Carbon balance of rubber (Hevea brasiliensis) plantations: A review of uncertainties at plot, landscape and production level

Rapid expansion of natural rubber plantations in South-East Asia and other regions has greatly altered ecosystem based carbon (C) stocks with potential impacts on climate change mitigation and future C trading opportunities. Therefore, reliable estimations of carbon sequestration and emission at the landscape level after land cover transition from forest, swidden agriculture and other land use types are needed. We reviewed studies on C stocks and dynamics in rubber plantations considering the contribution of aboveground and belowground biomass, soil organic matter, collected latex and other minor components. C sequestration occurred after conversion of arable land to rubber plantations while C losses usually prevailed if forest was converted to rubber. These general trends strongly depended on local climate conditions and soil properties as well as on topography. Non-traditional planting of rubber under subtropical conditions with a dryer climate and at high elevations (300-1000 m above sea level) decreased the C sequestration potential of rubber. We show how rotation length, rubber clone, and management strategies like tapping frequency or planting density affect C stocks, discuss the uncertainties in C stock estimation and highlight improved approaches. An important conclusion is that upscaling of C stocks and dynamics under different climate scenarios and rotation lengths to a regional level requires the use of time averaged C stocks. Enhanced remote sensing techniques can greatly improve C stock estimates at the regional level, allowing for an accounting of the variability caused by terrain and plantation properties. A partial life cycle assessment of rubber production revealed greenhouse gas emissions as a minor contribution when compared to land use change effects on plant and soil C stocks and C accumulation in latex, wood products and seed oil. The review highlights scantily explored topics and proposes directions for future studies, which should decrease uncertainties in C estimates in rubber dominated landscapes. (C) 2016 Elsevier B.V. All rights reserved.