JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.. Ecological Society of America is collaborating with JSTOR to digitize, preserve and extend access to Ecology.Abstract. The role of density dependence in the population dynamics of tropical trees has been a subject of considerable debate. Here, we present data on the demography of the edible palm Euterpe edulis, classified into seven size categories and monitored over three years. On average, each adult palm contributed 98 seedling recruits per year into the population. The pattern of mortality was similar to that of other palms, with mortality being highest among the smallest plants. Those plants with a diameter at soil level >20 mm had an annual mortality <7%. Density dependence was found to act only on the seedling stage of the life cycle. The probability of survival and transition of seedlings to the next size class were affected both by the density of seedlings and the presence of conspecific adults. Matrix modeling indicated that the true finite rate of population increase (X) was 1.28 and that the observed reverse "J"-shaped size distribution of plants was a consequence of the density dependence operating in the population. Elasticity analysis showed that the survival elements in the matrix contributed most to the value of X, and that the position of the transition matrix in growth-survival-fecundity (G-L-F) space was influenced by density. The matrix model incorporating density dependence predicted size distributions and densities approximating the maximum observed in the field. Spatial simulations indicated that the predictions from the matrix model relating to the size structure of plants are robust, but that the predictions of densities are sensitive to the precise spatial dynamics of the population.1979, Hubbell and Foster 1986), known as the community drift model. In contrast to this nonequilibrium model, it has been argued at the other extreme that density-dependent and distance-related recruitment may be responsible for the low mean densities of many tropical trees and the consequent high species diversity (Janzen 1970, Connell 1971. In this latter case, specialist herbivores, seed predators, and pathogens are envisaged as maintaining populations around low densities and high species diversity. Critics of the Janzen-Connell model have suggested that the strength of density dependence is insufficient to remove the clumped distributions of seedlings beneath parents (Hubbell 1980). The community drift model of tropical forest dynamics, on the other hand, has also been criticized on the basis that it fails to take into account high levels of compositional similarity in disjunct samples of forest (Terborgh et al. 1996). Moreover, Wills et al. (1997) have found that int...