Seed Dispersal by Wind and Plant Recruitment in Tropical Forests

An Interdisciplinary Investigation across Multiple Scales

This project develops and validates mechanistic models of seed dispersal for 50 wind-dispersed tree and liana species and uses those models to explain patterns of seed deposition and to investigate the consequences of dispersal for recruitment at the individual, population, and community levels. Seed dispersal is a key process in population and community dynamics, especially in diverse tropical forests where the rarity of adults of any one species insures that the seed rain of virtually all species is sparse and patchy. Yet ecologists remain largely ignorant of dispersal patterns and how they affect real populations and communities, and in particular of the magnitude and influence of long-distance dispersal.

The research builds on recent advances in atmospheric modeling and genetic techniques, using an interdisciplinary team with a history of successful collaboration to develop the next generation of atmospheric models of seed dispersal by wind, to test these models using long-term seed rain datasets and detailed genetic data, and to apply them to ecological questions. The study site, Barro Colorado Island in Panama, offers a unique combination of a highly diverse group of wind-dispersed plant species and extensive long-term and large-scale datasets on seed rain and seedling recruitment within mapped plots. The seed rain studies, together with available molecular markers for several species, provide unmatched opportunities to test the mechanistic models. The seedling recruitment studies, together with a 50-ha mapped plot and associated information on habitat heterogeneity and species traits, make it possible to investigate the implications of seed dispersal for individuals, populations, and communities in unprecedented detail, and to analyze how these implications vary with species’ dispersal abilities, regeneration habitat specificities, fecundities, and adult abundances.

The mechanistic seed dispersal models under development here are the first to incorporate the effects of spatial heterogeneity and complex topography, and will cover an unprecedented range of spatial and temporal scales. Micrometeorological data will be used together with landscape topography and foliage density to develop models of wind flow. Species-specific data on seed terminal velocities and on heights and conditions of seed release will be used to parameterize the dispersal model. To maintain the key advantage of mechanistic models of providing insights into the underlying processes while minimizing their major disadvantage of high computational costs of such models, we are also pioneering the development, validation, and application of corresponding analytical mechanistic models.

The proposed research draws on field data, experiments, and models at the interface between ecology, population genetics and computational fluid mechanics to address complex multi-scale ecological questions including: What are the most important factors explaining variation in seed shadows within and among species? To what degree is population seedling recruitment limited by the failure of seeds to arrive, and what are the overall impacts at the community level? Are competition-colonization trade-offs present, and if so, what is their influence on community dynamics? Which scales of seed dispersal are most important to genetic structure?

The models developed here will be broadly applicable to wind-dispersed species worldwide, and will constitute an important new tool for conservation and management. These models will be useful, for example, for predicting invasion rates of exotics, responses to range shifts, and seed movement and gene flow in increasingly fragmented landscapes. Activities planned to broaden the impact of the research include a published workshop on dispersal at multiple scales, a web site that will make public all data and models, training from undergraduate to postdoctoral levels, and educational outreach to broad audiences in the US and Panama.

This research is funded by Collaborative Research grants from the National Science Foundation’s Department of Environmental Biology (NSF DEB 0453445, 0453665, and 0453296) to the University of Minnesota, the Smithsonian Tropical Research Institute, and Duke University.

We are currently searching for an undergraduate student at a US university to fill an NSF-funded research position (REU) in Panama for February - April 2006. For further information, please see the job advertisement.