9714819 Steadman Analyses of modern landbird communities on South Pacific islands have made important contributions to biogeographic theory. In East Polynesia (Marquesas, Society, Tuamotu, Pitcairn, and Cook island groups), however, natural patterns of diversity and endemism among landbirds have been obscured by human-caused extinction's. Prehistoric East Polynesian avifaunas were much more homogeneous than modern avifaunas from the same islands in species richness (both between and within island groups), regardless of island area, elevation, or isolation. Most endemism was restricted to flightless species. Data on East Polynesian birds do not precede the arrival of humans at 2000 yr BP. The new research will examine the composition and stability of South Pacific bird communities over much longer time scales (up to 100,000 yr BP) by collecting and analyzing bone assemblages from precisely dated prehistoric sites in the less remote West Polynesian island groups of Tonga, Samoa, and Fiji. Older faunal records can be developed in West Polynesia because indigenous predatory birds have concentrated bones in caves there during pre-human times. The new data will provide an opportunity to estimate species richness, taxonomic diversity, and turnover (extinction/colonization) rates in West Polynesian landbirds both before and after human presence. It is predicted that, counter to classic biogeographic theory, variation in island area (10( - 104 km2), elevation (101-103 m), and isolation (0-800 km) had little influence on species richness or taxonomic diversity in West Polynesian avifaunas in pre-human times, and that turnover rates were very low. In other words, these bird communities were rich but stable under natural conditions, with extremely low rates of colonization and extinction. From a paleontological perspective, the existence of species on true oceanic islands may not be as ephemeral as is generally believed by those who study modern birds.
