The distributions of calcite rich horizons within dryland soils are commonly used as paleoclimate proxies. Comprehensive conceptual and mathematical models of calcite accumulation in soils are required to accurately interpret and calibrate these proxies. A conceptual model for calcite accumulation is already well established: As water percolates through a soil, it dissolves minerals, such as calcite, transporting the soluble minerals downward. As soil water is removed by evaporation and transpiration, the water solution becomes supersaturated resulting in precipitation of calcite at depth. The impacts of dynamic plant growth and microbial respiration have not yet been simulated in numerical models for calcite accumulation but are likely important because of their influence on variables governing calcite solubility. The soil hydraulic modeling software, HYDRUS-1D, simulates water and solute transfer through a soil column, accounting for variations in all previously studied variables (temperature, water content, soil pCO₂) while additionally simulating vegetation-soil interactions. Five separate sensitivity studies were conducted to determine the importance for calcite accumulation of 1) soil texture, 2) plant growth, 3) plant phenology, 4) atmospheric CO₂ concentrations, and 5) the proximal variables that control calcite dissolution and precipitation: soil CO₂, soil water content, and soil temperature. In each modeling simulation, calcite was leached from the top several cm and redistributed deeper in the soil after 20 years. Soils with courser texture yield deeper (+20cm), more diffuse calcite horizons, as did simulations with bare soil compared to vegetated soil. The phenology of plant communities (late spring versus late summer growth) resulted in soil calcite accumulation at temperatures differing by at least 10°C. Changes in atmospheric CO₂ concentrations do not affect the soil calcite distribution. Variations in concentration of soil CO₂, rather than soil water content, have the greatest direct effect on calcite solubility. The most significant time periods of annual accumulation also corresponded with positive water fluxes resulting from high matric potential at the surface. Transpiration and evaporation moisture sinks caused solution to travel upward from higher to lower soil CO₂ concentrations, causing CO₂ de-gassing and calcite accumulation. This pathway describes a new qualitative mechanism for soil calcite formation and should be included in the conceptual model. / text

Woodlot and natural woodland systems in the semi-arid regions in Tanzania are believed to have a high potential to sequester carbon (C) in their biomass and the soil which may qualify for C credits under the current voluntary C market schemes like, the REDD program. However, our understanding of the processes influencing storage and dynamics of C in soils under semi-arid agroforestry systems such as these woodlot systems is limited. This study evaluated C pools in soil and tree biomass in woodlot species of Albizia lebbeck, Leucaena leucocephala, Melia azedarach, and Gmelina arborea; and in farmland and ngitili systems. Synchrotron-based C K-edge x-ray absorption near-edge structure (XANES) spectroscopy was also used to study the influence of these land use systems on the soil organic matter (SOM) chemistry to understand the mechanisms of soil C changes. Soil samples were collected to 1 m depth and subsamples for each land use system to 0.4 m depth were fractionated into macroaggregates (2000-250 μm), microaggregates (250-53 μm), and silt and clay-sized aggregates (<53 μm) to provide information of C dynamics and stabilization in various land uses. SOC was analyzed in whole and soil aggregates and biomass C was estimated using developed biomass models from the literatures. Aboveground biomass carbon in the woodlots from the Kahama district ranged from 11.76 Mg C ha-1 to 24.40 Mg C ha-1. Based on the age of woodlots and the rate of carbon sequestration potential (CSP), Gmelina arborea had the highest rate of aboveground C sequestration (3.59 Mg C ha-1 year-1). The SOC stocks in whole soil for the land use systems from the two districts ranged from 43-67 Mg C ha-1. The degraded ngitili did not show a reduction in SOC stocks despite reducing aboveground biomass C stocks by 15.11 Mg C ha-1. SOC in the woodlots were found to be associated more with the micro and silt-and clay-sized aggregates than the macroaggregates, reflecting high stability of SOC in the woodlot systems. The XANES C K-edge spectra revealed the stabilization of recalcitrant aromatic C compounds in the silt and clay-sized aggregates. This study demonstrates the significant contributions of woodlots in biomass C accumulation as well as long-term SOC stabilization in soil fractions. Thus, these agroforestry practices hold promise to meet household energy needs while contributing to climate change mitigation and adaption.

The critical zone is the interface between abiotic and biotic constituents that spans from the vegetation canopy through the groundwater and represents an open system shaped by the climate, topography, and vegetation communities of a given environment. Four studies were completed to examine soil development, specifically mineral weathering and soil carbon storage, across semiarid sites spanning the Santa Catalina Mountain Critical Zone Observatory (SCM-CZO). The Santa Catalina Mountain Critical Zone Observatory is located along an environmental gradient in southern Arizona where co-varying climate and vegetation community properties have generated distinct changes in soil development across a relatively short distance (<20 miles). Soil, saprock, and parent rock were sampled on north-facing slopes from five climate-vegetation zones spanning desert scrub to mixed conifer forest. Within each climate-vegetation zone, samples were collected from two divergent summit and two convergent footslope landscape positions to account for topographic controls on mineral transformation. In the first study, the soil morphologic, physical, and chemical properties collected for all samples were combined with profile development indices to quantify soil variation with landscape position across the SCM-CZO. The results of this research demonstrated that climate and landscape position exert significant control on soil development in semiarid ecosystems, and that the profile development index is an effective tool for detecting these regional to hillslope scale variations in soil properties. The second study consisted of a cross-scale analysis of feldspar mineral transformation across the selected research sites to connect measures of pedon-scale soil development, depletions of feldspar and sodium in bulk soil, and elemental losses across feldspar grains at the microscale. Results indicated that greater soil development in the mixed conifer pedons corresponded to increased total feldspar and sodium losses. Desert scrub soils presented less evidence for feldspar transformation including lower profile development indices, gains in total feldspar percentages attributed to dust deposition, and less Na chemical depletion at the microscale. Greater soil development in convergent positions relative to adjacent divergent sites was consistent across all sites, with the highest degree of total feldspar depletion occurring in the conifer convergent locations. The third study focused on the physical distribution and mean residence time of soil organic carbon (SOC) in the SCM-CZO soils described for the first two studies. Surface (0-10 cm) and subsurface (30-40 cm) samples were collected from the aforementioned granitic regolith profiles. The soils were characterized using total C and N, δ¹³C, Δ¹⁴C, and radiocarbon derived mean residence time (MRT) estimates of bulk soil and physically separated C fractions to quantify SOC change with climate, vegetation, and landscape position. The results document a shift in SOC stabilization mechanisms across bioclimatically distinct ecosystems from mineral-associated SOC in the desert scrub soils to a mixture of mineral and occluded SOC in the conifer soils. Soils in the convergent landscapes concentrated the most SOC and typically exhibited the longest residence times across all locations. The fourth study examined the geochemical and mineralogical properties of the SCM-CZO soils across regional and hillslope scales of study to quantify soil development in semiarid environments. X-ray fluorescence and x-ray diffraction were used to characterize the elemental and mineralogical properties of the soils and parent material. Desert scrub dust samples were analyzed using x-ray fluorescence. The results indicate that mineral and base cation depletion were greatest in the convergent landscape positions at both sites and increased from the hot, moisture-limited desert scrub sites to the wetter, more productive conifer ecosystems. Enrichments in mica and select elements (i.e., Fe, Mg) suggested that dust deposition was a significant contributor to soil development across all sites. Geochemical estimates of dust fraction inputs confirmed this finding with dust composing up to 35% of the regolith material in the mixed conifer convergent soils. Clay mineral assemblage was dominated by halloysite and smectite minerals in the desert scrub site, reflecting complex climatic and mineral microtextural interactions in the dry, silica-rich desert environment. Clay minerals at the mixed conifer site exhibited the greatest degree of mineral transformation in the SCM, consisting of vermiculite, illite, kaolinite, and minor amounts of smectite and gibbsite. These findings confirm the interactive role of climate, vegetation, and landscape position in shaping the critical zone, where greater moisture availability and biological production are likely driving increased soil organic carbon storage and mineral weathering across various scales of study.

Tall fescue (Lolium arundinaceum (Schreb.) Darbysh.) is a cool-season perennial grass used in pastures throughout the Southeastern United States. The grass can harbor a fungal endophyte (Epichloë coenophiala) thought to provide the plant with enhanced resistance to biotic and abiotic stress. However, the alkaloids produced by the common variety of the endophyte cause severe animal health issues resulting in a considerable amount of research focused on eliminating the toxic class of alkaloids while retaining the positive abiotic and biotic stress tolerance attributes of the other alkaloids. In doing so, very little attention has been paid to the direct influence the fungal-plant symbiosis has on rhizosphere processes. Therefore, my objectives were to study the influence of this relationship on plant biomass production, root exudate composition, and soil biogeochemical processes using tall fescue cultivars PDF and 97TF1 without an endophyte (E-), or infected with the common toxic endophyte (CTE+), or with two novel endophytes (AR542E+, AR584E+). I found that root exudate composition and plant biomass production were influenced by endophyte status, tall fescue cultivar, and the interaction of cultivar and endophyte. Cluster analysis showed that the interaction between endophyte and cultivar results in a unique exudate profile. These interactions had a small but perceptible impact on soil microbial community structure and function with an equally small and perceptible impact on carbon and nitrogen cycling in soils from rhizobox and field sites. These studies represent the first comprehensive analysis of root exudate chemistry from common toxic and novel endophyte infected tall fescue cultivars and can be used to help explain in part the observed changes in C and N cycling and storage in pastures throughout the Southeast U.S..

Forested riparian zones, which function as a buffer between agricultural fields and streams, filter out contaminants and sediment from the fields thereby improving water quality, cool the water with shade from trees, stabilize the stream bank and provide habitat for wildlife. However, in many agricultural areas, riparian vegetation has been removed for crop production or pasture purposes. Riparian restoration or rehabilitation is a way of restoring riparian ecosystem functions. This study examines the effect of riparian rehabilitation via tree planting along a first-order creek in Southern Ontario, 25 years after rehabilitation. Litter input, soil quality parameters and soil CO2 production rates were determined for the rehabilitated riparian zone, a grass-forb riparian zone and a natural forest riparian zone. Total litter input was 480, 580 and 295 g m-2 y-1 for the rehabilitated riparian zone, grass riparian zone and forest riparian zone, respectively. Soil bulk density was higher and hydraulic conductivity was lower for the rehabilitated riparian zone compared to the grass riparian zone and forest riparian zone. The concentration and soil stock of organic carbon and total nitrogen was lowest for the rehabilitated riparian zone compared to the grass riparian zone and forest riparian zone which were similar. The effect of riparian zone on soil CO2 production rates varied over the season. From spring to mid-summer, rates were 167, 224 and 104 mg C m-2 h-1 for the rehabilitated riparian zone, grass riparian zone and forest riparian zone, respectively. Soil CO2 production rates did not differ significantly (p < 0.05) between riparian zones for late summer and fall sampling dates. Soil CO2 production rates were significantly negatively correlated with soil C/N and positively correlated with soil pH and litter input. Soil CO2 production rates were positively correlated with soil temperature (r = 0.32) and negatively correlated with soil moisture (r = -0.48). Of the three riparian zones, the natural forest riparian zone exhibited the least amount of seasonal fluctuation for soil CO2 production rates, soil moisture and temperature. Results from this research indicated that more time is needed before soil quality and soil CO2 production rates of the rehabilitated riparian zone reach values similar to the natural forest riparian zone.

Upland grasslands comprise a large proportion of the UK’s land area and are primarily used to graze sheep. These grasslands store large quantities of carbon (C). Changes in land use or climate could affect the ability of these soils to store C and the fluxes of other greenhouse gases associated with agricultural soils, nitrous oxide (N2O) and methane (CH4). Grazing substantially changes the cycling of C and nitrogen in grassland ecosystems, particularly through the deposition of rapidly degrading excreta, both dung and urine, on the soil. The major non-enteric greenhouse gas emissions associated with this type of extensive farming of ruminants are the emission of N2O and CH4 from soils affected by the animal’s excreta. This PhD project has investigated the cycling of sheep dung in two upland soils of different management regimes to investigate the effects imposed by the plant community. Dung incorporation was measured by capitalising on the natural difference in natural 13C abundance (δ13C ratios) between maize and native British vegetation, which permitted maize-derived sheep dung to be used as a 13C tracer of dung incorporation into soil. A physical and chemical soil fractionation methodology was used to isolate the distinct soil organic carbon (SOC) pools and ascertain the location of the dung C. There were differences between soils in dung C cycling, with more dung C being measured in semi-improved soils at experiment’s end. Throughout the one year timeframe of this experiment, most of the dung C was recovered in the particulate organic matter fraction. Changing the plant community did not have a measurable effect on dung C cycling within the experimental period. Urine patches in grazed pastures represent a major source of agriculture’s N2O emissions. The N2O, CH4 and CO2 fluxes from chambers treated with synthetic urine, synthetic urine and dung, or dung, and an untreated control in randomised block design at two sites were measured over one year. Relevant soil parameters were also measured at each sampling point. From this data N2O emission factors for sheep excreta at these sites were calculated. N2O emission factors were significantly different between sites, were different for dung and urine, and in all cases were less than the current default value used by countries utilising a Tier 1 methodology, according to the IPCC, to inventory N2O emissions derived from grazing livestock. Dietary manipulation has been proposed to increase certain components in urine that are thought to inhibit N2O emission with the aim of reducing livestock greenhouse gas emissions. One such urinary component is hippuric acid. Soil to which synthetic urine with incrementally increased quantities of hippuric acid were added were incubated, as were soils to which dung only and dung and synthetic urine had been added, as well as an untreated control. No significant effect of hippuric acid concentration was observed. N2O emissions from the dung only and dung and urine treatments were unusually high and surpassed those of the urine only treatments. This has been hypothesised to be due to fungal denitrification in the dung treatments or suppression of microbial activity due to ammonia toxicity in the urine-treated soils. The key conclusions from this PhD work are that the effect of dung deposition on SOC cycling may be quite small and appears to result in substitution of native SOC with dung C, rather than an increase in SOC; N2O emissions from sheep dung and urine deposition in semi-improved grasslands is likely to be very low and much lower than the current IPCC default value; and that in our incubation experiment there was no discernible impact of hippuric acid on N2O emissions, but it is possible that this is an experimental artefact.

Soil organic matter (SOM) decomposition and formation is an important climate feedback, with the potential to amplify or offset climate forcing. To understand the fate of soil carbon (C) stores and fluxes (i.e., soil respiration) under future climate it is necessary to investigate responses across spatial and temporal scales, from the ecosystem to the molecular level, from diurnal to decadal trends. Moreover, it is important to question the assumptions and paradigms that underlie apparently paradoxical evidence to reveal the true nature of soil-climate feedbacks. My dissertation includes research into the response of soil respiration in Pacific Northwest prairies to warming and wetting along a natural regional climate gradient (Chapter II), and then delves deeper into the mechanisms underlying SOM decomposition and formation, examining the temperature sensitivity of SOM decomposition of prairie soils that were experimentally warmed for ~2 yr, and a forest soil in which litter-inputs were manipulation for 20 yr (Chapter III), and finally testing soil C cycling dynamics, including mineral-associated C pools, decomposition dynamics, and the molecular nature of SOM itself, under litter-manipulation in order to understand the controls on SOM formation and mineralization (Chapter IV). This dissertation includes previously published and unpublished coauthored material; see the individual chapters for a list of co-authors, and description of contributions.

Doctor of Philosophy / Agronomy / Charles W. Rice / Substitution of cellulosic biofuel in place of gasoline or diesel could reduce greenhouse gas (GHG) emissions from transportation. However, emissions of nitrous oxide (N₂O) and changes in soil organic carbon (SOC) could have a large impact on the GHG balance of cellulosic biofuel, thus there is a need to quantify these responses in cellulosic biofuel crops. The objectives of this study were to: (i) measure changes in yield, SOC and microbial communities in potential cellulosic biofuel cropping systems (ii) measure and characterize the temporal variation in N₂O emissions from these systems (iii) characterize the yield and N₂O response of switchgrass to N fertilizer and to estimate the costs of production. Sweet sorghum, photoperiod-sensitive sorghum, and miscanthus yielded the highest aboveground biomass (20-32 Mg ha⁻¹). The perennial grasses sequestered SOC over 4 yrs, while SOC stocks did not change in the annual crops. Root stocks were 4-8 times higher in the perennial crops, suggesting greater belowground C inputs. Arbuscular mycorrhizal fungi (AMF) abundance and aggregate mean weight diameter were higher in the perennials. No consistent significant differences were found in N₂O emissions between crops, though miscanthus tended to have the lowest emissions. Most N₂O was emitted during large events of short duration (1-3 days) that occurred after high rainfall events with high soil NO₃₋. There was a weak relationship between IPCC Tier 1 N₂O estimates and measured emissions, and the IPCC method tended to underestimate emissions. The response of N₂O to N rate was nonlinear in 2 of 3 years. Fertilizer induced emission factor (EF) increased from 0.7% at 50 kg N ha⁻¹ to 2.6% at 150 kg N ha⁻¹. Switchgrass yields increased with N inputs up to 100-150 kg N ha⁻¹, but the critical N level for maximum yields decreased each year, suggesting N was being applied in excess at higher N rates. Yield-scaled costs of production were minimized at 100 kg N ha-1 ($70.91 Mg⁻¹). Together, these results show that crop selection and fertilizer management can have large impacts on the productivity and soil GHG emissions biofuel cropping systems.

Os objetivos do presente trabalho foram avaliar o potencial de acúmulo de carbono (C) em solos de cerrado sob diferentes usos florestais (Eucalyptus, Pinus versus vegetação natural) e as possíveis alterações ocorridas na ciclagem de nutrientes sob esse tipo de coberturas. Para isso, foram coletadas 30 amostras de solo (0-5, 10-25 e 35-50 cm) e de serrapilheira acumulada em plantios de Eucalyptus, Pinus e remanescentes de cerradão em quatro municípios do estado de SP, nas quais se determinou: pH, matéria orgânica (MO), C, macronutrientes e densidade, além da granulometria no solo. Os estoques de C do solo foram calculados através do ajuste e integração de equações exponenciais, obtendo-se valores entre 3,4 e 8,6 kgC.m-2.(na camada de 0 a 30 cm) e entre 5,7 e 11,3 kgC.m-2.(até 1m). Os resultados mostraram que a silvicultura de Eucalyptus e Pinus afeta o acúmulo de C e a ciclagem de nutrientes em áreas de cerrado. As alterações nos estoques de C ocorreram principalmente nos horizontes orgânicos e na camada superficial do solo em decorrência da substituição da MO original por outra de pior qualidade química. A influência do tipo de vegetação sobre o C da camada superficial do solo variou em função de características do sítio, verificando ganhos em alguns dos locais estudados, perdas em outros e ainda diferenças não significativas. Em profundidades maiores, o conteúdo de C mostrou-se fortemente relacionado com o teor de argila e diminuiu sob cultura de Eucalyptus e Pinus, sendo mais fortes as depleções sob Pinus. Nos plantios houve formação de horizontes orgânicos espessos, com concentrações de C elevadas. A concentração de nitrogênio (N), cálcio, magnésio e potássio do material aí acumulado foi menor do que nas áreas naturais, enquanto que a acidez e a relação C/N foram maiores. Isso pode inibir a decomposição, o que explicaria o maior armazenamento de C na serapilheira. A incorporação desse material ao solo implica em alterações da MO, que é um dos principais fatores de estruturação e fertilidade dos solos tropicais e foi afetada em quantidade e qualidade. As relações C/N quantificadas na camada superficial do solo foram significativamente maiores do que em áreas de vegetação natural, indicando substituição da MO nessa camada no tempo de vida dos plantios (~40 anos). Os resultados demonstram a ocorrência de alterações na qualidade química da MO na serapilheira e no solo superficial sob uso silvicultural, as quais podem originar maiores estoques e tempos de residência do C, mas também diminuições de recursos tróficos para a comunidade decompositora, com implicações no resto do ecossistema. Os resultados sugerem que a dinâmica do carbono do solo varia ao longo do perfil, sendo necessário esclarecer melhor os fatores que definem o carbono da camada superficial, maior em quantidade e mais sensível aos efeitos do manejo. / Effects on soil organic carbon storage potential and possible biogeochemical changes of established forest plantations were assessed in southeast Brazil, in Eucalyptus and Pinus plantations compared with natural areas of native dry forest (cerradão). 30 plots were randomly distributed for soil (0-5, 10-25 and 35-50 cm) and forest floor litter collection in mature plantations (~40 years old) and adjacent native forest. The design was replicated in 4 localities in Sao Paulo, southeast Brazil. Organic matter, organic carbon, macro nutrients, pH, density, and soil texture were determined. Soil organic carbon stocks were calculated through exponential equations adjustment and integration, values ranged from 3,4 to 8,6 kgC.m-2.(on the 0 and 30 cm layer) and from 5,7 to 11,3 kgC.m-2.(up to 1m). Soil organic carbon and biogeochemical features were affected under Eucalyptus and Pinus plantations. Changes in carbon stocks were stronger in organic layers and topsoil, due to the replacement of the original organic matter, causing chemical quality decrease. Effects of vegetation on topsoil organic carbon were site dependent, as plantations results in gains, losses and no remarkable differences between natural and forested areas. Eucalyptus and Pinus establishment led to organic carbon losses, which possibly conducted by soil disturbances at implementation of plantations. Carbon content was strongly related to clay at deeper layers, but not at shallower ones. Thicker organic layers with higher carbon content were found under plantations. Lower calcium, magnesium and potassium concentrations and higher acidity and C/N (carbon to nitrogen) ratios were measured at implanted forest floor litter layers. These changes could inhibit decomposition, explaining larger litter carbon storage. Soil organic matter is an important factor in maintaining tropical soil structure and fertility. It was affected by Eucalyptus and Pinus forestation. While litter organic mater is incorporated into the soil, it leads to soil organic matter chemical quality decreases. The topsoil C/N ratio measured in plantations was significantly higher than in natural vegetation areas, indicating organic matter replacement on this layer in plantations lifetime (~40 years). We concluded that litter and topsoil organic matter chemical properties were affected by forestation with Eucalyptus and Pinus, which could result in larger C stocks and residence times, but could also decrease trophic resources for decomposers, with implications on the whole ecosystem. The results suggest that soil carbon dynamics changes along the soil profile. Factors controlling surface soil carbon dynamics must be clarified further, as they contained high carbon amounts, the most sensible to management practices.

Malgré l'importance de la séquestration du carbone (C) dans les sols forestiers, il existe actuellement peu de projets Mécanisme pour un Développement Propore (MDP) qui tiennent compte de la gestion de ce compartiment dans l'atténuation du réchauffement climatique. La cause principale est que l'accréditation du carbone séquestré dans le sol représente de plus grands défis et des risques par rapport aux autres composantes des écosystèmes forestiers.Connaissant les difficultés économiques et les problèmes environnementaux impliqués dans l'adoption d'un tel projet et l'importance des forêts dans l'atténuation du changement climatique, l'objectif de cette étude était d'évaluer les performances des méthodes dedétermination des stocks de carbone du sol dans deux projets de reboisement, ainsi que pour leurs situations de références (c'est-à-dire utilisation des terres avant la plantation, des pâturages naturels, ainsi que la végétation native) comme base pour diminuer le rapport coût bénéfice des projets MDP dans le secteur forestier. Pour atteindre l'objectif principal, cestravaux de recherches ont consisté à l'étude de: (i) la variabilité spatiale du C du sol dans une zone de reboisement avec des espèces indigènes, établi dans le bassin amazonien dans le MatoGrosso à Cotriguaçu (Zone I) et une chronoséquence de plantations d'Eucalyptus, située dansl'état de São Paulo à Avare (zone II); (ii) la taille des parcelles et la distance idéale de séparation des échantillons à partir de l'étude de la dépendance spatiale du C, (iii)l'estimation de la teneur en C et la densité du sol (Ds) par spectroscopie dans le proche(NIRS) et moyen (MIRS) infrarouge, afin de réduire les coûts sans affecter la qualité des résultats analytiques : et (iv) du calcul des stocks de C du sol pour ces deux projets et l'estimation du bilan C du projet MDP menées dans la zone II, en utilisant l'outil EX-ACT («EX-Ante Carbon-balance Tool»). Les résultats ont confirmé l'existence d'une variabilité spatiale importante du C du sol, ainsi qu'une forte dépendance spatiale pour tous les traitements étudiés. L'analyse du nombre optimal d'échantillons de sol a montré que la collecte de cinq points par parcelle est aussi précise qu'un échantillonnage plus dense. La taille optimale des parcelles a été estimée de 361 à 841 m2 dans les plantations de la zone I et de 900 à 3721 m2 pour la Zone II. La performance de la spectroscopie MIRS et NIRS pourestimer la teneur en C des sols a été jugée très satisfaisante, surtout quand les modèles ont été étalonnés à partir de sous populations constituées de 10 à 50% de l'ensemble des données.Les résultats de l'estimation de Ds ont été légèrement moins satisfaisants que ceux pour le Cdu sol. Les stocks de C dans le sol calculés pour la zone I étaient supérieurs à ceux de la zone II. Si on ne considère que le compartiment du sol, il apparaît que le potentiel de génération de crédits C est plus important dans le reboisement à partir d'espèces indigènes dans un sol argileux que pour un reboisement d'eucalyptus dans un sol sableux. Le bilan C du projet de la zone I a montré un potentiel de séquestration de près de trois millions de tonnes d'équivalentCO2 en 40 ans. Il est espéré que cette étude qui montre des possibilités de réduction des coûts liés aux calculs des stocks de C du sol contribuera à une meilleure prise en compte de ce compartiment dans les projets MDP forestiers. / Considering the great importance of carbon sequestration (C) in forest soils, there are fewCDM projects that include this compartment as an agent of global warming mitigation. Thisoccurs because the quantification of soil C stocks represents a bigger challenge whencompared to other components of forest ecosystems. Considering the economic difficultiesand environmental issues involved in adopting this type of project and the importance offorests in mitigating climate change, the objective of this study was to evaluate theperformance of methods for obtaining soil C stocks in two forestry areas and their respectivebaselines (land use prior to planting, i.e. pastures and native vegetation) as a basis forreducing the cost-benefit ratio of CDM projects. To achieve the main objective, this researchwas composed of the following steps: (i) estimating the spatial variability of soil C in an areareforested with native species, established in Cotriguaçú, MT (Area I) and a Eucalyptuschronosequence, located in Avaré, SP (Area II), (ii) determining the optimal amount of soilsamples and the plot size from the soil C spatial dependence range in the reforestation areas,(iii) estimating soil C content and bulk density (BD) by Near and Mid Infrared ReflectanceSpectroscopy (NIRS and MIRS, respectively) to reduce analytical costs without affecting thequality of the results, and (iv) calculating soil C stocks in both areas and estimating the carbonbalance of a CDM Project conducted in Area II, using EX-ACT ("Ex-Ante Carbon BalanceTool"). The results confirmed the existence of significant soil C spatial variability in bothareas and a strong spatial dependence at all plots. The analysis of the optimal number of soilsamples indicated that the sampling procedure with five points per plot is as accurate asintensive sampling. The optimum size of plots ranged from 361-841 m2 at Area I plantationsand from 900-3721 m2 at Area II. The performance of MIRS and NIRS to estimate the soilcarbon content was very satisfactory, especially when the models were calibrated withamounts between 5-10% of the total data set. The estimations of BD were slightly less precisethan those of soil C content. The soil C stocks obtained at Area I were higher than Area II.Considering only the soil compartment, it is clear that the potential for C credit generation in areforestation with native species on a clayey soil is higher than in a reforestation witheucalyptus on a sandy soil. The C balance of the CDM project conducted in Area I is expectedto sequester almost three million tones of CO2 eq in 40 years. We hope this study contributesto the increased inclusion of soil in CDM projects, by confirming the feasibility of reducingthe costs associated with both sampling and analytical procedures.