Invasibility of tropical islands by introduced plants: partitioning the influence of isolation and propagule pressure.

All else being equal, more isolated islands should be more susceptible to invasion because their native species are derived from a smaller pool of colonists, and isolated islands may be missing key functional groups. Although some analyses seem to support this hypothesis, previous studies have not taken into account differences in the number of plant introductions made to different islands, which will affect invasibility estimates. Furthermore, previous studies have not assessed invasibility in terms of the rates at which introduced plant species attain different degrees invasion or naturalization. I compared the naturalization status of introduced plants on two pairs of Pacific island groups that are similar in most respects but that differ in their distances from a mainland. Then, to factor out differences in propagule pressure due to differing numbers of introductions, I compared the naturalization status only among shared introductions. In the first comparison, Hawai‘i (3700 km from a mainland) had three times more casual/weakly naturalized, naturalized and pest species than Taiwan (160 km from a mainland); however, roughly half (54%) of this difference can be attributed to a larger number of plant introductions to Hawai‘i. In the second comparison, Fiji (2500 km from a mainland) did not differ in susceptibility to invasion in comparison to New Caledonia (1000 km from a mainland); the latter two island groups appear to have experienced roughly similar propagule pressure, and they have similar invasibility. The rate at which naturalized species have become pests is similar for Hawai‘i and other island groups. The higher susceptibility of Hawai‘i to invasion is related to more species entering the earliest stages in the invasion process (more casual and weakly naturalized species), and these higher numbers are then maintained in the naturalized and pest pools. The number of indigenous (not endemic) species was significantly correlated with susceptibility to invasion across all four island groups. When islands share similar climates and habitat diversity, the number of indigenous species may be a better predictor of invasibility than indices of physical isolation because it is a composite measure of biological isolation.