Reducing seed predation by introduced rodents helps, but is insufficient, to prevent long-term decline of common forest trees

Understanding how novel biotic interactions can affect the persistence of plant and animal populations is of critical conservation concern. Rodents have been introduced to most oceanic islands and predate a wide range of seeds. However, little research has evaluated their effects on the long-term viability of plant populations. We assess the effects of seed predation by introduced rodents on the demography of three long-lived, common native trees in a Hawaiian forest. Using a 17-year demographic dataset, we assess if populations are viable under standard forest conservation practices; then we explore the potential effects of reducing seed predation by non-native organisms, reducing non-native understory cover, or reducing both, on population growth. Post-dispersal seed removal ranged from 40 to 98 % and increased as a function of understory cover. Survival was low for seedlings due to competition from non-native understory plants. Sixty-two percent of Pysdrax odorata seeds suffered pre-dispersal predation by an introduced moth and 72 % of Nestegis sandwicensis seeds were prematurely aborted. Integral projection models show that λ values under standard conservation were <1, indicating population decline. Reducing seed predation by introduced rodents had a strong positive effect on λ for Diospyros sandwicensis, but the effect was much smaller for the other two species. Seed predation by introduced rodents can drive slow population decline of common native trees even in areas actively managed for conservation. Further, reducing rodent predation alone is likely insufficient to ensure long-term persistence due to additional bottlenecks in viable seed production. More attention to common tree species on oceanic islands is critical.