Thursday, August 5, 2010: 4:20 PM
410, David L Lawrence Convention Center
Haldre S. Rogers, Biosciences, Rice University, Houston, TX, Janneke Hille Ris Lambers, Department of Biology, University of Washington, Seattle, WA and Joshua J. Tewksbury, Colorado Global Hub, Future Earth, Boulder, CO
Background/Question/Methods
Seed dispersal by animals is often cited as an important determinant of spatial structure in forest systems, yet few studies have directly linked the role of animal dispersal in seed shadows to the spatial structure of adult trees in forests. Our research on the importance of frugivorous birds in forest dynamics takes advantage of the mass extirpation of forest birds from the island of Guam caused by the introduction of the Brown Treesnake in addition to the presence of three nearby islands with comparable forests but intact bird communities. We measured seed rain for three native tree species on all four islands, and then used an inverse modeling approach to create dispersal kernels for islands with birds and Guam (no birds). Next, we took a representative spatial pattern of each species from islands with birds and apply the dispersal kernel from Guam (no birds) created from the seed rain data to parameterize spatial simulation models. This simulation allowed us to predict the spatial pattern for these three species in a scenario without bird dispersal. Finally, we compared this predicted spatial distribution to the actual current spatial patterns of these three tree species on Guam.
Results/Conclusions
Results for the seed rain traps indicate a significant difference in dispersal kernels between Guam and nearby islands with birds; as predicted, seeds fall closer to the parent tree on Guam than on islands with birds. The strength of the response varies by species, however, with one species demonstrating a negative relationship between seed rain and distance from parent tree on islands with birds, and a positive relationship on Guam. The impact of simulated bird loss on spatial structure is strongly dependent upon the strength of density dependence (a parameter set by us), with low density dependence resulting in high clumping, and high density dependence resulting in lack of recruitment. The actual spatial structure of these species on Guam is varied, but exhibits similar patterns to the simulated data. Collectively, these results suggest that birds can have large, but varied, impacts on recruitment of fleshy-fruited tree species.