Could the New Zealand storm-petrel still exist?


On 25th January 2003 a group of 12 birders were on a pelagic trip out of Whitianga, Coromandel Peninsula, North Island, New Zealand, organised by Wrybill Birding Tours, NZ. We were quite close to the Mercury Islands where the main target species of the trip – Pycroft’s petrel, Pterodroma pycrofti, breeds. We had been among large numbers of White-faced Storm-petrels – WFSP, Pelagodroma marina, for the previous hour or so when SS spotted a blackish storm-petrel with a gleaming white rump approaching the bow. Several other observers quickly got onto the bird and BS was able to take some photographs as the bird, flying strongly, made a couple of circuits in front of the vessel, before departing. It was in view for about one minute.

The bird was slightly smaller and slighter than the WFSPs nearby (seen within metres of each other) and appeared all-dark above, apart from the white rump, with a whitish belly and underwing. All those on board who saw the bird well were quite convinced that we had seen a Black-bellied Storm-petrel – BBSP, Fregetta tropica, and we all thought that we had seen a dark centre to the belly. This would have been an unusual sighting for the location and time of year, and everyone was pretty pleased with themselves – but the photographs threw us into confusion! The pictures clearly showed no trace of a dark mid-belly stripe, but rather a strangely streaked underside, which forced us to reconsider the ID.

Three species of black and white storm petrels have been encountered in New Zealand waters. The White-bellied Storm-petrel – WBSP, Fregetta grallaria, breeds on the subtropical Kermadec and Lord Howe Island groups and very rarely strays southward to NZ, while Wilson’s and BBSP breed to the south of New Zealand and pass on migration in spring and autumn (Wilson’s in some numbers and BBSP rarely). BBSP and WBSP have notoriously variable plumage (Murphy and Snyder 1952, Marchant & Higgins 1990) and variants have been also been described for Wilson’s Storm Petrel (Bourne 1987, Curtis 1988). The bird in the pictures matches nothing in the literature, so opinions were sought from New Zealand and Australian seabird experts, but the waters just became muddier by the day.

Initially, some thought it might be WBSP, but most decided that it must be BBSP. However, all those responding in favour of BBSP said they had not seen a bird with such strangely marked plumage. Eventually, it was suggested, almost in jest, that the Whitianga bird had a striking similarity to specimens of the supposedly extinct New Zealand Storm-petrel – NZSP, Oceanites maorianus. Now things were getting really interesting!


Elimination of extant Storm-petrel species

Clearly we are dealing with something unusual. We can readily eliminate most of the world’s storm-petrel species on the basis of size, shape or plumage. This leaves us with a handful of black-and-white Storm-petrel species to consider.

The bird’s body plumage resembles that of Elliot’s Storm-petrel, O. gracilis, but this little known species is extremely small, and would be dwarfed by WFSP (del Hoyo et al 1992). Furthermore, Elliot’s Storm-petrel has dark underwings. Similarly, based on size and structure, we can eliminate the little known central-Pacific White-throated Storm-petrel, Nesofregetta fuliginosa, which appears to be one of the largest of the Storm-petrels (del Hoyo et al 1992). More obvious contenders are BBSP and WBSP.

White-bellied Storm-petrel? We believe that we can eliminate WBSP by a number of features. Firstly, it should be as large as but more bulky than WFSP, but our bird was clearly slightly smaller and definitely more slender in direct comparison. Secondly, the extent of white/pale on the underwing should be much more pronounced than in our bird. Indeed any bird showing as much dark in the belly should be dark in the axillaries, not pale as ours was (Marchant & Higgins 1990). Thirdly, the upper-wing is quite wrong for WBSP. In fresh plumage they often have fine pale edges to the coverts but nothing that would be visible in the field (Marchant & Higgins 1990). They are not known to show a pronounced pale brownish panel on the upper-wing like our bird. Our bird clearly shows feet projecting well beyond the tail, which is often quoted as being a diagnostic feature separating the two Fregettas, (e.g. Marchant & Higgins 1990). However, comments by some experienced seabird observers now lead us to believe that this may be an unreliable character.

Black-bellied Storm-petrel? It should be noted that several well-respected birders think that our bird is ‘just’ a BBSP, but we are not so certain. In our opinion BBSP is probably precluded using a similar argument to WBSP, certainly regarding size and the upper wing pattern. Also, from photographs of this species, the underwing pattern of BBSP is almost always cleaner and always shows a dark distal ‘hook’ formed by dark tips to the median primary coverts. The under-tail coverts usually show far more black, but some birds without much black on the belly do have this area white. The ‘bleeding’ projections of black onto the belly are definitely unusual. Dark spotting on the flanks is found in a small proportion of BBSP (Murphy and Snyder 1952) but seldom to this degree (see below). The only known example of a bird with such plumage is the original specimen of ‘Pealea lineata’ (see below). And finally, the ‘jizz’ of the bird (even from the photos), although subjective, seems quite unlike the bulky, fat-headed, broad-winged appearance of the Fregettas. In all photos, even where the bird is not banking sharply, as in the accompanying photo, the wings are very pointed and long (there are more photos at our website).

Wilson’s Storm-petrel? An aberrant Wilson’s Storm-petrel could be a possibility. The upperparts of our bird look good for Wilson’s, having the pale panels on the upper wing that are usually present in this species. Melanism and partial albinism have both been recorded for this species (Bourne 1987, Curtis 1988) but the only specific references to albinism we have seen are to the three NZSP specimens! The Museum of New Zealand also has a partial albino Wilson’s skin, which has a smudgy greyish patch on the belly. Size of our bird would be at the upper extreme for Wilson’s (Marchant & Higgins 1990).


If this bird is not one of the extant Storm-petrels, then could it be a NZSP?

NZSP is currently only known from three very old specimens collected in the 1800s (one now in Tring and two in Paris) but even so, their story is confused by ‘Pealea lineata’. A strangely striped storm petrel was collected off Samoa by the American Exploring Expedition in 1842 and became known as ‘Pealea lineata’. Murphy and Snyder (1952) considered this specimen to be an aberrant BBSP, and that the three New Zealand specimens (which had previously been assigned to this form) were of a different species. Note that Marchant & Higgins (1990) draw an incorrect conclusion from Murphy and Snyder (1952). NZSP has also been considered as a sub-species of Wilson’s Storm-petrel (Harrison 1983).

Recent investigation of New Zealand sub-fossil deposits has revealed that the bones of several Storm-petrels do not belong to any recognised species. This naturally raises the question of the identity of NZSP again. Recently, IS has re-examined the old skins with a view to drawing comparisons with sub-fossil bones and clarify the form’s taxonomy. Although findings are preliminary, indications are that it may best be treated as a distinct species. Further, initial comparisons with sub-fossil material suggest they may be one and the same form.

The three NZSP skins are similar in showing a substantial amount of pale plumage under the wing. This varies from all white under-wing coverts on one individual, to another with the primary coverts and axillaries strongly washed with brown. They also have white flanks and belly, variably and irregularly streaked with dark sepia, and pale patches on the upper wing, although they vary in conspicuousness. The streaking on the lower breast varies in intensity, but in all three there are two black lines from the edges of the breast to the under-tail, which are either faint and broken, or broad and obvious. The mid line on the belly is white in all three specimens.

In each of the above characters there is at least one skin that matches the plumage of the bird in the photographs in detail. Measurements of the skins place them at the larger end of the range for Wilson’s Storm-petrel, so size is consistent with our bird. The pointed wings of our bird are also more consistent with Oceanites than Fregetta, which concurs with the likely taxonomic affinity of NZSP. The closeness of the plumage of our bird to the skins is uncanny. If NZSP was known to survive we’d be more comfortable claiming it had been seen and photographed.

Proof of survival would only come with the discovery of a live (or fresh) specimen, or from further sightings of birds at sea similar to the one photographed. It is quite possible that not all of the earlier records of ‘black and white’ Storm-petrels (BBSP and WBSP) from New Zealand waters have been correctly identified. It is hoped that by documenting this sighting and the difficulties surrounding the bird’s identification, observers of ‘black and white’ Storm-petrels will be more aware of the possibilities.

Is it so unlikely that a Storm-petrel could exist unseen or unreported in New Zealand waters for 150 years? Bourne (1965) reviewed a number of species of ‘missing’ petrels that had been rediscovered and several new species that had been found. He also made some predictions of other missing species that he thought might still be found. Several of them have indeed been rediscovered. Most relevant to New Zealand was the rediscovery, in 1979, of the Magenta Petrel (Chatham Island Taiko), Pterodroma magentae, which had not been seen since the collection of the original specimen in 1867. The breeding grounds of the relatively numerous Elliot’s Storm-petrel have not been found. With few observers covering the relatively huge coastline of New Zealand, and with so many offshore islands containing seabird colonies, the discovery of a tiny bird that only comes ashore at night might not be as outrageous as it first appears.

We are by no means certain about this bird’s identity, however, and would welcome further comment. We thank all those who have already taken the time to give their opinions. We shall update the argument on our website, as it evolves. Whatever this bird is – it certainly is an oddity.

The under parts of the two NZSP specimens held at the Paris Museum (Museum d’Histoire Naturelle de Paris), with the underwing of one of them exposed.

The under parts, upper parts and webs of the NZSP specimen held at the British Museum, London. It appears that the webs of this specimen have not been varnished, and so they appear to be naturally dark in colour.

Sav Saville and Brent Stephenson
Wrybill Birding Tours, NZ

Ian Southey


We would like to acknowledge the assistance of the following people in the preparation of this article – Rohan Clarke, Alan Tennyson, and Paul Scofield.


Bourne, W.R.P. 1965. The missing petrels. Bulletin of the British Ornithologists’ Club 85: 97-105.

Bourne, W.R.P. 1987. Parallel variation in the markings of Wilson’s and Leach’s Storm-petrels. Sea Swallow 36: 64.

Curtis, W.F. 1988. An example of melanism in Wilson’s Storm-petrel. Sea Swallow 37: 63.

del Hoyo, J., Elliott A. & Sargatal, J. 1992. Handbook of the Birds of the World. Barcelona: Lynx Edicions.

Harrison, P. 1983. Seabirds: an identification guide. Croom Helm Ltd., Beckenham, Kent.

Marchant, S. & Higgins, P.J. 1990. Handbook of Australian, New Zealand and Antarctic birds. Oxford University Press, Australia.

Murphy, R.C. & Snyder, J.P. 1952. The ‘Pealea’ phenomenon and other notes on Storm-petrels. American Museum Novitates 1596: 1-15.