Who eats what? Dispersal and distribution patterns of a vulnerable tropical tree species

Background/Question/Methods

Examining dispersal and distribution patterns of trees can help guide conservation strategies by increasing our understanding of why tropical forests have high diversity. The Janzen – Connell (J-C) hypothesis predicts that recruitment patterns of young life stages of tropical trees will be skewed away from reproductive trees and areas of high seed fall. To test the J-C hypothesis, we compared the spatial distribution of saplings with patterns of conspecific seed fall of Manilkara maxima, a threatened species in the Atlantic forest of southern Bahia, Brazil, a biodiversity hotspot. Five reproductively mature trees, > 25 m from the closest mature tree, from three study sites were selected to estimate seed fall. Fruit condition and seed fall were quantified over six months using 10 seed traps per tree. The GPS coordinates of reproductive trees, saplings and seed traps were recorded and mapped, and seed density zones were estimated by interpolation across the surface of the study sites using ArcMap v.10.2. We tested if the observed sapling distribution differed from the expected distribution of saplings using a chi-square test of distributions, assuming sapling density is proportional to seed density and all seeds contribute to the sapling life stage.  

Results/Conclusions

Vertebrate dispersers handled over 65% of the fruits (94% mammals, 6% birds). We found 49.6% of the fruits exhibited seed predation. The sapling distribution pattern of M. maxima was consistent with the spatially explicit predictions of the Janzen – Connell hypothesis. The areas with the lowest expected sapling densities, based on seed density distributions, had the highest observed sapling densities. The low sapling recruitment in high seed density zones suggests density dependent processes play a role in the recruitment pattern of this tropical tree. Potential mechanisms for this distribution pattern could be intraspecific or interspecific competition, as well as pathogen or predator-related mortality. Additionally, these data suggest that animal-mediated long distance seed dispersal is important for recruitment of M. maxima. Further, these data suggest conservation of the mammalian seed disperser species (e.g., golden-headed lion tamarin, Leontopithecus chrysomelas; yellow-breasted capuchin monkey, Cebus xanthosternos; and Guianan squirrel, Sciurus aestuans) is important for the long-term sustainability of this threatened tree species.