摘要
Tropical deforestation has reduced the extent of natural forests, which conserve biodiversity, provide essential resources to people, and reduce climate change by storing carbon. Forest conservation projects need tree species data to effectively manage biodiversity while greenhouse gas reduction programs require robust methods to estimate forest carbon. Here, we use field measurements, remote sensing, and Monte Carlo analyses to quantify tree biodiversity and aboveground carbon changes and uncertainties in 5200 km2 of Amazonian and Yungas rainforest and other land around the Parque Nacional Yanachaga-Chemill茅n and two other protected areas in the Selva Central, Peru. Field inventories of 17 ha found 438 tree species in 156 families. Field measurements of 10,838 trees and Monte Carlo analyses of uncertainties in measurements, allometric equations, wood density, and the carbon fraction of biomass showed that aboveground live carbon densities were 93 卤 39 Mg ha鈭? (mean 卤 95% confidence interval [CI]) in old-growth forest and 40 卤 10 Mg ha鈭? in secondary forest. Carbon density was significantly correlated to tree species richness (P < 0.0001). Supervised classification of Landsat images showed a 1989-2005 net deforestation rate of 0.3% y鈭?, reduction of forest cover from three-quarters of the area to two-thirds, and net degradation of additional forest equivalent to half the deforested area. A Monte Carlo analysis that combined carbon density and remote sensing uncertainties showed that forest changes caused statistically significant net emissions of 1.6 卤 0.4 million Mg carbon. Allometric equations and remote sensing accounted for most of the uncertainty. Multivariate statistical analyses showed that, of six factors examined, distance to roads most explained historical deforestation patterns. The protected areas experienced no net deforestation, very low degradation, and very low change close to roads. Projection of potential forest cover to 2021 indicates that a Reducing Emissions from Deforestation and Degradation (REDD ) project could avoid gross emissions of 2.8 卤 0.8 million Mg carbon. One-eighth of the area would be eligible for afforestation or reforestation under the Clean Development Mechanism (CDM), with credit for carbon storage occurring above a projected baseline gross reforestation rate of 1.8% y鈭?. These activities could conserve tropical forest carbon and biodiversity.