PARENT SESSION
Contributed Oral Session 146: Biogeochemistry: Nutrient Dynamics; Carbon Sequestration
Thursday, August 11, 1:30 PM - 5:00 PM, Meeting Room 524 C, Level 5, Palais des congrès de Montréal
Estimation of carbon fluxes and stocks and their uncertainty in a tropical forest of Colombia.
Sierra, Carlos*,1, Harmon, Mark1, del Valle, Jorge2, Orrego, Sergio 1, 2, Moreno, Flavio2, 1 Oregon State University, Corvallis, OR2 Universidad Nacional de Colombia Sede Medellín, Medellín, Colombia
ABSTRACT- A set of 33 permanent plots (0.1 ha) were used to estimate Total Carbon Stocks (TCS), Net Primary Production (NPP), Ecosystem Respiration (Re), and Net Ecosystem Production (NEP) in a tropical premontane forest in Colombia. Four main C pools were monitored: aboveground biomass (dbh > 1 cm), belowground biomass, necromass, and soil organic carbon to 4 m depth. Changes in above- and below-ground biomass as well as litter decomposition and soil respiration were monitored during one year to get estimates of NPP, Re, and NEP. Mean (± standard deviation) TCS was estimated as 407.5 ± 57.2 Mg C ha-1. Of this, soil organic carbon to 4 m depth represents 62%, total aboveground biomass 27%, belowground biomass 11%, and necromass 2%. For the period 2000-2001, mean NPP was estimated as 12.7 ± 0.8 Mg C ha-1 yr-1 and mean Re as -13.0 ± 2.4 Mg C ha-1 yr-1, which suggest that this forest was neutral in the net C exchange with the atmosphere for this particular period (NEP = -0.4 ± 1.7 Mg C ha-1 yr-1). Monte Carlo methods were used to assess the uncertainty of the estimates and their spatial variation. We found that the total uncertainty of TCS and NPP is mainly explained by the spatial variation of C pools between plots rather than measurement errors. Variability in both measurements and spatial heterogeneity in soil carbon stocks and fluxes, are the main sources of variation in our estimates of TCS and NEP. The 95% confidence interval for our estimate of NEP (-3.6, 2.9 Mg C ha-1 yr-1) is in close agreement with the spatial and temporal variation of NEP (95% confidence interval = -2.2, 2.6 Mg C ha-1 yr-1) simulated with the ecosystem carbon model StandCarb for an old-growth forest. We hypothesize that gap formation and regeneration cause much of the spatial variation in C fluxes observed in this study; however variation in these fluxes are balanced at the landscape scale.
Key words: Tropical forests, Forest carbon pools and fluxes, Uncertainty analysis