Successional patterns of mortality and growth of large trees in a Panamanian lowland forest

1. All trees ó 19.1 cm. d.b.h. (diameter at breast height) in three 1-ha forest plots in Panama were tagged and measured in 1975 and recensused in 1980 and 1988 to examine mortality, growth and recruitment. Plots ranged in age from old second-growth (plot 1), to disturbed primary forest of intermediate age (plot 2) to undisturbed primary forest perhaps 500 years old (plot 3). Common species were classified into two regeneration classes - gap-positive recruitment (GPR) or gap-neutral recruitment (GNR)-on the basis of other studies. 2. As plot age increased, tree density decreased slightly and forest composition changed from dominance of GPR to GNR trees. Species diversity was similar in plots 1 and 3 and lower in plot 2. Basal area per stem and total basal area increased with plot age. 3. Overall mortality rate was 2.0 percent per year. The mortality rate was 58 percent greter in the second period (1980-88) than in the first (1975-80). When all trees were considered, no effects of plot age or size on mortality were detected. However, when common species were considered according to their regeneration class, mortality exhibited significant regeneration class by size and regeneration class by plot interaction. Mortality of GPR trees increased with girth size and decreased with plot age. The converse was true for GNR trees. 4. Growth rate increased and relative growth rate decreased exponentially with tree size. After correction for the effects of size, GNR trees exhibited significantly faster. 5. We conclude that mortality patterns of large trees can vary with the age of the stand and are essentially uncorrelated with patterns exhibited by the same species at juvenile stages. Changes is mortality with size and stand age appeared dependent on regeneration class. A successional gradient is suggested from plots 1 to 3 whereby GNR species replace GPR species over time, though the number of large trees did not decrease significantly. Population studies confined only to stands of older or younger forest would probably fail to detect many successional effects noted in this study. The time scale at which mortality data are compiled can also effect results.