or not at all. Certain West African species, in particular, are river-birds (the pelicans Pelecanus onocrotalus and P. rufescens, the reed-cormorant Haliëtor africanus, the darter Anhinga rufa, the skimmer Rynchops flavirostris). The terns of the genus Chlidonias are primarily lake and marsh species throughout their range. In North America the gulls Larus pipixcan and L. philadelphia are purely inland species in the breeding season, and the tern Sterna forsteri and the pelican Pelecanus erythrorhynchos almost so. In South America the terns Phaëtusa simplex and Sterna superciliaris are purely river-species.
FIG. 2b Actual species. Arrows point to replacement species or to nearest ecological counterparts
One hundred and eleven species of primary and thirty-two of secondary sea-birds have been identified by competent observers at sea or on some shore in the North Atlantic since 1800: a total of one hundred and forty-one. Of these one primary sea-bird, Alca impennis the great auk, and one secondary sea-bird, Camptorhynchus labradorius the Labrador duck, are now extinct. Of the survivors, eighty-two primary and thirty secondary sea-birds actually nest, or have nested, on or near a North Atlantic or Mediterranean shore or a shore of that part of the Arctic (north of the Circle) that communicates directly with the North Atlantic (this brings in six arctic species: ivory-gull, Ross’s gull, little auk, white-billed northern diver, brent-goose and Steller’s eider). Two further species (Larus pipixcan and L. philadelphia, see table) are purely inland breeders.
Most remarkably, the number breeding on the Old World and New World sides is almost exactly the same. We can derive the following summary of breeding-species from the Appendix; the totals include the two North American purely inland species, and the two extinct species. Doubtful (“?” in the Appendix) and casual cases are deliberately included—most of them are from tropical West Africa north of the equator where the breeding of the species in question seems likely but, owing to the scanty exploration of the coast, is not formally proved.
We can see that if we add the six purely arctic breeders to those species which are common to both east and west sides of the North Atlantic, we have fifty-five, out of a total of 116, or about half. Of the remaining sixty-one species, 24 breed only on the west side of the North Atlantic, four on the west side and in the Arctic; and six purely on the east side, and seven on the east side and in the Arctic. Those on the east side include four ‘sea-birds’ which breed in the Mediterranean area but not in the North Atlantic (the crested pelican, pigmy cormorant, Mediterranean black-headed gull and the lesser crested tern).
The general conclusion is of considerable ecological interest, showing how exactly the sea-bird communities of both sides reflect one another. Although only about two-thirds of the members of the community on one side of the Atlantic are found in that of the other, the species comprising the remaining third ‘balance each other’ and occupy very much the same ‘niches’ or places in nature. Opposite species which pair off by occupying similar niches are grouped together in the list in the Appendix, see here.
A NOTE ON NON-BREEDERS AND CASUAL WANDERERS
Apart from these 116 breeders, the limbo of twenty-six primary sea-birds and one secondary sea-bird (the spectacled eider Somateriafischeri, which has been recorded twice in Norway, though it breeds on the other side of the Polar Basin) consists of casual wanderers, with three remarkable exceptions. These are all tubenoses (two shearwaters and a storm-petrel) which breed in the southern hemisphere but which cross the equator in large numbers to ‘winter.’ The most familiar of these in Britain is the Tristan great shearwater Puffinus gravis, which is rather similar, and certainly closely related to the heavier North Atlantic or Cory’s shearwater, P. diomedea. Incidentally we suggest confusion between the two would be reduced if P. diomedea were consistently called ‘North Atlantic shearwater’ and P. gravis ‘Tristan great shearwater’—not just ‘great shearwater.’
The Tristan great shearwater nests only on Nightingale and Inaccessible Islands, in the Tristan da Cunha group; possibly a few may survive on Tristan itself. The population remains vast, though ‘farmed’ by the Tristan islanders, and an annual penetration of the North Atlantic by off-season birds has put the species on the list of regular and expected visitors to both West Atlantic and East Atlantic waters, as well as some arctic waters of Greenland. The northward movement reaches the North Atlantic in May, mostly on the west side at first, but odd birds appear in Irish and west British waters in June and have even been seen then in the Skagerak; one of us saw a few already at Rockall in mid-May (1949), and they were abundant there and in moult by late June (1948).
The Tristan great shearwater seldom comes close to land, and it is never common in British waters within sight of shore; but some distance to sea off west England, Ireland and the Hebrides it is always present in July and August; and some elements usually penetrate northabout into the North Sea, descending to the latitude of Yorkshire. The Tristan great shearwater is much more common than the Northern Atlantic shearwater in our seas; indeed, the Mediterranean race of the latter P. d. diomedea, and Cory’s race P. d. borealis, have each only once been taken ashore in Britain, although birds which may have been of Cory’s subspecies have several times been seen at the entrance of the Channel. The only Scottish record is of one, seen at sea close to Aberdeen on 10 September 1947, by R. N. Winnall. Normally as Wynne-Edwards and Rankin and Duffey have shown, Puffinus diomedea does not get much farther north in the Atlantic than 50°N., and that at about 30°W. It is much more common on the North American coast than on that of Britain, although this coast is much farther from its base; ‘they seem to arrive on our coasts early in August,’ writes Bent, ‘and spend the next three months with us, mainly between Cape Cod and Long Island Sound.’ The Tristan great shearwater also probably reaches its greatest abundance on the North American coast, particularly in the area of the Newfoundland Banks, where it is known as the ‘hagdon’; from here it extends every season along the coast of Labrador to Greenland;—it has been recorded near Iceland.
The other southern hemisphere shearwater that regularly visits North Atlantic waters is Puffinus griseus, the sooty shearwater. It is much rarer than the Tristan great shearwater, though it has been seen in British waters regularly enough to be classed as an autumn visitor. It breeds in New Zealand and its islands, in southern South America and its islands, and the Falkland Islands (in places many miles inland), and ranges the Pacific as well as the Atlantic; its Atlantic population is low compared with that of the other southern shearwater. Unlike the Tristan great shearwater, it probably makes its way into the North Sea by the Channel; and it is regular in small numbers in the Western approaches. At Rockall on 17 May 1949 J.F. saw none, but from 18 to 27 June 1948 R.M.L. found them always present there, singly and up to eight together, that is in the proportion of about one to a hundred hagdons. On the Newfoundland Banks, where it is in the same proportion, the fishermen called it the haglet. It reaches Greenland and Icelandic waters, and has been seen once as far north as Bear Island.
The storm-petrel from the south is Wilson’s petrel Oceanites oceanicus, which nests in vast numbers in the antarctic continent and on the southern islands of South Shetland, South Orkney, South Georgia, Falkland, Tierra del Fuego and Kerguelen. It disperses into, and across, the Equator in the Atlantic, Indian and Pacific Oceans. Wilson’s petrel has been the subject of an exhaustive monograph by Brian Roberts (1940), who mapped the dispersal in the Atlantic month by month (Fig. 29). Records north of the Equator are only irregular and sporadic between November and March, but in April the species is spread widely over the western half of the North Atlantic as far as Cape Cod. In May the petrels spread eastwards reaching from Cape Cod across the Atlantic towards Portugal and the Bay of Biscay, off which there is quite