Seed and microsite limitations of large-seeded, zoochorous trees in tropical forest restoration plantations in northern Thailand

ABSTRACT: On deforested or degraded land, planting mixtures of native forest tree species facilitates establishment of incoming tree seedling species (i.e. "species recruitment") by rapidly re-establishing canopy cover. However, delayed colonization of plantations by large-seeded, zoochorous (LSZ) tree species can affect the species composition, structure and function of the regenerating forest. The study investigated whether microsite conditions limit establishment (germination and early seedling survival) of LSZ tree species in three 13-year-old, seasonally dry, tropical forest restoration plantations in northern Thailand. We conducted a seed-sowing experiment, which combined five LSZ tree species with four microsite treatments in a split-plot design. All five tree species were previously absent from the understory of the plantations, despite seed sources being present in nearby natural forest. The four treatments simulated potential microsites that naturally-dispersed seeds may encounter. They included seed deposition on the surface of soil and on leaf litter, as well as seed burial beneath soil and beneath leaf litter. We fenced the experimental areas to prevent seed predation and to focus on the environmental effects of microsites on the different stages of seedling establishment. Following seed sowing, we measured germination and seedling survival, mean height and stem diameter, for 26 months.

Microsite treatments did not significantly affect germination and seedling survival, providing strong evidence that establishment of the tested species was not limited by the applied microsite environments. Furthermore, although differences in germination and survival among species were significant, high overall germination (37%) and 26-month seedling survival (58.5%), irrespective of microsite treatment, suggested that environmental conditions in the 13-year-old plantations supported germination and early seedling establishment. Furthermore, microsite treatments did not correlate with seedling height or stem diameter at 20 months, suggesting that initial germination microsites have no effect on seedling growth and robustness.

Therefore, the hypothesis that inadequate seed dispersal or low seed availability limits seedling-recruitment of LSZ tree species into forest restoration plots was supported. Successful establishment of LSZ seedlings from sown seeds in this experiment suggested that direct seeding, beneath the canopy of restoration plantations, may be an effective way to offset seed-dispersal limitations in restoration plantations. Furthermore, the absence of microsite-treatment effects suggested that broadcast sowing of seeds may be a simple way to recolonize closed-canopy restoration plantations with LSZ tree species.