Publications Archive

Little is known of the causes of mortality in captive-bred brown teal (Anas chlorotis) released to the wild. To test whether feeding difficulties have contributed to the poor survival of released birds, we developed a method to detect starvation using the wing fat content of brown teal. We extracted the lipids from 4 outer wing components of 17 intact brown teal carcasses. The lipid content of each component reflected the birds’ nutritional condition (based on body mass and size, and visible fat). Lipids were also extracted from the outer wing components of 7 partial brown teal carcasses, 6 of which were from captive-bred released birds whose cause of death could not be determined. All of the released teal were found to have been in very poor nutritional condition immediately before death, implicating starvation as a factor in their deaths. Improving the feeding regime of captive-bred brown teal (pre- and post-release) is likely to increase their survival.

Bellbird breeding activity was monitored on Tiritiri Matangi Is in 1977 and 1978. The density of bellbirds breeding within the study area was 2.13 pairs ha-1. Resighting of banded adult birds in consecutive years was 69% and less than half the pair bonds were maintained into the following season. Breeding extended from Oct to Jan, with a peak in Nov. Most nests were built in tree ferns, and the mean clutch size was 3.6 eggs. Incubation and nestling periods were c. 12 and 14 days, respectively. The female bellbird was significantly more active than the male in care of the nestlings. Measurements of chick head and wing length were correlated with age. Nest success was 44%, and daily survival rate 97%. Predation, desertion, and exposure contributed to nest failure, and re-nesting was common. The mean number of nests female-1 season-1 was 1.3, while mean reproductive success was 2 chicks fledged. Young were independent 2 weeks after fledging, and moulted to adult plumage in their 1st year. The dynamics of the Tiritiri Matangi population, when compared with other populations, suggest that breeding behaviour is a flexible response to environmental factors, and will vary between populations depending on the level of predation, competition and habitat quality.

A quantitative survey of the bird community of Kaitoke wetland, Great Barrier Is., New Zealand between May 1998 and July 2000 using 5-minute counts recorded 33 species, most of which occurred in less than 10% of counts. The commonest species were North Island fernbud (Bowdleria punctata), fantail (Rhipidura fuliginosa), silvereye (Zosterops lateralis), grey warbler (Geygone igata), welcome swallow (Hirundo neoxena) and (collectively) the exotic finches, yellowhammer (Emberiza citrinella), chaffinch (Fringilla coelebs) and goldfinch (Carduelis carduelis). Native wetland species also recorded were spotless crake (Porzana tabuensis), banded rail (Rallus phillipensis), Australasian bittern (Botaurus poiciloptilus), pukeko (Porphyrio melanotus) and Australasian harrier (Circus approximans). Distributional analysis of the commonest species (those occurring in more than 10% of counts) showed most had some association with a particular vegetation type(s), while few showed any change in conspicuousness with season. There was little apparent movement of buds associated with phenology of the main wetland plant species. The sigruficance of Kaitoke wetland in providing habitat for a range of native wetland bud species is recognised. The current threat to this ecosystem from introduced pests and development pressure, and the paucity of data available on native wetland bird species to inform conservation management, is discussed.

The breeding success of northern New Zealand dotterels (Charadrius obscurus aquilonius) on Matakana Island, Bay of Plenty, was determined over 8 seasons (1992/93-1999/2000) in managed and unmanaged areas. Management to enhance breeding success included shifting nests to reduce the risk of flooding during spring tides and storms, and reducing predator populations of brushtail possum (Trichosurus vulpecula), feral house cat (Felis catus), Norway rat (Rattus norvegicus), stoat (Mustela ermines), and southern black-backed gull (Larus dominicanus) at dotterel breeding areas. Measures taken to reduce the incidence of disturbance by people on breeding dotterels included erecting fences around nesting areas, and speaking to members of the public about the plight of the species. The number of breeding pairs on Matakana Island fluctuated between 19 and 31 during the 8 seasons. Overall, 35.1% of 276 nesting attempts resulted in broods hatching. The main causes of nest failure during incubation were flooding by high tides or storms, and predation. The proportion of nests in which 1 or more eggs hatched was fairly stable during the 1993/94 to 1997/98 seasons at 26.1-33.3%, but was 68.0 and 51.3% in the last 2 years. This marked improvement in nesting success was attributed to the increased duration of pest control: starting before dotterel breeding began and continuing until most broods had fledged. Nesting success during incubation in managed habitat (47.5%) was significantly greater than in managed habitat (19.5%). Overall, 52.6% of chicks fledged. The number of chicks fledged per season (5 – 33), and fledglings produced per breeding pair (0.26 – 1.08) increased through the study In conclusion, we make suggestions to further promote the conservation of the northern New Zealand dotterel.

The brown teal (Anas chlorotis) is a threatened duck endemic to New Zealand, whose single remaining stronghold is on Great Barrier Island. During a study at Awana from 1985 to 1987 it was concluded that the population was stable. We evaluated this proposition by analysing counts made at flocking sites between 1985 and 2001, examining their demography, and by developing an age-specific demographic model to predict the likely persistence time of the population. Numbers of teal at flocking sites declined between 1985 and 2001 though demographic variables did not change except for duckling survival rate, which increased from 26% during 1996 to 54% during 1999. Previous predictions of stability were based on an incorrect model and our model suggests that the brown teal population will halve every 4.1 years. The decline observed since 1987 could be halted by improving adult survival rate.

The reproductive biology of red-crowned parakeets (Cyanoramphus novaezelandiae novaezelandiae) was studied on Little Barrier Island (Hauturu) during the 1989/90 and 1990/91 breeding seasons. Nests sited in cavities of pohutukawa (Metrosideros excelsa) and puriri (Vitex lucens) trees accounted for all but 1 nest which was found in a hole in the ground. Breeding activity was recorded from Nov to Mar with peak egg laying in Dec. Mean clutch size was 7.0 eggs (range 4-9). Replacement clutches were rare and only followed the loss of young broods. Female parakeets were entirely responsible for incubation that lasted about 23 days. The nestling period averaged 40.8 days (range 32-49) with about a 30% probability of an egg resulting in a fledgling. The main cause of nestling mortality was starvation. Infestation of nests by mites and predation of nestlings were also occasionally recorded. Post-fledging mortality was high, particularly in the 1st 2 weeks, and was attributed mainly to aerial predators such as morepork (Ninox novaeseelandiae). Aspects of nest and roost site characteristics, brood reduction, mortality, and vulnerability to predation are discussed in relation to the conservation of the species.

More than 35,000 southern royal albatrosses (Diomedea epomophora) were banded on Campbell Island from 1941 to 1998. Recoveries of 2187 birds while breeding on Campbell Island during 1994-98 included 54 (2.5%) that were injured by their bands; over all years, 195 (3.4%) injured birds and 225 others with bands fitted incorrectly were reported. Injury rates were higher for birds banded as chicks (7%) than adults (0.5%). Untrained volunteer banders from the island’s meteorological station banded up to 5200 birds annually, and in some years bands were not closed properly. The partially open bands eventually embedded in the leg or ankle, crippling the birds. Six annual banding cohorts were responsible for 83% of injuries and almost half (n = 90) came from the 1979 cohort. Banding quality improved after 1982 and only two injured birds have been found from more recent cohorts. The band’s large circumference relative to its thickness may have contributed to it springing open with time, so a stronger band is recommended. For animal welfare reasons, a band repair operation should be conducted. If nothing is done, the situation will improve over the next 20-30 years as birds die, but regular band maintenance would prevent future problems.

The aim of this study was to associate the distribution of waders (Charadriiformes) with physical attributes of estuaries and map the distribution of estuaries that have the potential to support large diverse wader populations. Maximum wader abundance and the maximum number of species recorded (1983-1994) at 94 estuaries on the North Island, New Zealand were related to physical attributes of the estuaries using General Linear Modelling techniques. Parameters describing the estuary type, area, tidal influence, catchment area, catchment rainfall, intertidal area and temperature were used as independent variables. Anaiysis revealed strong positive relationships between the size and diversity of a wader population and estuary area. A negative relationship between catchment runoff and wader abundance and diversity existed in some estuaries. The model predicted that relatively small estuaries (100 to 999 ha) could support large, diverse wader populations if catchment runoff was low. A map of estuaries predicted by the model to support large diverse populations of wading birds is presented. This study highlights the value of large estuaries in New Zealand in maintaining large populations and high species diversity of migratory and nonmigratory waders.

Dispersal within New Zealand of Australasian shoveler (Anas rhynchotis), banded as duckhngs in Otago (n=489) and Southland (n=392) during 1971-1979, was determined from the locations at which 180 were shot by hunters. There were no statistically sigruficant differences in recovery distributions of Otago and Southland birds either when recovered in their year of banding (y-o-b) or in all subsequent years combined (later). About 50% of total recoveries were made in the y-o-b and 2-thirds of these from within 200 km of the banding site. Recoveries in later years were more widely distributed than those made in the y-o-b. North Island recoveries were 28% of total recoveries and were from most large coastal and lowland wetlands as far as Northland, 1400 km from the banding site. Recovery distributions of ducklings were not sigruficantly different from those previously determined for moulting adults banded in the same areas. However, ducklings in their 2nd year of life appear to be more distantly dispersed from their natal sites than during their 1st year or are adults from their moulting sites. We speculate that long distance dispersal may be undertaken mostly by birds that fail to breed in their natal regions in their 1st year of life, and that dispersing birds may become irregular breeders at varying distant locations.