Below is the Executive Summary of The GEFSOC Project Technical Report. It outlines the major findings of The Project including estimates of SOC stocks and stock changes for The Brazilian Amazon, Jordan, Kenya and The Indian part of the Indo-Gangetic Plains made using the 3 methods included in the GEFSOC System (Century, RothC and the IPCC method). The full report is available for download from the Output page of this site. Please give the following citation when quoting any of the results given below:

Milne, E., Easter, M., Cerri, C.E.P, Paustian, K. and Williams, S. (Eds.) 2005 Technical Report of GEF Co-financed Project No. GFL-2740-02-4381. Assessment of Soil Organic Carbon Stocks and Change at National Scale. Available from http://www.nrel.colostate.edu/projects/gefsoc-uk

Soil organic carbon (SOC) is of local importance as it determines ecosystem and agro-ecosystem function, influencing soil fertility, water holding capacity and many other functions. It is also of global importance because of its role in the global carbon cycle and therefore, the part it plays in the mitigation or worsening of atmospheric levels of greenhouse gases (GHGs). Past long-term experimental studies have shown that soil organic C is highly sensitive to changes in land use, with changes from native ecosystems such as forest or grassland to agricultural systems almost always resulting in a loss of SOC (Jenkinson and Rayner 1977, Paul et al., 1997). Likewise, the way in which land is managed following land use change has also been shown to affect SOC stocks. We therefore have the opportunity in the future to adopt land use and land management strategies that lead to greater C storage in the soil, thereby mitigating GHG effects and improving soil fertility. Maximising this opportunity will require the formulation of policy at the national and sub-national scale. A need therefore exists for a generically applicable system for estimating current soil organic carbon stocks and likely changes under future land use change scenarios, at the national and sub-national scale.

The biggest changes in land use and SOC are likely to occur in tropical areas, which (along side arid areas) are also the areas for which we have the least data and the least understanding of SOM processes. In general present understanding of SOC stocks and changes is severely skewed in favour of temperate areas. Some studies have used models linked to spatial data to estimate SOC stock changes at the sub-national and national scale for areas of USA and Europe (Paustian 1997; Falloon 1998) and at the watershed scale for areas of Mexico and Cuba (Ponce Hernandez, 2004). However, a need remains for a model based system that is generically applicable, encompassing as wide a range of soil types, climates and land uses as possible. This project has developed the Global Environment Facility Soil Organic Carbon  (GEFSOC) Modelling System in response to this need. Such a system can be used by funding agencies such as The Global Environment Facility, to allow them to determine likely effects of national and sub-national scale projects on SOC stocks and by the Intergovernmental Panel on Climate Change (IPCC) to improve their guidelines on SOC stock estimation.

The GEFSOC Modelling System was developed using data from 4 contrasting eco-regions, The Brazilian Amazon, Jordan, Kenya and the Indian part of the Indo-Gangetic Plains. These regions were chosen for their contrasting climates and soil types and the fact that they provide examples of those areas underrepresented by current soil carbon models. Each region was treated as a case study, being coordinated by an in-country institution, with a common method being employed in all. The method involved five stages; 1) The evaluation and parameterisation of the two SOC models being used (RothC and Century) for their performance in the case study conditions, 2) Collation and formatting of national and sub-national scale data sets of soils, climate and land use for each of the case studies, 3) The development of a system linking 2 SOC models and the IPCC method to spatially explicit data via a geographic information system (GIS), 4) Assessment of current SOC stocks and comparison of these estimates with outcome from existing methods, 5) Development of plausible land use change scenarios for each of the case study areas and assessment of SOC stock changes under these scenarios.

A number of data sources and approaches were used for the evaluation of RothC and Century, depending on the availability of data in the different case study areas. These included long term experimental data sets and data from land use change chronosequences. The applicability of the models to case study country conditions was improved in a number of ways including; -1) The development of or improvement of crop and tree input parameters, 2) The improvement of Century to enhance model performance in double and triple cropping systems and flooded systems such as rice and 3) The improvement of the decomposition model in Century to allow better performance at high temperatures. In addition, a system was developed which allowed both RothC and Century to be run in tandem and to allow RothC to run on LINUX and UNIX operating systems.

Under The GEFSOC Project, national and sub-national scale data sets have been collated and formatted for soils, climate, land use (historical and current) and land management for The Brazilian Amazon, the Indian Indo-Gangetic Plains, Jordan and Kenya. All of these data sets are now in a comparable, standardised format. The GEFSOC project used existing data sets from a variety of sources with the idea of consolidating fragmented data into an easily accessible format that could be used by a generic modelling system. For soils, each case study country now has a complete Soil and Terrain (SOTER) database. Each country now has a database of land use and land management (both current and historical) compiled from national statistical reports (formerly in paper format), research reports and global data sets. Each country also has a comparable climate database, put together from a variety of sources using different approaches, but with a common output. These datasets will be useful for compilation of national inventories of GHGs, other research projects and agricultural, biodiversity and land degradation assessments.

Many different factors affected SOC dynamics and therefore, SOC stocks in each of the case study countries. Land use in the case studies varied widely, ranging from very recent change from native vegetation in The Brazilian Amazon, to centuries old agricultural systems in the IGP. Soil C stock estimates and their distribution between major land use types and ecological zones showed some differences between the three models (Century, RothC and IPCC) and the type and degree of agreement also varied between the countries.  For the Brazilian Amazon, Century estimated higher SOC stocks (~ 32,500 Tg) than either RothC or IPCC, with both estimating current soil C stocks at ~  27,000 Tg. Map and survey based estimates were ~ 21,000 to 25,000 Tg C.  In terms of total stocks (Tg) native vegetation and degraded pasture were the highest contributors to the total, however on a per unit area basis (Tonnes ha^-1), native vegetation and well-managed pasture were more important. For Kenya, Roth C and Century predicted similar SOC stocks of ca. 1400-1500 Tg C, although there were discrepancies between predictions for different land use classes. IPCC results were 25% higher (~2000Tg), which were comparable given the greater depth represented (0-30 cm). In Jordan SOC stocks were low, as expected for such an arid country. Stocks in 2000 were 102 and 66 Tg for RothC and Century respectively, although distribution between land uses was similar for both models.  These values are comparable with estimates made for Jordan using the SOTER mapping approach (76-78 Tg).  In the Indian IGP, SOC stocks estimates for 2000 were 1325 Tg C (0-20 cm depth) according to Century and almost the same 1381 for the IPCC method. However, results were comparable on a total volume basis as the IPCC method accounts for a larger soil volume (0-30 cm). With Century in the IGP, SOC stock estimates followed land area distribution, with the rice-wheat system being responsible for 50% of land area and SOC stocks alike.

Land use and land management changes for each of the case study countries were estimated from 2000 to 2030 based on the FAO projections of changes in cropping area and crop production (FAO 2002). In general, the two simulation models (Century and RothC) predict overall losses of soil C for all the countries, for the future projections (2000-2030).  This reflects both continued land use change in Brazil and Kenya as well as current and projected land management changes, within a specific land use.  In contrast, the IPCC method predicts a lower degree of change than the simulation models.

In the Brazilian Amazon, The system estimated that if current deforestation rates continue, 4200 Tg of SOC will have been lost by 2030 (compared to the 1990 value). Therefore, stopping (or at least reducing) deforestation is an obvious recommendation from an SOC point of view, in addition to all of the other benefits of forest conservation. In the Indian IGP, projected future SOC losses were mainly associated with the intensification of cropping and the disappearance of fallow periods in cropping rotations. Intensification of cropping can only benefit SOC if it is accompanied by appropriate tillage practices, therefore it is recommended that the projected intensification in the IGP be accompanied by reduced tillage practices. In Jordan, again, future SOC stock reductions were estimated by both models in The GEFSOC system. These reductions were due almost exclusively to degradation of rangeland. Reducing stocking rates and avoiding over grazing are recommended in order to avoid SOC stock depletion and possible desertification in Jordan. Similarly, for Kenya, projected SOC stock reductions between 2000 and 2030 were attributed mainly to one land use change, conversion of grasslands and savannah to subsistence agriculture. Therefore, preservation of natural grasslands and savannah is likely to benefit Kenyan national SOC stocks in the future.

The GEFSOC Modelling System and outputs from The GEFSOC Project have great potential for use by a variety of end-users. The system can be used by The GEF to provide information on the likely impacts of national and sub-national scale projects (that propose land use/management change) on soil carbon stocks and changes. It can also be used by other research projects, land use planners or government departments wishing to answer questions on the impact of land use change on SOC stocks and GHGs inventories. Although originally designed for use at the national and sub-national scale, The GEFSOC system in essence, can be applied to any complex system land use system involving large amounts of data. It could therefore be applied at the watershed scale, or other scales, providing adequate datasets were available.

The GEFSOC Modelling System will continue to be developed and improved. Ideas for development include adaptation of the system to allow consideration of soil erosion and salinisation on SOC stocks, inclusion of further detailed land management information for tropical subsistence farming systems, consideration of soil inorganic carbon stocks and modification of outputs to include trace GHGs emissions.

SOURCE:

Milne, E., Easter, M., Cerri, C.E.P, Paustian, K. and Williams, S. (Eds.) 2005 Technical Report of GEF Co-financed Project No. GFL-2740-02-4381. Assessment of Soil Organic Carbon Stocks and Change at National Scale. Available from http://www.nrel.colostate.edu/projects/gefsoc-uk

References

Falloon, P.D., Smith, P., Smith, J.U., Szabó, J., Coleman, K., Marshall, S., 1998. Regional estimates of carbon sequestration potential: linking the Rothamsted carbon model to GIS databases. Biology and Fertility of Soils 27, 236-241.

FAO, 2002. World agriculture: towards 2015/2030. Food and Agriculture Organization of the United Nations. Rome, pp.97.

Jenkinson D.S and Rayner J.H. 1977. The turnover of soil organic matter in some of the Rothamsted Classical Experiments. Soil Science 123, 298-305

Paul, E.A., Paustian K., Elliott, E.T and C.V. Cole.  1997. Soil Organic Matter in Temperate Agroecosystems.  CRC Press, New York

Paustian, K., Levine, E., Post, W.M., Ryzhova, I.M., 1997. The use of models to integrate information and understanding of soil C at the regional scale. Geoderma 79, 227-260.

Ponce Hernandez, R. 2004. Assessing carbon stocks and modelling win-win scenarios of carbon sequestration through land use change. Food and Agriculture Organisation of the United Nations, Rome, Italy