Abstract:Land-use change may be carbon sources or sinks that play vital roles in the terrestrial carbon cycle. This paper reviewed the effects of land-use change on the carbon cycle of forests, grasslands, and agricultural ecosystems based on researches carried out in China and worldwide. Globally, the effects of the conversion of forest to cropland and grassland on the global carbon cycle dominate the carbon emissions from deforestation. Regional differences in carbon emissions due to transformation of forests into croplands and grasslands reflect on the regional climate. The carbon emissions per/hm2 from tropical forest converted into cropland and grassland were 151 and 120 tC/hm2 higher than the emissions from temperate or boreal forests, respectively. Furthermore, land-use change could promote forest carbon storage through reforestation and improved forest management. Significant differences exist in the potential carbon sinks created through change in land use. Tropical humid and semi-humid regions have greater potential for carbon sequestration than temperate regions, while the potential for carbon sequestration in arid regions is relatively small. Reclamation of grassland to cropland is one of the main human activities that affects the carbon stock of grassland ecosystems, causing the soil carbon stock to fall by 59%. When a forest or grassland is converted to cropland, carbon stocks in both vegetation and soil are reduced as well as the ecosystem's carbon storage capacity. With the expansion of cities, the conversion of cropland to land for construction further increases carbon emissions. Currently, researchers employ several methods when analyzing the effects of land-use change including remote sensing and models based on remote sensing, statistical estimation, ecosystem models, and the coupling models of land-use change and ecosystem. Temporal land-use change data can be obtained from remote sensing data, as well as the data related to temperature, soil moisture, and vegetation structure parameters. These data can be used to drive remote sensing models such as the Carnegie-Ames-Stanford Approach (CASA) and Global Production Efficiency Modeling (GLO-PEM). The bookkeeping model is a typical statistical model and is easy to use, while the simulation results cannot provide sufficient accuracy. Ecosystem models include models of static and dynamic types. CENTURY, CASA, and BIOME are believed to be the most widely used static models, while currently the Lund-Potsdam-Jena Dynamic Global Vegetation Model (LPJ-DGVM) and Terrestrial Ecosystem Model (TEM) are the most widely used dynamic models. Although ecosystem models can simulate the carbon balance of an ecosystem quite well, the current models do not consider the impact of land-use change on the carbon cycle and could be applied only at certain geographical scales. Therefore, a land-use model should be coupled with ecosystem models, such as the Integrated Model to Assess the Global Environment (IMAGE) and the Lund-Potsdam-Jena managed Land Dynamic Global Vegetation, Agriculture, and Water Balance Model (LPJmL). Although the research methods have been continually refined and improved, uncertainties remain in both the data and the models. Therefore, unified statistical observation methods should be established to reduce the uncertainties within various datasets, which will improve the accuracy of simulation results. Additionally, we need to improve the use of land-use models coupled with ecosystem models, which is expected to be the main research direction of this field. Integrated technology systems using multi-scale land-use change and ecosystem data should be established while considering social, economic, and environmental factors that drive carbon storage and emission. Furthermore, we need to carry out research designed to aid in the optimization of land-use layout with the target of reducing carbon emissions to provide more valuable information and sound policy recommendations.