Currently, there is little information about soil organic carbon (SOC) stocks and changes in Mediterranean areas at a regional scale. We modelled an area of 95 269 km2 in northeast Spain using the Global Environmental Facility Soil Organic Carbon (GEFSOC) system to predict SOC stocks and changes in pasture, forest and agricultural soils. The spatial distribution of the different land-use categories and their change over time was obtained by using the Corine database and official Spanish statistics on land use from 1926 to 2007. The model predicted the largest current SOC stock in forest soils at 578 Tg C. Agricultural soils were the second largest SOC reservoir, containing 244 Tg C. During the last 30 years, the model predicted a total SOC gain in the 0–30-cm soil layer of 34 Tg C. Forest and grassland-pasture soils had a decline in their stored SOC of 5 and 3 Tg C, respectively, because of the reduction in the soil surface occupied by both classes. The greatest SOC gain was predicted in agricultural soils with 42 Tg C caused by changes in management, which led to increases in C inputs. Although model uncertainty was not quantified, some hypothetical assumptions about the initialization and parameterization of the model could be potential sources of uncertainty. Our simulations predicted that in northeast Spain soil management has contributed to the sequestration of substantial amounts of atmospheric CO2 during the last 30 years. More research is needed in order to study the potential role of soils as atmospheric CO2 sinks under different managements and climatic conditions.

Carbon (C) added to soil as organic matter in crop residues and carbon emitted to the atmosphere as CO2 in soil respiration are key determinants of the C balance in cropland ecosystems. We used complete and comprehensive county-level yields and area data to estimate and analyze the spatial and temporal variability of regional and national scale residue C inputs, net primary productivity (NPP), and C stocks in US croplands from 1982 to 1997. Annual residue C inputs were highest in the North Central and Central and Northern Plains regions that comprise ~70% of US cropland. Average residue C inputs ranged from 1.8 (Delta States) to 3.0 (North Central region) Mg C ha−1 year−1, and average NPP ranged from 3.1 (Delta States) to 5.4 (Far West region) Mg C ha−1 year−1. Residue C inputs tended to be inversely proportional to the mean growing season temperature. A quadratic relationship incorporating the growing season mean temperature and total precipitation closely predicted the variation in residue C inputs in the North Central region and Central and Northern Plains. We analyzed the soil C balance using the crop residue database and the Introductory Carbon Balance regional Model (ICBMr). Soil C stocks (0–20 cm) on permanent cropland ranged between 3.07 and 3.1 Pg during the study period, with an average increase of ~4 Tg C year−1, during the 1990s. Interannual variability in soil C stocks ranged from 0 to 20 Tg C (across a mean C stock of 3.08 ± 0.01 Pg) during the study period; interannual variability in residue C inputs varied between 1 and 43 Tg C (across a mean input of 220 ± 19 Tg). Such interannual variation has implications for national estimates of CO2 emissions from cropland soils needed for implementation of greenhouse gas (GHG) mitigation strategies involving agriculture.

Two long-term fertilizer experiments (LTFE) datasets were used to evaluate the performance of the Century ecosystem model in contrasting regions of India viz Mohanpur (humid) and Akola (semi-arid) with mean annual rainfall of 1619 mm and 793 mm, respectively. Mohanpur grew rice and wheat in rotation for 19 years since 1986. Akola grew sorghum and wheat in rotation for 9 years since 1988. Both these experiments involved treatments with different doses of inorganic (fertilizer) and organic (farm yard manure, paddy straw, and green manure) inputs. The model closely resembled measured SOC level for all the treatments in Mohanpur LTFE. For Akola, it tended to overestimate treatment effects at the end of the experimental period. At the humid site (Mohanpur) modelled data simulated measured data reasonably well for all treatments, with control and treatments with fertilizer alone and in combination with organic inputs showing the best agreement (RMSE 1-3). At the semi-arid site Century performed well for the early years, but lower during the end of the experiment. The comparison between measured and modelled yields for the three crops (rice, wheat and sorghum) showed a reasonably good correlation (r = 0.8). (C) 2010 Elsevier B.V. All rights reserved.

Land use, land use change and forestry activities play an important role in determining whether soil is a sink or source of atmospheric carbon dioxide (CO2). The effects of land use change on greenhouse gases and climate change are receiving greater attention in many developing countries. We simulated changes in soil organic carbon (SOC) pool over 100 years (1950–2050) under managed dense Shorea forest (DS), rainfed upland (Bari) and irrigated low land (Khet) of Pokhare Khola, a mid-hill watershed of Nepal, using the Century model. The model was calibrated and parameterized on the basis of biophysical and weather data of the watershed and field measurements made in September/October 2004. Historic C pool data were calculated using the IPCC inventory guidelines for the year 1976 and 1989. The Century model performed well in representing the effects of various land management systems on SOC in the Pokhare Khola watershed. There was a dramatic loss of SOC following the removal of tree biomass and conversion to agriculture. However, it was shown that partial recovery of lost SOC is possible through better forest management, especially if erosion is abated. In the cultivated land we found better recovery by addition of farm yard manure and fertilizer. In all the land use systems, the SOC recovery was lower under the climate change scenario than under prevailing climate.

This article was submitted without an abstract, please refer to the full-text PDF file.

Soil organic carbon (SOC) plays a vital role in ecosystem function, determining soil fertility, water holding capacity and susceptibility to land degradation. In addition, SOC is related to atmospheric CO, levels with soils having the potential for C release or sequestration, depending on land use, land management and climate. The United Nations Convention on Climate Change and its Kyoto Protocol, and other United Nations Conventions to Combat Desertification and on Biodiversity all recognize the importance of SOC and point to the need for quantification of SOC stocks and changes. An understanding of SOC stocks and changes at the national and regional scale is necessary to further our understanding of the global C cycle, to assess the responses of terrestrial ecosystems to climate change and to aid policy makers in making land use/management decisions. Several studies have considered SOC stocks at the plot scale, but these are site specific and of limited value in making inferences about larger areas. Some studies have used empirical methods to estimate SOC stocks and changes at the regional scale, but such studies are limited in their ability to project future changes, and most have been carried out using temperate data sets. The computational method outlined by the Intergovernmental Panel on Climate Change (IPCC) has been used to estimate SOC stock changes at the regional scale in several studies, including a recent study considering five contrasting eco regions. This 'one step' approach fails to account for the dynamic manner in which SOC changes are likely to occur following changes in land use and land management.A dynamic modelling approach allows estimates to be made in a manner that accounts for the underlying processes leading to SOC change. Ecosystem models, designed for site scale applications can be linked to spatial databases, giving spatially explicit results that allow geographic areas of change in SOC stocks to be identified. Some studies have used variations on this approach to estimate SOC stock changes at the sub-national and national scale for areas of the USA and Europe and at the watershed scale for areas of Mexico and Cuba. However, a need remained for a national and regional scale, spatially explicit system that is generically applicable and can be applied to as wide a range of soil types, climates and land uses as possible. The Global Environment Facility Soil Organic Carbon (GEFSOC) Modelling System was developed in response to this need. The GEFSOC system allows estimates of SOC stocks and changes to be made for diverse conditions, providing essential information for countries wishing to take part in an emerging C market, and bringing us closer to an understanding of the future role of soils in the global C cycle. (C) 2007 Elsevier B.V. All rights reserved.

The GEFSOC Project developed a system for estimating soil carbon (C) stocks and changes at the national and sub-national scale. As part of the development of the system, the Century ecosystem model was evaluated for its ability to simulate soil organic C (SOC) changes in environmental conditions in the Indo-Gangetic Plains, India (IGP). Two long-term fertilizer trials (LTFT), with all necessary parameters needed to run Century, were used for this purpose: a jute (Corchorus capsularis L.), rice (Oryza sativa L.) and wheat (Triticum aestivum L.) trial at Barrackpore, West Bengal, and a rice–wheat trial at Ludhiana, Punjab. The trials represent two contrasting climates of the IGP, viz. semi-arid, dry with mean annual rainfall (MAR) of <800mm and humid with >1600mm. Both trials involved several different treatments with different organic and inorganic fertilizer inputs. In general, the model tended to overestimate treatment effects by approximately 15%. At the semi-arid site, modelled data simulated actual data reasonably well for all treatments, with the control and chemical N+farm yard manure showing the best agreement (RMSE=7). At the humid site, Century performed less well. This could have been due to a range of factors including site history. During the study, Century was calibrated to simulate crop yields for the two sites considered using data from across the Indian IGP. However, further adjustments may improve model performance at these sites and others in the IGP. The availability of more long-term experimental data sets (especially those involving flooded lowland rice and triple cropping systems from the IGP) for testing and validation is critical to the application of the model's predictive capabilities for this area of the Indian sub-continent.

The GEFSOC soil carbon modelling system was built to provide interdisciplinary teams of scientists, natural resource managers and policy analysts (who have the appropriate computing skills) with the necessary tools to conduct regional-scale soil carbon (C) inventories. It allows users to assess the effects of land use change on soil organic C (SOC) stocks, soil fertility and the potential for soil C sequestration. The tool was developed in conjunction with case-studies of land use and management impacts on SOC in Brazil, Jordan, Kenya and India, which represent a diversity of land use and land management patterns and are countries where sustaining soil organic matter and fertility for food security is an on-going problem. The tool was designed to run using two common desktop computers, connected via a local area network. It utilizes open-source software that is freely available. All new software and user interfaces developed for the tool are available in an open source environment allowing users to examine system details, suggest improvements or write additional modules to interface with the system. The tool incorporates three widely used models for estimating soil C dynamics: (1) the Century ecosystem model; (2) the RothC soil C decomposition model; and (3) the Intergovernmental Panel on Climate Change (IPCC) method for assessing soil C at regional scales. The tool interacts with a Soil and Terrain Digital Database (SOTER) built for the specific country or region the user intends to model. A demonstration of the tool and results from an assessment of land use change in a sample region of North America are presented.

The Global Environment Facility co-financed Soil Organic Carbon (GEFSOC) Project developed a comprehensive modelling system for predicting soil organic carbon (SOC) stocks and changes over time. This research is an effort to predict SOC stocks and changes for the Indian, Indo-Gangetic Plains (IGP), an area with a predominantly rice (Oryza sativa)–wheat (Triticum aestivum) cropping system, using the GEFSOC Modelling System and to compare output with stocks generated using mapping approaches based on soil survey data. The GEFSOC Modelling System predicts an estimated SOC stock for the IGP, India of 1.27, 1.32 and 1.27 Pg for 1990, 2000 and 2030, respectively, in the top 20 cm of soil. The SOC stock using a mapping approach based on soil survey data was 0.66 and 0.88 Pg for 1980 and 2000, respectively. The SOC stock estimated using the GEFSOC Modelling System is higher than the stock estimated using the mapping approach. This is due to the fact that while the GEFSOC System accounts for variation in crop input data (crop management), the soil mapping approach only considers regional variation in soil texture and wetness. The trend of overall change in the modelled SOC stock estimates shows that the IGP, India may have reached an equilibrium following 30–40 years of the Green Revolution. This can be seen in the SOC stock change rates. Various different estimation methods show SOC stocks of 0.57–1.44 Pg C for the study area. The trend of overall change in C stock assessed from the soil survey data indicates that the soils of the IGP, India may store a projected 1.1 Pg of C in 2030.

Soil organic carbon (SOC) is vital for ecosystem and agro-ecosystem function. Any sustainable land management strategy should, therefore, include a consideration of long-term effects on SOC. In the future, we have the opportunity to adopt land management strategies that lead to greater C storage in the soil. However, to do so, we need consistent estimates of SOC stocks and changes under varying land use and climate change scenarios. A Global Environment Facility (GEF) project developed a generically applicable system (the GEFSOC Modelling System) for making such estimates. The system links two dynamic SOC models, designed for site scale applications (Century and RothC) and an empirical method, to spatial databases, giving spatially explicit results that allow geographic areas of change in SOC stocks to be identified. The system was developed using data from four contrasting eco-regions (The Brazilian Amazon, Jordan, Kenya and the Indian part of the Indo-Gangetic Plains). These areas were chosen, as they are located in regions previously underrepresented by soil C models. The system was then used to estimates SOC stocks and changes between 1990 and 2030 under likely land use change scenarios in each of the four regions. Losses in SOC of between 5 and 16 % were projected for each of the four areas over a 30-year period (2000-2030), driven by a range of factors including deforestation, overgrazing and conversion of grazing land to agriculture. Implications for sustainable land management and future land use policy are discussed for The Brazilian Amazon, Jordan, Kenya and the Indian Indo-Gangetic Plains.

RothC and Century are two of the most widely used soil organic matter (SOM) models. However there are few examples of specific parameterisation of these models for environmental conditions in East Africa. The aim of this study was therefore, to evaluate the ability of RothC and the Century to estimate changes in soil organic carbon (SOC) resulting from varying land use/management practices for the climate and soil conditions found in Kenya. The study used climate, soils and crop data from a long term experiment (1976–2001) carried out at The Kabete site at The Kenya National Agricultural Research Laboratories (NARL, located in a semi-humid region) and data from a 13 year experiment carried out in Machang’a (Embu District, located in a semi-arid region). The NARL experiment included various fertiliser (0, 60 and 120kg of N and P2O5ha−1), farmyard manure (FYM—5 and 10tha−1) and plant residue treatments, in a variety of combinations. The Machang’a experiment involved a fertiliser (51kgNha−1) and a FYM (0, 5 and 10tha−1) treatment with both monocropping and intercropping. At Kabete both models showed a fair to good fit to measured data, although Century simulations for treatments with high levels of FYM were better than those without. At the Machang’a site with monocrops, both models showed a fair to good fit to measured data for all treatments. However, the fit of both models (especially RothC) to measured data for intercropping treatments at Machang’a was much poorer. Further model development for intercrop systems is recommended. Both models can be useful tools in soil C predictions, provided time series of measured soil C and crop production data are available for validating model performance against local or regional agricultural crops.

Under the United Nations Framework Convention on Climate Change (UNFCCC), Non-Annex 1 countries such as Kenya are obliged to report green house gas (GHG) emissions from all sources where possible, including those from soils as a result of changes in land use or land management. At present, the convention encourages countries to estimate emissions using the most advanced methods possible, given the country circumstances and resources. Estimates of soil organic carbon (SOC) stocks and changes were made for Kenya using the Global Environment Facility Soil Organic Carbon (GEFSOC) Modelling System. The tool conducts analysis using three methods: (1) the Century general ecosystem model; (2) the RothC soil C decomposition model; and (3) the Intergovernmental Panel on Climate Change (IPCC) method for assessing soil C at regional scales. The required datasets included: land use history, monthly mean precipitation, monthly mean minimum and maximum temperatures for all the agro-climatic zones of Kenya and historical vegetation cover. Soil C stocks of 1.4–2.0 Pg (0–20 cm), compared well with a Soil and Terrain (SOTER) based approach that estimated ∼1.8–2.0 Pg (0–30 cm). In 1990 48% of the country had SOC stocks of <18 t C ha−1 and 20% of the country had SOC stocks of 18–30 t C ha−1, whereas in 2000 56% of the country had SOC stocks of <18 t C ha−1 and 31% of the country had SOC stocks of 18–30 t C ha−1. Conversion of natural vegetation to annual crops led to the greatest soil C losses. Simulations suggest that soil C losses remain substantial throughout the modelling period of 1990–2030. All three methods involved in the GEFSOC System estimated that there would be a net loss of soil C between 2000 and 2030 in Kenya. The decline was more marked with RothC than with Century or the IPCC method. In non-hydric soils the SOC change rates were more pronounced in high sandy soils compared to high clay soils in most land use systems.

Estimates of soil organic carbon (SOC) stocks and changes under different land use systems can help determine vulnerability to land degradation. Such information is important for countries in arid areas with high susceptibility to desertification. SOC stocks, and predicted changes between 2000 and 2030, were determined at the national scale for Jordan using The Global Environment Facility Soil Organic Carbon (GEFSOC) Modelling System. For the purpose of this study, Jordan was divided into three natural regions (The Jordan Valley, the Uplands and the Badia) and three developmental regions (North, Middle and South). Based on this division, Jordan was divided into five zones (based on the dominant land use): the Jordan Valley, the North Uplands, the Middle Uplands, the South Uplands and the Badia. This information was merged using GIS, along with a map of rainfall isohyets, to produce a map with 498 polygons. Each of these was given a unique ID, a land management unit identifier and was characterized in terms of its dominant soil type. Historical land use data, current land use and future land use change scenarios were also assembled, forming major inputs of the modelling system.

Currently we have little understanding of the impacts of land use change on soil C stocks in the Brazilian Amazon. Such information is needed to determine impacts on the global C cycle and the sustainability of agricultural systems that are replacing native forest. The aim of this study was to predict soil carbon stocks and changes in the Brazilian Amazon during the period between 2000 and 2030, using the GEFSOC soil carbon (C) modelling system. In order to do so, we devised current and future land use scenarios for the Brazilian Amazon, taking into account: (i) deforestation rates from the past three decades, (ii) census data on land use from 1940 to 2000, including the expansion and intensification of agriculture in the region, (iii) available information on management practices, primarily related to well managed pasture versus degraded pasture and conventional systems versus no-tillage systems for soybean (Glycine max) and (iv) FAO predictions on agricultural land use and land use changes for the years 2015 and 2030. The land use scenarios were integrated with spatially explicit soils data (SOTER database), climate, potential natural vegetation and land management units using the recently developed GEFSOC soil C modelling system. Results are presented in map, table and graph form for the entire Brazilian Amazon for the current situation (1990 and 2000) and the future (2015 and 2030). Results include soil organic C (SOC) stocks and SOC stock change rates estimated by three methods: (i) the Century ecosystem model, (ii) the Rothamsted C model and (iii) the intergovernmental panel on climate change (IPCC) method for assessing soil C at regional scale. In addition, we show estimated values of above and belowground biomass for native vegetation, pasture and soybean. The results on regional SOC stocks compare reasonably well with those based on mapping approaches. The GEFSOC system provided a means of efficiently handling complex interactions among biotic-edapho-climatic conditions (>363,000 combinations) in a very large area (∼500Mha) such as the Brazilian Amazon. All of the methods used showed a decline in SOC stock for the period studied; Century and RothC simulated values for 2030 being about 7% lower than those in 1990. Values from Century and RothC (30,430 and 25,000Tg for the 0–20cm layer for the Brazilian Amazon region were higher than those obtained from the IPCC system (23,400Tg in the 0–30cm layer). Finally, our results can help understand the major biogeochemical cycles that influence soil fertility and help devise management strategies that enhance the sustainability of these areas and thus slow further deforestation.