We present an annotated working list of the bird species breeding in New Zealand (luring the late Pleistocene and Holocene, up to the time of human contact. New Zealand is defined as including the three main islands and the surrounding smaller islands, plus outlying island groups from Norfolk Island in the northwest, the Kermadec, Chatham, Bounty, Antipodes, Campbell, Auckland,:Snares, to Macquarie Islands, but excluding islands south of Macquarie Island and the Ross Dependency. Inclusions or exclusions of species from the list were based on specified criteria. We include only species with a breeding population and not vagrants that occur in New Zealand but which breed elsewhere. Species with validly published names were included if there was fossil evidence for a breeding population before human contact. Species with a breeding population at the time of European contact were included unless contrary evidence from the fossil record indicates that they actually colonised after human settlement. Species without a fossil record were included if a breeding population exists on a relatively undisturbed island within the New Zealand archipelago as defined above. Species now present on the main islands were excluded if they are absent from all well-documented fossil faunas. Species were excluded from the breeding fauna and treated as vagrants where sustained breeding has not been demonstrated. The phylogenetic species concept is applied both to fossil and to living taxa. The late Quaternary fossil record of birds in New Zealand is excellent, and the contribution of extinct taxa to the total list is understood at least as well as that of the surviving taxa. Many taxa presently recognised at subspecific level are treated here as full species. Twelve extinct species whose former presence is known from fossil evidence, but for which no description has been published, are listed under informal species designations. Taxonomic considerations limited the extent to which the main list could reflect present understanding of the diversity of the avifauna; some undescribed species are at present subsumed under one species name. Where previous taxonomic publications provide precedence, available names at the species-level have been used. A supplementary hypothetical species list includes all nomenclatural changes signalled in extensive annotations to the main list. In this list we accept 245 species in 110 genera representing 46 families; 176 species were endemic to the archipelago. Preliminary biogeographic analyses based on the composition of the supplementary list show that there were four separate regional faunas: a northern subtropical fauna (Norfolk, Kermadecs); the major fauna of the main islands (North, South, Stewart, and offshore islands); a Chathams fauna (Chatham Islands only); and a subantarctic fauna on the southern islands. Species with wider distributions formed link groups. The origin and compositions of the regional avifaunas and their endemic species differ with their geographic position, climate, and proximity to source faunas. Instances of speciation in groups such as the Coenocorypha snipe and Petroica flycatchers, and adaptive radiations in groups including moa and acanthisittid wrens, show that there are many avenues for research on the rate of evolution in island and mainland populations of New Zealand birds and that there are large gaps in knowledge of even common taxa. A brief case study demonstrates the inadequacies of using species lists that do not include Holocene fossil species. Species-area curves based on the total fauna differ substantially from those developed in previous studies based on incomplete, or biased, lists. Pleistocene glaciations caused the pattern of distribution of species on the main islands to change in concert with vegetation changes. Other possible effects include the elimination of warm climate species early in the cooling phase more than 1 million years ago, the speciation in groups including waders and parrots as new habitats (e.g., braided riverbeds and alpine areas) appeared, and the appearance regularly during the Pleistocene of islands that were potential staging points for colonisation of the Chatham Islands. For at least the past 100 000 years, until 2000 years ago, the fauna appears to have been very stable in composition, despite strong cyclic fluctuations in climate and vegetation. The effects of extinctions within the past 2000 years on the composition of the present fauna include the elimination of most of the endemic taxa from all but the subantarctic faunas. Only 169 species of the original late Holocene breeding fauna survive. The extinctions have resulted in a strong bias towards marine and coastal taxa in the present avifauna, in contrast to the balanced representation of terrestrial and marine species in the Pleistocene and Holocene fauna. The importance of systematic studies and the determination of the status of island populations to conservation and basic ornithological research is emphasised. The systematic status of many New Zealand birds is poorly known at present.
