, bandedthere with a white thermoplasticand a metal band, and releasedon November30, 1957. The bird stayedat the stationfor 3 days, after which it was not seenagain. Three more were flown there from McMurdo Sound on December 3, 1957, where Remington,station glaciologist,banded and releasedthem the same day. They took off immediatelytowards the BeardmoreGlacier, from which directionthey had beenflown. None have beenseen as yet. Stoner Haven, biologist at the McMurdo Station, and Evan Young. New Zealandbiologist at Scott Station,took a bandedbreeding bird from each of two nestsat Cape Royds on January 3, 1960, and had themflown approximately 350 milessouthward to the NAAF Beardmore Vol.1961XXXII EKLUNt•,theSottth-Polar St•'ua [213

Glacier summer base and released. One of these birds was observed back at its nest on January 14. A paired skua (USFWS 617-00781), with one youngin its nestnear WilkesStation, was sprayed with yellowenamel on both wings,taken 51 milesinland to the IcecapStation on December2, 1957,and rdeased at 2000 hours. On December 4 at 1045 hours it was observed back at its nest with the mate. Populations No systematicpopulation study had previouslybeen conducted over an extensiveskua-nesting area. Visual estimateshave been made of singlenesting grounds, but if many birds occupythe area the counts are subjectto considerableerror. Capturingskuas with a cannonnet, as discussed previously, presented an excellentopportunity to obtain systematiccensus data within the Windmill Islands. The skuas thus taken consisted of an unselective group of both banded and unbandedbirds, in much the same way that widespreadshooting of waterfowl resultsin an unselectivereturn of bothbanded and unbandedducks and geese.Such data can be used to determinepopulations over a given area throughuse of the so-called LincolnIndex fLincoln,1930, and Allee.et al., 1949,pp. 268-269). The formula for this index is expressedas the ratio of the total unbandedpopulation (x) to the total bandedpopulation, which is equal to the ratio of thetotal unbanded captures to the totalbanded recaptures. The total unbandedpopulation plus the total bandedbirds thus equals the total population. The cannonnet was fired 10 timesduring Februaryand March of 1957, and 21 times during the next nestingseason, for a total of 31 shots.To providea representativeor randomsample of bandedand unbandedbirds, these shots were made in nine differentareas throughoutthe Windmills,from Holl Islandin the southernpart of the group to the mostnortheasterly island. Theseareas are shownon Plate III. In this manner a total of 258 banded and unbanded adult skuas were captured,of which54 had bandsand 204 had not. Elevenbanded and 52 unbandedwere netted during the 1956-57 season,and 43 banded and 152 unbandedwere taken in the followingseason. Five banded birdswere captured two or moretimes, but after the first capturewere considered "dead" and were not thereafter counted. A total of 449 adultskuas were banded throughout the islandsduring the first two yearsof the study,of which 277 were taken during the 1956-57 seasonand 222 the next year. Juvenileswere not considered in the census,because banding observations gave no evidencethat those bandedthe firstyear returned to the areathe following year, and those markedthe secondyear were not flying whenthe net was shot. In usingthe formula,it wasconsidered valid to groupbanded birds of both yearstogether. Also.in the final determinationa minor correction factor must be appliedto the adult birds bandedthe first year becauseof natural mortality to theseduring the 7 monthsbetween seasonswhen the net was not shot. Basedon an assumedaverage age of 10 years,the correctionfactor would be approximately5 percenton the 277 birds. Applicationof the LincolnIndex to the abovefigures, i.e., 214] EKLUND,The South-Polar Shua Bird-BandingOctober

X (total unbandedpopulation) __ 204 (total unbandedcaptures) 499 l,total banded) 54 (total banded recaptures) showsa total of 1,885, plus 499, or 2,384 birds. Consideringnatural mortality to bandedbirds betweenseasons of shootingthe net, this figure is roundedto 2,300 as being representativeof the adult skua populationin the Windmill Islands. The 54 bandedrecaptures represent 10.8 percentof the total number banded(499), and the 258 total capturedin the cannonnet represents 11.2 percentof the total estimatedpopulation. This is considereda representativesample. With exceptionof the minorland areasof DavisIslets, 65 milesfrom the Windmills. and Merritt Islet, there are no lands westward of the Windmill Group until Bowman Island is reached,approximately 220 miles distant. From the mostnortheasterly of the Windmillseastward alongthe coastthere are likewiseno land areas,except for the six small BalaenaIslets, until the Henry Islets are reached295 miles distant. In a 74-miletrip alongthis sectionof the coastonly four pairs of nesting skuaswere seenand thesewere at a small Adelie penguinrockcry on the Balaena Islets. This indicates a total of more than 500 miles of coastline in which the Windmill Islandsconstitute the only major land area to provide nestinghabitat for the skua. Basedon the estimatethat over 90 percent of thetotal nesting population in thisstretch of thecoast would probably be in the Windmills,it is concludedthat the total adult skuapopulation during the nestingseason along this 500-milesection is not more than 2.500 birds. Skua Populationsin Relation to Penguins It has been estimatedthat at least 95 percent of the skua nesting habitatin the Windmill Islandsis near Adeliepenguin rookeries. This proportionmight prevail also in otherregions under natural conditions, that is, where the birds are not influencedby a human habitation makingit possiblefor them to obtain suchfood as garbageor seals killed for dog food. Austin i1957, p. 12). on the basisof visitsto severalrookeries in the RossSea sectorof Antarcticaduring the summerof 1955-56,has suggestedthere is a ratio of 1 pair of skuasto about 100 pairs of penguins.As he state•. sucha ratio will doubtlessvary with the size of the area,since small penguin rookeries provide more fringe habitat for skuanesting than do rookeriesin large solid blocks. Whileunloading operations were taking place at CapeHallett, it was necessaryfor us on December30, 1956, to removeall penguinsfrom an area of 2.06 acresto providespace for stationconstruction. This ground,precisely measured by the engineers,was fencedprior to the removalof the penguins. The penguinsremoved included 3,318 juveniles.Not all the adult birds were in the rookeryat the time, sincemany were feedingat sea. Basedon observations of young produced at a rookeryin theWindmills, it was estimatedthere would have been a minimum of 4,900 adult Adeliepenguins in these2.06 acres.The observationsfurther indicated that about20 percentof the adultseither were unmated or that some pairsdid not produceyoung. At the time of the removal,therefore, Vol. 1961XXXII EKLUND,the So,th-•'ot,,. S•,,a [215 there would have been about 1.7 youngper successfulbreeding pair. The sizeof the hook-shapedrookery at Hallett was determinedby planimeteringthe limits of the nestingarea on a 2-foot contourmap, on the scaleof 1 inch to 100 feet. It was calculatedthat the rookery coveredslightly more than 55 acres. Using the above data it was estimatedthere would be about 130,000 adult penguins,or a total populationof approximately218,000 usingthe entire area at that time of the year. It wasconservatively estimated that approximately150 pairs of skuas nestedat Cape Hallett. Allowing an additional20 percentfor non- breeedingskuas there would thus be an adult skua-penguinratio of about1:360. Sucha populationwould be understandable,considering the relativelylarge, solid block in which the penguinsnested. Also, a large percentageof the fringe area runs closelyalong the sea,and this was not usedby skuasfor nesting. Possiblyfor thesereasons most of the birdsnest together in a "skuary"close to the baseof a mountain at the southwesternedge of the penguinarea. Sladen•]958, p. 66) estimatedthat a populationof over 100,000 Adeliepenguins supported about 20 breedingand somenonbreeding brownskuas (C. lonnbergi)at HopeBay in Palmerland(latitude 63øS., longitude57øW.). In the Windmill Islands there were 16 Adelie penguin rookeries. Most of thesewere relatively small, the largsthaving an estimatedmaxi- mumpopulation of slightlyover 22,000 adult birds. To determinepopulation fluctuations within a penguinrookery, in which birds are continuallycoming and going,a countof adultswas made every 48 hours throughoutthe seasonat rookery A on Clark Peninsula.The rather small maximumpopulation of 3,450 birds, con- sisting of small groupingswithin the area, made the count fairly accurate..All the 15 other rookeries within the Windmills were counted at least once. From knowledgeof the percentageof the maximum numberon any givenday in the rookeryunder detailed study, together with the assumptionthat all rookeriesin the Windmills had similar fluctuationsthroughout the season,it was then possibleto apply a correctionfactor to all other rookery countsfor the days thesewere taken. Basedon the abovestudies the total breedingpopulation of Adelie penguinswithin the Windmill Group was estimatedat 155,000. With an estimatedtotal of approximately2,300 adultskuas for the samearea, the ratio of skuasto adult penguinswould be about 1:70. The ratios for C. maccormickithus seem to vary from an estimated1:360 to 1.70, and theseappear to be governedby the size and compositionof the penguin rookeries. Occurrenceo/ First-Year Birds There is no evidence that skuas return to the same area the first seasonfollowing that of hatching,but somereturn as nonbreedersthe secondyear. Thisis basedon Austin's 1955-56 banding of 52 juveniles, and on our 1956-57markings of 35 similarbirds at Hallett Stationand 46 at Wilkes Station. None of these birds at either area have been reportedobserved during 1958-50season. An immaturebird bandedat Wilkest USFWS617-003281 on February 2, 1957,was recovered at the Bird-Banding 216] EKLUNq),TheSouth-Polar Skua October

USSR Mirny Station on March 19, 1958. Penncy captured a two-yearbird (USFWS617-00284) at Wilkeson April 11, 1959,which was banded there as a chick on February 8, 1957. Reid (1961) observedthree 24-month-old nonbreeding birds in 1960-1961at Ilallett whichhe had bandedas juvenilesduring the 1958-59 season.

SEASONAL MIGRATIONS Banding Returns It has been shown in the discussionon breeding-sitetenacity that skuasbanded at the McMurdo, Hallett, and Wilkes and various FIDS Stationsreturn to the sameareas in successiveyears. A total of over 2,500 skuasbanded in six consecutiveseasons has now madeit possible to alsolearn muchabout their movementsaway from the areaswhere they were banded. The first recoveryof a skuaaway from the placeof bandingwas by Kirpichnikovfrom aboard the USSR icebreakerOB. This bird. an adult (USFWS 617-00567), was bandedas well as dyed red at Wilkes on March26, 1957,and recoveredin the packice 120 mileswestward on February19, 1958 (Shevareva,1959). This recordmight be an indicationthat the bird was a nonbreederthat had stayedin the pack duringthe nestingseason. The Sovietsalso recovered another (USFWS 61%00464) in the samegeneral area •latitude65 ø 08' $.. longitude 111ø 15' E.), 80 milesfrom WilkesStation, on February4, 1959. It had beenbanded at Wilkes on February 24, 1957. Another adult skua (USFWS 617-00566), banded at the same time and placeas that of the 1958 Sovietrecovery, was collectedby Dear- born at McMurdo Sound,on January3, 1959. A similarrecord of an adult (USFWS 617-00359) banded at Wilkes on February 21, 1957, was a recoveryat ScottStation in McMurdoSound on December20, 1958, by Caughley,Sladen, and Penney.The movementof both these birdswas probably eastward around the continent,which would representa flight of about 1,900 nauticalmiles. Dearborn,on January3, 1959, recoveredan adult (USFWS 617- 52517) at McMurdo Sound that had been banded at Little America, 375 milesto the eastward,on January6, 1957. An adult banded at Midgley Island East, near Wilkes Station (USFWS 617-00660), on November 22, 1957, was recovered by De Gerlacheand Frazier, of the Belgium Antarctic Expedition,near their IGY station(latitude 70 ø 26' S., longitude23 ø 19' E.), on Janu- ary 26, 1958. Assuminga westwardflight alongthe coast.this bird traveledapproximately 2,400 miles. Furthermore,it was obviouslya nonbreeder. Another adult (USFWS 617-00634) banded at Wilkes on November 17, 1957, was recoveredat the French Dumont D'Urville Station,approximately 720 miles eastward,on December23, 1959. Dingle reportedrecovery of a skua (MOSKWA C 66942) at Aus- tralia's Davis Station in February 1958, which had presumablybeen bandedat the USSR Mirny Station in 1956-57, and Braunstefferre- covered two more (MOSKWA C 66915 and C 66969) at this station in February 1959. This would representan eastwardflight along the continent of at least 400 miles. Vol.1961xxxH ES:LUN,),TheSouth-Polar Skua [2 ] 7 218] EKLUND,The South-Polar Stcua Bird-Banding,October

Rubin. U.S. exchangescientist stationed at Mirny during 1958-59, reportedrecovery of two Wilkes-bandedskuas there by Kuznitsov.The first return (USFWS 617-00424) was taken on February 26. 1958. and was of an adult banded at Wilkes Station on February 22. 1957. The secondbird (USFWS 617-003281,banded as a nestlingon Febru- ary 2, 1957, was taken on March 19, 1958. This is the first return madein Antarcticaof an immaturesouth-polar skua away from the placeof banding,and is further indicationof a circumpolarmovement of the nonbreeders. Dewart, U.S. exchangescientist with the Soviets in 1960-61, reported that Korotkevichhad collecteda green-banded Wilkes skua at Mirny in April 1960. Two recoveriesby Stonehouseand Freeman at the British FIDS base on StoningtonIsland, in Palmerland, which had been banded there during Americanoccupation in 1941, have been reportedin the dis- cussion on age. There have been many observationsof skuas at various stations in Palmerlandwhich have returned to the area of banding. As shown in Table IV, birds have been reportedrecovered away from the place wherethey were banded,and this indicatesa generalnorth-south movement alongthe peninsula.However, of the more than 900 skuasbanded by the British, Argentines,and Chileansin this area, it is peculiarthat there have been no returns which would indicate an eastward or westward movementalong the Antarctic coast away from Palmerland. Neither have any of the more than 1,600 skuasbanded in other parts of the continentbeen seenon the peninsula. A recoveryin Australiaof a Wilkes-bandedbird representsthe first wintertime return (Eklund, 1959). This was also the first record of C. maccormickiin Australia,although a specimentaken in Queenscliff. Victoria,Australia, identified as Megalestrismaccormicki (Campbell, 1924, p. 77), was later correctly identified as Catharactaantarctica •Anon., 1924, p. 119'i. The recovery (USFSW 617-00501 and with a greenband also3was bandedon March 1, 1957, and found deadby the children of Mr. L. S. Barrett on the southern shore of Hardwicke Bay, York Peninsula,in SouthAustralia, on May 19, 1958 •Plate I). In reportingit, Mr. Barrett statedthe bird had been dead probably lessthan 7 daysand that this wouldcoincide with a stormperiod of gale-forcewinds from the west,which apparently had originatedin the polar or sub-polarregions. Records of dead birds such as those recovered in New Zealand. and the Australianrecovery, would not indicatea normal range for the speciessince the birds were apparently carried and exhausted by strong winds. It is believedthat stormsoriginating in Antarcticacause con- siderablemortality and wouldresult in the greatestsource of records of C. maccormickioutside that area. On the basis of this, as well as the failure thus far to obtain sufficient recoveries outside the south polar regions,no definiteconclusions can be made on the normal winter range of the species.Probably it is confinedto within the extreme northernmostlimits of packice duringthe maximumcold seasonfrom Junethrough August as shownon Plate I. Recordsof recoveriesaway from the areaof bandingare shownin Table IV. Vol. XXXII 1961 ]•KLUND,The So.th-Polar Skua [219 Bird-Banding 220] EKnUN,,The South-Polar Skua October

BandingObservations In addition to the band recoveries, the movements of skuas have been followedthrough observation of the coloredbands. The red bandsof two birds bandedat CapeHallett during the 1956-57season were observedat McMurdo Sound by Cadwaladeron December20, 1957, and by Miller on November20, 1958. Theserepresent a south- ward movementof approximately380 miles. Sladen observed a skua at McMurdo Sound on December 20, 1958, which had been marked with a blue band at Little America during the 1956-57 or 1957-58 seasons. This indicates a westward movement, probablyalong the RossShelf Ice, of about375 miles. The sameob- serversaw a green-bandedbird at CapeEvans in McMurdoSound on December 13, 1958, which had been marked at Wilkes in either the 1956-57 or 1957-58 seasons.Salmon 11959) reported seeinga blue- bandedskua from Little Americaat Hallett Stationon January6, 1959, and if this bird followed the coast it would representa west-north flight of about 700 miles. Penneyreported that a blue-bandedskua was seenin November1960 by a trail party near the Wilkes Icecap Station (Site 2}, 51 miles inland, at an elevation of 4,200 feet. Ingham (1959) reportedobservations by Norris and Evans of two Wilkes green-bandedskuas at Australia's Davis Station, 850 miles westward,in January1959, and Black saw anotherthere in February 1960. A Wilkes bird was seen at the Mawson Station, 1,200 miles westward,on January 5, 1959. Observationsof skuasaway from the area of bandingare shownin Table V. CONCLUSIONS The resultsof studieson the south-polarskua, conductedover the past five years,have led to the followingconclusions with respectto the life historyand distributionof the species: 1. Site tenacityis a breedingtrait of the south-polarskua. 2. The south-polarskua has a pair-bondrelationship and will remate and reoccupythe sameterritory in successiveyears. 3. Ninety-five percent or more of all skua-nestingsites in the Windmill Islands are confined to the vicinity of Adelie penguin rookeries. 4. The incubationperiod of the south-polarskua eggsaverages 30 days,and the incubatingtemperatures taken over a nine-dayperiod averaged96.6 ø F. and rangedfrom 103.5ø to 87ø. 5. A detailedstudy of 40 nestsindicated a nestingsuccess of 70 percent,a 39 percentmortality of incubatingeggs, and a 23 percentloss of chicksup to 4 weeksof age. 6. The daily feedingrange of the south-polarskua may be up to at least 8 miles from the nest site. 7. The south-polarskua has a homing instinct,as shownby its breeding-sitetenacity as well as by homingexperiments wherein breeding birds returnedto their nest sitesfrom variouspoints of release. 8. The total populationof adult skuaswithin an approximate500- mile sector of the Wilkes Land Coast is about 2,500. Vol. 19SlXXXII EKLUND,The SoutIi-Polar Skua [221

9. The normalwinter rangeof the south-polarskua has not yet been determinedconclusively. but it is probablyconfined to the limits of the Antarcticpack ice. 10. Distributionof the south-polarskua is circumpolar,and returns and observationsfrom over 2,500 bandedbirds showthat its migratory movementsare north and south as well as east and west along the coasts. The circumpolarflights appear to be made primarily by nonbreedingbirds during the summer. 11. South-polarskuas do not return to the area wherehatched until at least the secondyear of age, and they return to nest for the first time at this samearea at five yearsof age.

LITERA TURE CITED ALLEE, W. C., EMERSON,A. E., PARK, O., PAnK T., and SCHMIDT,K. P. 1949. Principles of animal ecology. W. B. SaundersCo., Phila. and London. 837 pp. ANONYMOUS. 1924. A correction. (Australian sea birds.) Emu 24(2): 81-160, Oct. AUSTIN, O. L., JR. 1957. Notes on banding birds in Antarctica, and on the Adelie penguin colonies of the Ross Sea sector. Bird-Banding 28(1): 1-26, Jan. BAGSHAWE,T. W. 1939. Two men in the Antarctic. Cambridge Univ. Press, London, 283 pp. BENDER,JAMES A. 1959. Perso,•al correspondence,U.S. Army Snow Ice and Permafrost Research Establishment, Wilmette, II1., Feb. 19. CAMPBELL,A. J. 1924. Australian sea birds. Emu 24(1-2): 1-80, July. CLARK,W. E. 1906. On the birds of •he South Orkney Islands. Ornith. Results of the Scottish Nat'l Ant. Exp., Ibis, Vol. 6, 8th Series, No. XXI, p. 182, Jan. CROCKFOnD,J. A., and 'CANNON,R. W. 1956 (revised). A modification of the cannon-projectednet banding trap. Ga. State Game & Fish Comm., Fed. Aid Proj. 14-D, Atlanta. DILL, H. H., and THORNSBERRY,W. H. 1950. A cannon projected net trap for capturing waterf,owl. Journ. Wild. Mgt. 14(2): 132-137, April. EKLUND, C. R. 1942. Body tem,peraturesof Antarctic birds. Auk 59(4): 544-548, Oct. 1945. Condensedornithology report, East Base. In Rpts. on Sci. Re- suits U.S. Ant. Serv., 1939-41. Proc. Amer. Philos. Soc., Phila., 89(1): 303. 1957a. A distribution study of the souvh-polarskua. Antarctic, N. Z. Ant. Soc., Wellington, N. Z. 1 (6): 127-128,2VIarch. '1957b. A distribution study of the south-polar skua. The Ring No. 11: 223-225, JMay. 1957c. Distribution study of the south-polarskua So. Afric. Ornith. Soc. and Wkwatersrand Bird Club, The Bokmakierie 9t2): 46-47, Oct. --. 1958a. A distribution study of the south-polarskua. Notornis, Quar- terly Bull. Ornith. Soc.N. Z. 7 (8): 234-235,April. 1958b. A distribution study of Me south-polar skua. So. Austral. Ornith. 22 (5-6): 71, July. 1958c. A distribution study of the souS-polar ,skua. Emu 58(2):98- 100, ,May. 1959. South-polarskua recoveryin Australia. Emu 59(3): 158, Aug. --, and CItARLTON,F. E. 1958. Measuring the temperatures of incubating penguin eggs. Research Rev., Off. Naval Res., pp. 1-6, Dec. Reprinted in Amer. Sci. 47(1): 80-86, March 1959. ETCHECOPAR,R. D., and PREVOST,J. 1954. Donnees oologiques sur l'avifaune de Terre Adetie. Erpeds. Antarctiques en Terre Adelie, 1949-1953.Expeds. Polaires FrancRises,Paris. Note Ornithlologique12, p. 243. EVANS,C. 1951. A method of color marking young waterfowl. Journ. Wild. Mgt. 15(1): 101-103, Jan. Fart.a, R. A. 1937. B,A.N.Z. Ant. Res. Exp. 1929-31. Rpts.--Series B, 2, Birds, pp. 247-250. 222 ] EKLUNU,The South-Polar Skt,a BirdOctober -Banding

1940. Occurrence of 5IcCormicks skua on the coast of New Zealand. Emu'40(2): 119-120. FLEMING, C. A. 1953. A fourth New Zealand record of the Antarctic skua, Stercorarius skua maccormicki (Saunders). Notornis, Bull. Ornith. Soc N. Z. 5(6): 182-183, Oct. FRIEDMANN,H. 1945. Birds of the U.S. Antarctic Service Expedition, 1939-41 In Rpts. on Sci. Results U.S. Ant. Serv. 1939-41. Proc. Amer. Philos. Soc. Phila., 89(1): 305-313, April. GAIN, L. 1914. Deuxieme expedition antarctique francaisc 1908-1910. Oiseaux antarctiques, 2, Masson et Cie, Paris, 200 pp. HAMILTON, J. E. 1934. The sub-antarctic forms of the great skua (Catharacta skua skua). Discovery Rpts., Cambridge Univ. Press, London, 9: 163-175. INGHAM, S. E. 1959. Personal correspondence,Biological Secretary, Ant. Div., Dept. Extern. Affairs, 187 Collins St., Melbourne, C. 1 Victoria, Australia. March 18. ISHIDA, TAMOTSU,and SnIMIZU, Hmoa•Ia. 1958. Resistanceto air flow through snow layers, (Part 1). Trans. 60, U.S. Army Snow Ice and Permafrost ResearchEstablishment, Wilmette, Ilk, 111.1-8, Oct. LAPINA, 1. J. 1958. A short summary of ornithological work conducted by Soviet scientists in the Antarctic from 1955-58. Personal correspondence, Sci. Secr't, Soviet Ant. Comm., M. Haritonyevsky 4, Moscow, TSENCR, U.S.S.R., 18 Nov. LINCOLN,F. C. 1930. Calculating waterfowl abundanceon the basis of banding returns. U.S. Dept. Agri. Circ. 118: 1-4, May. McCoRMICK, R. 1884. Voyages of discovery in the Arctic and Antarctic seas. S. Low, Marston, Searle, and Rivington, London, 1: 3-432. MUltpHY, R. C. 1936. Oceanic birds of South America. Amer. Mus. Nat. His., N.Y., Vols. 1 and 2, 1,245 pp. OATES,E. W. 1901. Catal,ogueof birds' eggs. Brit. 5Ius., 1,252 pp. PENNEY, R. L. 1961. Wilkes Station zoological report Feb. 1959 to Jan. 1961, 31 pp., unpublished. PONTII•G, H. G. 1921. The great white south. Gerald Duckworth & Co., Ltd., London, 299 pp., 132 pls. REID, BRIAN. 1961. Personal correspondence.Geological survey, D.S.I.R., Ro- torua, New Zealand, June 19. ROMEa, ALFl•EDSHERWOOU. 1955. The vertebrate body (2d Ed.) W. B. Saund- ers Co., Phila. and London, 644 pp. ROI•I•E, FINN. 1945. The main southern sledge journey from East ,Base, Palmer- land, Antarctica. In Rpts. on Sci. Results of the U.S. Ant. Serv., 1939-41. Proc. Amer. Philos. Soc., Phila., 89 (1): 13-22. RUUEI•SUORF,W. J., and VAN DEUSEN,R. D. 1955. A suggested technique for making waterfowl neckbandsusing rubber-styrene. Paper at 17th Ana. Mid- west Wildlife Meeting, Dec. 12-14, 1955. Purdue Univ., Lafayette, Ind. SALMON,K. J. 1959. Hallen Station Leader's Rpt. for period Jan. 1958-Jan. 1959, Nat'l Acad. Sci., Wash., D.C. SALYER,J. C. 1955. Improvements in the cannon net trap. USDI, Fish and Wildlife Serv., Wash., D.C., Wildlife Mgt. Series 12. SAUNDERS,H. 1893. Bull. Brit. Ornith. Club, edited by R. Bowdler Sharpe. 3 (13): XII. SCOTLAI•u,'C. D. 1959. The use of A•merican IGY thermoplastic skua rings at British Antarctic bases in the Falkland Islands Dependencies. Un.pu,b. lkshed FIDS report, 7 pp. Oct. 26. SnEVaaEVa,T. 1959. Correspondencewith U.S. Fish and Wildlife Serv. from U.S. S. R. Acad. Sci. Connnission for Nature Protection, 33 Lenin-Ave., Moscow B-71, March 16. SLADEN,W. J. L. 1958. The Pygoscelid penguins: I, H. FIDS, Sci. Rpt. 17, Her Majesties Sta. Off., London, 97 pp. SLADEN,W. J. L., and TICKELL,W. L.N. 1958. Antarctic bird bandingby the Falkland Islands DependenciesSurvey. Bird-Banding29(1): 1-26, Jan. STONEHOUSE,B. 1956. The brown skua Catharacta skua lonnbergi (Mathews) South Georgia. FIDS, Sci. Rpts. 14. Her MajestiesSta. Off., Lond,on,25 pp. TIe,EIleEn, N. 1953. The herring gulfs world. Collins, London,255 pp. TRESSLER,W. L. 1959. Personal correspondence,Wilkes Station, Antarctica, Feb. 27. Vol. 1961XXXII HAILMAN,Laughi•g Gull Chicks [223

VAN RooY, ,M. ]P. 1957. Meteorology of the Antarctic. Weather Bureau Dept. Transport, ¸ov't Printing Office, Pretoria, So. Africa, pp. 95-102. WALTER,W.g. 1943. Some experimentson the senseof smell in birds. Archiv. Needand, Physiol. de l'Homme et des animaux. N'ord•HollandscheVitgevcr-, Maatschappij, Amsterdam, vol. 27: 1-73, May. WILSON,E.A. 1907. National Antarctic Expedition 1901-1904. Nat. His. Vol. II Zoology, pt. II Aves, Brit. Mus., London, pp. 1-121. WINSTON,'F. A. 1954. Status, movement, and management of the mourning dove in Florida. Fla. Game & Fresh Water Fish Comm., Tech. Bull. 2, Fed. Aid Proj. W-22-R, Tallahassee. Polar Branch, Army Research Office, Washington, D.C. Received June, 1959 (revised to June, 1961)

AGE OF LAUGHING GULL CHICKS INDICATED BY TARSAL LENGTH By JACK P. HAILMAN This paper investigatesthe possibilityof predictingthe age of a LaughingGull (Larus atricilla) chick from measurementsof its tarsus. Often bird bandersand field ornithologistsdesire to know the age of a bird in the field for populationand mortality studiesor other reasons. I desiredto know agesof chicksfor two reasons:t 11 to analyzebanding results(Hailman, 1960). and •2} to assemblegroups of wild chicksof a certain age in order to comparetheir behavior with that of hand- reared chicksof the sameaverage age l in prep.). Age Criteria In LaughingGulls. color pattern of plumageand color of softparts do not changeappreciably prior to flight.and the eggtooth disappears at varying rates,so that thesecharacteristics cannot be usedto determine early agesvery accurately.Total bodyweight, which increases with age. dependspartly uponhow recentlychicks have fed and seemsto be too variableto be usedin age determinationin gull chicks(Goethe, 1955]: furthermore,it is generallyimpracticable to carry an accuratebalance or scalein the field. The problemrequires a quantitativemeasurement that does not demandelaborate equipment. Tarsal length was tried because(1) it dependslargely upon the lengthof the tarsometatarsus bone,which is rigid, and (2) it canbe easilymeasured with a simple rule, withoutthe use of calipers. Chicks The 18 chicksused in this study were hatchedfrom the eggsin an incubatorduring June, 1960. The eggswere collectedfrom a colony of over 1000 pairs, locatedon an islandin PamlicoSound, North Carolina.Tarsus measurements were taken about twice a day,the exact time beingnoted. The time of hatchingis knownto within 15 minutes on l0 birds and to within 2 hourson the remaining8. As part of behaviorexperiments to be reportedelsewhere, the chickswere raised in dark boxesand force-fedto satiationthree or more timesper day. It is impossibleto assessthe effectsof hand-raisingupon the growth rate of the chicks'tarsi. All appearedto be growingnormally in all