The Chatham Island Taiko Pterodroma magentae (Ayes: Procellariidae) was rediscovered on 1 January 1978 in its traditional breeding grounds in the Tuku Gully in the south west of main Chatham Island. The history of efforts to rediscover Pterodroma magentae (known as the Magenta Petrel) and to identify the mysterious Chatham Island Taiko are described. The identity of the Chatham Island Taiko and the Magenta Petrel is confirmed. Aspects of behaviour, habitat, morphology, and ecology of the species are reviewed. To date, 42 Taiko have been banded; 12 recaptures suggest a population of over 100 birds. Night observations of flying birds and times of burrow use indicate that the breeding season extends from mid-September to mid-May. Efforts to discover the breeding grounds before the assistance of radio telemetry are summarised.
 
 

    On 22 July 1867, personnel from the Italian research vessel Magenta collected a large petrel when the ship was at 39o38’ S 125o58’ W, south of the Tubuai Islands in the South Pacific Ocean (Gigliolo & Salvadori 1868). Similar birds were sighted on 3 August 1867 at 32o23’ S 92o39’ W, south of Easter Island. Further birds were sighted on 31 August at 26o07’ 5, 88o50’ W, north of the Juan Fernandez Islands.

    The specimen collected was described as a new species, the Magenta Petrel Pterodroma magentae (Giglioli & Salvadori 1868, as Aestrelata magentae). A description was published in English a year later (Giglioli & Salvadori 1869). The umque type was deposited in the Turin University Museum. The specific name commemorated the first Italian warship to circumnavigate the world. Salvin (1876) agreed that the Magenta Petrel was a distinct species
— Oestrelata magentae — but it remained somewhat mysterious because further specimens were not forthcoming.
Fleming (1939) recorded reports of a large petrel breeding on the farm of H.G. Blyth on the south-western coast of the main Chatham Island. Known to local people as the Chatham Island Taiko, it had been collected as a “mutton bird” by Moriori and Maori until 1908.

    For the 25 years after Fleming’s paper, speculation about the identity of the unknown petrel continued, with its being identified successively as Pterodroma axillaris, P. inexpectata, and P. neglecta (Fleming 1944, 1953). Fleming himself concluded that it was none of these species but a Pterodroma similar to P. macroptera.
 
 

    W.R.P. Bourne located and examined the type of P. magentae in the Turin Museum in 1956. It was still in good condition despite being in a bomb-damaged attic. Bourne later worked on Pterodroma bones from the Chathams in the H.O. Forbes collection discovered by Dawson (1958) in the British Museum (Natural History). Bourne’s comparisons led him to postulate that the Chatham Island Taiko could be the lost Magenta Petrel Pterodroma magentae that had not been sighted or collected since the original collection (Bourne 1964).

    As a schoolboy, I worked with R.J. Scarlett on the osteological collection of the Canterbury Museum. I established that many unidentified petrel bones from numerous subfossil dune deposits and Motion middens in the Chathams could relate to the Taiko as it was described to me in a letter by H.G. Blyth (see Appendix). Blyth (1952) described three colonies of Taiko on his property, two of which disappeared during his 45-year tenure (1905-1950) of land at Tuku. One colony then remained on an inaccessible bush ridge, known as Taiko Hilt. Its existence was confirmed by landowners, the late Manual Tuanui, Ron Seymour, and David Holmes.

    During an expedition to the Chathams in January 1970,1 collected details of strange nocturnal calls heard by these same local farmers. On 3 January 1973, during a further expedition, four birds with dark bodies and white breasts that matched the descriptions of the Taiko and Magenta Petrel were attracted to a bright light powered by a portable petrol-driven generator. The birds did not land when dazzled by a spotlight, so I was unable to confirm the identification. Further expeditions in 1974-75 and May 1976 were also unsuccessful — no examples were captured. However, during an expedition in the summer of 1977/78, extra lights and improved equipment resulted in the capture of two birds on 1 January 1978. The birds were measured, weighed, photographed, and released. Comparison of the dimensions and photographs with the original description confirmed that both were Pterodroma magentae. This paper documents the prolonged search that led to the rediscovery of this enigmatic species and presents data that confirm the identity of Chatham Island Taiko and the Magenta Petrel.
 
 

The search locality was selected on the Tuku-a-tamatea River within the “Moriori Block” based on details included in H.G. Blyth’s letter (24 June 1952), the accounts from land owners Manual Tuanui, Ron Seymour, and David Holmes of nocturnal bird activity in the Tuku River (pers. comm. to DEQ, and the history of the “Moriori Block” purchased by Blyth. This “square mile” was allocated to the Solomon family when land settlement claims were settled last century. The Morioris travelled between their main settlement at Manakau Point, near Owenga, to collect eels, fish, seals and birds. The diary of surveyor Percy Smith for 1868 mentioned two camps used for mutton birding on the southwest coast. One was said to be near the Tuku River (Smith 1868). Recent archaeological investigations confirmed that Taiko were collected near the Tuku at Point Durham (Sutton & Marshall 1977).
 

Catching Taiko
The Tuku River light observation station and Taiko Camp were established within the boundaries of the original Moriori Block. The light observation station was set up 2 km from the mouth of the Tuku River, on a ridge extending into the valley from the south bank, at about 80 m above the river. Vegetation at the site was bracken fern Pteridium esculentum and regenerating Dracophyllum arboreum trees which have tough, long leaves and flexible branchlets. The vegetation not only cushioned each bird’s heavy landing, but also impeded those that tried to fly off again. The landing site was sheltered and confined and the relatively still air provided poor conditions for the birds to sustain controlled flight, so assisting their capture.

    The light station was equipped with two lights, one on a slope behind the ridge and a second lower down, in a sheltered gully where Taiko could be induced to land. Each light system consisted of a 270 W No.1 magnesium photoflood bulb in a concave 0.4 m diameter white reflector with an armoured glass cover, on a 1 rn-high stand. The lights were placed on the ridgetop, and directed at a 450 angle towards the valley. The lights were controlled by a switching system that allowed the photofloods to be alternated and Taiko attracted by the lights could be manoeuvred closer to the sheltered landing site.

    Once birds had been attracted, spotlights (Quartz-halogen bulbs powered by 12 V lead-acid motorcycle or car batteries) were used to confuse Taiko. This caused them to land or flutter into the vegetation. Two spotlights were used where possible, one behind and up the slope from the main light, the other closer to and below it. If possible, a third person was present to catch the Taiko when it landed.

    During expeditions before 1979, four loudspeakers were placed at points in an area of 0.5 km2. Tape-recordings of calls of related species of Pterodroma were played through one speaker at a time, controlled by a switching system. The widely-spread lights and speakers were designed to create the illusion of a colony of several pairs.

    The main lights and tape-recorders were powered by a Honda 1000 W petrol-driven generator which was sheltered in a small sound-proofed shed about 100 m from the lights. The generator was also used to charge the wet cell batteries for the spotlights.

    The lights were more effective during the period of little or no moon, from last quarter to first quarter, especially those nights with foggy, misty weather, and low cloud. During the rest of the lunar cycle, nights with thick mist and heavy showery rain were just as effective. The station was normally attended from dusk to dawn. Once captured, each Taiko was kept individually in a darkened box (440 L x 330 W x 260 D), with ventilation holes. After the bird had been weighed, measured, banded, and examined for moult and brood patch, it was checked for lice and photographed. Normally the Taiko was released at the light observation site, but in adverse weather from the cliff tops at Point Gap (Figure 1). The birds were measured and weighed by various people, as available. Standard measurements were taken with vernier calipers (culmen, tarsus, mid-toe and claw) to + 0.1 mm, or steel rules (wing and tail), usually to ± 1mm. Birds were weighed by Pesola balance, to either +5 or ± 1 g, except for one measurement.
 
 

Figure  1 - Location of catching site and other features on southwestern part of Chatham Island.  Inset: Area in relation to other sites mentioned in text.
 
 

Although searching for individual burrows started in January 1970 and continued during all expeditions up to, and including, the 1977/78, the main objective was to locate Blyth’s “Taiko Hill”. After its rediscovery, a more comprehensive burrow search would then be made of steep hillsides, banks and ridges of the Tuku River and its tributaries.

    Outline of search activities 1969-70: L. Edlin and DEC searched bushed. ridges adjacent to pasture. 1972-73: L. Edlin and W. Campbell searched the bluff opposite Taiko Hill, DEC and 1. Anton (deceased), and others, investigated ridges near the light station. 1974-75: A. Gordon and L. Howell searched ridges on the south side of Tuku Gully towards Taiko Hill and prominent points nearby; M. Imber searched Taiko Hill and the bluff opposite it. 1976: expedition personnel continued searching the same areas. 1977-78: ridges 1.6 km inland from Taiko Hill were investigated. 1978-79:
Tuku Valley and the valley leading to Tixnihonga. 1980-81: tent camps were established at Taiko Hill and Timihonga, night work alternated with intensive burrow searching during the day. Tracks were made from the two camps to main peaks in the area — peaks 175, 281, and 251 were ground-searched. 1982-86: After 13 years of unsuccessful searching for burrows I began collaborating with M. Imber on a radio-telemetric tracking effort (Imber et al. 1994). While radio-telemetry equipment was developed and trailed, burrow searching was limited to revisiting promising areas identified previously in the lower Tuku and no active burrows were found.
 
 

The Chatham Island Taiko — Magenta Petrel
The first two Chatham Island Taiko were captured at the main lights on 1 January 1978. The birds’ coloration agreed with the original description. Further, colour photographs were compared with the colour prints of the Turin specimen (courtesy of Dr P. Passerin d’Entreves of the Museo ed Institute do Zoologia Systematica de Universita d’Torino) and the lithographs in Godman (1907-19 10) and Salvin (1876). Allowing for deterioration resulting from its age, and for its being poorly mounted, the Turin specimen of P. magentae is clearly identifiable with the captured specimens of the Chatham Island Taiko. Bourne’s (1964) conjecture that the Magenta Petrel and the Chatham Island Taiko were the same species was confirmed.

    The main points of comparison were: the short, strongly hooked beak; white areas about forehead and throat; indication of white stripe above eye; and the general distribution of dark areas. Measurements of the Turin specimen and of a series of live birds are given in Table 1.
 

TABLE 1 — Standard dimensions (mm) and body mass (g) of live Chatham Taiko. L, length; D, depth; W, width; Tarsus, tarsus length; MTC, mid-toe and claw. Type, type specimen (Turin), measurements by C. Joanin (pers. comm.)

Description
The following description is based on notes and colour photographs taken of living birds. No differences were identified that could be used to separate sexes.

    Adult: Forehead, crown, hind neck dark brown. White scalloped feathers on forehead merge into dark brown crown; extent and amount of white variable on some individuals. Sooty black around eye, more intense in front. Line of white tipped feathers above the eye, more pronounced in some birds. Chin and throat mainly brown, lighter than crown and neck. Chin with variable patch of whitish feathers, merging into brown of throat. White tipped feathers from white chin patch to lower gape. Dark brown band -extending across fore neck to upper breast, where brown feathers lighten and merge into white breast and belly. Upper parts brown-black; wing coverts, tail, sides of body paler. White flanks with variable number of dark brown, sometimes mottled, feathers (highlighted in Keuleman’s lithographs of the type specimen in Goodman). Undertail coverts white, flecked with grey. Under wing mostly brownish black; greater primary and coverts brownish grey with highly reflective sheen, in some lighting sheen appears almost white.
    Eye brown. Beak lead grey to black. Legs and feet pink, outer half of toes and webs black.

Population size and distribution
In total, 42 birds were seen or caught and banded at the light sites from 1 January 1978 to January 1992. There have been 12 recaptures at the lights, which suggests that the population was greater than 100 birds (Lincoln Index, modified according to observations in Simons (1984)). Despite searches (umber et al. 1994), no birds have been found ashore elsewhere in the Chatham Islands.

Band numbers and recoveries of individual birds, 1978 — January 1992. 1978: A, 1 Jan;
B, 1 Jan; E-127201, 1 Dec, recaptured 7 Nov 1980; E-127202, 3 Dec; E-127203, 3 Dec.
1980: E-127204, 1 Nov; E-127205, S Nov, recaptured 10 Oct 1983. 1982: E-127206, 7
Oct; E-127207, 9 Oct; E-127208, 12 Oct, recaptured 21 Nov 1982 and 28 Nov 1983;
E-127209, 15 Oct; E-127210, 13 Nov; E-127211, 15 Nov; E-127212, 21 Nov; E-127213,
11 Dec, recaptured 15 Oct 1985; E-127214, 15 Dec; E-127215, 15 Dec. 1983: E-127216,
1 Nov, recaptured 11 Dec 1985; E-127217, 1 Nov; E-127218, 7 Nov; E-127219, 9 Nov,
recaptured 10 Dec 1983; E-127220, 24 Nov; E-127221, 4 Dec; E-127222, 11 Dec, recaptured
15 Nov 1987. 1984: E-127223, 5 Jan. 1985: E-127224, 2 Jan; E-127225, 14 Oct; E-127226,
17 Oct; E-127227, 3 Nov; E-127228, 12 Dec. 1986: E-127229, 11 Jan, recaptured 28 Sep
1990. 1987: E-127230, 1 Jan, recaptured 11 Nov 1987; E-127231, 10 Oct; E-127232, 14
Oct; E-127233, 16 Oct, recaptured 31 Dec 1991; E-127234, 17 Oct; E-127235, 17 Oct;
E-127236, 25 Oct; E-127237, 1 Nov; E-127238, 10 Nov; E-127239, 14 Nov; E-127240,
9 Dec. 1992: E-127242, 10 Jan.

Fossil distribution Chatham Island Taiko Pterodroma magentae bones have been found in many Moriori middens and in natural deposits on the main island at Kaingaroa, in the sand dunes from Waitangi to Te One, at Wairau Creek, above Red Bluff, at Maunganui, west of Owenga, at Lake Pateriki, at Mananea, and along all 16 km of Long Beach.

    Many Taiko bones were recovered from archaeological sites between Waihora and Waitangi on the south west coast during excavations conducted by the Anthropology Department, Otago University. The above records confirm that this species was widespread in prehistoric times and was an important item of food for Polynesian inhabitants of the Chathams.
 
 

Relationship to other species of Pterodroma
The three species of lice (Phthiraptera) recorded from the Taiko (see Pilgrim & Palma 1982: 9) have also been recorded from the White-headed Petrel Pterodroma lessonii, the Grey-faced Petrel Pterodroma macroptera, and the Atlantic Petrel Pterodroma incerta. The distribution of species of Mallophaga (feather lice) can be used to elucidate relationships among bird species because lice often co-evolve with their hosts. In this instance, the evidence from the lice shows a possible relationship between the Taiko and the species of Pterodroma mentioned above. This conclusion is only tentative because of the limited number of lice samples available. Further collecting will help to confirm known records and may well uncover more species.

Distribution, and observations at sea
The Chatham Island Taiko has been found to occur widely as recent fossils in middens and “sand blows” throughout the main Chatham Island, but has never been recorded on other islands in the group. The breeding population has probably always been concentrated on the southwestern region of the southern tableland.
    In breeding and feeding distribution, the Taiko occupies a position intermediate between the subtropical Grey-faced Petrel and the subantarctic White-headed Petrel. The Chathani Islands lie within subantarctic waters in winter, but within sub-tropical waters in summer when the subtropical convergence lies to the south of the archipelago (Deacon 1937). The original observations of Giglioli and Salvadori from the Magenta were of individuals foraging in winter in the area of the subtropical convergence to the east of the Chathams.
    Recently there have been a number of observations at sea (e.g., Bourne & Dixon 1975, Rogers 1980). One previously unpublished report is that by Sir Peter Scott who reported sighting a Taiko from the Lindblad Explorer on 5 February 1982, 72 miles north of Kaingaroa, main Chatham Island (P. Scott, pers. comm.).
    Some of these sightings are controversial (Imber 1980, Fades & Rogers 1982). Field identification of this species is difficult because the Softpluinaged Petrel Pterodroma mollis has a similar plumage pattern, and is found over the same range as the Taiko (Imbcr 1980). Fades & Rogers (1982) summarised the main identification features of both species. Unless closely related species are present for comparison and good lighting conditions are available, positive identification may be impossible. Further study of Taiko flight patterns and field characteristics at or near the breeding grounds will be necessary so that the “jizz” of this species can be clarified to help observers to distinguish between these two similar petrels.

My observations of this species at night with a bright white light revealed it to be an agile fast-flying bird of a uniform black and white plumage with highly reflective outer primaries. No other features appeared to be prominent or consistent. The variable white face markings were observed only at close quarters and with good lighting.

The Taiko as a muttonbird
Blyth (1952, see Appendix) mentioned that Morioris collected muttonbirds from an inaccessible bush ridge on his property but that these activities had ceased about 1900. Smith (1868) described campsites in the bush used during muttonbirding trips. When Moriori land issues were settled by the Maori Land Courts, one square mile was allocated at Tuku to the Morioris as a Reserve on which to collect eels and birds. A track through the bush and across the clears linked the Moriori Block with their main property at Manakau Point, Owenga on the East Coast. During Blyth’s 45 years tenure, an elderly Moriori, Riwai, lived in a tree-fern pole hut near the mouth of the Tuku. Riwai, with Charles Seymour and Tommy Solomon, made annual trips to the Taiko breeding ground, the last being in 1903 when 300 birda were collected (D. Holmes, pers. comm.). The Moriori Block on which Taiko Camp and the light observation site are situated was purchased by Blyth about 1908 and incorporated into his property which is now owned by B. and L. Tuanui of Awatotara.

    Moriori mutton-birding practices were described in detail in a paper by Skinner & Baucke (1928). The highly organised and ritualised, annual process matched events recorded by other authors. The Tchaik fledglings (Moriori name for Taiko) were collected in a cruel way with a lingering death which missionaries modified to a more humane method. Conservation of the burrows was of prime importance to the way fledglings were collected. Although large numbers of fledglings were collected annually for food by the Moriori (Skinner & Baucke 1928), and later by the Maori inhabitants of the Chathams, the species was still common at the time of European discovery.

    Archaeological evidence describing the collecting of the Taiko, and the bird’s place in Moriori culture, were described by Sutton & Marshall (1977). They excavated and analysed material from Waihora Mound, 5 km northwest of Tuku. Taiko bones contributed 53.22% and 49.0 1% of total bone mass in two middens. Other middens on the coast contained smaller percentages of Taiko bone, with amounts decreasing progressively at sites to the north (Sutton & Marshall 1977). This supported the interpretation that the main concentration of the Taiko breeding population was on the southwestern coast. Bones found in middens throughout the island may have been carried there as food, or were remains of birds cast ashore and included in natural bone deposits.

Reasons for decline in numbers
Following the discovery of the Chatham Islands by W.R. Broughton of HMS Chatham in 1791, the islands became an important centre for whaling and sealing ships. Cattle, pigs, horses, and sheep were introduced, but soon became feral. Cats and two species of rats (Rattus norvegicus and R. rattus) also arrived in the European period, but the Polynesian rat R. exulans had arrived with the Morioris.

The inaccessible southwestern coast apparently suffered little from Maori and European settlement until the turn of this century, except for the occasional birding campsite and bush gardens. The pristine state of the vegetation is described in the diaries of Percy Smith from notes made during his field trips to the south coast in 1868 (Smith 1868). Blyth (1952, Appendix) attributed the disappearance of two colonies of Taiko to “the march of civilization”.
The destruction of the primeval vegetation and the pastoral development of the Otonga Block which includes the traditional breeding grounds of the Taiko was outlined by Begg (1977). Pastoral development has added another dimension to the pressure on a declining Taiko population by concentrating the feral cattle, sheep, pigs and cats in the southern block.

    The Weka Gallirallus australis (Rahhidae) was introduced to the main island from the South Island of New Zealand in 1905 (Oliver 1955). Wekas are known to catch and kill Pterodroma petrels and have been implicated in the destruction of the colony of Cook’s Petrel Pterodroma cookii on Codfish Island (O’Brien 1990) The Brush-tailed Possum Trichosurus vulpecula was introduced in 1911 (King 1990) and it also maybe associated with the decline of the Taiko, as it is known to take eggs and chicks of birds (Brown et al. 1993). The major factor(s) that have led to the decline of the Taiko population which existed in primeval times will probably never be known, but one or all of the factors listed above are probably major contenders.

Vernacular name of the Taiko
With the establishment of the identity of the Chatham Island Taiko as Pterodroma magentae, a minor point of contention is which of the available vernacular names should be adopted for general usage. Godman (1910) referred to it as Gigliohi’s Fulmar, a name which has not stood the test of time. Bourne repeatedly used Magenta Petrel, which seems to be the popular choice at present. M. Pietro Passerin d’ Entreves (Turin University Museum) has pointed out that the Pirocorvetta Magenta was named after the battle of Magenta in 1859, itself an allusion to the blood shed in the baffle. Kinsky (1970) listed it as PterQdroma magentae (Giglioli and Salvadori, 1868) Chatham Island Taiko, but in Turbott (1990) the vernacular name is given as “Magenta Petrel (Chatham Island Taiko)”. I prefer to follow Kinsky (1970), especially for usage in New Zealand and the Chatham Islands because of the link to the cultural history of the bird.
 
 

I thank all sponsors and supporters of Chatham Island Taiko Expeditions over 25 years. The Taiko Research Project received financial assistance from the following: Ornithological Society of New Zealand Project Assistance Fund; Department of Conservation — Canterbury Conservancy; Environmental Council; Federation of the Acclimatization Societies of New Zealand; Mobil NZ Environmental Grants; National Geographic Society of America; National Provident Fund; New Zealand Lottery Grants Board; New Zealand Wildlife Service; Royal Forest and Bird Protection Society of New Zealand, Inc.; World Wide Fund for Nature/New Zealand. For the support and interest shown during the developing phase of this research I am most grateful to the following: Early research *H.G. Blyth, Dr W.R.P. Bourne, P. Passerin d’Entreves, *5fr Robert Falla, *5fr Charles Fleming, Warren B. King, Christian Jouanin, M.D. Murray, Mrs M. Ringer, George Watson. Support on the Chathams *Manuel, Evelyn, Bruce and Liz Tuanui, David Holmes, Robert Holmes, Ron Seymour. Expedition support Maxj Clode, Reg Cotter, Ruth Crockett, Lawrence Edlin, *Alastair Gordon, Audrey Gordon, Laurie and Alison Howell, Hugh Robertson, and Russell Thomas.
Assistance with this paper Brian D. Bell, Ruth Crockett, Barrie Heather, Michael Imber. Ricardo Palma kindly examined and commented on the Mallophaga. The National Library of New Zealand kindly allowed access to the MS diary of Percy Smith. I am grateful for measurements of the Tunn specimen provided for me by W.R.P. Bourne and C. Jouanin. Sincere thanks to Terry Conaghan for drawing the map, and to Sheryl Feetau for typing the MS.
*Deceased
 
 

70 Purii St

CHRISTCHURCH W.1
Yours just of May 21st to hand for which I thank you. I am afraid if Dr Falla cannot tell you much about the “Taiko”, your chance with me is even less. I will happy to tell you what I know from my own observations although I’ve not made a study of the Birdlife of the Chathani Islands. I lived there about 40 years and still have a Run there situated at the S.S.W. end of the Island. Gap Point is on my property which runs along the Coast for about 4 miles.

I do not know if the “Taiko” is in N.Z. I’ve not heard of it, if it is. I don’t know if you have ever seen the Muttonbird or “akivakive” (I do not know if either of these Maori words are spelt right).

The “Taiko” is very like a muttonbird to look at but it is smaller and has a white breast, it’s habits are those of the Muttonbird with the exception that it nests inland, burrowing into a peat hillside. I have never heard of the “Taiko” being on any place other than mine on the island, although of course it is quite possible that it nests in other places. It is safe to say I think that it is found in no other area than in the south portion of the island. There was at one time 3 rookeries on my farm, but shall I say the advance of civiLisation has driven them out of,two. The remaining one was situated on a bill “Timihonga” about 2 miles inland. I say “was” because I do not know if it is still there, for in my 45 Years of ownership, I’ve never been onto the “Hill” and I do not think it has been visited during the last 50 years. I have been told that the Old Morioris used to make annual visits to the Hill, how they got there remains a mystery, as it is an area comprising a very thick and rough Bush and swampy ground.
Personally I think the rookery would still be there for in their habit of burrowing they penetrate into a hill much further than the ordinary muttonbird. I had several of these latter birds on the Farm but with Pigs, Cats, Wekas and Penguins have helped to destroy these, as usually they burrow only a short distance into a cliff face which is always in peat, but the peat here is only a thin overlay on the Rock face usually. Referring to the Mutton Bird these are the same as in New Zealand, some say the Island ones are a bit larger.