From 10 to 16 August 1991, we surveyed the complex of sounds making up Breaksea and Dusky Sounds, the Acheron Passage and Wet Jacket Arm, Fiordland, for Fiordland Crested Penguins. Two islands in Doubtful Sound were also checked. Minimum total estimates were 24 nests and 62 penguins in Doubtful Sound and 47 nests and-106 penguins in the Breaksea/Dusky complex. Penguins known to be on Breaksea I. and adjacent islands were not surveyed. Relatively more penguins were found nesting in caves or under rock overhangs and fewer were found in dugouts under trees than were found in the more northern sounds in 1990. No penguins were found on the many islands in Dusky Sound formed from steep-sided granite domes. The results support our suggestion made in 1990 that there are fewer than 1000 nests for the species annually.
In 1988, 3603 kilometres of the coast of New Zealand were patrolled and 7545 dead seabirds were found as part of the Beach Patrol Scheme. Two new species were the Tahiti Petrel (Pseudobulweria rostrata) and the Masked Booby (Sula serrator dactylatra). Unusual finds were a Black-bellied Storm Petrel (Fregetta tropica) and a White Tern (Gygis alba). A summary is given of the coastal and monthly distributions for 10 species of Puffinus found between 1943 and 1988. Of these shearwaters, the Sooty Shearwater (P. griseus) was found most frequently overall, mainly in May and November-December. This review of beach-wrecked shearwaters indicates that the information derived from the Beach Patrol Scheme for common seabird species relates reasonably well to the movements of these species about New Zealand coasts.
Differences in the time of onset of breeding, morphology, egg size, plumage and vocalisations of Dark-rumped Petrels (Pterodroma phaeopygia phaeopygia) were investigated on four islands in the Galapagos. Comparisons were made with P. p. sandwichensis in Hawaii. In Galapagos, breeding cycles differed among islands, and on San Cristobal there were two populations that bred at different times. On Floreana, colonies at different altitudes bred at different times. Eggs were laid on Santiago over four consecutive months; on Santa Cruz the egg-laying period was shorter. Analyses of morphological measurements and notional volume separated Galapagos Dark-rumped Petrels into three groups. Birds on Santa Cruz and those breeding in the middle of the year on San Cristobal were the smallest; birds on Santiago and those breeding at the end of the year on San Cristobal were of intermediate size; and those on Floreana were the largest. There was a similar size trend in the breadth and volume of eggs. No relationship was found between variable plumage patterns on head and chest or between plumage and island populations. Evidence is presented that supports sexual dimorphism in vocalisations, and it is suggested that males make Sweet calls and females make Coarse calls. There were statistically significant interisland differences among Sweet calls and among Coarse calls. Dialects probably exist within the archipelago. Calls had either one or two introductory syllables. When present, the second introductory syllable was very similar to the single introductory syllable, and these may serve the same function. Discriminant analysis of Sweet calls correctly classified 82.296 of these into island of origin. A similar analysis of morphology correctly classified 58.6% of birds from five populations. A theoretical combination of these two analyses indicates a potential classification rate of 92.6%. Although there are differences among Galapagos populations, there is not yet sufficient evidence to warrant subspecific status. Vocalisations of the Hawaiian birds were quite different from those in Galapagos, and Galapagos birds were bigger. Dark-rumped Petrels in Galapagos and Hawaii might be more distant taxonomically than currently recognised and they may be different species.
White-faced Herons (Egretta novaehollandiae novaehollandiae) captured a wide variety of prey from pastures with transitory surface water. A total of 357 prey items were recorded, mainly from regurgitated pellets but also from observations of birds feeding. Major prey groups were insects, tadpole shrimps, earthworms, tadpoles and frogs. The diet depended on the seasonal availability of prey. In spring, earthworms, cosmopolitan diving beetles, tadpole shrimps, tadpoles and frogs were major prey. During the drier conditions of summer and autumn, terrestrial insects such as flies, damselflies, and grasshoppers became important. Earthworms and tadpole shrimps were the main prey in winter.
A survey of Banded Dotterel on Adams, Enderby, Ewing, Rose and part of the main Auckland Island was made in November 1989. Nesting was recorded for the first time on Enderby Island (8 nests) and 11 nests were found on Adams Island. The behaviour of dotterels on both islands indicated the presence of many more nests. Some dotterels moved at night from fellfield and moorland breeding grounds to the beaches of Adams, Rose and Ewing Islands. A total of 730 birds was counted, most on Enderby Island (440) and Adams Island (273). Previously, the total was considered to be 100-200 birds. Our high count is probably the result of an increase in breeding habitat following drastic vegetation modification by fire and introduced mammals on Enderby Island during the 19th century. The population has probably been rising for at least 20 years, but this has been overlooked because there has been no previous opportunity to count dotterels during their breeding season, when they are sedentary.
The first known breeding of the Black-winged Petrel (Pterodroma nigripennis) on Mangere Island was studied in 1987-88 from the pre-laying period to part way through the chick rearing period. The first returning bird was noted on 30 November; eight incubated eggs were found, calculated to have been laid in mid-January. The eggs hatched in late February – early March; fledglings were calculated to have left in late May. The breeding cycle was about 2 weeks later than at more northerly breeding sites. The approximate mean incubation shift was 13.5 days. The colours of chicks are described in detail. Only two of the five chicks that hatched were still alive by the end of the study. Interference at the nest site by other petrel species apparently caused some nesting failures. No birds were found undergoing active wing or tail moult. The brood patches of most adults examined were largely bare in late January and February and were refeathering in March. The birds were active only at night.