Seed disperser loss affects community assembly and forest structure

Background/Question/Methods: Small-scale disturbances in forests provide niches for pioneer species and set the stage for changes in forest composition. The rate of disturbance also determines forest structure, or the proportion of forest in a disturbed state. While dispersal has theoretically been linked to regeneration after disturbance, we lack a good understanding of how dispersal affects community assembly and recovery after disturbance. In tropical forests, 75-90% of trees are vertebrate dispersed including a large proportion of pioneer species, thus, vertebrate frugivores likely affect gap regeneration by facilitating colonization of fleshy-fruited species and many pioneers. We propose that frugivore loss will lead to reduced colonization by fleshy-fruited pioneer species, reducing diversity. In addition, with fewer pioneer species, we hypothesize that the pace of recovery will be slowed, ultimately affecting forest structure. The Mariana Islands offer a unique opportunity to examine the effect of frugivores on gap dynamics. Most tree species, including all pioneers, rely on frugivorous vertebrates to disperse their seeds, but an invasive snake was introduced to one island (Guam) and caused the functional extirpation of all native vertebrate frugivores. In contrast, the nearby islands of Saipan and Rota support similar forest to Guam yet still have frugivores. We created experimental treefall gaps across three islands, and used seed additions or seedling removals in a portion of the gaps to either simulated dispersal or lack of dispersal. We monitored the seedling community and gap closure for 2 1/2 years.

Results/Conclusions We found a higher diversity of seedlings regenerating in gaps on islands with dispersers than on Guam, likely due to a larger available species pool with the presence of dispersers. Gap regeneration with dispersers was faster than on Guam (without dispersers) primarily because gaps on Saipan contained more pioneer species seedlings. Preliminary results suggest gaps are closing more quickly in the presence of dispersers (real or simulated), suggesting that a forest without frugivores will be have a more open structure. This is consistent with the observation that there are two to four times more natural gaps on Guam than on islands with dispersers. We provide the first evidence that vertebrate seed dispersal can influence the physical structure of forests, with gap regeneration slowing in its absence. These results suggest that the ongoing loss of biotic dispersers could result in forests that have lower diversity and more open forest structure.