In spite of recent progress, the composition, structure and dynamics of species-rich tropical forests, and the demography of tropical trees, remain largely unknown. Few long-term studies of such tropical forests exist that have population samples adequate for life history analyses of individual tree species, or for community-level analyses. This research will monitor the fate (growth, survival, and recuitement) of over a quarter of a million individuallly tagged, mapped, and identified trees, saplings, and shrubs of 306 species in a 50 ha permanent forest plot on Barro Colorado Island, Panama. Two censuses of the plot have been completed, the first between 1980 and 1982, and the second in 1985. This 1990 recensus is particularly important for a number of reasons. During the first census interval, the worst drought in more than 60 years occurred in central Panama, causing heavy mortality in many tree species in the plot as well as shifts in relative species abundance. No comparable drought has occurred during this census interval. Thus it will be possible to assess the community-level effects of a major episodic climatic event on longer-term patterns of tree mortality, growth, and recuitment. The recensus will also reveal whether the direction of population change for each species has remained consistent over the past decade. It will also enable an improved analyses of tree growth and survival measures current performance of each tree as a function of not only of its current local biotic and physical environment but also as a function of its past performance and recent environmental history. For the first time, the extent to which individual tree growth rates are temporally autocorrelated can be tested. According to one hypothesis, only the fastest growing individuals survive to reach the canopy. If this is true, then canopy adults may be much younger than ages estimated from averaged and randomized size- specific growth rates in the population--the usual projection metnods. Finally, but not least, the recensus results are needed because the present plot data, which are used by a sizeable number of scientists, are no longer accurate. Based on preliminary stem counts, it is estimated that the plot has already experienced about 10-12% total turnover in plants with diameters greater than one centimeter.
