Breeding systems of naturalized versus indigenous species provide support for Baker's law on Pohnpei island

The factors that facilitate successful colonization of islands should be especially evident where the establishment filter is strongest. Colonizers of small, remote oceanic islands should be initially rare, extremely mate-limited and often without pollinators. Hence, plant communities on such islands should reflect an establishment history in which young 'naturalized' species are most likely to display self-compatibility and autonomous selfing, whereas 'indigenous' species may exhibit more diverse reproductive strategies. To test this prediction, we characterized breeding systems of 28 species on Pohnpei, in the Federated States of Micronesia, a group of remote Pacific islands that are considered a global biodiversity hotspot. Three families with both naturalized and indigenous species were selected-Fabaceae, Malvaceae and Melastomataceae. Measurements included field observations of dichogamy/herkogamy and floral attraction traits, pollen:ovule (P:O) ratios and experimental hand-pollinations for self-compatibility and pollen limitation. Phylogenetic generalized least squares analyses tested for trait correlations between naturalized and indigenous species. Flowers of all 28 species were bisexual, and pollinator attraction features were common. Pollen:ovule ratios ranged from 9 to 557 (median = 87), and all 11 hand-pollinated species were self-compatible. All species had >5 ovules and <3500 pollen grains per flower. Indigenous species did not differ significantly from naturalized species for any trait. There is a dearth of data from remote islands bearing on the question of establishment history. In this study, we inferred all species to have some degree of autogamy and indigenous species were no more likely than naturalized species to display outcrossing mechanisms. On Pohnpei, high ovule numbers, and the inaccessibility of wind pollination and obligate outcrossing strategies, reflect the importance of retaining reproductive assurance mechanisms in the face of pollinator uncertainty.