Between January 1965 and March 1967 4,926 observations were made on 20 species of petrel feeding in the Southern Ocean from the research ship USNS Eltanin. Most observations were made at night while the ship was stopped on oceanographic research stations. Observations were made possible by bright decklights, following birds with the powerful bridge lights, or under moonlight. Eleven feeding methods were distinguished. Most common were surface seizing [49.1%: used by 14 species), dipping (25.2%: 9 species), and surface plunging (c. 0.6%: 6 species). Seven species foraged entirely at night, and five fed by day only. Food recorded was chiefly crustaceans and squid. The submergence time and prey-handling time for some species are also given.
Over 27 months, 335 Brown Teal (Anas aucklandica chlorotis) were trapped and banded on Great Barrier Island. Four trapping methods were used to trap both solitary and flocking birds in all seasons. The efficiency of the trapping methods is compared, and the study areas and banding scheme are described. The computer program used to generate the colour band combinations is included as an appendix.
Yellowheads (Mohoua ochrocephala) have disappeared from many parts of the South Island. The northernmost self-sustaining populations now occur in Arthur’s Pass National Park, where we surveyed Yellowheads in the Hawdon River Valley during their breeding in 1983 and 1984. Survey techniques were developed for assessing Yellowhead numbers at low densities. In 1983, 11 breeding pairs were found in 500 ha of forest, and in 1984, 21 breeding pairs were found in 2900 ha. About 70% of the birds occurred as breeding pairs, and the rest were itinerant non-breeders and helpers at nests. Yellowheads were spread throughout the valley but were more common at low altitudes. Breeding pairs preferred red beech (Nothofagus fusca) forests. There is some evidence that the nationwide decline of the species is continuing in the Hawdon River Valley.
Yellowheads (Mohoua ochrocephala) were observed between November 1983 and May 1984 in the Hawdon River Valley, Arthur’s Pass National Park. Of 11 breeding pairs, at least three had a third adult associated with them. Three pairs were watched in detail, and two nests were found, both in holes in live beech trees. The maximum breeding range was 7 ha. Young fledged in late December, and so laying had occurred in early November and hatching in late November-early December. There was no indication of double broods
or renesting. Family groups of adults and juveniles remained in their breeding ranges for up to 2.5 days after fledging and then ranged throughout the forest. Parental care continued for about 55 days. From mid- January family groups began to come together for several hours and flocks of up to 25 Yellowheads were seen. In April and May flocks of 6-12 Yellowheads were seen, the birds apparently remaining together all day. Yellowheads were often in mixed species flocks.
Spotless Crakes (Porzana tabuensis) and Marsh Crakes (Porzana pusilla) were studied by R. Lavers, April 1971-July 1972, and by G. Kaufmann, September-December 1982, at Pukepuke Lagoon, Manawatu. Eleven Spotless Crake nests were found. Eggs were laid between 30 August and 19 December. Earlier nests usually contained 3 eggs; later nests contained 4-5 eggs, and the eggs were larger. Copulation and major calls are described. One male incubated 40°/o, the female 60%, of the 35 hours observed. Eggs of 4 nests were eaten by predators. Two Marsh Crake nests were found, containing 4 and 6 eggs. Both nests were unsuccessful. Analysis of museum skins highlighted similarities in bill size and structure, suggesting that interspecific competition occurs.
An accidental bird poisoning in May 1971 in the Obornjaca region, Yugoslavia, exterminated a Starling colony. The food in 186 chick and 56 adult stomachs was analysed: more than 95% of the insects eaten were Orthoptera and Coleoptera, including many species harmful to agriculture. Differences in diet between the chicks and their parents may be explained
by time of day or by the location in which adult birds foraged.
The breeding of the Black Petrel (Procellaria parkinsoni) on Little Barrier Island was studied during 1971-75 in 22 study burrows and then reviewed at about 2-yearly intervals. Predation by feral cats affected the population most, causing the number of breeding and non-breeding birds associated with study burrows to decline from 39 in 1971-72 to 14 in 1976-77. Cats were eliminated between 1977
and 1980. By 1982-83 further attrition due to poor recruitment had stopped. The breeding season of the Black Petrel is from October to July. Eggs are laid from about 10 November to about 20 January but mainly in early December. Pre-laying activities are brief, but other phases of the breeding cycle are not. Incubation shifts and the nestling period may be long because of limitations of the food supply. The major breeding place is Great Barrier Island. Surveys there during chick-rearing in 1977 and 1978 revealed very little predation and relatively high breeding success. The cause of this is discussed.
The study group comprised territorial pairs and non-territorial adults and sub-adults of weka (Gallirallus australis). Pairs were maintained all the year and occupied minimum areas of between 2.6 ha and 15.8 ha. Most non-territorials were younger than 18 months. All wekas that obtained territorial positions in the study period were under 2 years old. Most pairs bred successfully at least once per year and one bred successfully three times in the same year, raising nine young. Recruitment exceeded losses in the territorial group. In the 1985 season, 37.5% of the territorial birds were lost through death or displacement.
The hakawai was a ‘mystery bird’ formerly found on islands off Stewart Island; although never seen, its startling call was heard at night. The call of the hakawai, and its distribution and decline are described. The hakawai (under several spelling variations) is widely mentioned in myths and legends of the Maori throughout New Zealand; these records are summarised and the various theories for the hakawai’s identity are discussed. Evidence for non-vocal aerial displaying by New Zealand snipe (Coenocorypha) is presented. The hypothesis that the hakawai was an aerial display of Stewart Island Snipe (C. aucklandica iredalei) was investigated by comparing the distribution and decline of snipe with that of the hakawai, and by playing a tape recording of an aerial display of Chatham Island Snipe (C. pusilla) to people who had heard the hakawai. These data support the hakawai = snipe hypothesis. The historical distribution of Stewart Island Snipe included Big South Cape, Pukeweka, Solomon, Pourama, Jacky Lee, Herekopare, Ruapuke and Green Islands in the last 100 years. The extinction of snipe on these islands is attributed to introductions of ship rats (Rattus rattus, two islands), weka (Gallirallus australis, four islands) and a combination of weka and cats (Felis catus, two islands). It is proposed that subfossil remains of Coenocorypha from the North Island and the South Island be referred to C. a. barrierensis Oliver 1955 and C. a. iredalei Rothschild 1921 respectively.