Where the penguins breed
Plot every known penguin breeding site in Antarctica and the first thing the map says is that they are not in Antarctica — they are in one corner of it. Cut the continent into 36 ten-degree wedges of longitude, a grid chosen before anything was counted, and three adjacent wedges hold 510 of the 729 sites. That is not really a fact about crowding. It is a fact about which birds live where: four of the six species have at least 99% of their sites inside that window, and the two that do not are the two this site already measures. The emperor breeds in 34 of the 36 sectors.
Four of the six never leave that window
The same 729 sites, drawn six times. Each frame marks one species in orange against every other site in grey — and the frames do not look like six versions of one distribution. The chinstrap, gentoo, king and macaroni each breed in between 2 and 4 of the 36 sectors, and put 561 of their 563 sites inside the window. The Adélie and the emperor go round: 23 and 34 sectors, with 43% and 12% of their sites in the crowded one. Two ways of being an Antarctic penguin, and the map separates them without being asked to.
The same thing as a count
Six species against the 36 sectors, shaded by how many sites each holds. A picture and a tally of the same data have to agree, so both are here. The single fullest sector on the continent is 60°W to 50°W, which holds 210 sites — very nearly as many as the 33 sectors outside the window hold between them, which is 219.
How far south each one goes
Longitude is where the argument is, but latitude sorts them just as cleanly. The most southerly breeding site of any penguin is Gould at 77.7°S, an emperor colony; the most southerly site of any species that needs bare rock is Cape Barne at 77.6°S. At the other end the macaroni penguin stops at 64.1°S and has never been recorded breeding further south than that. The emperor is the only one of the six whose middle half of sites lies entirely south of the Antarctic Circle.
All six, counted
| Species | Binomial | Sites | In the window | Sectors of 36 | Northernmost | Southernmost | Shares a site |
|---|---|---|---|---|---|---|---|
| Adélie penguin | Pygoscelis adeliae | 287 | 124 | 23 | 60.6°S | 77.6°S | 28% |
| Chinstrap penguin | Pygoscelis antarcticus | 369 | 367 | 4 | 60.6°S | 67.9°S | 31% |
| Gentoo penguin | Pygoscelis papua | 142 | 142 | 3 | 60.6°S | 65.3°S | 61% |
| Emperor penguin | Aptenodytes forsteri | 68 | 8 | 34 | 64.5°S | 77.7°S | 10% |
| King penguin | Aptenodytes patagonicus | 8 | 8 | 2 | 61.1°S | 65.1°S | 100% |
| Macaroni penguin | Eudyptes chrysolophus | 44 | 44 | 3 | 60.7°S | 64.1°S | 98% |
A site can carry more than one species, so the site column sums to more than 729: 147 of the 729 sites have two or more species breeding at them and 8 have four. The window is 70°W to 40°W.
The emperor is the only one that does not share
Read the last column of that table on its own and it says something the map does not. Five of the six species turn up beside a neighbour routinely — the macaroni at 98% of its sites, the king at 100%, the gentoo at 61%. The emperor shares at 7 of 68, and every one of those seven is shared with the Adélie. The reason is not competition, it is the substrate: the other five need bare rock that clears of snow in spring, there is very little of it, and the same headlands turn up in every species list. Emperors breed on fast ice — frozen sea, attached to the shore, which re-forms somewhere slightly different every year — so they are not in the queue for rock at all.
The eight sites where four species breed together are all inside the same crowded window: Hannah Point, Jacobs Ladder, Llano Point, North Point, Point Thomas (Rakusa Point), Stinker Point, Stranger Point and The Green Glen.
By region
The 21 regions the source records, ranked, with each one's species. The first six rows hold 512 of the 729 sites; the last six hold 16 between them.
| Region | Sites | Adélie | Chinstrap | Gentoo | Emperor | King | Macaroni |
|---|---|---|---|---|---|---|---|
| South Shetland Islands | 144 | 10 | 131 | 34 | — | 1 | 12 |
| Central-west Antarctic Peninsula | 110 | 15 | 64 | 59 | — | 3 | 1 |
| South Orkney Islands | 96 | 22 | 91 | 8 | — | — | 9 |
| Victoria Land | 61 | 52 | 2 | — | 10 | — | — |
| Elephant Island | 54 | 3 | 53 | 9 | — | 4 | 22 |
| Northeast Antarctic Peninsula | 47 | 40 | 2 | 27 | 2 | — | — |
| Southwest Antarctic Peninsula | 31 | 31 | 7 | — | 1 | — | — |
| Wilkes Land | 31 | 25 | — | — | 6 | — | — |
| Mac. Robertson Land | 30 | 27 | — | — | 3 | — | — |
| Queen Maud Land | 26 | 16 | — | — | 10 | — | — |
| Northwest Antarctic Peninsula | 22 | 2 | 19 | 5 | — | — | — |
| Marie Byrd Land | 20 | 13 | — | — | 8 | — | — |
| Princess Elizabeth Land | 19 | 14 | — | — | 5 | — | — |
| Enderby Land | 12 | 10 | — | — | 2 | — | — |
| Palmer Land | 10 | 3 | — | — | 7 | — | — |
| Queen Mary Land | 4 | 2 | — | — | 3 | — | — |
| Ellsworth Land | 4 | 1 | — | — | 3 | — | — |
| Coats Land | 4 | — | — | — | 4 | — | — |
| Kemp Land | 2 | — | — | — | 2 | — | — |
| Queen Elizabeth Land | 1 | — | — | — | 1 | — | — |
| Adelie Land | 1 | 1 | — | — | 1 | — | — |
What this map is not
It is not a map of how many penguins there are. There are no abundance figures anywhere on this site and there will not be until they can be defended: our own guano detector correlates at r = −0.16 against ground counts across eight colonies, which is no signal at all. A dot here is a site where breeding has been recorded, and a big colony and a small one are the same dot.
It is not a map of equal effort either, and the difference matters for one of the two claims on this page. That four species are confined to the Peninsula and the Scotia arc is a statement about their range and is not at risk from survey effort — nobody has to visit East Antarctica to know a gentoo is not breeding there, because the two species that do breed there have been censused from orbit right around the continent. But the density of sites inside the window is another matter: the Peninsula is the most-visited coast in Antarctica, and 144 sites in the South Shetlands against 31 in Wilkes Land is partly a map of where the ships go. Read the window as a range boundary, not as a headcount.
And it is a record of presence, not of absence. Three of the 729 sites carry no species in the source at all and are drawn but not counted in any species figure above. A site that has been abandoned is still in the record until somebody goes back and says so — which for most of this coastline has not happened, and is exactly why the ice around a colony is worth measuring even when the colony itself cannot be counted.
The two that go all the way round
The emperor and the Adélie are the two species this map shows spread around the whole continent, and they are the two whose habitat this site can actually measure. Every one of the 68 emperor colonies above has a page carrying 46 years of the sea ice around it, when that ice retreats and how long the season lasts.
Method
Positions and species composition for every known breeding site south of 60°S, from the Antarctic Penguin Biogeography Project via the MAPPPD archive at the Antarctic Biodiversity Portal. Sectors are fixed ten-degree bands of longitude starting at 180°W, chosen before anything was counted; the three quoted are simply the three fullest and they happen to be adjacent, which the generator asserts rather than assumes. Latitudes and longitudes are the source's own, projected to EPSG:3031 for drawing — Web Mercator cannot render this continent at all. As an independent check on the window, the two CCAMLR statistical subareas covering the Peninsula and the South Orkneys hold 504 of the 729 sites, a boundary drawn by a treaty body rather than by us and agreeing with the sector count to within 6. Every figure on this page is counted at build time from the site list; none of it is typed in.