Effects of Predation and Light on Seedling Establishment in Gustavia Superba

Abstract

Large—seeded trees whose seeds are dispersed by mammals often have high levels of seed dispersal but high seed losses due to predation. Consequently, seedling establishment is influenced both by the impact of mammalian seed predators as well as by the ability of the seedling to survive where the seed is dispersed. I examined the relative roles of seed/seedling predation by mammals and of the light environment in seedling establishment in three populations of the tropical tree Gustavia superba. One population was located in semideciduous forest on Gigante Peninsula and two were on Barro Colorado Island (BCI), Panama. On Gigante Peninsula, mammals that feed on Gustavia fruit and seeds were much less abundant than on BCI. My objectives were: (1) to quantify seedling densities in each population, (2) to measure the intensity of seed predation, and (3) to determine the relative importance of seed/seedling predation and light in seedling survival and growth. Seedling density was significantly greater on the Gigante Peninsula than on BCI. This pattern was influenced by postdispersal seed predation that was much lower on the Gigante Peninsula than on BCI. Seedling transplant experiments, which measured the effects of site, light condition (gap, edge, understory), and protection from mammals (tall cage, short cage, no cage), showed that protection from mammals significantly increased all three measures of seedling performance: seedling survival, seedling height, and overall seedling success (survival times growth). In addition, as expected from differences in mammal abundance among sites, the three sites differed in the extent to which each measure of seedling performance was affected by seedling protection. Seedling survival, growth, and overall success were significantly greater on the Gigante Peninsula than on BCI. However, even on the Gigante Peninsula, seedlings benefitted significantly by protection from mammals. All three measures of seedling performance were greater in light gaps than in the understory, but seedlings survived well in the understory at all three sites. I conclude that seedling density was determined primarily by mammalian predation on both seeds and seedlings, although seedling establishment was influenced also by site and light conditions. Large seeds size, which probably increases vulnerability to mammalian see predators, facilitates the establishment of seedlings under low light conditions.