Guilds, structure and dynamics in the forest on Barro Colorado Island, Panama

The structural and dynamical characteristics of 56 species of free-standing woody plants in 50 ha of tropical forest on Barro Colorado Island, Panama, were studied to test the hypothesis that tropical forest tree species can be organized into guilds, or groups of species with similar characteristics. The overall structure of the forest was analyzed by studying basal area and diameter distributions at several scales. Regressions of tree height, height to crown base, and crown area of the species were calculated as a function of bole diameter on a logarithmic scale for each species. In general, all three parameters were highly correlated with diameter. Thus, the aboveground height of most trees and the size and shape of their crowns can be predicted from their trunk diameter.

Growth, recruitment and mortality rates for each species were calculated under different environmental conditions. In the observation period which experienced a one-year drought, basal area had less of an effect on the rates studied than the length of time between observations. In the period with normal climatic conditions, basal area had the opposite effect on the rates than was expected. Growth and recruitment were higher and mortality rates were lower in subplots with higher basal area than in subplots with lower basal area.

The structural and dynamic characteristics of the species were used in a hierarchical, agglomorative cluster analysis. The species were grouped into 5 categories. A PCA of the data supported the separation of several of the clusters in the tree diagram, but indicated that some clusters were more similar than the tree diagram suggested. The division of species into clusters and the characteristics of the clusters generally agreed with published accounts of several species' life history strategies. The distinction between the clusters was tested using the diameter distributions of the clusters and of several species in the clusters. The results suggested that neither a model of forest organization in which the species can be grouped into guilds nor one in which each species occupies a unique niche represented the competitive relationship between the species in the forest. Among canopy trees there may be some guilds, however, among the subcanopy species there was little difference between species of the same growth form.

Note: Abstract extracted from PDF file via OCR.