FIGURE 1 - South-west Chatham Island, showing places mentioned in text, predator tramlines positions where feral cats were caught in April- May 1991 and September-October 1992.
 
 

    In September 1990, a 12 km circuit track (the Southern Loop) was cut from the coastward boundary of the Tuku Nature Reserve (Figure 1) to allow a more extensive area to be trapped. Another track, 9 km long, was developed around the northern burrows (the Northern Loop), partly within the Reserve and partly on land owned by the Daymond family of Waitangi. Leg-hold traps were spaced at irregular intervals of 50-100 m along these tracks. In 1991/92, 390 traps were in use. The trapping circuits passed within 50 m of all known burrows but also far southwards in the Tuku River catchment so as to protect possible, but undiscovered, burrows there.

    Traplines were checked, cleared, and re-set daily by two people, one to each circuit, during two 4- to 9-week sessions (September-November and January-May) each breeding season. Baits were changed as demanded by their state of decay. An advantage of the manufactured baits was that they lasted for weeks if kept reasonably dry. Numbers of traps in operation and kills made were recorded daily.

    Various baits were used. In 1987/88 food scraps were used until Possums and Weka Gallirallus australis were caught; then parts of their corpses were used for bait. Weka was more attractive to feral cats. The same baits, as well as a paste consisting largely of sardines, were used in 1988/89. In 1989/90 and 1990/91, to cater for the large numbers of traps in use, and the need to replace baits frequently, fish skeletons, normally used as bait for crayfish Jasus edwardsii, were used.
In 199 1/92 and 1992/93, a manufactured pig bait (fishmeal/cereal, polymer-bound, in cylindrical biscuit form, Du Pont, USA), under trial by NZ Forest Research Institute Ltd as a bait for cats, was used. The biscuit was suspended above each trap in a wire-mesh cage to prevent theft by rats; a plastic cover reduced rain damage. It appears to have been successful.

    Because Taiko breeding failed repeatedly in Taiko Stream Valley (see Table 2: burrows Tuku 1 and 10) possibly as a result of rat predation, brodifacoum poison baits were laid immediately next to the burrows from the 1990/9 1 breeding season. A grid of permanent poison bait stations was set up during 1991/92 and this was replaced by a more extensive grid with improved bait holders in 1992/93. Possums and Weka were also poisoned by the baits.

    Each burrow was examined to see if it was feasible to make an opening for occasional access to the nest chamber. All except one were too deep or too protected by tree roots. On 29/9/90, muted calls of a Taiko led to the nest chamber of Tuku 10 burrow being found: it was opened and then sealed for subsequent inspection (Taylor 1991). Although the nest was about 2.8 m from the entrance, it was only 0.3 m ,under the surface. Even with ready access to this nest, the inspection lid was opened as infrequently as possible. Twigs or leaves placed over the burrow entrance sufficed for most observations of bird movements.

    Taiko burrows were checked at about monthly intervals, or more frequently near the estimated time of laying and fledging. Before 1990, screening of burrow entrances with leaves or twigs began in October. In 1990, Taylor (1991) found that activity at burrows began in late September. More intensive observations of burrows were made from 25 October to 13 November 1988, 17 September to 3 October 1990 and on 14 January and 8 February 1991 (Taylor 1991); 14 January to 5 February 1992; and 26 October to 12 November 1992 (Taylor & Imber unpublished data).

    All burrows were examined annually between 18 April and 5 May. They were screened, if necessary, to find whether a fledgling had been reared. All fledglings caught were banded. During the survey in early 1992, a ground search by seven people failed to find any further breeding burrows. They did find a possible new burrow near Tuku 1 burrow, a disused burrow in Taiko Valley, and another (much older) burrow in an adjacent valley.

    Taiko were enticed out of their burrows several times at night by “warwhooping” vocalisations (Tennyson & Taylor 1990). A Taiko on the ground also responded to these calls.

TABLE 1 — Number, and number/l00 trap nights (N/l00), of predators trapped and killed around burrows of Chatham Island Taiko and in adjacent Tuku Nature Reserve, August-May 1987/88-1992/93.

The results of trapping over the first six years are shown in Table 1. Trapping was first directed against feral cats because there was extensive evidence of them in Taiko Stream Valley when the first burrow was found there in 1987 (Imber a al. 1994). During the first widespread trapping in 1989/90, traps were often set in areas less favoured by cats, so fewer cats were caught per 100 trap-nights (Table 1). When extensive trapping began, more cats were caught along the tracks near the coastal boundary of the Reserve than inland (Figure 1), suggesting that cats were more common in the scrublands adjacent to farmland. Although the number of feral cats killed per 100 trap-nights, and as a proportion of all kills, declined, the total number killed annually increased with the greater effort in 1989/90 and remained relatively constant until 1992/93.

    At first, all rats caught were identified as Ship (Bush or Black) Rats Rattus rattus. However, after a dead Kiore (Pacific or Polynesian Rat) R. exulans was positively identified on the Southern Loop track in 1992, it was
concluded that a proportion of the rats killed previously had been Kiore. -Until 1992, Kiore had been thought to be extinct on Chatham Islands.
    Weka are endemic to the New Zealand mainland, and were introduced to the chatham Islands from the South Island in 1905 (Turbott 1990). As it was an introduced bird and a potential threat to Taiko, particularly if it could enter a burrow, we considered it necessary to kill them.
    Hedgehogs Erinaceus europaeus eat eggs (King 1990), but they were trapped along the seaward boundary of the Reserve and not in the bush. Their spread along the coast to the study area from Waitangi (introduced early 20th Century) was relatively recent.
 

Productivity

Although (to May 1993), 12 burrows were known to have been visited by Taiko since November 1987, an egg had been laid in only four of these over that period. The inferred or known breeding results in these burrows are shown in Table 2. Only breeding attempts in North 1 and Tuku 4 burrows were successful.

    In 1993, of the other eight burrows visited, two were no longer in use, five were at an early stage of being dug (all first active in 1992), and one was in the final stages of being dug (or re-dug, probably after a period of disuse). Breeding may have been imminent in that burrow.

No adult Taiko was found dead from predation around the burrows from 1987 to 1993. Cessation of activity at two burrows may have resulted from adult mortality at sea, birds shifting to another burrow, or interference (possibly by a dog or possum, from evidence of digging by downward scratching) at one burrow.

Breeding biology
Mating and the pre-laying exodus   The earliest sighting of a Tatko visiting the breeding grounds — passing over the light used to catch Taiko by the Tuku Valley — in any year was on 15 September 1980 (R. Cotter, pers. comm.). In the 1990/9 1 breeding season, Taylor (1991) made the first observations of Taiko at their burrows at the time of mating. At first inspection on 18 September, GAT considered that the burrows had not been visited that breeding season. Taiko first visited one of the three main burrows on 24 September; all three had been visited by 30 September.

TABLE 2 — Breeding attempts in four Taiko breeding burrows over 6 years, 1987-93

    At North 1 burrow, a pair spent at least 2 and 3 October continuously in the burrow, during which time copulation probably occurred. The pair occupying Tuku 10 was probably in the burrow on 29 September, though only one was caught, because GAT heard muted calling of Taiko in response to the physical disturbance of people passing. Grey-faced Petrels Pterodroma macroptera gouldi call in this way only when there is a pair in the burrow (MJI, pers. obs.). Screens at all three breeding burrows were disturbed often between 1 and 15 October. Pairs seem to mate around 1 October, but occasionally as late as 15 October.

    Taiko visited the breeding burrows in the month after mating. These birds were probably males (cf. Imber 1976), but may also have been other Taiko prospecting for burrows. Tuku 10 burrow was not visited from 16 October to 18 November, but the burrow had been reoccupied and laid in by 28 November (Taylor 1991). However, in 1992 both Tuku 1 and Tuku 4 were visited by the breeding males between 30 October and 3 November.

    Females of the closely related Grey-faced Petrel departed on their prelaying exodus immediately after mating, and males usually visited again after mating, before leaving on their pre-laying exodus (Imber 1976). It therefore seems that the pre-laying exodus of female Taiko lasted about 40-55 days, and that of male Taiko between 25 and 45 days.

    Laying Laying dates were determined, approximately, by the date of first disturbance of the burrow mouth screen after the pre-laying exodus. However, in Grey-faced Petrels, the male, rather than the laying female, sometimes returns first (Imber 1976). In 1988, screens were displaced at the three main burrows on 24, 26, and 27 November. The egg in Tuku 10 burrow was just starred (first stage of pipping) on 14 January 1991 (Taylor 1991). Allowing 5 days to hatching (cf. Imber 1976), and possibly 54 days for incubation, the egg was probably laid about 25 November.

Hatching The only information on hatching was for the egg in Tuku 10 burrow. It was starred on 14 January 1991 and hatched about 18-19 January.  The egg measured 64.8 x 47.0 mm and weighed 66 g on 14 January. On 2 February 1992, Tuku 10 contained a chick about 10 days old.  Chick development The chick in Tuku 10 was examined on 8 February 1991, at about 20 days old. It weighed 240 g; bill length was 23.5 mm (Taylor 1991). The down was brownish-grey above and light grey on the underside.  The bill was black; the tarsus and foot flesh-coloured but greyish-black on the distal part of toes and webs, and on all of the outer toe.

    Fledging and departure of young Fledging dates at North 1 burrow were: probably about 5-10/5/88; 5-10/5/89; after 1/5/90; 4/5/91; 28/4-2/5/92; 25-27/4/93. At Tuku 4 it was after 5/5/93. From hatching in mid to late January, to fledging about 4 May, would give a fledging period of about 105 days, similar to that of White-headed Petrels Pterodroma lessonii (Warham 1967). Fledgling Taiko were as strictly nocturnal in their activity above ground as were the adults, and spent about 7 to 15 nights sometimes outside the burrow before leaving. They were usually very cautious when emerging, and often waited just inside the burrow for hours.
 
 

IMBER, M.J. 1976. Breeding biology of the Grey-faced Petrel Pterodroma macroptera gouldi. Ibis 118: 51-64.
IMBER, M.J.; CROCKETT, D.E.; GORDON, A.H.; BEST, H.A.; DOUGLAS, M.E.; COTTER, R.N. 1994. Finding the burrows of Chatham Island Taiko Pterodroma magentae by radio telemetry. Notornis 41 (Supplement): 69-96.
KING, C.M. (Ed.) 1990. Handbook of New Zealand Mammals. Auckland, Collins. 600 p. TAYLOR, G.A. 1991. Report on the Chatham Island Taiko and Chathani Island Petrel
recovery programmes (1990/91). Threatened Species 0cc. Publ. 2, 26 pp. Wellington, Department of Conservation.
TENNYSON, A.J.D.; TAYLOR, G.A. 1990. Behaviour of Pterodroma petrels m response to “war.whoops”. Notornis 37: 12 1-128.
TURBOTT, E.G. (Convener). 1990. Checklist of the Birds of New Zealand and the Ross Dependency, Antarctica. 3rd ed. Auckland, Random Century and Ornithological Society of NZ. 247 p.
WARHAM, J. 1967. The White-headed Petrel, Pterodroma lessoni, at Macquarie Island. Emu 67: 1-22.