Uncertainty in the biomass of Amazonian forests: An example from Rondônia, Brazil

摘要: Abstract A critical factor in estimating the contribution of tropical deforestation to nutrient mobilization and CO2 build-up atmosphere is amount biomass available burn. The data for Brazil, a major site deforestation, are few uncertain accuracy. Recent international agreements, however, require national inventories sources sinks atmospheric greenhouse gases; such will need better estimates their uncertainties. To provide additional information on uncertainty forest structure southwestern Amazonia, region active we measured 1988 diameter, bole canopy heights 474 trees covering total 1 ha (10 000 m2) Ecological Station Samuel Hydroelectric Reservoir Rondonia (845′S, 63°23′W). Using allometric equations based destructively sampled trees, estimated largest component, standing alive aboveground (SAAB), as 285 Mg (dry weight) ha−1. Fallen trunks litter were 30 10 ha−1, respectively. sum these components, 325 an underestimate because roots, vines, shrubs, small was not measured. Measurement error SAAB ± 20%, 57 ha−1 about mean (95% confidence interval), derived by Monte Carlo simulation. distribution among highly skewed: 3% contain 50% SAAB. For forests similar distributions, sampling units typically used (less than 2000 usually produce significantly different from those larger units. Based subsamples our data, 1000 m2 or smaller had at least 75% chance being outside interval global (228–342 ha−1) improve program should focus emergent large dominant contributors biomass.