Spatial and temporal variability of carbon dynamics in a tropical forest of Colombia

Graduation date: 2006Despite the importance of tropical forest ecosystems in the global carbon cycle, there have been few studies of carbon dynamics in this biome. The magnitude of carbon stocks in the tropics and their changes over time are poorly known since ground-based observations are lacking. In this study, total carbon stocks (TCS) and net ecosystem production (NEP) were quantified for tropical forests of the Porce region of Colombia. A modeling exercise was also performed to analyze the effects of population and community processes on carbon dynamics at the ecosystem level.\ud A set of 110 permanent plots were used to estimate TCS and its uncertainty in primary and secondary forests. In primary forests, mean TCS were estimated to be 383.7 ± 43.0 Mg C ha-1 (± standard error). Of this amount, soil organic carbon to 4 m depth represented 59%, total aboveground biomass 29%, total belowground biomass 10%, and necromass 2%. In secondary forests, TCS was 228.2 ± 11.5 Mg C ha-1. Of this store, soil organic carbon to 4 m depth accounted for 84%, total aboveground biomass represented only 9%, total belowground biomass 5%, and total necromass 1%. Based on the uncertainty analysis of TCS estimates, the variability associated with the spatial variation\ud of C pools between plots was higher than measurement errors within plots. A larger variability was observed in primary than in secondary forests and this difference might be explained by gap dynamics.\ud Net ecosystem production was measured in primary forests in a set of 33 permanent plots from 2000 to 2002 in two, one-year intervals. Uncertainty analysis indicated that NEP ranged between -4.03 and 2.22 Mg C ha-1 yr-1 for the two intervals. This range was compared to a priori defined range of natural variation (-1.5 and 1.5 Mg C ha-1 yr-1) estimated from the ecosystem model STANDCARB. The observed variation in NEP did not provide sufficient evidence to reject the hypothesis that the ecosystem was within its expected natural range.\ud Simulations using the STANDCARB model showed that at the population level, the processes of colonization and mortality can limit the maximum biomass achieved during a successional sequence. Colonization can introduce lags during the initiation of succession and mortality can have important effects on annual variation in carbon stores. Community dynamics, defined as the replacement of species during succession, altered the mixture of species over time. When species had different ecosystem parameters, such as growth and mortality rates, community dynamics caused non-linear patterns of carbon accumulation. These patterns could not be reproduced using a single species with the average of parameters of a multi-species simulation or by using the more abundant species in the simulations