Modeling impacts of management alternatives on soil carbon storage of farmland in Northwest China

• Main Authors: F. Zhang, C. Li, Z. Wang, H. Wu
• Format: Article
• Language: English
• Published: Copernicus Publications 2006-01-01
• Series: Biogeosciences
• Online Access: http://www.biogeosciences.net/3/451/2006/bg-3-451-2006.pdf
• ISSN: 1726-4170; 1726-4189

Long-term losses of soil organic carbon (SOC) have been observed in many agriculture lands in Northwest China, one of the regions with the longest cultivation history in the world. The decline of SOC contents not only impaired the soil fertility but also increased the amount of carbon dioxide (CO₂) emitted from terrestrial ecosystems into the atmosphere. However, quantifying the SOC losses at regional scale has long been remaining unsolved. A process-based model, <b>D</b>e<b>n</b>itrification-<b>D</b>e<b>c</b>omposition or DNDC, was adopted in the study to quantify impacts of farming management practices on SOC dynamics for a selected region, Shaanxi Province. The selected domain, with 3 million hectares of cropland across different climatic and farming management regimes, is representative for the major agricultural areas in Northwest China. The DNDC model was tested against long-term SOC dynamics observed at five agricultural sites in China. The agreement between the observed and modeled results indicate that DNDC was capable of capturing patterns and magnitudes of SOC changes across the climate zones, soil types, and management regimes in China. To quantify SOC dynamics for Shaanxi, we constructed a GIS database to hold all of the DNDC-required input information (e.g., weather, soil properties, crop acreage, and farming practices) for all the farmland in the studied domain. Sensitivity tests indicated the spatial heterogeneity of soil properties, especially initial SOC content, was the major source of uncertainty for the modeled SOC dynamics at regional scale. The Most Sensitive Factor (MSF) method was employed in the study to quantify the uncertainties produced from the upscaling process. The results from the regional simulations for Shaanxi indicated that (1) the overall 3 million hectares of farmland in the province was a source of atmospheric CO₂ with 0.5 Tg C yr^&minus;1 lost in 2000 under the current farming management conditions, (2) an increase in the rate of crop residue incorporation from 15% to 50% or 90% converted the farmland from the C source to a C sink with 0.7 or 2.1 Tg C yr^&minus;1 sequestered, respectively, and (3) applying farmyard manure at a rate of 500 kg C ha^&minus;1 also converted the cropland into a weak C sink with 0.2 Tg C sequestered in the farmland in Shaanxi. In the studied domain, crop residue and farmyard manure used to be the major sources of soil nutrients during the centuries-long cultivation history but were gradually abandoned since 1950s when synthetic fertilizers became available. The results from this modeling study suggest that recovery of the traditional farming practices (i.e., residue incorporation and manure amendment) be a feasible approach to substantially improve the soil C status in the farmland of Shaanxi. The values of C gain or loss from this study for Shaanxi could vary if the modeled domain shifts to other provinces in Northwest China although the general conclusion may remain.