Sources & licences

This guide is built on public sources and on media that are under a Creative Commons licence or in the public domain. Below, first where the data come from — the species list, the tree, the measurements, the habitats and the range maps — and then the full list of photographs and recordings, with author, licence and source page for every file. Non-commercial (NC) licences are admitted; no-derivatives (ND) licences are not. For this region a number of choices differ from those in the European and North American guides; those differences are named below.

Where the data come from

The guide rests on a handful of public data sources. They are gathered here, with what they supply and on what terms.

• AviListCC BY 4.0

  The arrangement and sequence of orders, families, genera and species, the scientific names and the Red List category. AviList is the global checklist that since 2025 has been maintained jointly by the major checklists — the IOC list, Clements and BirdLife — so that there is one list instead of three. This guide follows version 2025b, including where it departs from GBIF.

• B10K and the PhyloCode

  The clades above the level of the order, which appear in the tree and in the side menu. They follow the family-genome tree of the Bird 10,000 Genomes project (Stiller et al. 2024) and carry names established under the PhyloCode — a naming that labels a branch of the tree rather than a rank.

• AVONETCC BY 4.0

  The wing length and the mass of every species. Eleven measured traits for all 11,009 bird species in the world, built from measurements of 90,020 individual birds in museums and in the field (Tobias et al. 2022). One source for all species, so that two birds in this guide can genuinely be compared.

• The Australian Bird Guide

  The length of the bird, for the whole area in which the species appears. Menkhorst, Rogers, Clarke, Davies, Marsack and Franklin give a range and measure birds in the field; it is the modern standard for Australasia and the tradition this guide holds to.

• New Zealand Birds Online

  The length of the New Zealand endemics, which appear in no Australian field guide. The site is maintained by Te Papa Tongarewa together with Birds New Zealand. Which of the two sources supplied a length is recorded per species in the data file.

• GBIFCC BY 4.0

  The species list, the season and the density on the range map. GBIF gathers records from museums, ringing stations and observer networks into one searchable whole; for this region the great majority arrives through eBird and Birdata. This guide uses GBIF not only for the map but also for the question of which species belong in it.

• GRIISCC BY 4.0

  Which species in Australia, New Zealand and on Norfolk Island have been introduced. The Global Register of Introduced and Invasive Species is compiled country by country and published through GBIF; this guide uses the Australian and the New Zealand list, checked by hand.

• NVIS — Major Vegetation GroupsCC BY 4.0

  The backbone of the habitat list on the Australian side. The National Vegetation Information System divides the continent's vegetation into major groups — eucalypt forest, mallee, hummock grassland, acacia shrubland — and so supplies the words that a European or American classification does not have.

• Land Cover DatabaseCC BY 4.0

  The same for the New Zealand side. Manaaki Whenua's LCDB distinguishes, among others, southern beech forest, tussock grassland and the bare shingle of the South Island's braided rivers, all three habitats with birds of their own.

• Natural Earthpublic domain

  The coastlines, the national borders and the boundaries of the states, territories and regions on which the range map is drawn, at 1:10 million.

• Wikimedia CommonsCreative Commons or public domain

  All photographs: of the birds, of the habitats and of the front page. Every file carries CC0, public domain, CC BY, CC BY-SA, CC BY-NC or CC BY-NC-SA; no-derivatives (ND) licences are not admitted. Author, licence and source page are listed below for each file.

• xeno-cantoCreative Commons or public domain

  The sound recordings. They play at xeno-canto and are not held on this site. Only recordings under CC0, public domain, CC BY, CC BY-SA, CC BY-NC or CC BY-NC-SA have been used.

Readers' photographs

Anyone with a photograph of a free-living wild bird that the guide still lacks can submit it under a Creative Commons licence. The rules and the form are on a page of their own. Submit a photograph

Which birds are in it

The map settles the guide. The window runs from Steep Point in Western Australia to beyond the Chathams and the Kermadecs, and from the Torres Strait to Macquarie Island — 6,956 by 5,243 kilometres. That window is also the boundary of the list. Christmas Island and the Cocos Islands fall outside it: Australian territory, but they lie against Java and their birds are Asian. New Guinea, Timor, the Solomons, New Caledonia, Fiji and Vanuatu are drawn as land, because a map is a rectangle; their birds do not belong in the guide.

The guide covers Australia, New Zealand and Norfolk Island. Of these it holds everything that occurs annually — breeders, winter visitors and passage birds, however scarce — plus vagrants with more than ten records. Extinct species are not included; they are named in the text of the family pages.

Here the species list was measured, not adopted. For North America there was the ABA checklist, with a committee doing the judging and a rarity code per species; for this region no such list is available to download. So it was measured from GBIF, over ten full years: 2015 through 2024, and not 2016–2025, because the passing of eBird data to GBIF lags behind — in September 2026 fewer than a fifth as many Australian records stood for 2025 as for 2024, and a half-delivered year counts in the rule as a bad year.

The rule: a species seen in eight of the ten years with at least ten records a year — three for Norfolk, which is far less visited — occurs annually and goes straight onto the list. The rest go through the ten-record rule: more than ten occurrences, an occurrence being a distinct combination of year, month and quarter-degree cell. Counting records would turn a single overwintering vagrant with three hundred sightings into a common bird; counting occurrences does not.

The outcome, from 999 species resolved: 849 on the list — 781 native, 32 vagrant and 36 introduced — and 150 dropped.

What that measure cannot do. It measures presence, not origin: an escaped cage bird that hung about the same spot for eight years passes the rule as easily as a breeder. And the reverse: a thinly reported island bird sometimes fails the threshold, not because it is absent but because nobody goes there. The borderline cases were therefore worked through by hand: fourteen species were admitted after all — the Snares Snipe, the Chatham Petrel, the Kakapo, the Princess Parrot — and ten escaped cage birds were kept out, from the Mandarin Duck to the Canary. Each of those twenty-four decisions is recorded with its reason in data/soortenlijst.json. Nothing happens silently.

Which name belongs to which bird

The guide follows AviList v2025b, including where it departs from GBIF. For this region the two diverge much further than elsewhere: at roughly a hundred of the thousand measured species, against 58 of 731 in the European guide. For Australasia the GBIF backbone still largely follows the arrangement of before the major revisions — Megalurus is Cincloramphus and Poodytes in AviList, Chrysococcyx is Chalcites, and the White-bellied Sea-Eagle sits under Haliaeetus in GBIF and under Icthyophaga in AviList.

A change of genus is visible and therefore fixable. The second case is not: GBIF sometimes carries a name that AviList also knows, but for a different bird. Then the name is right, the bird is wrong, and no error is ever raised. The Rainbow Lorikeet stood under Trichoglossus haematodus — that is the lorikeet of New Guinea; the bird that flies here is T. moluccanus, and it accounts for 1,167,407 of the 1,168,319 records under that name.

Such cases were tracked down by asking, for every measured GBIF species, for its subspecies and synonyms, and checking whether any of those names is a species in AviList; each branch was then counted for how many records it holds in this region. That produced twenty-six cases, ten of them turning on a name that belongs to another bird. They are listed with their counts in data/splitsingen.json.

Measurements: wing, mass and length

The wing length and the mass in the box beside each bird come from one source: AVONET. The wing length is the ringer's measure: from the carpal joint — the bend halfway along the wing — to the tip of the longest primary. The mass is the mean of the birds weighed; a bird's weight varies widely over the year, so one figure says more here than a range from another source.

The length of the bird does not come from AVONET but from the field-guide tradition, and for this region there are of necessity two of them. The Australian Bird Guide by Menkhorst and others covers everything that occurs in Australia. The New Zealand endemics appear in no Australian field guide; their length comes from New Zealand Birds Online. The boundary therefore runs along the range and not along the species, and which source supplied a measurement is recorded per species in the data file. Two traditions is one more than we allow ourselves elsewhere; the alternative was no length at all for several hundred species.

Wingspan is not given. It is recorded in no series of measurements on living birds and would have to be estimated from wing length with a ratio that differs by family; that calibration has not been done for this region.

• Tobias, J.A. et al. (2022), AVONET: morphological, ecological and geographical data for all birds. Ecology Letters 25, 581–597. CC BY 4.0.
• Menkhorst, P., Rogers, D., Clarke, R., Davies, J., Marsack, P. & Franklin, K., The Australian Bird Guide, CSIRO Publishing.
• New Zealand Birds Online, Museum of New Zealand Te Papa Tongarewa with Birds New Zealand.

The Red List categories

The status in each species' summary box is the global assessment of the IUCN Red List, as carried by AviList — not an Australian, New Zealand or other national list. A species listed under Australia's Environment Protection and Biodiversity Conservation Act, or on New Zealand's Department of Conservation Threat Classification System, can be “least concern” globally, and the reverse also occurs. On the cards in the overview the code is shown only for species outside the lowest category: were it everywhere, the threatened bird would be the one that no longer stood out.

The IUCN assesses every species against fixed criteria: rate of decline, size of range, population size and probability of extinction. The nine categories, from lowest to highest risk:

LC Least concern

Assessed and not in any of the threatened categories: usually a widespread and numerous species.

NT Near threatened

Does not currently meet the thresholds for Vulnerable but is close to them, or is expected to meet them in the near future.

VU Vulnerable

High risk of extinction in the wild: for example a decline of 30 percent or more over ten years or three generations, or fewer than ten thousand mature individuals.

EN Endangered

Very high risk: for example a decline of 50 percent or more, or fewer than 2,500 mature individuals.

CR Critically endangered

Extremely high risk: for example a decline of 80 percent or more, or fewer than 250 mature individuals.

EW Extinct in the wild

Survives only in captivity or as an introduced population outside its original range.

EX Extinct

There is no reasonable doubt that the last individual has died.

DD Data deficient

There is too little information on distribution or population to assess the risk.

NE Not evaluated

Not yet assessed against the criteria.

• IUCN (2012), IUCN Red List Categories and Criteria, version 3.1, second edition. The thresholds above are the outline; the full criteria are in that document.

The Red List

The status behind each species name is the IUCN Red List category, taken from AviList rather than fetched separately from the IUCN, so that the name and the status always rest on the same species concept. That is more than a formality here: where AviList splits a species and the IUCN does not, or the other way round, a separately fetched category would land on the wrong bird.

For sixty-four species AviList leaves the category blank — it reads not evaluated. Fifty-four of those have in fact been assessed by the IUCN, and that assessment is taken from the GBIF Red List mirror, which carries the IUCN taxon number with it. Every species records in the data file where its category comes from. The emu was the clearest case: AviList gives it no category, while the IUCN assessed it in 2018 as of least concern.

Species in category EX (extinct) or EW (extinct in the wild) are not in this guide; the checking program refuses them. For this region that is a heavier sieve than elsewhere.

The range map

The map has two layers of different origin, and the difference matters enough to be stated here.

Which seasonal colour a state, territory or region receives — resident, breeding, winter or passage — is measured here, not set. Each unit is counted month by month and calibrated against all bird records in that same unit, over 2006–2026. The seasons stand on their head relative to the other two guides: summer is December and January, winter is June and July. A program that follows the north here turns every Australian breeder into a winter visitor, and that looks plausible enough to pass unnoticed. Nine small islands — Lord Howe, Macquarie, Norfolk, the Kermadecs, the Three Kings, the Antipodes, Auckland, Campbell and the Snares — have no code of their own in the boundary set used and cannot be measured; their season is set by hand.

How dark that colour is comes from the GBIF records: counted per 0.7° cell, divided by all bird records in that same cell, and averaged per unit. That division is needed because GBIF counts observers and not birds: without it the map is a map of where birdwatchers live.

Dividing by a small number gives nonsense, so the number of thinly covered cells was counted beforehand. Of the 1,505 land cells in the map window, 262 — 17.4 per cent — fall below the threshold of two hundred records. That is well under the limit of one third above which such maps become dishonest, so the density layer is admissible here. The thin cells lie where you would expect: the interior of Western Australia.

The map unit is the state, territory or region, not the country. There are forty-five drawable units, eight of them surrounding countries drawn only as background. One warning belongs with that: Western Australia is 2.5 million square kilometres. One colour for that whole slab says almost nothing — a species confined to the south-western corner colours an area the size of Western Europe. The density layer therefore carries more of the story here than in the other two guides, and where the state colour and the density contradict each other, the text wins: the block Range and movements says what the map cannot.

The projection is Lambert azimuthal equal-area about 148°E and 32°S, at 640 by 482 pixels. The window crosses the date line — the Chathams lie at 176°W — so the map selects on angular distance from the centre of the projection rather than on a rectangle of longitude and latitude, which would have a seam there.

Introduced species

Which species have been introduced comes from GRIIS, the Global Register of Introduced and Invasive Species, through GBIF. New Zealand is the clearest case: Blackbird, Chaffinch, Song Thrush, House Sparrow, Starling and Dunnock are established and common there, and a guide that leaves them out is useless in the field. A species native in Australia and released in New Zealand is marked differently for each country.

There is a trap in it. The Australian GRIIS list records not only species from outside but also movements within Australia: a bird released in Tasmania or on Kangaroo Island appears on it, even though it is thoroughly native on the mainland. Taken over unfiltered, that made the Emu, the Laughing Kookaburra and the Superb Lyrebird introduced species. Of the sixty-eight species on the lists, thirty-two were therefore returned by hand to native status, each with its reason; thirty-six remain marked as introduced.

The habitats

Every species is tied to the habitats in which it actually rests and forages, at most seven. Ten main groups from sea to city, and forty-eight formations below them.

The classification is one made specifically for this region. Neither the European (EUNIS) nor the North American (USNVC) system fits mallee, spinifex, gibber, brigalow, monsoon forest, tussock grassland, braided river or southern beech forest — and these are not edge cases but the vegetation of most of the continent. The list leans on the Major Vegetation Groups of the Australian NVIS and on the New Zealand Land Cover Database, and is written towards use in the field: the question is where a bird sits, not which plant species grows there.

The scientific name

At the foot of every species page stands the full zoological name line: the species name, followed by who first described the bird and in what year. If that author and year stand in brackets, the species has since been placed in a genus other than the one it was described in — the brackets are not a detail but a requirement of the naming code. In this region they stand often: many Australian birds were described in the nineteenth century and have moved since.

In italics are the genus, the species and the subspecies; the author, the year and the words subsp. and var. are not. The genus is given in full at first mention and only abbreviated thereafter. The guide's checking program tests those rules at every build.

The guide follows the arrangement and the naming of AviList. Author, year, the original name and the type locality come from its extended edition. The Dutch names follow the list of the Netherlands Ornithologists' Union where one exists, and otherwise AviList; the Spanish follow SEO/BirdLife where one exists, and otherwise AviList. For part of the species in this region no settled name exists in either language.

• AviList Core Team (2025), AviList: The Global Avian Checklist, version 2025b, extended edition.
• Jobling, J.A., The Helm Dictionary of Scientific Bird Names, cited through the species texts of Wikipedia.

Arrangement and sequence

The arrangement and sequence of orders, families, genera and species follow AviList. Above the level of the order stand clades: these follow the family-genome tree of the B10K project and carry names from the PhyloCode. Where the genetic support for a branch is weak, that branch is marked as such in the tree.

• AviList v2025b
• Stiller et al. 2024, Nature 629, 851–860 (B10K)

Texts

The texts on this site were written for this guide, equally in all three languages. Factual data on measurements, range and status can be traced back to the sources above.

Typefaces

The running text is set in IBM Plex Sans, the headings in Cormorant Garamond and the small labels in IBM Plex Mono — the letters of vorstman.com. Only the title of each page is set in Antonio. All four are freely available under the SIL Open Font License.