Extinction of birds in Eastern Polynesia: A Review of the Record, and Comparisons with other Pacific Island Groups

TitleExtinction of birds in Eastern Polynesia: A Review of the Record, and Comparisons with other Pacific Island Groups
Publication TypeJournal Article
Year of Publication1989
AuthorsSteadman, DW
JournalJournal of Archaeological Science
Volume16
Pagination177-205
Keywordsanthropology biology
Abstract
The study of bones from archaeological sites in Eastern Polynesia has revealed much late Holocene extinction of birds. Because this extinction has been found in all Eastern Polynesian archipelagos where bird bones are part of the archaeological record (Marquesas, Society, Pitcairn, and Cook island groups), similar levels of extinction are likely to be found in other Eastern Polynesian island groups (Line, Tuamotu, Gambier, Austral, Easter), if the evidence is sought. Human impact is the most plausible explanation for these extinctions, which begin immediately after peopling of the islands about 2000 years ago and diminish only after the avifaunas are largely depleted.
The sites yielding bones of extinct birds include limestone caves, volcanic rockshelters, calcareous beach sands, and organic estuarine sediments. Most of the fossil avifaunas are biased toward large species because traditional techniques of archaeological excavation have recovered very few bones of small birds such as swifts, kingfishers, and passerines. A rich record of small species has been recovered from the few sites where screens of 116 in. mesh have been used.
More species of Eastern Polynesian landbirds have become extinct since human arrival than survive in the region today. Extirpation, which refers to loss of individual island populations of extant species, is another major contributor to the loss of avian diversity in Eastern Polynesia; nearly all species of birds that have survived, whether seabirds or landbirds, no longer occur on most of the islands that once made up their natural (= pre-human) range. Fossils from archaeological sites have extended the ranges of many extant species by hundreds to thousands of kilometers. Such species, traditionally regarded as endemic to one or several islands, were widespread until human influences caused their disappearance from island after island.
Petrels and shearwaters have undergone the greatest losses among Eastern Polynesian seabirds, although storm-petrels, boobies, frigatebirds, gulls, terns, herons, and ducks have also been affected. The losses of Eastern Polynesian landbirds are dominated by extinctions of rails (mostly flightless species confined to single islands), and the extinctions or extirpations of pigeons, parrots, passerines, and, to a lesser extent, swifts and kingfishers.
The fossil record of birds in Western Polynesia (Niue, Tonga, Samoa, Wallis and Futuna, Tokelau, Tuvalu) is very limited. From Tonga, an extirpated petrel and shearwater, two extinct megapodes, and three extinct pigeons are known from Lifuka, and an extirpated shearwater, tern, rail, ground-dove, and two or three passerines have been found on 'Eua. An extinct pigeon has been described from Wallis. From Lakeba, in the Lau Group of Fiji (culturally Melanesian, faunally Western Polynesian), an extinct megapode and pigeon are known. There is no fossil record from Vanuatu or the Solomon Islands except for the very small Polynesian outlier islands of Tikopia and Anuta, where archaeological remains include extirpated shearwaters, boobies, terns, megapode, and rail. An extensive late Holocene fossil assemblage from New Caledonia includes extinct or extirpated snipe, rails, kagu, hawks, magapodes, pigeons, owls, and owlet-nightjar. The evidence from Tonga, Wallis, Fiji, and New Caledonia suggests that the Western Polynesian-Melanesian fossil record, when known more fully, may disclose as much ormore extinction as in Eastern Polynesia. There is no fossil record of birds from Micronesia.
The loss of avian diversity in Eastern Polynesia is comparable in severity to that recorded from late Holocene sites (archaeological and non-archaeological) in the two major outlying Polynesian island groups, Hawaii and New Zealand. As in Eastern Polynesia, most or all Holocene extinction in the highly endemic avifaunas of Hawaii and New Zealand postdates the arrival of people during the past two millennia. Humans seem to be responsible for these extinctions.
The late Holocene fossil record of the Galapagos Islands differs from those in Eastern Polynesia, Hawaii, or New Zealand in the relative abundance of reptiles (tortoises, lizards, snakes) and mammals (bats, rodents). Most fossil deposits in the Galapagos consist of bony prey-remains of the Galapagos barn owl (Tyto punctatissima). (Although the ecologically similar common barn owl, T. alba, occurs through much of Western Polynesia and Melanesia, only on 'Eua have T. alba-derived fossil deposits been collected.) Other Galapagos fossil deposits are derived from natural trap activity; unlike in Polynesia, none is anthropogenic.
As in Eastern Polynesia, there are no clear-out cases of vertebrate extinction in the Galapagos that predate the arrival of people. Most avian losses in the Galapagos involve local populations rather than entire species. The extinction of birds in the Galapagos is less extensive than in Eastern Polynesia, Hawaii, or New Zealand, for the following reasons:

(1) the period of human occupation is shorter (200 versus 2000 or 3000 years);

(2) human populations have been much lower because of the less hospitable terrain, including many islands that have never been inhabited;

(3) Galapagos birds never have been a major source of food for humans;

(4) the Galapagos avifauna has been given legal protection for the past three decades;

(5) introduced birds are absent on most islands;

(6) introduced mammals and plants are scarce or absent on many islands.

Most Galapagos extinctions have occurred only on the particular islands where the level of human impact (habitat alterations, introduced mammals) has approached that of many Polynesian islands.
URLhttp://www.sciencedirect.com/science/article/pii/0305440389900654
DOI10.1016/0305-4403(89)90065-4