Mar 1, 1982 - Editor - B. D. HEATHER, 10 Jocelyn Crescent, Silverstream ..... Reports on the collection of birds made during the voyage of the H.M.S..
NOTORNIS Journal of the Ornithological Society of New Zealand
Volume 29
Part 1
March 1982
OFFICERS 1981 - 82 President -R. B. SIBSON, 26 Entrican Avenue, Auckland 5 Vice-president B. BROWN, 39 Red Hill Road, Papakura Editor -B. D. HEATHER, 10 Jocelyn Crescent, Silverstream Treasurer - G. M. H. PETERSON, P.O. Box 22230, Auckland Secretary - R. S. SLACK, c/o Royal Society of NZ, P.O. Box 12249, Wellington Council Members: BEN D. BELL, 45 Gurney Road, Belmont, Lower Hutt BRIAN D. BELL, 9 Ferry Road, Seatoun, Wellington P. C. BULL, 131A Waterloo Road, Lower Hutt D. E. CROCKETT, 21 McMillan Avenue, Kamo, Whangarei P. D. GAZE, Ecology Division, DSIR, Private Bag, Nelson S. M. REED, 4 Mamaku Street, Auckland 5 P. M. SAGAR, 38A Yardley Street, Christchurch 4 Conveners and Organisers: Rare Birds Committee: Secretary, J. E. SQUIRE, 135 Tirohanga Road, Lower Hutt Beach Patrol: R. G. POWLESLAND, Wildlife Service, Dept. of Internal Affairs, Private Bag, Wellington Card Committee: R. N. THOMAS, 25 Ravenswood Drive, Forest Hill, Auckland 10 Librarian: A. J. GOODWIN, R.D. 1, Clevedon Nest Records: D. E. CROCKETT Recording (including material for Classified Summarised Notes) : D. F. BOOTH, 16 Valdese Rise, Browns Bay, Auckland 10 S.W.Pacific Islands Records: J. L. MOORE, 32 Brook St, Lower Hutt Assistant Editor: A. BLACKBURN, 10 Score- Road, Gisborne Reviews Editor: D. H. BRATHWAITE, P.O. Box 31022 Ilam, Christchurch 4 Editor of OSNZ news: P. SAGAR, 38A Yardley St, Christchurch 4
-
SUBSCRIPTIONS AND MEMBERSHIP Annual Subscription: Ordinary member $16; Husband & wife members $24; Junior member (under 20) $12; Life Member $320; Family member (one Notornis per household) being other family of a member in the same household as a member $8; Institution $32; Overseas member and overseas institution $5.00 extra (postage). Subscri tions are for the calendar year of first joining and are renewed gy invoice each January. Please pay promptly to ensure receiving Notornis and OSNZ News. Applications for membership, changes of address and resignations should be sent to the Treasurer. Exchanges and library subscriptions should be sent to the Treasurer. Editorial matters ONLY should be sent to the Editor. [Registered with the GPO Gisborne as a publication]
ISSN 0029- 4470
CONTENTS
.
BECKON. W. N Breeding Record of the Whistling Dove of Kadavu. Fiji .............................. ............ ONLEY. D
.
.
Spotless Crake on Aorangi. Poor Knights Islands
.
..
WARHAM. J WILSON. G J Size of Sooty Shearwater Population at Snares Islands. New Zealand .................. CHILD. P. Additions to the Avifauna of Kiribati and Tuvalu
.
NUGENT. G Sexing Black-backed Gulls from External Measurements ..........................................
. .
VEITCH. C R Seabirds found dead in New Zealand in 1980 Classified Summarised Notes
........................
......
Short Notes BROWN. B Terek Sandpiper feeding like Avocet ...... ...... JENKINS. P . Wood Duck in Marlborough ............ ...... HAWKINS. J M. Gannets feeding behind trawler ...... ...... CHILD. P . Far inland sighting of Welcome Swallow ..... ...... SIBSON. R . B Arctic Tern in Manukau Harbour ...... ...... HAWKINS. J M Reef Heron on Nelson Haven ...... ...... HOWELL. P. A . G . HARRISON. K C. A Turnstone-Banded ...... Dotterel relationship .................. ...... ONLEY. D Nomenclature of Spotless Crake ...... ...... ...... BROWN. B Unusual feeding of Red-billed Gulls ............ BLACKBURN, A 1927 Record of Laughing Owl ............
.
.
. . .
. .
.
.
.
Obituary . BOURNE. W . R P Averil Margaret Lysaght (1905-1981)
. .
R i n t d by TE RAU PRESS LTD., F n l Stmt. Girbonw
......
REGIONAL REPRESENTATIVES FAR NORTH: D. E. Crockett, 21 McMillan Ave., Kamo, Whangaroi. NORTHLAND: Ph. 50954. AUCKLAND: S. Reed, 4 Mamaku St.. Auckland 5. Ph. 547784. SOUTH AUCKLAND: Beth ~ r o w n j39 Red Hill Road, Papakura. Ph. 2988157. WAIKATO: B. Sedd~n,11 Grey Street, Cambridge. Ph. 7761. BAY OF PLENTY: R. M. Weston, 250 River Road, Kawerau. Ph. 8357. VOLCANIC PLATEAU: J. G. Innes, Loop Road, Okareka. Ph. 53365. Rotorua. GISBORNE/WAIROA: J. C. Henley, 9 Mason St., Gisborne. Ph. 81581. TARANAKI: D. G. Medway, 25A Norman Street, New Plymouth MANAWATU: L. J. Davies, 390A Botanical Road, Palmerston North. WANGANUI: Ph. 87680. HAWKES BAY: N. B. Mackenzie, 10 Bumess Road, Greenmeadows. Phone 447426. WAIRARAPA: C. Scadden, 15 Madden Place, Masterton. Ph. 86423. WELLINGTON: A. H. Gollop, 6 Dawn Grove, Akatarawa, Upper Hutt. Phone 268749. NELSON: J. Hawkins, 772 Atawhai Drive, Nelson. Ph. 520151. MARLBOROUGH: P. Jenkins, 234 Howick Rd., Blenheim. Ph. 83775. CANTERBURY: P. M. Sagar, 38A Yardley Street, Christchurch 4. Phone 429720. WEST COAST: C. S. Lauder, 9 Winnie Street, Greymouth. Ph. 6349. OTAGO: R. F. Smith, 8 Butler Street, Maori Hill, Dunedin. Ph. 740672. Phone 740672. SOUTHLAND: R. R. Sutton, Lorneville, No. 4 R.D., Invercargill. Ph. 358230.
{
{
LITERATURE AVAILABLE From all bookshops: A field guide to the birds of New Zealand, by R. A. Falla, $13.95 R. B. Sibson and E. G. Turbott, new ed. From B. D. Heather, 10 Jocelyn Crescent, Silverstream: $1.50 A biology of birds, by B. D. Heather. From H. Hagen, 53 Minnehaha Street, Titirangi, Auckland 7: Back numbers of 'Notornis': Parts of Vol. 1, 50c each; Vols. 2-13, $1.00 per part; Vols. 14-21, $1.50 per part; Vols. 22-25, $2.00 per part; Vols. 26-, $3.00 per part; all plus postage (10% in NZ) . Reports and bulletins (1939-1942) $2.00 OSNZ Library catalogue (1976 ed) 17 pp. $0.55 Bandin reports, Nos 8-14, 55c each. ~ e r m a f e cExpedition, 1964, by A. T. Edgar. $0.50 Guide to Identification of Shearwaters and Petrels in New Zealand waters (Auckland Museum), J. P. Croxall $0.55 Amendments & Additions to 1970 Checklist $2.00 From P.O. Box 12397, Wellington North: Bird distribution in NZ. A provisional atlas $6.00 From B. D. Bell, 9 Ferv Road, Seatoun, Wellington: $6.00 OSNZ tie (mid-grey with Notornis motifs).
NOTORNIS is the journal of the Ornithological Society of New Zealand (Inc.)
Editor: B. D. Heather, 10 Jocelyn Crescent, SILVERSTREAM
VOLUME 29
PART 1
MARCH, 1982
A BREEDING RECORD OF THE WHISTLING DOVE OF KADAVU, FIJI By WILLIAM
N. BECKON
ABSTRACT The first known record of the nesting of the Whistling Dove (Ptilinopus layardi) is presented, along with notes on its diet and very un-dove-like whistled calls. The fact that only the female was seen brooding the young may be of some significance to an understanding of the evolution of extreme sexual dichromatism in the Golden Dove Group to which the Whistling Dove belongs. INTRODUCTION The Whistling Dove*, Ptilinopus layardi, is confined to the island of Kadavu and the nearby smaller island of Ono, in the south of the Fiji Group. It is one cf the distinctive " Golden Dove Group" of three geographically representative fruit doves, all of which are endemic to Fiji. The other two members of the group are the Golden Dove (P. luteovirens) and the Orange Dove (P. vicfor). The group was originally classified in a genus of its own, Chrysoenas Hartlaub (1854), but in recent years it has generally been considered to be a subgroup of the genus Ptilinopus (Amadon 1943: 5, 6; Cain 1954: 273, 278; Goodwin 1970: 332). The Whistling Dove is the most " primitive " I have chosen to call this species the Whistling Dove because that name appears to be the most distinctive and most authentic of the English names which have been applied to this bird (Martin 1940: 5). As far as I know, the name Velvet Dove originated with Mayr (1945: 1301, who implied that he had some doubt that the dove gives a whistled call. Goodwin (1970: 378) followed Wood (1926: 116) i n using the name Yellow-headed Dove.
2
BECKON
NOTORNIS 29
of the group. While sexual dichromatism is very pronounced in the other two species, in the Whistling Dove, adult male plumage is more like that of females and juveniles and that of other doves of the genus Pfilinopus. See Sibson (1972) for Belcher's painting of both male and female. Little has been published about the Whistling Dove since it was originally described in 1875 by E. L. Layard, the British administrator of the newly ceded colony of Fiji. Layard noted that his male specimens, collected in August, had "enormously developed " testes. He concluded that they were breeding at that time. NESTING About noon on 28 September, 1977, I was watching a male Whistling Dove giving his peculiar whistle call in the top of a nearby tree when I stepped back and inadvertently flushed another Whistling Dove off a nest just behind me. The nest was about 3 metres above the ground in forest on a low rounded ridge near the edge of a longabandoned Fijian teitei, or garden, in the interior of the south-western part of Kadavu, about 3 km south of Richmond Bay (Fig. 1 ) . The nest was a loose thin platform, about 10 to 12 cm in diameter, constructed mainly of twig-like pieces of the vine that grew on nearby tree trunks (Fig. 2). The nest was so sparsely built that one could see right through it from below. It contained a single nestling and a fragment of white shell. A larger piece of pure white shell lay on the ground directly below the nest.
FIGURE 1 .- Location of the Whistling Dove nest in the hterior of the mountainous island of Kadavu
FIGURE 2 - Whistling Dove nest, photographed from directly below, after death of hatchling. A 1-foot (30.5 c m ) ruler across the nest provides a scale.
4
BECKON
NOTORNIS 29
Apparently the nestling had just hatched. The membrane inside the eggshell was still moist, although it had not rained that day. The hatchling lay prostrate; its down was damp and matted. The next morning the nestling held its head erect and its light-coloured scanty down was dry despite the misty rain that was then falling. The nest was situated where two thin branches came close together. It was supported mainly by the long petioles of the large leaves of the tree, which a Fijian schoolboy identified as vecea (Fig. 3 ) . When I first flushed the bird from the nest, it flew off with a loud clapping of wings, like the sound of a Feral Rock Pigeon taking off, but I did not see it well enough to determine its sex. However, during my subsequent visits to the nest throughout the week, I never definitely identified a male either on the nest or within 10 metres of it. When I returned to the site at 2.45 p.m., a female was sitting on the nest. She sat quietly for the next 39 minutes while I watched from about 10 metres away. When I then began approaching the nest, she sat tight until I was directly under the nest. She then flew off without such loud clapping of wings as I had heard the first time. The hatchling gave one soft peep as the female left. On my next eight visits to the nest site over the following two
FIGURE 3 - Whistling Dove nest seen from below and to one side of the nest
1982
WHISTLING DOVE OF KADAVU
5
days, I found the female sitting on the nest four times, the nest was unattended on three occasions, and once, at 6.45 p.m., I was unable to determine definitely the sex of the brooding bird, since darkness had fallen. 29 Sept. 2.45 p.m. Nest unattended, dove whistling nearby 4.05 p.m. Female brooding 5.00 p.m. Female brooding 6.45 p.m. Dove brooding, sex undetermined 30 Sept. 9.00 a.m. Nest unattended, dove whistling nearby 10.15 a.m. Female brooding 1.45 p.m. Female brooding 3.00 p.m. Nest unattended, dove whistling nearby Over this period the female became progressively more wary and quiet in leaving the nest, until by 1 October she was apparently slipping off the nest while I approached, without my seeing her. leaving only the swaying nest as evidence that she had just been there. Thereafter I did not again see a brooding bird on the nest, although I continued to pass near the nest several times each day. O n 3 October I found the nestling cn the ground beneath the nest, still alive. After unsuccessfully attempting to feed it slightly diluted sweetened condensed milk, I returned it to the nest. The next day I found it still alive in the nest. On 5 October the nestling was dead in the nest. The small piece of eggshell still remained in the nest where I had first seen it. When I matched this to the other larger piece, a nearly complete eggshell was formed measuring 23.5 mm by 33 mm. CAL,LS Throughout this period, whether or not the female bird was sitting on the nest, I generally heard the male call in the canopy about 10 to 20 metres away. The call was a loud clear rising whistle immediately followed by a softer sound which I can only (inadequately) describe 2:s a falling tinkle (Fig. 4 ) . The whistle was easy to imitate closely, but the tinkle was not humanly reproduceable. The call could be heard at almost any time of the day, from all over the forest. It is probably the same call described by Wood (126: 116) as a " double whistle, the second sound being much less distinct or loud." More recent authors have described the call simply as a double whistle" (Blackburn 1971: 157; Sibson l972), a description which poorly fits the calls I heard. The male sometimes gave another distinctive call. It sounded somewhat similar to a small, high-speed electric motor starting up, immediately followed by a very high squeaky whistle, falling in pitch, reminiscent of the calls of the Polynesian Starling (Aplonis tabuensis). "
DIET Layard (1875: 151) reported that the Whistling Dove "fed on the berries of a species of banian." Wood (1926: 116) described its
6
BECKON
NOTORNIS 29
food as berries, or small pea-like fruit." Finsch (1877: 736) found a " fruit about the size of a cherry, with a large hard stone " in the stomach of one specimen. I saw Whistling Doves in the top of a tree (identified by Fijian boys as fugiri) feeding on the fruit, which grew in clusters. The fruit were 2.5 to 3 cm in diameter. Each one apparently opened in situ, exposing three carpels, each filled by two large brown seeds with a small amount of attached reddish-orange pulp. The Whistling Dove did not swallow the whole fruit, but plucked out and swallowed the contents of each carpel. Peale's Pigeons (Ducula latrans) fed on the fruit in the same manner. "
DISCUSSION The fact that I saw only the female at the nest may be only a coincidence, but on the other hand it may possibly be part of the answer to a baffling puzzle regarding the evolution of the Golden Dove Group. This group includes the most extreme cases in the Columbiformes of sexual dichromatism, with the development of brilliant plumage in the male. Yet, as far as 1 have been able to determine, these doves do not very well fit any proposed explanation of the evolution of such dichromatism unless they are quite exceptional among the Columbiformes in breeding behaviour, the males having little o r no role at the nest (Beckon: in manuscript). If further observations prove that males do participate in incubating eggs and feeding young, like other better-known pigeons and doves, then the puzzle remains unsolved.
FIGURE 4 - Approximate diagrammatic sonogram of the most common call of male Whistling Doves in south-western Kadavu in SeptemberOctober
1982
WHISTLING DOVE OF KADAVU
7
ACKNOWLEDGEMENTS Special thanks are due to my wife, Ruth Beckon, for her help in the field and for editing this paper. I am indebted to the teachers of Richmond High School, Kadavu, for their hospitality; and especially to Paul Howland and to Glenda, Ellen and Jenny for their generous logistic support. Thanks are also due to Sivisivi of Nasegai Village, Kadavu, who helped me find the Whistling Dove nest and who provided Fijian names for the plants I have mentioned. LITERATURE CITED AMADON, D. 1943. Birds collected during the Whitney South Sea Expedition. 52: Notes on some non-passer~negenera 3. Am. Mus. Nov. 1237: 1-22. BECKON, W. N. I n manuscript. p prediction of polygyny or promiscuity i n the Golden Dove Group of Fiji. BELCHER, W. J. 1929. Fragmentary notes on bird life i n the Fijis. Condor 31: 19-20. BLACKBURN, A. 1971. Some notes on Fijian birds. Notornis 18 (3): 147-174. CAIN, A. J. 1954. Subdivisions of the genus Ftilinopus (Aves, Columbae). Bull. Br. Mus. (Nat. Hist.) Zoology 2 (8): 267-284. FINSCH, 0. 1877. Reports on the collection of birds made during the voyage of the H.M.S. On the birds of Tongatabu, the F i j i Islands, Api (New Challenger," no. IV. Hebrides), and Tahiti. Proc. ZcoI. Soc. London. 1877: 726-742. GOODWIN D. 1970. Pigeons and doves of the world. London: British Museum [Nat. Hist.). LAYARD, E. L. 1875. Descriptions of some supposed new species of birds from the F i j i Islands. Proc. Zcol. Soc. London 1875: 149-151. MARTIN, A. H. 1940. The b ~ r d sof F i j i . Trans. F i j i Soc. Sci. Ind. 1: 4-7. MAYR E. 1945. Birds of the southwest Pacific. New York: Macmillan. SIBSON, R. B. 1972. Birds of Fiji in colour. Paintings by W. J. Belcher. Auckland: Collins. WOOD, C. 1926. Field observations. In WOOD, C.; WETMORE, A. A collection of birds from the F i j i Islands, Pt. 3. Ibis (12th ser.) 1: 91-136.
WILLIAM N. BECKON, 44503 21st Street West, Lancaster, California 93534, USA
SHORT NOTE TEREK SANDPIPER FEEDING LIKE AN AVOCET At Karaka shellbanks, Manukau Harbour. on 22 April 1978, J. A. Brown and I sat quietly on the outer bank watching waders. Before us was an area of firm wet mud with many little pools and patches of shells where birds were feeding on a falling tide. Among them was a Terek Sandpiper (Tringa cimrea) working busily towards us. We were able to watch it for about 30 minutes at distances as short as 20-30 m. These are wary birds in New Zealand, and so we were fortunate to have such an opportunity. The weather was fine and calm with clear visibility and 1 was able to note the following feeding methods: Stood still, then ran to pick flies from surface of water with sometimes an extra run as prey escaped. Picked randomly from surface of mud. Jumped to take flying insects. Probed in water about 3 cm deep for ? crabs, which were shaken and turned before being eaten. Bird was distant at this time.
8
SHORT NOTE
NOTORNIS 29
5 . Used bill in continuous side-to-side sweeps in a small shelly pool. No indication as to what prey was taken. Water half-way up bird's legs. Since 1978 I have looked in vain for a repetition of the last method until at Access Bay, Firth of Thames, on 17 November 1981 a Terek Sandpiper, which had been picking at random over thick wet mud and runnels, carried out side-to-side sweeping in muddy water for 20 to 30 seconds. It then resumed picking as before. At the time I was watching through a x30 telescope with the bird about 50 m distant. The side-to-side passes of the bill presented a continuous almost stirring motion, the right and left movements being rhythmic and deliberate. The bird then dropped into a deep runnel and was gone from my view. Bill sweeping seems to be used infrequently. Whenever possible I have watched this species, which visits New Zealand in very small numbers annually, since 1964. The bill seems to be admirably adapted for such use. This feeding method is similar to that of the Australian Avocet (Recurvirostra novaehollandiae) , which I have watched at sewage treatment ponds in Australia at Alice Springs in April 1979 and at Werribee, Victoria, in October 1980. I have found no record in the literature of T. cinerea feeding by side-to-side bill sweeping. Ali & Ripley (1969, Handbook o f the birds of India and Pakistan, vol. 2 ) say " When feeding probes with its bill deep into the squelchy ooze almost to the eyes so that the base of the lower mandible and chin usually become bare of feathers due to the constant attrition. Often a morsel is shaken in water and carefully washed before being swallowed . . . . Whether the peculiar upturned shape of the bill has any special adaptive significance remains to be studied." O n the Avocet ( A . avocetfa), they describe feeding thus: " The birds run about and feed on squelchy mudflats or by wading in shallow water. The curious upturned bill is directed obliquely against the ground - like a hockey stick held in play - the curved part skimming the surface. It is then moved back and forth in a rotatory skimming motion churning up the bottom mud along with minute organisms which are collected in the mandibles." The bill of a Terek Sandpiper seems less abruptly upcurved than that of an Avocet and may not be as effective in food gathering in the side sweeping method. This may explain the Terek Sandpiper's infrequent use of a method of feeding for which it appears well suited. BETH BROWN, 39 Red Hill Road, Papakura
THE SPOTLESS CRAKE (Porzana tabuensis) O N AORANGI, POOR KNIGHTS ISLANDS By DEREK ONLEY
ABSTRACT Fewer than 50 adult Spotless Crakes plus 24 chicks and four juveniles were found on Aorangi in February and March 1980. They were confined to the valley and preferred a low mixed forest which is being replaced by a less favoured habitat. Habitat preference is discussed and related to changes in numbers in the past. Clutch size is lower than on the mainland and chick production probably lower. The breeding cycle is long. Some observations on the adult-chick relationship are presented. INTRODUCTION I was on Aorangi, Poor Knights Islands, from 19 February to 9 March 1980. The Poor Knights are a group of about 23 islands and stacks lying in subtropical water 20 km off the east coast of Northland, near the edge of the continental shelf. The rock is rhyolitic breccia of Miocene age. The region has a high annual mean temperature of over 13 "C, with a summer mean cf 20 "C, varying by only a few degrees. Rainfall is probably similar to that of the Mokohinau Islands which have a mean rainfall of 1000 mm (Whitaker 1968). The two main islands, Aorangi and Tawhiti Rahi, were intensively cultivated by Maoris until 1823, when intertribal war resulted in depopulation. The only introduced mammals were pigs, present on Aorangi until their extermination in 1936. Aorangi consists cf a broad north-south running valley in the northern part, bounded by steep cliffs to the west and a sharp peak to the east. The southern part of the island rises to the highest point, Oneho Hill, in broad complex ridges. Figure 1 shows the main peaks and ridges. Most of Aorangi is now covered in forest and the approximate distribution of the main types is shown in Figure I . The following description of the main forest types is* based on work done by G. N. Park and is supplemented by my own notes.
Mixed low forest A low, 3-4 m high canopy is dominated by tawapou, Coprosma macrccurpu and Myrsine divaricuta, forming together 60-80% of the canopy cover. Curmiclmelia forms up to 10% of the canopy with mahoe occasionally as important. Kohekohe forms 15-20% of the canopy in places. Macropiper excelsum is an important understorey
10
ONLEY
NOTORNIS 29
element. The slopes are gentle and the ground is sparsely vegetated and has thick litter and occasional boulders. The canopy is often broken and the probability of whau gap development is high. A whau gap forms when a windfall opens the canopy; whau grows rapidly to fill the gap temporarily. These open places have a thick ground cover of bush grasses and frequent whau to 3 metres and kohekohe, flax, Carmichaelia, Macropiper excelsum and Myrsine divaricata occurring in varying amounts to 1 metre high.
Karo-dominated mixed low forest The canopy is low, 4-5 m, with karo forming about 50% and tawapou, C~prosmamacrocarpa and mahoe forming another 20% of the canopy cover. Overall the ground surface is steep but flat terraced areas are frequent at the old pa site near Urupa Point. Litter accumulation is thick on flat terraces only, with scree on steeper slopes. Macropiper excelsum is frequent in the understorey and some supplejack thickets occur. Tall tawapou-kohekohe forest Tawapou forms 20-60°6 of the canopy and kohekohe 15-30%. Large pohutukawa trees form up to 20% of the canopy in some places, especially just west of Tatua Peak. Canopy height is 6-14 m. Kohekohe. tawapou, karaka, Coprosma macrocarpa, Myrsine divaricata, mahoe and pigeonwood are the main subcanopy species. Lower levels are generally open, although Macropiper excelsum and supplejack form thickets. The ground is open, the litter cover is thick, and slopes are moderate or gentle. The probability of whau gap development is medium. The ground in these gaps is strewn with dead wood, covered with bush grass, a Polygonum species, thistles and nightshade. Whau up to 7 metres high is przsent with, frequently, a thick growth of Myrsine divaricata, Coprosma macrocarpa and mahoe up to 1.5 metres. There is often a fringe of thick Macropiper excelsum and supplejack. Pohutukawa (Broadleaf) forest The canopy is almost exclusively pohutukawa (over 90%) at heights up to 12 m. Mahoe and Coprosma macrocarpa are important elements in the subcanopy (up to 20% each), while the understorey and ground layers are rather sparse. Slopes are steep. Litter accumulation varies considerably. The probability of whau gap development is very low. Kanuka-pohutukawa forest Kanuka forms over 50% of the canopy and pohutukawa about 40%. Canopy height decreases towards the summit of Oneho Hill. Pohutukawa is important ir. the subcanopy (c. 15%) along with Myrsine divaricata and Coprasmu macrocarpa (c. 20% each) away from the summit. The shrub layer is 0.5-1.5 metres high and thick. The ground, in contrast to other forest types, is thickly covered with
1982
SPOTLESS CRAKE
11
nArid
-
0
100
ZOO 300
Karo dominated mixed low fomst
Mixed tow
Fbhutuknwa fomt
-- - -
Main ridges
o
Chick
A
Ldependent juvenile
FIGURE 1 - Distribution and territories of the Spotless Crake on Aorangi, Poor Knights, February-March 1980.
12
NOTORNIS 29
ONLEY
Blechnum capense, dead kanuka, Metrosideros perforata and mosses near the summit and on southern slopes. Litter accumulation is noticeably less than in other forest types. The probability of whau gap development is medium to high only on the lower slopes where the kanuka is large and old. CENSUS METHODS I covered the island several times during our visit and plotted, as accurately as possible in the thick bush, the position of all crakes that were seen or heard. I could soon recognise family groups by the size and number of young and other birds by leg colour (which varied from dull brownish orange to bright orange-red) and behaviour. Adults, chicks and immatures could be distinguished in the field. The positions of the birds are shown in Figure 1. Approximate territory boundaries are shown in Figure 1. The solid lines show well-defined territories where the birds could be individually recognised, occurred regularly, and showed territorial defence where territories adjoined. Dotted lines show territory boundaries that were not completely mapped because the birds were difficult to recognise as individuals or were only infrequently seen. RESULTS I recorded a total of 44 adults, 4 independent juveniles and 24 dependent young. Figure 1 shows 22 definite and 5 possible territories. Territories were occupied by several combinations of birds (Table 1). TABLE 1 - Composition of Spotless Crake groups occupying territories on Aorangi (excluding territories that were possible only) Number of territories 2 adults with young
1 adult with young
3 adults with young 2 adults 1 adult
6
1 II sI 1
I cannot claim a complete precise census, mainly because of the difficulty of finding my position in the thick bush and the lack of a good large-scale map, but I would be surprised if there were more than 50 or less than 40 adult birds on the island. Young were more easily counted as they usually all appeared after 5 or 10 minutes' quiet waiting and the behaviour of adults with young was distinctive. Independent juveniles did not call and often remained well hidden until the last moment (before the boot descended), and so the number of juveniles may well be an underestimate.
13
SPOTLESS CRAKE
DISTRIBUTION AND HABITAT PREFERENCES During our visit Spotless Crakes were found in Puweto Valley, apart from three territories near Urupa Point and a single record of a bird seen west of Oneho Hill (Fig. 1 ) . Within the valley crakes were most numerous in mixed low forest and less so in tawapou-kohekohe forest. Those at Urupa Point occurred in karo-dominant mixed low forest, but none occurred in a similar forest west of the valley cliffs. Apart from one record, no crakes were seen in pohutukawa (broadleaf) forest or kanukapohutukawa forest. Table 2 shows the numbers of adults and definite territories recorded in each vegetation type. Territories were closer together in mixed low forest than in tawapou-kohekohe forest and closely pac,ked in the small part of the karo forest occupied. Territorial disputes were observed in 15 of 17 territories in mixed low forest, in all three in karo-dominated forest, and in none of the six in tawapou-kohekohe forest. TABLE 2 - Number of adults and definite territories in each vegetation type on Aorangi No. adults
$ total
Mixed low f o r e s t
28
64
Tawapou-kohekohe f o r e s t
10
23
Karo, mixed low f o r e s t
5
11
Pohu tukawa ( b r o a d l e a f ) f o r e s t
1
2
Kanuka-pohutukawa
0
0
forest
Totals
adults
44
DISCUSSION Habitat preferences Buddle (1941 and 1951) observed that Spotless Crakes ate worms, spiders, beetles and insect larvae " from the litter on the forest floor. All my feeding observations were of birds turning over litter in search of such prey. The preference for the valley floor forests may be due to the quality and quantity of litter under these forests and the food therein. Litter derived from a canopy of 90% pohutukawa and 50% kanuka-pohutukawa will certainly be different from that of the valley forests and may have a different or scarcer invertebrate fauna. The kanuka-pohutukawa forest of the Oneho Hill area has a thick ground cover of Blechnum capense, poor litter "
14
ONLEY
NOTORNIS 29
accumulation and extensive areas of mosses, all of which suggest that ground litter foraging would be less successful than in the valley floor. Within the valley crakes favoured flatter areas, especially those near the stream, along Maori terrace edges and on the wide shoulders of small ridges. Those territories in karo-dominated forest near Urupa Point, although on an overall steep slope, included areas of flatter Maori terracing. Litter accumulation is generally higher on flatter sites and the mobility of litter on steep slopes prevents the build-up of high invertebrate numbers. A11 but four of the territories included some thick low cover, even if only a few metres square, consisting of bush grasses, a Polygonum species and introduced plants over dead fallen trees. This low cover was invariably associated with whau gaps. Crakes were most numerous in areas where whau gap development was most extensive (mixed low forest), less numerous where whau gap development was medium (tawapou-kohekohe forest) and absent where gap development was low (pohutukawa forest). The only place where whau gap development was medium to high and crakes were absent was tall kanukapohutukawa forest on the lower slopes of Oneho Hill. Buddle (1941 and 1951) mentioned the association of nesting crakes with bush grass. During my visit adults with small young were found near or in thick ground cover. So although they forage in the open, some thick cover in the territory is usually required for nesting. Possible predators of eggs and young on Aorangi are tuataras, harriers and kingfishers. ( I have seen both harriers and kingfishers taking small Pukeko chicks on the mainland.) Roy Bell (unpubl. ms. Diary 1908-1911) noted that Spotless Crakes on Raoul Island preferred to nest in swampy areas, emerging to drier open habitats after the young had hatched. He suggested this may be a response to predation by rats. O n mainland New Zealand, where there are yet more predators, crakes are always associated with thick marshy cover. With increasing predator pressure there seems to be an increasing trend for crakes to live in areas of thick cover. Changes in cruke numbers on Aorangi The first mention of Spotless Crakes on Aorangi was by Hamilton (1925), who in 1924 saw one bird, which suggests that numbers were low. Not until December 1938 was any estimate of numbers attempted, when Fleming (1941) estimated 12 pairs on Aorangi. He and his party, which included Buddle, camped on the southern part of the island and visited Puweto Valley and the northern coast only once, and so actual numbers could well have been higher (Fleming, pers. comm.). Two years later, in November 1940, Buddle (1941) found crakes well distributed over the greater part of the island (he mentions nests in sedges at high altitude) with the highest concentrations being near the saddle and throughout the valley. He estimated 100 pairs and said that the evident increase in the numbers of the Crake was "
1982
SPOTLESS CRAKE
15
hardly believable." Although it is apparent that crakes were increasing rapidly at this time and had an extensive range on the island, the evidence for an increase from 12 to 100 pairs cannot be taken literally. Although Chambers (1955) mentioned six pairs and Kinsky & Sibson (1959) " quickly located " six pairs in January 1956, both these visits were short and not designed to count crakes, and so actual numbers were certainly higher. Bartle (pers. comm.) estimated that numbers were constant between December 1963 and January 1978 at about 16 pairs. My estimate of 22 territories would seem to be in line with Bartle's estimates for the 1960s and 1970s, suggesting that little change has taken place in recent years. Briefly; numbers appeared to be low in the 1920s, with a rapid increase in the late 1930s to very high numbers in 1940. By the mid-1950s, numbers had probably fallen to a level that remained throughout the 1960s and 1970s. Rapid and extensive habitat changes have taken place on Aorangi since Maori occupation ceased. In 1823 most of the island was clear and all forest trees are post-Maori except for scattered large pohutukawas and a group of puriris and associated trees under the valley cliffs (G. N. Park, pers. comm.). When the Maoris left, pigs remained, ranging freely over the island and restricting regeneration, although low mixed forest, pohutukawa forest and kanuka forest probably grew slowly. Hamilton visited the island during this stage, noting only one crake. Pigs not only restricted regeneration of low mixed forest but prevented, by constantly digging and turning over the ground, the accumulation of litter and build-up of invertebrates, destroyed low nesting cover and probably directly interfered with nesting crakes. After the pigs were removed in 1936, a low mixed forest developed throughout the valley, favouring the rapid increase in crake numbers suggested by Buddle 4 years after the extermination. Buddle's reason for the large increase in crake numbers was that the birds favoured the extensive bush grass that grew after the pigs were exterminated. The ill effects of pigs on the crakes had been more on their favoured habitat and litter feeding than on their nesting habitat. The low mixed forest is gradually being replaced by tawapoukohekohe forest, and the decline in crake numbers by the 1950s might be attributed to the reduction in their favoured habitat. This replacement is expected to continue and crake numbers can be expected to decline in the future. A further increase in Buller's Shearwaters, which have doubled since 1964 (Bartle, pers. comm.), may also affect the crakes as their digging disperses much of the leaf litter. BREEDING This section brings together notes from my visit to Aorangi on breeding and behaviour of crakes and such scattered references as I could find in the National Museum, Wellington.
16
ONLEY
NOTORNIS 29
Territory During our visit the breeding season was well advanced, with newly hatched to independent young present. Birds with and without young were strongly territorial. Boundary disputes consisted of one bird calling a short sharp trill preceded by a low quiet ooh ooh ooh. The non-calling bird left, often rapidly, and there was no pursuit. Occasionally the ooh ooh noise was enough to deter another bird, and several times a crake that had moved into a territory occupied by a bird with young would rapidly depart when the territory holder approached without any aggressive behaviour. It seems likely that boundaries between territories were well known and well defined because boundary disputes of any intensity were unusual during our visit, bzing most frequent in the densely populated areas near the stream. Territorial calls are apparently more common earlier in the season, and chasing is frequent (Bartle, pers. comm.). Territory size could be estimated only in the area near the stream, where boundaries between pairs abutted and the smallest territory was 45 x 50 metres, but birds ranged more widely in the tawapou-kohekohe forest. Birds with small young were closely associated with small areas of low, thick cover, often not moving more than 10 metres. Bryant & Amos (1949) estimated that territory size near Melbourne, Australia, was as small as 50 sq. ft. (c. 5 m2). Hadden (1970) at Waingaro, New Zealand, estimated the cover in the nesting territory to be no more than 12 yards square (c. 120 m2), although some open land might also be used. These areas of cover are comparable to areas of thick ground cover within the territories on Aorangi. TABLE 3 - Frequency of different clutch sizes in New Zealand and Australia Reference
1982
SPOTLESS CRAKE
17
Breeding season On Aorangi, Buddle (1941) suggested that egg laying began in October and continued at least until early December. In February 1980 1 found newly hatched chicks in four crake territories, and so laying continues until late January. On mainland New Zealand, Hadden (1970) found that egg laying started at the end of August and continued until mid-November, whereas Fraser (1972) found nests with eggs in mid-January and suggested that laying took place from August to January. Roy Bell on Raoul Island (1908-1911) said that nests were built from the first week in September but that laying did not start until mid-October and continued until 10 December. Most birds laid between 10 and 20 November. He suggested that later nesting took place in high water years. In Australia, dates from Bryant & Amos (1949) and Fletcher (1914) suggest that the breeding season is similar, ranging from late September to late January. Fletcher also suggested that laying may continue into February. depending on seasonal rains. Clutch size The commonest clutch size is three, with two frequent on Aorangi and the Kermadec Islands, and four frequent in Australia (Table 3 ) . Although the sample is small there is a suggestion that clutches are larger on mainland New Zealand than on the islands and larger in Australia than in New Zealand. Australian handbooks also support the idea of a larger clutch size for Australia and Tasmania than for New Zealand - 4-6 (Macdonald 1973), 4-7 (Frith 1969), 4-6 (Readers Digest 1976). Smaller clutches on islands have been demonstrated for many species (Bull & Whitaker 1975). Buddle also suggested a reduction in clutch size from three to two later in the season on Aorangi. Incubation Two birds. presumably both parents, incubate (Hadden, in OSNZ nest records). From the date of the last egg to hatching inclusive is 21 days (Hadden, in OSNZ nest reccrds) . Fraser (1972) gave c. 19 days for incubation (not defined) and Fletcher (1914) gave 18 days. Young The young stay in the nest for up to 2 days (Hadden, in OSNZ nest records, Buddle, 1941). Roy Bell said that on Raoul Island the young - leave the nest on hatching for other nests on dry land. Information on the rate of growth of young after leaving the nest is sparse. Roy Bell says that the young grow very slowly; those hatching at the end of December were only half-grown by April.
18
ONLEY
NOTORNIS 29
During my 18 days on Aorangi, it was difficult to see any growth changes in the chicks. Ruddle estimated the earliest nesting on Aorangi would be in October, and newly independent young were present during our visit in February. Obviously, the adults attend the chicks for a long period, probably 4-5 months. Newly hatched chicks on Aorangi during our visit at the end of February would therefore not reach independence until June or 1uly. Spotless Crakes take a long time to raise young. Nest building, preceded by an unknown time when territories are established, starts several weeks before laying. Incubation takes 3 weeks and the young may be accompanied for 4-5 months before becoming independent a total of 5-6 months. Egg laying from August to late January suggests that some part of the population is involved in raising chicks for most of the year. The immature plumage was overall browner than that of the adult, lacking the blue sheen on the head and breast of the adult and the golden brown on the back of the adult. The chin was white with whitish striations on the breast, and legs were a light pinkish brown but rather variable, occasionally still dark as in the downy chicks. Immatures were noticeably smaller. An immature female in the National Museum weighed 28 g in contrast to adult female weights of 43.0 g and 43.2 g. Breeding success Apart from newly hatched chicks, the young seen during our visit to Aorangi were impossible to age, but they were divided into four categories on the following basis: 1 . Small - weak twittering call, not very mobile, staying close to the adult, occasionally being brooded. 2. Medium - single thin r i s i ~ gcall, constantly repeated. Downy. Mobile, following adult often at a distance of 20-30 metres. 3. Old - fully feathered but still accompanying adult. No call. 4. Independent - immature plumage, not associated with an adult.
Table 4 shows the numbers of young seen on Aorangi in February and March 1980. The large number of " medium " young seen is to be expected as this category probably includes chicks of the greatest age range. A very rough measure of breeding success can be derived from these figures. It is unlikely that all the young seen would reach independence, and two " medium" young from a group of three were lost and two newly hatched young disappeared during our visit. Also the small number of " old " young seen were singles. If 20 of the young reached independence and if all 44 adults counted had bred, less than one half a young per adult would reach independence in a year.
1982
SPOTLESS CRAKE
19
TABLE 4 - Frequency of different-sized groups of young seen on Aorangi, February-March 1 9 8 0
Bartle's records for the 1960s and this study suggest that the population of crakes is stable at about 40 individuals. A production of 20 young a year suggests that mortality or movement accounts for about one-third of the over-winter population of adults and juveniles if numbers are to remain stable.
Attendance of the young by the adults During our visit to Aorangi the number of adults attending young varied according to the age of the juvenile birds. Small newly hatched young were always accompanied by two adults. During the medium " stage one adult stops accompanying the young, and so by the time the young are fully feathered, only one adult is in attendance. There was one record of three adults accompanying young. The assistance of I-year-old birds in feeding downy young has been noted for the Tasmanian Native Hrn (Ridpath 1968). The behaviour of young and adults throws some light on this seeming lack of parental attention. Small young were not very mobile and stayed close to one parent and were occasionally brooded. The other adult sounded the alarm, at which the young stopped calling, and investigated intruders. As the young became older they foraged widely in the open forest floor, called constantly, and often moved "
20
ONLEY
NOTORNIS 29
20-30 metres away from the adult. The adults did not call to the young and did not respond to intense calling by the young by returning to them. It seemed, in fact, as though the young followed the adult round the territory. At an alarm call from the adult the young would run rapidly for several metres, sometimes managing to hide and at other times quickly giving up the attempt. Often they continued to call loudly while the adult was calling the alarm. Fully feathered young accompanied the adult at distances of up to 40 metres and were silent. They seldom heeded the adults' alarm calls and continued feeding unconcerned. So, after the initial stage, when there is some division in the adults' roles attending the young, the adults' role is minimal and there seems little need for two adults to look after the young. In two cases the adult not accompanying the young was still in the breeding territory, and one of them was actively defending the territory against a neighbouring crake. In six other cases of a single adult attending young no other adult could be found in the territory. One territory was definitely occupied by a single adult and three other less well defined territories had a single adult only. Whether these birds were the missing parents is not known. CONCLUSION Although Spotless Crakes are being found in many places on mainland New Zealand with the help of taped calls, the reduction in their marshland habitat is a cause for concern. Aspects of this study illustrate the danger in New Zealand of assuming that our island sanctuaries and reserves are a substitute for adequate mainland reserves. Island systems, especially those of smaller islands, are very specialised, and processes on islands may be very different. This study suggests that for Spotless Crakes, and possibly other species, islands such as Aorangi are not ideal refuges because 1. The population is small - between 40 and 50 individuals. 2. The crake reproductive rate may be slow - clutches are smaller than cn the mainland and less than one chick per pair is raised to independence for each breeding attempt. 3 . Their favoured habitat is gradually changing as the forest matures. A small population and a low reproductive rate mean that the birds are susceptible to catastrophes such as bad weather or habitat destruction and cannot easily recover their numbers. With the changing habitat, the occupation of Aorangi by Spotless Crakes may well be temporary. ACKNOWLEDGEMENTS Thanks to Sandy Bartle for organising the trip to Aorangi and to those who also shared the motor camp, cave, tent and island; Shane Cotter, Paul Sagar and Toy Sagar. The Department of Lands and
1982
SPOTLESS CRAKE
21
Survey gave me financial support. Paul Sagar and Sandy Bartle read previous manuscripts. LITERATURE
CITED
BELL,
R. Unpubl. MS Diary 1908-1911. Alexander Turnbull Library. Extracted i n Files of W. R. 6 . Oliver. National Museum of New Zealand. BRYANT, C. E.; AMOS, B. 1949. Notes on the crakes of the genus Porzana around Melbourne. Victoria. Emu 48: 249-275. BUDDLE G. A. 1941. The birds of the Poor K ~ i g h t s . Emu 41: 56-68. BUDDLE: G. A. 1951. B i r d secrets of New Zealind. A. H. & A. W. Reed. BULL, P. C.; WHITAKER, A. H. 1975. The Amphibians reptiles birds and mammals. pp. 231-276 i n Kuschel, G. (ed) Biogeography a n d ' ~ o l o g y 'in New Zealand. The Hague: W. Junk. CHAMBERS, B. S. 1955. Birds of the Poor Knights Islands. Tane 17: 66-67. READERS DIGEST BOOK OF AUSTRALIAN BIROS. 1976. Sydney: Readers Digest Services. FLEMING, C. A. 1941. Annual rsport of the Ornithological Society of New Zealand. NO. 1 . FLETCHER, J. A. 1914. Field notes on the Spotless Crake (Porzana immaculata). Emu 13: 197.22 FRASER E. 1972. Some notes on Spotless Crake. Notornis 19: 87-88. FRITH, 'H. J. (Ed.) 1969. Birds i n the Auqtralian high country. Sydney: A. H. & A. W. Reed. HADDEN, D. 1970. Notes on the Spotless Crake i n the Waingaro District. Notornis 17: 200-213. HAMILTON, H. 1925. Birds of the Poor Knights and Hen Islands. NZ J. Sci. Tech. 8: 15-18. KINSKY, F. C.; SIBSON, R. 6 . 1959. Notes on the birds of the Poor Knights Islands. Notornis 13: 132-133. MacDONALD, J. D. 1973. Birds of Australia. Sydney: A. H: & A. W. R*. RIDPATH, M. G. 1964. The Tasmanian Native Hen. AustralIan Natural H ~ s t o r y 14: 346-350. WHITAKER, A. H The lizards of the Poor Knights Islands. NZ J. Sci. 11: 623-651.
GLOSSARY OF ANIMAL AND PLANT NAMES Tuatara Buller's Shearwater Harrier Pukeko Tasmanian Native Hen Kingfisher Flax Supplejack Pigeonwood Kawakawa Mahoe Karo Kanuka Pohutukawa Whau Karaka Kohekohe Tawapou
Puriri Bush grass - 2 species
Sphenodon punctatus Puffinus bulleri Circus approximans Porphyrio porphyrio Gallinula mortierii Halcyon sancta Blechnum capense Phormium tenax Rhipogonum scandens Hedycarya arborea Macropiper excelsum Melicytus ramiflorus Pittosporum crassifoliurn Leptospermum ericoides Metrosideros excelsa Metrosideros perforata Elttelea srborescens Corynocarpus laevigatus Dysoxylum spectabile Planchonella novo-zeurlandica Myrsine divaricata Coprosma macrocarpa Vitex lucens Oplisuenus undulatifolius Microlsena stipoida
D. J . O N L E Y , Madman's Farm, Little Totara River, Private Bag. Westport
SHORT NOTE
SHORT NOTES WOOD DUCK IN MARLBOROUGH During the second weekend of the duck-shooting season in April 1980, Wayne Johnston of Blenheim was shooting with his father, Max, at the edge of the Wairau River near the mouth of the Waikakaho. There, some Catchment Board stop banks between the Tua Marina Track and the main river form ponds surrounded by willows. Decoys had been set out on the pond about 9 to 10 a.m. on this sunny morning, when a duck came in against the light towards Wayne. Max was better sighted across the prevailing light and was impressed by the speed of the wing beat, but Wayne's quick shot was fatal. The duck was so unusual that they handed it to the local Acclimatisation Society Ranger. Since then the Ranger has heard reports of a strange lightcoloured duck seen in the Spring Creek area before the shooting season. The two areas are only about 4 miles apart. The duck has been identified by Brian Bell as a male Australian Wood Duck or Maned Goose (Chenont-tta jubata), a small goose-like duck. The Annotated Checklist mentions only two records of stragglers from Glendhu, Wanaka (1910) and Orawia, Southland (1944), while the New guide to the birds of New Zealand gives a further two records, both in 1944, from Linwood and Wairaki in Southland. PAULINE JENKINS, 234 Howick Road, Blenheim
GANNETS FEEDING BEHIND TRAWLER During the 1981 w i ~ t e r ,while we were trawling in Tasman Bay, there were up to 100 Australasian Gannets (Sula bassana serrafor) within sight of the boat at any one time. As usual they were taking no notice of the boat but sitting on the water or feeding. As we were picking up a morning trawl on 6 June, two adult Gannets joined the Shy Mollymawks (Diomedea caufa), Cape Pigeons (Daption capense) and Black-backed Gulls (Larus dominicanus) that were feeding on the offal and small moribund fish which fall out of the cod end as it comes on board. The Gannets were diving for both the offal and small fish, spitting out the offal and swallowing the fish. We noticed this behaviour several times during the trip and each time there were only one or two Gannets. This is the only time in several years that we have seen or heard of Gannets feeding in this manner. Usually they cast an eye at the boat and the accompanying birds and then continue on their way. J. M. HAWKINS, 772 Atawhai Drive, Nelson
THE SIZE OF THE SOOTY SHEARWATER POPULATION AT THE SNARES ISLANDS, NEW ZEALAND* By JOHN WARHAM and GRAHAM J. WILSON ABSTRACT The size of the Sooty Shearwater (Pufinus griseus) population at the Snares Islands was estimated by counting burrows in the main vegetation types. Some 3 287 000 burrows were calculated for Main Island, the highest densities being 1.9/m2 in Poa meadows, with 1.2/m2 under the trees of the Oleuria forest. Most burrows were occupied but data on rates of occupation by breeding birds were not satisfactory. Assuming a 75% occupancy rate. we get a total population of about 2 750 000 burrow-holding peirs on the 328 ha of the two largest islands. INTRODUCTION In recent years there has been a marked increase in attempts to estimate seabird populations. Much of this effort has been to establish base-line information in the face of threats or potential threats posed by new developments in the exploitation of marine resources. Such threats exist in the southern hemisphere as in the northern one, and with the prospect of further explcration oil drilling in the ' Great South Basin,' phosphate mining on the Chatham Rise and similar projects in New Zealand seas, there is a clear need for information on the sizes of New Zealand seabird populations. Such information is also needed to evaluate the mortality occurring in the northern hemisphere, to which many southern species migrate after breeding. For example, King et ul. (1979) ~stimatedthat the shearwaters Pufinus griseus and P. tenuirostris comprised 27% of the annual mortality of 214 000 - 715 000 seabirds killed by the Pixific cffshore salmon gill-net fishery. More recent data (Ainley et cl., in press) suggest that these mortality estimates should be doubled. Contamination by chlorinated hydrocarbons and PCBs while in northern waters may also be important - summary in Bourne (1976). Probably the most important of New Zealand seabirds in terms of cumbers and biomass is the Sooty Shearwater, or New Zealand Muttonbird (Pujrfinus griseus). This breeds in small numbers as far north as Three Kings Islands (34"S), more commonly in Cook Strait (41°S), but its centre of abundance is south of the South Island - in Foveaux Strait and on islands around Stewart Island. The Snares Islands at 4 8 3 166"E have big populations. Further south at Antipodes, " University of Canterbury Snares Islands Expeditions Paper No. 45
24
WARHAM & WILSON
NOTORNIS 29
Campbell, the Auckland Group and Macquarie Islands only quite small colonies occur. The Sooty Shearwater population at The Snares appears to be in a very healthy state with no evidence of the derelict burrows that are seen at petrel colonies in decline. Many eggs are laid on the surface and abandoned. We found that these averaged narrower than those laid underground and were presumably produced by young birds (Serventy 1967): such surface eggs appear to be typical of shearwater colonies in good fettle. On the two largest islands in the group, Main Island (280 ha) and Broughton Island (48 h a ) , the whole of the peaty soil is burrowed, except where the soil is too thin, areas where the bare peat has been eroded by wind and rain, where the ground is occupied by colonies of Snares Crested Penguins (Eudyptes robustus) or where the soil becomes waterloggzd and burrows flooded after heavy rain. For the purpose of our census only four vegetation types needed consideration - minor types were poorly burrowed and 38 ha of the Main Island and 13 ha of Broughton Island were bare of soil. The deep peat supports a central forest of Olearia lyallii forming an almost closed canopy at about 6 m. This is surrounded by meadows of Poa tussocks which extend outwards to the coastal cliffs. There are small patches of forest dominated by the tree Senecio stewartiae and thickets of the coastal bush Hebe ellipfica. These vegetation types are rather clearly defined and with the help of a series of stereo-aerial photographs and a ground survey, a detailed vegetation map of the Main Island was prepared by C. H. Hay in 1970. This shows the planar areas for each vegetation type. In 1974 a similar map of Broughton Island was made by H. A. Best. Conditions for burrowing vary with vegetation type. Beneath Olearia the ground is clear of understorey and leaf litter as the shearwaters remove all loose vegetation for their nest linings but the soil is interlaced with stout roots which hinder burrowing. Similar conditions apply to the Senecio areas, but these trees tend to occur on rather waterlogged soils. The Poa tussocks are of two types. Poa tennantiana plants are about 0.6 m across and 0.6 m high and form meadows surrounding the central forest. These meadows are themselves surrounded by a zone of Pca astonii whose plants tend to form large stools 1-1.5 m high and 0.5 m diameter and extend down steep coastal slopes wherever there is enough soil. This grass does well on quite shallow soils but on steep slopes offers fewer opportunities for burrowing shearwaters, although it is suitable for smaller species of petrels. METHODS To lessen damage through breakthrough in the extensively burrowed ground, while permitting ready movement around the island, an access track was cleared and marked at the start of the study (see map in Warham, 1967).
1982
SOOTY SHEARWATER
25
An estimate of the number of burrows was attempted by sampling burrow densities in the four vegetation types and computing the total number of burrows. On level or near-level ground burrow densities in the two Poa zones were found to be similar and the figures for these have been aggregated. In the summers of 1969/70 and 1970/71 a series of 10 m x 10 m quadrats was pegged out and the number of used burrow entrances counted. Quadrats were located at random within the various zones and burrow entrances counted inside the tape and marked with a short stick as counted. Those on the boundaries were included if the nest chambers were clearly within the quadrat. Used burrows were those whose entrances were free from obstructions, cobwebs and the like, but the only instances of disuse were some burrows in dense Hebe which were blocked with sticks and whose birds had clearly been excluded by the rapid growth of the bushes since the previous season. The planar areas for each vegetation type were measured from the vegetation map by planimeter. Most parts of the Main Island were sampled, and in the Olearia zone some quadrats were close to forest margins and others towards the centre of the island. In the small area of Senecio the few quadrats examined were not chosen randomly but selected to include a highdensity area, a poorly drained one, one near the Olearia-Senecio boundary and one at random, to show all the extremes in this vegetation type. No allowance was made for slope in computing burrow densities because, although the ground slopes gently to the east and is dissected by shallow gullies (see Fleming et al. 1953), checks on gully sides showed that in planar projection burrow densities were not greater than on more level ground: such slopes tended to be wet and burrow densities low. On gently sloping ground burrow densities were higher than on flat ground - an effect of better drainage perhaps - but variations between quadrats were far greater than could be accounted for by slope alone. Attempts were made to get information on burrow occupancy from three kinds of data: (a) from the contents of nest chambers accidentally broken into between 3 December 1970 and 24 February 1971, (b) from the disturbance of fences of fine wires set across burrow entrances and checked each morning from 4 to 22 January 1971 while hatching was in progress and (c) from the state of 100 burrows chosen at random under the Olearia forest and examined daily during April 1972 by C. 1. and D. S. Horning. Examination was by feeling an occupant by hand or with a long wire, by noting fresh digging, by finding nestling down in the tunnel and by seeing birds entering burrows. RESULTS The total number of burrows on Main Island was estimated at 3 287 000. See Table I . The highest burrow densities of nearly 2/m2 occurred among tusscck grass. This is probably because burrowing is
TABLE 1 - Sooty Shearwater burrow densities on M a i n Island, Snares Islands Vegetation tYPf
Burrows/100
n
m2
+ 1 S.E.
Olearia forest
22
116.1+.7.10
Poa t u s s o c k -
15
194.8+
Senecio f o r e s t
4
Hebe elliptica -
10
5.83
45.8 68.3
+ 11.48
Area covered vegetation (ha)
E s t i m a t e d no. of burrows*
Range
by
40
-
155
146.80
1 704 000
1 496,000
159
-
240
76.51
1 490 000
1 4 0 1 000
76
3.32
15 000
11.45
78 000
5 1 900
238.08
3 287 000
2 963 900
20 10
-
129
TOTAL
F i g u r e s rounded t o t h e n e a r e s t 1 , 0 0 0
95% c o n f i d e n c e i n t e r v a l s *t
'i' + 2 standard e r r o r s -
-
s
1 913 000
$
1 580 000
Lp
9 1 300
0
3 599 300
z
in z
1982
SOOTY SHEARWATER
27
easier there with less obstruction from roots. Variability between quadrats was also least in the tussock, probably because the plants themselves are very evenly distributed and of rather constant size. Densities fcir the Senecio forest were low, partly owing to the damp ground conditions. The great variability in the counts in the Hebe quadrats directly reflected their cpenness, with dense, nearly impenetrable thickets being almost unused. Figures for Broughton Island are less reliable as no ground sampling was done there, but the shearwater flocks over this island appeared as dense as those over the Main Island. If we assume similar burrow densities in similar vegetation, the 13.5 ha of Olearia/Senecio forest would have 131 900 burrows, the 14.3 ha of Poa 235 600 burrows and the 2.6 ha of Hcbe 14 800 burrows, some 382 000 burrows in all for that island. Some previous figures for shearwater burrow densities have been published. Rowan (1952) estimated ~ e s densities t of the Greater Shearwater (Pufinus grcvis) at Nightingale Island, Tristan da Cunha, from 15 sample counts in three vegetation types. She found the highest density of 1.5 nests per square yard (1.8/m2) in short tussock and sedge, falling to about 1 per square yard (1.2/m2) in very heavy tussock 2-2.5 m high, and only about 0.6 nests per square yard (0.7/m2) under groves of Phylica trees. She estimated an average burrow density of 1.2/m2. The same figure was gained for a colony of the Short-tailed Shearwater (P. t.muirosfris) from 10 sample counts in the two predominant vegetation types at Cat Island, Tasmania (Warham 1960). Recently Skira & Wapstra (1980) estimated much lower burrow densities (0.210.81/m2) at a number of Puffinus tenuirosfris colonies using a line transect method. For this same Australian species Norman & Harris (1981) reported rather similar densities at Rabbit Island, Victoria, of 0.3-0.6/m2 but Harris & Bode (1981) estimated lower densities of burrows (0.05-0.47/m2) at the small and rather disturbed colonies in Phillip Island, Victoria. At his small colony of 295 nests of Sooty Shearwaters spread over 420 m2 Richdale (1963, p. 94) calculated the burrow density at 0.7/m2. Wormell (1976) counted occupied burrows of Manx Shearwaters (P. puffinus) on Rhum in 36 sample plots, finding a mean density of 150 burrows per square chain (0.37/m2), but densities there are low, partly because the birds burrow under boulders (Bourne, pers. comm.) . Very recently Floyd & Swanson ( i n press) have estimated breeding success and population size and density of the Wedge-tailed Shearwater ( P . pacificus) at Muttonbird Island, New South Wales, and compered their findings with those from other Australian colonies of that species. These authors also found that burrow densities varied with vegetation type, their mean being 0.51/m2 with 25-34% of nests occupied.
28
WARHAM & WILSON
NOTORNIS 29
None of these figures approaches the burrow density of nearly 2/m2 calculated for Poa tussock at The Snares. The figures for the total numbers are also smaller. Rowan (1952) estimated 2 million Greater Shearwater burrows on Nightingale's 200 ha, but thought that this was an underestimate. Elliott (1970) examined the same species on Gough Island, computing some 600 000 square yards (501 600 m2) as being burrowed, found about 1 chick per square yard (1.2/m2) and hence a breeding population of 600 000 pairs. Warham (1960) calculated that there were 250 000 burrows of P. terzuirostris on Cat Island, Bass Strait (49 h a ) , and pointed out that there were much larger colonies on bigger islands in the area. Wormell's estimate for occupied burrows on Rhum, stated to be the largest colony of the Manx Shearwater in Britain, was 116 100. The main problem in estimating population sizes by burrow counts lies in the difficulty of relating burrow numbers to the numbers of birds. At the Snares Islands it is quite impossible to establish occupancy by direct inspection for many burrows have long twisting tunnels leading to chambers 0.5-1 m below ground and quite hidden from view. The difficulties are exacerbated by the sensitivity of Sooty Shearwaters to interference. Birds disturbed on eggs frequently desert, and regular handling such as seems possible with Manx Shearwaters is not feasible. None of our attempts to determine occupancy rates was entirely satisfactory. Of 54 nesting chambers accidentally broken into during other work, 30, or 56%, contained eggs or chicks. Fourteen of the others were empty by day and 10 contained one or two adults without egg or chick. This latter group would have included non-breeders and failed breeders. In all, 74% of nests were occupied by day. All 31 fenced burrows examined over a 20-day period had their fences disturbed. At 23 burrows the fences were displaced on 10-18 days (mean 14.2 days); the rest were displaced much less often (0-9 days), mean 4.4 days. If entry about every other day is taken to mean occupancy, then 74% of the sample were occupied during that period. However, some of the more disturbed burrows could have been visited by prospecting non-breeders while some of the less disturbed ones could have held incubating birds, as these often fast for 9 days during their incubating stints (pers. obs.) . More useful are the data from the 100 burrows examined in April 1972, of which at least 86 had occupants. This figure is significant because by April non-breeding and failed breeders have left the island. Richdale (1963, p. SO), for example, found no 'unemployed' birds ashore by day after 8 March and believed that the regimes of such birds were similar to those found for P. tenuirosfris (see Serventy 1967). Disturbances in April must therefore have been caused by parents or by chicks. The probable source of error here, of unknown magnitude, is due to chicks that failed to reach the sea and sheltered in another burrow at daybreak. Skira & Wapstra's figures for Short-tailed Shear-
1982
SOGTY SHEARWATER
29
water chicks in burrows during mid- to late March are much lower, their highest rates of 51.2 and 64.7% being at colonies not subjected to muttonbird harvesting. An occupancy rate of 86% towards the end of the breeding season is very high. However, there was indirect evidence for a high level of burrow use at these islands because (a) large numbers of eggs were laid on the surface and abandoned, which seems unlikely to occur if nesting chambers were readily available, (b) disused or cobwebbed burrows were very seldom seen, (c) even unsuitable ground susceptible to flooding still attracted some burrowers and (d) physical competition occurred for burrows with the eviction of intruders and some fighting. Occupants defended an area around the entrance and this may have acted as a spacing-out mechanism. Deterring others from digging too close to existing nests would reduce the risk of burrow collapse. Nevertheless, we regard none of these figures for burrow occupancy 2s being sufficiently reliable, the sizes of the samples being much tco small. Clearly a simple method for determining the presence of birds underground is needed. Fibre-optics are almost useless in the tortuous and obstructed tunnels and the most promising technique we are now investigating is to detect petrels by their heartbeats or breathing using a sensitive microphone. The huge size of some of the southern shearwater populations has long been known. During his remarkable circumnavigation of Tasmania in 1798, Matthew Flinders (1801) described an encounter with a flock of P. tenuirostris which he calculated as containing 151 500 000 birds and requiring 186 geographic square miles of ground for their burrows - he allowed a square yard for each. Although Flinders' arithmetic was faulty - the correct figure on his data is 132 000 000 birds (Campbell 1900). the total is still impressive. So is the size of The Snares Island Sooty Shearwater population. Even at a 75% occupancy rate, some 2 750 000 pairs must be based on Main and Broughton Islands together with an unknown number of burrowless non-breeders. At a mean weight of 800 g (SD 74.5 g; n = 154), a 75% occupancy would represent a biomass of 4 400 000 kg. The density of birds is remarkable too, for the above biomass requires only 328 ha of ground for nesting, while, in addition to the shearwaters, some hundreds of thousands of other petrels and of penguins also breed there. Indeed, this small area supports a bird population similar in size to that of the whole of the seabird population of Britain and Ireland - some 3 million pairs (Cramp et al., 1974) and although the shearwaters can travel far for food, the surrounding seas must be highly productive to support such a biomass. In turn, the nutrients brought ashore are believed to be important for the plants (Fineran 1969) and doubtless rainwater run-off fertilises the surrounding seas, although we have no data on that. With populations of such magnitude it is not surprising that
30
WARHAM
€t
WILSON
NOTORNIS 29
the southern shearwaters P. gratlis, tenuirostris and griseus can come to dominate the northern waters into which they migrate during the southern winter. For example, Wiens & Scott (1975), using a simulation model approach, calculated that the Sooty Shearwater is the major seabird consumer during its fall passage off the Oregon coast, taking s anchovies during its 2 months' stay. some 24 000 metric t ~ n of ACKNOWLEDGEMENTS This work was supported by the Nufiield Foundation and the New Zealand University Grants Committee. W. R. P. Bourne, P. C. Harper, M. P. Harris and C. L. McLay kindly commented on early drafts of the paper, and we are grateful for the survey work done by H. A. Best and C. H. Hay. LITERATURE CITED AINLEY, D. G.; DE GANGE, A. R.; JONES, L. L.; BEACH, R. J. ( i n press). Mortality of seabirds i n hiah seas salmon ail1 nets. Fisherv Bull BOURNE, W. R. p. 1976. Seabjrds and polluiion. Pp. 403-502 i n Marine Pollution ( R . Johnston, Ed.). London: Academic Press. CAMPBELL, A. J. 1900. Nests and eggs of Australian birds. Shefield, privately printed. CRAMP. S.: BOURNE. W. R. P.: SAUNDERS. D. 1974. The seabirds of Britain and Ireland. iondon: ~ o l l i n s ' . ELLIOTF, C. C. H. 1970. Ecological considerations and the possible significance of weight variations i n the chicks of the Great Shearwater on Gough Island. Ostrich, Suppl. 8:
--- - - . 385.7'26 .
FINERAN, B. A. 1969. The flora of the Snares Islands, New Zealand. Trans. Roy. Soc. NZ Bot. 3: 237-270. FLEMING, C. A.; REED, J. J.; HARRIS, W. F. 1953. The geology of the Snares Islands. D.S.I.R. Caoe Exo. Ser. Eull. 13: 1-42 FLINDERS, M.' 1801. Observations on the coasts of Van Diemen's Land, on Bass's Strait and its Islands, and on part of the coasts of New South Wales. London. FLOYD, R. B.; SWANSON, N. M . ( i n press). Wedge-tailed Shearwaters on Muttonbird Island. an estimate of the breeding success and the breeding population. Emu. HARRIS, M . B.; BODE, K. G. 1981. Populations of L i t t l e Penguins, Short-tailed Shearwaters and other seabirds on P h i l l i p Island, Victoria, 1978. Emu 81: 20-28. KING, W. 8.; BROWN, R. G. B.; SANGER, G. A. 1979. Mortality t o marine birds through Pp. 195-199 i n Conservatio~ of marine birds of northern North commercial fishing. America (J. C. Bartonek and D. N. Nettl?ship, Eds). US Fish & W i l d l . Serv., W i l d l . Res. Rep. 11. Washington, DC. NORMAN, F. I.; HARRIS, M . P. 1981. Some recent changes i n the flora and avifauna of Rabbit Island, Wilson's Promontory, Victoria. Proc. Roy. Soc. Victoria 92: 209-212. RICHDALE, L . E. 1963. Biology of the Sooty Shearwater Puffinus griseus. Proc. Z w l . Soc. Lond. 141: 1-117. ROWAN, M . K. 1952. The Greater Shearwater Puffinus gravis at its breeding grounds. Ibis 94: 97-121. SERVENTY, D. L. 1967. Asrjects of the population ecology of the Short-tailed Shearwater Puffinus t e n u i r o s t r i s . Po. i 6 5 - l s 0 i n Proc. 14th Inter. Ornithol. Conar. .ID. W. Snow. Ed.). Oxford: Blackwell. SKIRA, I . J.; WAPSTRA, J. E. 1980. Occupation of burrows as a means of estimating the harvest of Short-tailed Shearwaters i n Tasmania. Emu 80: 233-238. WARHAM, J. 1960 Some aspects of breeding behaviour i n the Short-tailed Shearwater. Emu 60: 75-87. WARHAM, J. 1967. Snares Islands birds. Notornis 14: 122.139. WORMELL P. 1976. The Manx Shearwaters of Rhum. Scot. Birds 9: 103-118. WIENS, J.'A.; SCOTT, J. M . 1975. Model estimation of energy flow i n Oregon coastal seabird populations. Condor 77: 439-452.
JOHN WARHAM & GRAHAM WILSON, Zoology Department, University o f Canterbury, Christchurch 1, New Zealand.
ADDITIONS TO THE AVIFAUNA OF KIRIBATI AND TUVALU By PETER CHILD After an absence of 25 years my wife and I visited Kiribati (formerly the Gilbert Islands) from 17 July to 14 August 1981. The msin ornithological purpose of this trip was to visit the two northernmost atolls of Butaritari and Makin, which I had not had the opportunity to include during my tour of duty in 1953-56 (Child 1960). We spent from 31 July to 10 August in the north (visiting Makin from 3 to 6 August), the rest of the time being on the central administrative atoll of Tarawa. The tremendous influx of population to Tarawa (estimated 25 000 on the s ~ u t h e r nside of the atoll) has made this less worthwhile for birdwatching. However, most previously recorded species could still be found there; in particular, the Crested Tern seemed to have increased considerably, and we estimated at least 120, mainly on the lagoon side of Betio. Constant human disturbance on the ocean reefs and lagoon shallows at low tide resulted in only small numbers of Arctic migrants being recorded. One advantage (compared with Tuvalu) is that the Gilbertese people are not allowed rifle permits. Also, the Kiribati government seems to be alive to conservation requirements, ss is evidenced by their enlightened policies and protection for the Christmas Island bird colonies. They have an active Department of Natural Resources. Butaritari is a very large atoll, being some 30 km by 20 km, with many square kilometres of sand and mudflats at low tide; we did not visit the extensive tidal areas near Bikati in the north-west, nor Kuma in the far east. In the north-east there is a fascinating group of about ten tiny low-lying islets collectively known as Ubrantakoto, which apparently has not previously had an ornithological visit. This group is separated from the southernmost tip of Makin by only 3 km of open ocean. We made the crossing from Makin by launch on 5 August. The islets are sparsely covered in mangrove and three other species of low shrubs (the tzllest being about 6 m ) . They have beaches of coral pebbles and, being devoid of coconuts, are seldom visited, thus forming natural bird sanctuaries. They harboured thousands of breeding Black Noddy (but no Browns), two small colonies of Black-naped Tern (totalling about 70-80 birds), hundreds of breeding White Tern and a thriving heronry of 50-60 adults and young (with several disused nests seen in the shrubs). We also recorded a few roosting Lesser Frigates, an immature Crested Tern and an assortment of the common Arctic migrants. I t is also likely that the Sooty Tern breeds there.
32
CHILD
NOTORNIS 29
Because Makin has not previously had an ornithological visit (Amerson 1969), a completz summary is given for that island. (Note that Amerson called Butaritari " Makin " and Makin he called Little Makin"). Map details for Kiribati atolls are depicted in Amerson's bulletin. New species recorded for Kiribati were Rock Pigeon, Blue-grey Noddy, Asiatic Whimbrel and Asiatic Black-tailed Godwit. ROCK PIGEON Columba livia On 28/7/81, at Betio (Tarawa), about 20 feral pigeons were roosting (mid-afternoon) on a shed at the wharf area. A local informant thought there were upwards of 40 altogether, always in this vicinity where they survived on scraps from shipping cargoes. He also thought some had been here since before World War 11. BLUE-GREY NODDY Procelsterna cerulea On 1/8/81, at Butaritari, a single bird of this species flew in close to us on the ocean reef and then out to sea again. ASIATIC WHIMRREI, Numenius phaeopus variegatus Although Amerson lists a " Whimbrel " for four Gilbert Islands the subspecies is not stated, but is most likely to be the Asiatic. On the lagoon mudflats near Butaritari village we recorded the largest group of 33 on the incoming tide on 9 August, and estimated that at least 50 were present on this atoll during the period of our visit. They are much less approachable than the Bristle-thighed Curlew (N. tahitiensis), which was also seen every day in small numbers. "
ASIATIC BLACK-TAILED GODWIT Limosa limosa melanuroides A single bird was sighted near seven Bar-tailed Godwits ( L . lapponica) on 31 July on the Butaritari mudflats. Slightly slimmer than the Bar-tails, it had the pale axillaries diagnostic of this subspecies. It was not seen again on our other visits to this area, although the Bar-tails were present throughout the whole period. RED-FOOTED BOOBY Sula sula A bird banded in the Phoenix group and found dead at Tanimainiku, Butaritari, earlier in 1981 proved to be 18 years old. A Bird-list for Makin, 3-6 August 1981 Makin atoll (3"20fN 173"00'E), the northern limit of the Kiribati chain, consists of a string of five main islets lying north-south, without a lagoon, and three tiny islets on the far northern edge of the reef. In the centre of the northern islet (Makin) is an interesting tidal and muddy ponded area which attracted Black-naped Terns, Crested Terns, Reef Herons and several species of waders, including the only Whimbrel recorded. Most of the likely habitats were examined during our stay, and, as there is no lagoon, we could cover almost the whole perimeter of the reefs and beaches. Waders are rather limited and scattered because of the lack of lagoon flats and of extensive areas of stable productive sand of mudflat.
1982
AVIFAUNA OF KIRIBATI AND TUVALU
33
The weather was mostly fine and hot, partly cloudy, with a light to moderate easterly; heavy rain fell late afternoon on the 3rd. LESSER FRIGATE BIRD Fregata nriel (Visitant, non-breeding) One male beyond the reef of Makin islet on 4th was seen to chase a Common Noddy until the latter disgorged some food. O n 5th, one seen over the ocean south of Onne. GREAT FRIGATE BIRD Fregata minor (Visitant, non-breeding) Four males beyond the reef of Makin islet on the 4th. In addition, six unidentified Frigates passed overhead while we were travelling north on Makin islet on a tractor on the 4th. It is possible that they breed on Ubrantakoto, northern islets of Butaritari. WHITE TERN Gygis alba (Resident, breeding) Common (in hundreds) throughout. Breeding (eggs and downy chicks), especially on breadfruit, pandanus and 'ren' (Messerschmidia) trees. Small groups coming and going from the ocean all day; flocks fishing outside the reefs. BLACK-NAPED TERN Sterna sumatrana (Resident, possibly breeding) Uncommon; total seen 9. One carried a fish from the ponded area to a juvenile on a sandbar nearby. A possible breeding site is the coral beach in the far north of Makin islet. Two breeding colonies were seen on islets of nearby Ubrantakoto, Butaritari. CRESTED TERN Sterna bergii (Visitant, non-breeding) Uncommon; total seen 15. Four in various stages of moult sitting on low rocks on a beach on the 4th. Eleven roosting and fishing in the ponded area on 6th. Unlikely to breed here, but probably does so on Ubrantakoto. SOOTY TERN Sterna fuscafa (Visitant, non-breeding) Only 1 sighted, off the south coast of Onne on the 5th. This species possibly breeds on the nearby islets of Ubrantakoto. BROWN NODDY Anous stolidus (Common Noddy). Resident, breeding) Common (in tens, possibly hundreds, throughout). Coming and going from the ocean all day; breeding mainly on coconut and pandanus; seemed to be eggs only. BLACK NODDY Anous fenuirostris (Resident, breeding) Abundant (thousands) throughout. Nesting throughout, especially on coconut, breadfruit and ' ren.' Seemed to be on eggs only. Coming and going from the ocean all day, with many ffocks at sea fishing. REEF HERON Egrefta sacra (Resident, probably breeding) Common on reefs throughout; total counted: 41. Of these 23 were white phase, 16 grey and 2 mottled. May breed on low shrubs in the far north; certainly breeds on nearby Ubrantakoto. TURNSTONE Arenaria interpres (Arctic migrant) Scattered throughout low-tide reefs and beaches. Total counted
34
CHILD
NOTORNIS 29
128, largest group seen was 31. Various states of plumage, including a few in full breeding dress. WANDERING TATTLER Tringa incana (Arctic migrant) Scattered throughout low-tide reefs and beaches, but not very common. Total counted 33; largest group 6, otherwise mostly individuals. SIBERIAN TATTLER Tringa hrevipes (Arctic migrant) A group of five among rocks on the north dead coral on the 4th; flew off with curlews when disturbed. On the same day, one perched on a dead tree in the ponded area and called loudly. No others recorded. PACIFIC GOLDEN PLOVER Pluvialis fulva (Arctic migrant) Fairly common throughout low-tide reefs and beaches, and grassy areas. Total counted 18; largest group 4, otherwise mostly singles. In various stages of plumage, including a few in full breeding dress. BRISTLE-THIGHED CURLEW Numenius tahitietzsis (Arctic migrant) Two at the passage south of Makin islet on the 3rd; a group of nine resting among dead coral in the far north on the 4th; two single birds in the ponded area later on the 4th. ASIATIC WHIMBREL Numenius phaeopus variegafus (Arctic migrant) A single bird of this subspecies was stalked in the ponded area on the 4th; on taking flight it called loudly. LONG-TAILED CUCKOO Eudynamys tailerrsis ( N Z migrant) Only one bird heard. This is close to their northern limits of migration - a few reach the Marshalls each season (Amerson 1969). JUNGLE FOWL Gallus gallus (Resident, breeding) Scattered throughout vegetated areas.
After an absence of 25 years my wife and I visited Tuvalu (formerly the Ellice Islands) from 15 May to 17 July 1981. Most of the time was spent on the administrative centre of Funafuti, with short visits to the outer atolls of Vaitupu and Nukufetau, 5 days on Nukulaelae and 2 weeks on Nui. New species recorded for Tuvalu were Christmas Shearwater, Audubon's Shearwater, Red-footed Booby, Asiatic Whimbrel, Sanderling, Siberian Tattler and Ringed Plover. In addition, the first known breeding colony (for this group) of the Red-footed Booby was found on Nukulaelae. Somewhat surprising was the large number (at least several hundreds) of Turnstones present throughout this period of the Arctic summer. Smaller numbers of other Arctic migrants also added interest. Because of the growing human population influx to the capital and increasing numbers of rifle permits being issued, the shooting of birds for food is a matter for concern on Funafuti. Seabird numbers seemed to have fallen drastically there in 25 years, although this was not noticeable on the other atolls visited. The existing protective legislation is very out-dated, confusing and difficult to enforce.
1982
AVIFAUNA OF KIRIBATI AND TUVALU
35
CHRISTMAS SHEARWATER Puffinus nativitatis 29/5/81: One bird resting on the sea in the SE passage of Funafuti's lagoon took off leisurely as we approached by yacht. AUDUBON'S SHEARWATER Puffinus lherminieri 3/6/81: One flying low between waves at sea between Vaitupu and Nukufetau. 21/6/81: One outside the reef at Funafuti. 17/7/81: One seen from a ship just north of Nanumea. RED-FOOTED BOOBY Sula Sula 9/6/81: A tame bird, taken as a nestling and now 9 years old, was shown to us at Nukulaelae. Another semi-tame one, 3 years old, was said to be a female as it was bringing sticks to a breadfruit tree in the village - in preparation for nesting? 10/6/81: On the islet of Niuoko (Nukulaelae) we were shown a roosting/nesting area; at 5 p.m. 40-50 birds were wheeling above; the total was said to reach 200-300. Various colour phases were noted. We saw 5 birds sitting on flimsy nests of twigs c.20 m up in a tall puka " (Pisonia grandis) tree, evidently on eggs. Other trees said to be occupied were not examined. This is the only known breeding colony in Tuvalu, and was believed by our guide to be only 15-20 years old. ( I was not told of any colony during my last visit to Tuvalu in 1956). Birds are sometimes shot or taken as nestlings for food. One morning we saw five immature birds which had been killed by hitting them with a stick as they flew low over an open fishing launch. FRIGATE BIRDS Fregafa spp. We recorded both Great and Lesser in various places, the Lesser being much the more common. At Niuoku islet on Nukulaelae, an estimated 300-400 Lesser came in to roost alongside the Red-footed Booby area. There are still no known breeding places in Tuvalu. BLACK-NAPED TERN Sterna sumatrana Common at Funafuti, but at Nui we saw only two. Sometimes the locals confuse it with the White Tern (Gygis alba) and the same vernacular name is used. On 17/5/81 at Funafuti, we noted one with a blue plastic wing streamer - a P.O.B.S.P. marking? "
TURNSTONE Arenaria inferpres Birds in various stages of eclipse plumage, and a few in breeding dress, were common thrcughout our stay on all atolls visited. O n Nui, where we covered all habitats the most thoroughly, we estimated c.350-400 Turnstones present on lagoon beaches and ocean reefs. On 22/6/81, betweec Nukufetau and Nui a single bird flew around the ship for about 5 minutes, twice landing c n the railing, before heading further south-west. (Compare john lenkins' hitch-hiker: Notornis 18: 130.) ASIATIC WHIMBREL Nurnenius phaeopus variegatus Almost daily at Nui between 22 June and 8 July we studied three birds of this subspecies on a lagoon-side mudflat. Detailed comparisons of body features and behaviour were made with Bristle-
36
CHILD
NOTORNIS 29
thighed Curlews present in the same area. We later concluded that the Whimbrel, being a much more wary bird, prefers the very extensive low-tide mudflats and lagoon-side sandbars of larger atolls (such as in Kiribati). When disturbed too much they tend to fly over the coconut fringe and " hide " among the dead coral of the ocean reefs. No other Whimbrels were seen on our visits to other atolls of Tuvalu. SANDERLING Calidris alba 29/5/81: One feeding on a sandbar on the edge of the outgoing tide at Funafala, the southernmost islet of Funafuti. Still present in the same spot next morning. 27/6/81: A group of four on dead coral of the ocean-side reef at Nui resting at high tide. We watched this group for several days: mostly they fed on small sandylmuddy flats at the base of the main sand-beach on the ocean-side (exposed at low tide). 1/7/81: Three on the lagoon-side mudflat at Nui - could have been from the above group because on 4 July we found four here, one of which tended to be a " loner." Calls recorded were wheef-weet (sharp and short); a short cheep in flight; and sharp highpitched chittering flight calls chickity-chick (repeated several times). SIBERIAN TATTLER Tringa brevipes We did not examine closely the numerous Tattlers seen at all places visited. However, this species was definitely present in very small numbers; at Nui on 3/7/81 we found a group of three busiIy feeding on the exposed low-tide ocean reef. They kept together (and apart from Wandering Tattlers) whenever disturbed and were also more difficult to approach. Pee-wee calls were heard in flight. RINGED PLOVER Charadrius hiaticula On 24/6/81 at Nui on the lagoon mudflat so favoured by the Whimbrels, Curlews, Bar-tailed Godwits, Turnstones, Golden Plovers and Tattlers, we discovered this tiny plover busily feeding among a tight group of Turnstones. Later it tended to keep to itself, and when disturbed made short flights of less than 50 metres. When flushed it uttered a sharp chee-ip. We studied this bird on the three succeeding days; on another occasion its flight call was a sharp cher-wee three times. We could not find it again after 27 June. LITERATURE CITED AMERSON, A. BINION. Jr. 1969. Ornithology ~f the Marshall & Gilbert Islands. Atoll Res. Bull. No. 127. Smithsonian Inst. Washington. CHILC), P. 1960. Birds of the Gilbert & Ellice Islands Colony. Atoll Res. Bull. No. 74. Pacific SEience Board, Washington.
PETER CHILD, 10 Royal Terrace, Alexandra
SEXING BLACK-BACKED GULLS FROM EXTERNAL MEASUREMENTS By G. NUGENT ABSTRACT Six body measurements were taken from 283 adult and subadult Black-backed Gulls (Larus dominicanus) in Auckland, New Zealand. Sex was determined in 158 of these by dissection or chromosomal methods. Using measurements from these 158 birds a classification function was derived and used to assign sexes to the remaining 125 gulls. Discriminant analyses were then made on the measurements from all 283 birds to describe the sexual size dimorphism accurately and to derive a simple classification function for the routine sexing of birds in the field. 1NTRODUCTION Gulls are typically monomorphic in plumage patterns and external genitalia, but are sexually dimorphic in body size (e.g. L. argentafus, L. fuscus, Harris & Hope-Jones 1969; L. glaucescens, L. hyperboreus, Ingolfsson 1969; L. dominicanus, Kinsky 1963). A number of workers (Mills 1971 for L. novaehollandiae; Shugart 1977 for L. delawarensis; Ryder 1978 for L. argenfatus; Threlfall & Jewer 1978 for L. argentatus) have been able to predict sex correctly in a high percentage (i.e. >95%) of cases by using two or more measurements in a classification function. Such a function is derived from a discriminant analysis of measurement data, the analytical theory and tests for statistical significance being described by Nie et al. 1975. In this study, discriminant analyses were used to describe the sexual size dimorphism of the Southern Black-backed Gull (L. dominicanus) and give a classification function for routine sexing of this species using two external measurements. METHODS A sample of 283 Black-backed Gulls over 1 year of age was obtained between 15 February and 24 June 1980 from three sites in the Auckland metropolitan area. Potential sampling biases in the sex and size of the birds were minimised by using an unbaited cannon-net on flocks of resting gulls at or near high tide and after 11.00 a.m., when feeding activity was generally low. Six body measurements were taken from each bird: head length, gape, and bill depth (as in Fig. 1) measured to the nearest 0.1 mm; wing chord to the nearest 0.5 cm; tarsus length to the nearest 0.1 mm; and body weight to the nearest 10 g as in Baldwyn et al. (1931).
38
NOTORNIS 29
NUGENT
Thirty-one birds killed for gut analysis were sexed by dissection. A further 127 birds were sexed by chromosomal analysis of growing feather tissue (see Nugent 1981). Preliminary data confirmed that male and female size distributions did not overlap at the extremes of the size range. Very large and very small birds could therefore be assumed to be male and female respectively. The time-consuming and expensive chromosomal sexing was therefore carried out on all birds within the overlap region of male and female size distributions, but only on one-third of those at the extremes. All these latter birds were proven to be of the assumed sex. A classification function derived from the measurements of the " known sex " birds correctly predicted sex for 98.7% of the 158 such birds. The so-far unsexed birds were then assigned a sex using this preliminary function. As these " assigned sex " birds were in the main those at the extremes of the size range, no more than 1% of them are likely to have been incorrectly sexed. To describe the sexual dimorphism of this species accurately required that the data from " known " and " assigned-sex " birds be pooled. This data was subjected to discriminant analysis to distinguish the most dimorphic measurements. A classification function using two measurements was then derived for routine field use. All analyses were carried out using the sub-program Discriminant " in SPSS (Nie et a[. 1975). "
RESULTS The overall difference in body size between male and female Black-backed Gulls is highly significant (6-measurement discriminant analysis, Wilks lambda = 0.1568, Chi square = 537, df = 6,
