Closing the seed dispersal loop for Guettarda viburnoides (Rub.): Connecting patterns of avian seed dispersal with population growth in a Neotropical savanna.

摘要

Seed dispersal by vertebrates in the tropics is a key ecological process therefore, understanding its consequences for plant fitness is a central question of plant population ecology. Little is known, however, of how frugivore activity translates into demographic and evolutionary consequences for plants. In this study, I connect data on frugivore activity, post-dispersal seed fate and plant population demography using stage-specific demographic modeling, to examine the role of individual dispersers for plant population dynamics of the fleshy-fruited Neotropical tree Guettarda viburnoides (Rubiaceae) in northeastern Bolivian savannas.In chapter one, I examine avian frugivory and seed dispersal of G. viburnoides, focusing on the quantitative and qualitative aspects of disperser effectiveness. The endocarps of this plant are dispersed primarily by two species: Cyanocorax cyanomelas and Pteroglossus castanotis, which I identify as the quantitatively important dispersers (QID) of G. viburnoides. These two species differ in several qualitative aspects of seed dispersal: (1) They select fruits of different sizes (2) they differ in their fruit handling treatment, which in turn affects the probability of seedling emergence, the temporal pattern of emergence, and the number of emerged seedlings per endocarp and (3) they differ in their landscape patterns of seed deposition. These results suggest that C. cyanomelas and P. castanotis differ in the quality of seed dispersal services they provide to G. viburnoides.In chapter two, I analyze how habitat affects the post-dispersal seed fate of G. viburnoides. The results show that habitat strongly affects seed dispersal, seed predation, and seedling emergence. Additionally, uncoupling among different life-stages occurs in all habitats and there is spatial discordance between the seed rain and sapling recruitment patterns. This discordance suggests that, in certain years at least, habitat available for recruitment of G. viburnoides in this landscape is limited. The results also reveal high inter-annual variability in the strength of post-dispersal processes, which leads to a constant shifting of habitat "suitability" from one year to the next that is, it is context-dependent.Finally, in chapter three, I determine the individual effect of seed dispersal by C. cyanomelas and P. castanotis on population growth of G. viburnoides, and assess the degree to which these species may be ecologically redundant. The results demonstrate that seed dispersal by C. cyanomelas leads to positive population growth of G. viburnoides, whereas seed dispersal by P. castanotis has a detrimental effect on the population growth of this species. Therefore, C. cyanomelas is a key species for the long-term persistence of G. viburnoides, and the loss of this dispersal agent would not be compensated by the dispersal services provided by P. castanotis. Ultimately, the integration of frugivore activity with plant demography using models such as this one are important for plant ecology because they enable us to close the "seed dispersal loop" and gain a better understanding of the demographic consequences of seed dispersal by different dispersal agents.