This area is currently under construction . . .
Determinants of relative species abundance in tropical tree communities
Determining the causes of commonness and rarity in ecological communities is essential for understanding how communities are structured and has important implications for biodiversity conservation. I am using data from the Center for Tropical Forest Science's network of large forest dynamics plots located throughout the tropics to examine the determinants of species abundance in highly diverse tree communities, and to test whether processes shaping relative abundance patterns differ among forests. Click here to read about results published in Science!
Effect of water availability on tropical tree species dynamics and distributions
Species coexistence in diverse tropical forests
For my dissertation research, I studied spatial patterns of seedling distributions and dynamics on Barro Colorado Island (BCI), Panama in order to test hypotheses concerning species coexistence in tropical forests (Comita 2006, Comita et al. 2007a, Comita et al. 2007b, Comita & Hubbell 2009). I found that seedling survival was negatively affected by the local density of conspecific neighbors, consistent with the Janzen-Connell hypothesis. However, local scale negative density dependence did not result in the expected negative relationship between seedling survival and species abundance at the community level (i.e. a community compensatory trend), due in part to variation in shade-tolerance and in the strength of density dependence among species (Comita & Hubbell 2009). This study was part of an ongoing, NSF-funded project with Dr. Steve Hubbell in which we are monitoring seedling dynamics on an annual basis in 20,000 1x1 m plots located throughout the BCI 50-ha Forest Dynamics Plot.
Effects of large-scale natural and human disturbance on tropical forest regeneration
Previously, I held an Earth Institute postdoctoral fellowship at Columbia University. Working with Dr. Maria Uriarte in the Department of Ecology, Evolution and Environmental Biology (E3B), I examined how forests regenerate following large natural disturbances, namely hurricanes, and how regeneration pathways and post‐ disturbance community composition change as a result of human alterations to landscapes. For this project, I utilized a 15‐year record of tree and seedling dynamics in the 16‐ha Luquillo Forest Dynamics Plot in the Luquillo mountains of northeastern Puerto Rico. The plot encompasses areas of differing intensities of past human disturbance and has been hit by two major hurricanes since 1989. Thus, the site provides an excellent opportunity to study forest regeneration and community stability following large-scale natural and human disturbance. I found shifts in the drivers of seedling survival over time following hurricane disturbance (Comita et al. 2009), and documented interactive effects of hurricane disturbance and land use history on seedling dynamics and forest regeneration at this site (Comita et al. 2010).
Cross-site comparisons of seedling dynamics
Since 2004, I have been collaborating with Drs. Margaret Metz, Yu-Yun Chen, and Natalia Norden on a comparative study of seedling ecology in tropical forest plots in Central and South America and Southeast Asia. Our goal is to test whether the factors driving seedling dynamics are similar in tropical forests worldwide. Thus far, we have analyzed rates of seedling recruitment, survival and growth in forest plots at Yasuni Scientific Station (Ecuador), Barro Colorado Island (Panama), Pasoh Forest Reserve (Malaysia) and Nouragues Biological Station (French Guiana) (Metz et al. 2008). We organized a symposium on cross-site studies of seed and seedling ecology at the 2008 Association for Tropical Biology and Conservation annual meeting in Paramaribo, Suriname.
Impacts of long-term research on vegetation dynamics and community composition
In collaboration with REU student Greg Goldsmith, I tested whether the intense research activity in the BCI plot over the past 25 years has affected the dynamics and composition of the vegetation being studied. We found no significant differences between the density, size-class distribution and composition of the seedling layer inside versus outside the plot, suggesting that long-term effects of research activity in the BCI plot are negligible (Goldsmith et al. 2006). However, we did detect short-term impacts on seedling dynamics immediately following an intense research event for some growth forms, as well as significant effects of research trails (Comita et al. 2008, Comita & Goldsmith 2009).