COS 94-6
Changes in abundance and seedling survival of Terminalia carolinensis, an endemic tropical freshwater wetland tree on Kosrae, Micronesia

Wednesday, August 12, 2015: 3:20 PM
349, Baltimore Convention Center
Silke Buschmann, School of Forestry, Northern Arizona University, Flagstaff, AZ
James A. Allen, School of Forestry, Northern Arizona University, Flagstaff, AZ
Ken W. Krauss, National Wetlands Research Center, U.S. Geological Survey, Lafayette, LA
Erick E. Waguk, Kosrae Island Resource Management Authority, Kosrae, Federated States of Micronesia, Micronesia
Freshwater wetlands are a major component of the rainforests of the Caroline Islands, Micronesia. Among the tallest canopy trees on the island of Kosrae is Terminalia carolinensis, an endemic species that dominates the valleys between coastal mangrove forests and steep mountain slopes. However, regeneration of this impressive tree is low, and competition for light is high, preventing T. carolinensis seedlings from growing tall.
We asked whether regeneration of T. carolinensis is sufficient to sustain its populations, what role light level and soil type play to sustain regeneration of seedlings, and if different genotypes can be matched to distinct populations and habitat characteristics.
We hypothesized (1) higher abundance of mature T. carolinensis trees on wet soils resulting from its higher tolerance of root water saturation, (2) a positive relationship between seedling size and gap occurrence as Terminalia species generally favor high light environments.
We collected data on forest stand conditions in five watersheds across Kosrae. Our sites varied in soil type and levels of disturbance (e.g. human impact through agroforestry practices). We identified each tree to species level and recorded DBH for 1268 trees. We counted 1686
T. carolinensis seedlings and recorded size and habitat characteristics for fifty-three percent of these.

We found (1) the healthiest juvenile and mature T. carolinensis trees to grow well on both wet low-light sites and drier high-light sites. Different soil types were not the main driver of tree growth; instead, the level of disturbance between sites was most important. We found that (2) T. carolinensis seeds germinated well in high-light sites of medium disturbance; however, seedling survival was very low. We suspect its low competitive ability with other species to be the reason for this finding. Seedlings generally grew better near dead wood and agroforestry species, possibly due to increased nutrient input from decomposition and to nurse plant interactions. Overall, abundance of T. carolinensis trees decreased on all but one site over the last decade. We also recorded shifts in species abundance that illustrate ongoing forest succession.
The overall decrease of T. carolinensis trees ties directly into ongoing debates on the loss of biodiversity in Oceania, a biodiversity hotspot; and stresses the importance of understanding patterns of forest succession on oceanic islands. Our study is the first to report changes in tree abundance and species succession for a Micronesian freshwater forest and highlights the potential threats to successful regeneration of an endemic tree species.