Dissertations / Theses on the topic 'Soil Organic Carbon model calibration'

The evaluation of the soil spatial variability using a fast, robust and cheap tool is one of the key steps towards the implementation of Precision Agriculture (PA) successfully. Soil organic carbon (OC), soil total nitrogen (TN) and soil moisture content (MC) are needed to be monitored for both agriculture and environmental applications. The literature has proven that visible and near infrared (vis-NIR) spectroscopy to be a quick, cheap and robust tool to acquire information about key soil properties simultaneously with relatively high accuracy. The on-line vis-NIR measurement accuracy depends largely on the quality of calibration models. In order to establish robust calibration models for OC, TN and MC valid for few selected European farms, several factors affecting model accuracy have been studied. Nonlinear calibration techniques, e.g. artificial neural network (ANN) combined with partial least squares regression (PLSR) has provided better calibration accuracy than the linear PLSR or principal component regression analysis (PCR) alone. It was also found that effects of sample concentration statistics, including the range or standard derivation and the number of samples used for model calibration are substantial, which should be taking into account carefully. Soil MC, texture and their interaction effects are other principle factors affecting the in situ and on-line vis-NIR measurement accuracy. This study confirmed that MC is the main negative effect, whereas soil clay content plays a positive role. The general calibration models developed for soil OC, TN and MC for farms in European were validated using a previously developed vis-NIR on-line measurement system equipped with a wider vis-NIR spectrophotometer (305 – 2200 nm) than the previous version. The validation results showed this wider range on-line vis-NIR system can acquire larger than 1500 data point per ha with a very good measurement accuracy for TN and OC and excellent accuracy for MC. The validation also showed that spiking few target field samples into the general calibration models is an effective and efficient approach for upgrading the implementation of the on-line vis-NIR sensor for measurement in new fields in the selected European farms.

The fate of the carbon and nitrogen in dairy farm effluent (DFE) applied onto soil was investigated through laboratory experiments and field lysimeter studies. They resulted in the development and testing of a complex carbon (C) and nitrogen (N) simulation model (CaNS-Eff) of the soil-plant-microbial system. To minimise the risk of contamination of surface waters, regulatory authorities in New Zealand promote irrigation onto land as the preferred treatment method for DFE. The allowable annual loading rates for DFE, as defined in statutory regional plans are based on annual N balance calculations, comparing N inputs to outputs from the farming system. Little information is available, however, to assess the effects that these loading rates have on the receiving environment. It is this need, to understand the fate of land-applied DFE and develop a tool to describe the process, that is addressed in this research. The microbially mediated net N mineralisation from DFE takes a central role in the turnover of DFE, as the total N in DFE is dominated by organic N. In a laboratory experiment, where DFE was applied at the standard farm loading rate of 68 kg N ha⁻¹, the net C mineralisation from the DFE was finished 13 days after application and represented 30% of the applied C, with no net N mineralisation being measured by Day 113. The soluble fraction of DFE appeared to have a microbial availability similar to that of glucose. The low and gradually changing respiration rate measured from DFE indicated a semi-continuous substrate supply to the microbial biomass, reflecting the complex nature and broad range of C compounds in DFE. The repeated application of DFE will gradually enhance the mineralisable fraction of the total soil organic N and in the long term increase net N mineralisation. To address the lack of data on the fate of faecal-N in DFE, a ¹⁵N-labelled faecal component of DFE was applied under two different water treatments onto intact soil cores with pasture growing on them. At the end of 255 days, approximately 2% of the applied faecal ¹⁵N had been leached, 11 % was in plant material, 11 % was still as effluent on the surface, and 40% remained in the soil (39% as organic N). Unmeasured gaseous losses and physical losses from the soil surface of the cores supposedly account for the remaining ¹⁵N (approximately 36%). Separate analysis of the total and ammonium nitrogen contents and ¹⁵N enrichments of the DFE and filtered sub-samples (0.5 mm, 0.2µm) showed that the faecal-N fraction was not labelled homogeneously. Due to this heterogeneity, which was exacerbated by the filtration of DFE on the soil surface, it was difficult to calculate the turnover of the total faecal-N fraction based on ¹⁵N results. By making a simplifying assumption about the enrichment of the ¹⁵N in the DFE that infiltrated the soil, the contribution from DFE-N to all plant available N fractions including soil inorganic N was estimated to have been approximately 11 % of the applied DFE-N. An initial two-year study investigating the feasibility of manipulating soil water conditions through controlled drainage to enhance denitrification from irrigated DFE was extended a further two years for this thesis project. The resulting four-year data set provided the opportunity to evaluate the sustainability of DFE application onto land, an extended data set against which to test the adequacy of CaNS-Eff, and to identify the key processes in the fate of DFE irrigated onto soil under field conditions. In the final year of DFE irrigation, 1554 kg N ha⁻¹ of DFE-N was applied onto the lysimeters, with the main removal mechanism being pasture uptake (700 kg N ha⁻¹ yr⁻¹ removed). An average of 193 kg N ha⁻¹ yr⁻¹ was leached, with 80% of this being organic N. The nitrate leaching decreased with increasing soil moisture conditions through controlled drainage. At the high DFE loading rate used, the total soil C and N, pH and the microbial biomass increased at different rates over the four years. The long-term sustainability of the application of DFE can only be maintained when the supply of inorganic N is matched by the demand of the pasture. The complex simulation model (CaNS-Eff) of the soil-plant-microbial system was developed to describe the transport and transformations of C and N components in effluents applied onto the soil. The model addresses the shortcomings in existing models and simulates the transport, adsorption and filtration of both dissolved and particulate components of an effluent. The soil matrix is divided into mobile and immobile flow domains with convective flow of solutes occurring in the mobile fraction only. Diffusion is considered to occur between the micropore and mesopore domains both between and within a soil layer, allowing dissolved material to move into the immobile zone. To select an appropriate sub-model to simulate the water fluxes within CaNS-Eff, the measured drainage volumes and water table heights from the lysimeters were compared to simulated values over four years. Two different modelling approaches were compared, a simpler water balance model, DRAINMOD, and a solution to Richards' equation, SWIM. Both models provided excellent estimation of the total amount of drainage and water table height. The greatest errors in drainage volume were associated with rain events over the summer and autumn, when antecedent soil conditions were driest. When soil water and interlayer fluxes are required at small time steps such as during infiltration under DFE-irrigation, SWIM's more mechanistic approach offered more flexibility and consequently was the sub-model selected to use within CaNS-Eff. Measured bromide leaching from the lysimeters showed that on average 18% of the bromide from an irrigation event bypassed the soil matrix and was leached in the initial drainage event. This bypass mechanism accounted for the high amount of organic N leached under DFE-irrigation onto these soils and a description of this bypass process needed to be included in CaNS-Eff. Between 80 and 90% of the N and C leached from the lysimeters was particulate (> 0.2 µm in size), demonstrating the need to describe transport of particulate material in CaNS-Eff. The filtration behaviour of four soil horizons was measured by characterising the size of C material in a DFE, applying this DFE onto intact soil cores, and collecting and analyzing the resulting leachate using the same size characterisation. After two water flushes, an average of 34% of the applied DFE-C was leached through the top 0-50 mm soil cores, with a corresponding amount of 27% being leached from the 50-150 mm soil cores. Most of the C leaching occurred during the initial DFE application onto the soil. To simulate the transport and leaching of particulate C, a sub-model was developed and parameterised that describes the movement of the effluent in terms of filtering and trapping the C within a soil horizon and then washing it out with subsequent flow events. The microbial availability of the various organic fractions within the soil system are described in CaNS-Eff by availability spectra of multiple first-order decay functions. The simulation of microbial dynamics is based on actual consumption of available C for three microbial biomass populations: heterotrophs, nitrifiers and denitrifiers. The respiration level of a population is controlled by the amount of C that is available to that population. This respiration rate can vary between low level maintenance requirements, when very little substrate is available, and higher levels when excess substrate is available to an actively growing population. The plant component is described as both above and below-ground fractions of a rye grass-clover pasture. The parameter set used in CaNS-Eff to simulate the fate of DFE irrigated onto the conventionally drained lysimeter treatments over three years with a subsequent 10 months non-irrigation period was derived from own laboratory studies, field measurements, experimental literature data and published model studies. As no systematic calibration exercise was undertaken to optimise these parameters, the parameter set should be considered as "initial best estimates" and not as a calibrated data set on which a full validation of CaNS-Eff could be based. Over the 42 months of simulation, the cumulative drainage from CaNS-Eff for the conventionally drained DFE lysimeter was always within the 95% CI of the measured value. On the basis of individual drainage bulking periods, CaNS-Eff was able to explain 92% of the variation in the measured drainage volumes. On an event basis the accuracy of the simulated water filled pore space (WFPS) was better than that of the drainage volume, with an average of 70% of the simulated WFPS values being within the 95% CI for the soil layers investigated, compared to 44% for the drainage volumes. Overall the hydrological component of CaNS-Eff, which is based on the SWIM model, could be considered as satisfactory for the purposes of predicting the soil water status and drainage volume from the conventionally drained lysimeter treatment for this study. The simulated cumulative nitrate leaching of 4.7 g NO₃-N m⁻² over the 42 months of lysimeter operation was in good agreement to the measured amount of 3.0 (± 2.7) g NO₃-N m⁻². Similarly, the total simulated ammonium leaching of 2.7g NH₄- N m⁻² was very close to the measured amount of 2.5 (± 1.35) g NH₄- N m⁻² , however the dynamics were not as close to the measured values as with the nitrate leaching. The simulated amount of organic N leached was approximately double that measured, and most of the difference originated from the simulated de-adsorption of the dissolved fraction of organic N during the l0-month period after the final DFE irrigation. The 305 g C m⁻² of simulated particulate C leached was close to the measured amount of 224 g C m⁻² over the 31 months of simulation. The dissolved C fraction was substantially over-predicted. There was good agreement in the non-adsorbed and particulate fractions of the leached C and N in DFE. However, the isothermic behaviour of the adsorbed pools indicated that a non-reversible component needed to be introduced or that the dynamics of the de-adsorption needed to be improved. Taking into account that the parameters were not calibrated but only "initial best estimates", the agreement in the dynamics and the absolute amounts between the measured and simulated values of leached C and N demonstrated that CaNS-Eff contains an adequate description of the leaching processes following DFE irrigation onto the soil. The simulated pasture N production was in reasonable agreement with the measured data. The simulated dynamics and amounts of microbial biomass in the topsoil layers were in good agreement with the measured data. This is an important result as the soil microbial biomass is the key transformation station for organic materials. Excepting the topsoil layer, the simulated total C and N dynamics were close to the measured values. The model predicted an accumulation of C and N in the topsoil layer as expected, but not measured. Although no measurements were available to compare the dynamics and amounts of the soil NO₃-N and NH₄-N, the simulated values appear realistic for an effluent treatment site and are consistent with measured pasture data. Considering the large amount of total N and C applied onto the lysimeters over the 42 months of operation (4 t ha⁻¹ of N and 42 t ha⁻¹0f C), the various forms of C and N in dissolved and particulate DFE as well as in returned pasture, and that the parameters used in the test have not been calibrated, the simulated values from CaNS-Eff compared satisfactorily to the measured data.

Soil organic carbon (SOC) concentrations and greenhouse gas (GHG) emissions are not uniform across the landscape, but assemble in "hotspots" in specific areas. These differences are mainly driven by human-induced activities such as agricultural management. 40-50% of the Earth's land surface is under agricultural land-use, for instance cropland, managed grassland and permanent crops including agro-forestry and bio-energy crops. Furthermore, 62% of the global soil C stock is SOC and the soil stores more than 3 times more C than the atmosphere. Thus, C sequestration in agricultural soil has a potentially important role in increasing SOC storage and GHG mitigation, and there is considerable interest in understanding the effects of agricultural management on SOC and GHG fluxes in both grasslands and croplands, in order to better assess the uncertainty and vulnerability of terrestrial SOC reservoirs. For the sake of discovering the agricultural management practices relating to the effective and sustainable C sequestration in agricultural lands in Europe, simulating future terrestrial C stocks and GHG budgets under varied agricultural management systems in major European ecosystems is essential. Using models is a useful method with the purpose of this and abundant studies have carried out. However, many model results have not been validated with reliable observed long-term data, while other studies have reported a strong impact of model initialisation on model result. Nevertheless, predictions of annual to decadal variability in the European terrestrial C and GHG ressources largely rely on model results. Consequently, finding the most appropriate and comprehensive model initialisation method for obtaining reliable model simulations became important, especially for process-based ecosystem models. In recent years, Zimmermann et al. (2007) have succeed in initialising the Rothamsted Carbon model (RothC) using a physical and chemical soil fractionation method. For that reason, we hypothesised that measured detailed SOC data would be useful to initialise ecosystem models, and this hypothesis should be tested for different process-based models and agricultural land-use and management. (...)

Soil organic matter (SOM) is the largest terrestrial pool of organic carbon (C), and potential carbon-climate feedbacks involving SOM decomposition could exacerbate anthropogenic climate change. Despite the importance of SOM in the global C cycle, our understanding of the controls on SOM stabilization and decomposition is still developing, and as such, SOM dynamics are a source of major uncertainty in current Earth system models (ESMs), which reduces the effectiveness of these models in predicting the efficacy of climate change mitigation strategies. To improve our understanding of controls on SOM decomposition at scales relevant to such modeling efforts, A and upper B horizon soil samples from 22 National Ecological Observatory Network (NEON) sites spanning the conterminous U.S. were incubated for 52 weeks under conditions representing site-specific mean summer temperature and horizon-specific field capacity (-33 kPa) water potential. Cumulative CO2 respired was periodically measured and normalized by soil organic C content to obtain cumulative specific respiration (CSR). A two-pool decomposition model was fitted to the CSR data to calculate decomposition rates of fast- (kfast) and slow-cycling pools (kslow). Post-LASSO best subsets multiple linear regression was used to construct horizon-specific models of significant predictors for CSR, kfast, and kslow. Significant predictors for all three response variables consisted mostly of proximal factors related to clay-sized fraction mineralogy and SOM composition. Non-crystalline minerals and lower SOM lability negatively affected CSR for both A and B horizons. Significant predictors for decomposition rates varied by horizon and pool. B horizon decomposition rates were positively influenced by nitrogen (N) availability, while an index of pyrogenic C had a negative effect on kfast in both horizons. These results reinforce the recognized need to explicitly represent SOM stabilization via interactions with non-crystalline minerals in ESMs, and they also suggest that increased N inputs could enhance SOM decomposition in the subsoil, highlighting another mechanism beyond shifts in temperature and precipitation regimes that could alter SOM decomposition rates.<br>Master of Science

Soil organic matter (SOM) is important to the environment. Its carbon content is the largest reservoir in the global terrestrial ecosystem and decomposition of it emits greenhouse gases including CO2, CH4 and N2O. The current status of GB SOM is under debate because recent observation programmes suggested different findings. Independent and parallel computer simulations of SOM dynamics could provide useful information for the science debate and that is the objective of this thesis. A newly coupled land surface -- SOM dynamics model, JULES-ECOSSE, was applied for the computer simulations in this Ph.D study. The details of this coupled model are described in Chap. 2. Before the simulation studies in Chap. 6, three other studies were done to (1) derive a new algebraic method to spin up the content of SOM pools such that their equilibrium values can be found efficiently; (2) evaluate the model's capability in simulating various gas fluxes using the observed data from the NitroEurope project. This evaluation study provided important information about how well the model works in different aspects; (3) find better performing parameter sets for GB vegetation using a factorial experiment as sensitivity analysis followed by a multi-objective calibration scheme. After the aforementioned studies, the model was applied in Chap. 6 to answer three environmental questions. The simulations suggest that climate change over 1978-2007 had minor impacts on GB SOM, however the future impacts of climate change could potentially be big. The exact magnitudes vary between ecosystems and will also depend on the representative concentration pathway that the world will follow. Inclusion of other observed environmental changes (i.e. changing age structure and composition in woodlands, nitrogen eutrophication effect, changing soil pH and its impact on DOC mobility, soil erosion, non-equilibrium status of SOM) could better match observed changes with simulated changes. This suggests that they could be the candidates explaining the recent observed trends in GB SOM.

I denna rapport undersöks hur mängden organiskt kol, TOC (Totalt organiskt kol), varierar i tid och rum i Mälarens avrinningsområde, samt vad det är som styr TOC-halten i Mälaren. Det är viktigt att förstå dynamiken hos TOC i Mälaren och i dess avrinningsområde eftersom ökat TOC i vattnet påverkar vattenkvaliteten och orsakar problem vid beredning av dricksvatten. TOC kan bland annat reagera med klor/UV-ljus och bilda cancerframkallande ämnen. Det kan också öka antal mikrober i vattnets distributionssystem. Arbetet omfattar analys av samband mellan elementen, transportberäkningar per ytenhet av elementen till Mälaren och en modelleringsansats för ett av avrinningsområdena. Rapporten innehåller även en jämförelse mellan de olika vattenföringsmodellerna samt uppmätt vattenföring för analys av eventuella systematiska skillnader mellan dessa som påverkar beräkningen av TOC och de andra elementens transport till Mälaren. Analysen av sambanden mellan variablerna TOC (mg/l), kaliumpermanganat förbrukning (KMnO4, mg/l), absorbans_F (F=filtrerad), järn (mg/l), mangan (mg/l) och SO4_IC (sulfat mätt med hjälp av jonkromatografi, mg/l), visade att vissa av dessa variabler är korrelerade med varandra. TOC mot KMnO4 och TOC mot absorbans_F hade de bästa anpassningarna med respektive R2- värden 0,65 och 0,59 och p-värden &lt;0,001. Årsnederbörd är positivt korrelerad med TOC per ytenhet för Kolbäcksån med R2-värde 0,63 och p-värde &lt;0,01, vilket innebär att sambandet är signifikant. Ökad årsnederbörd leder till ökad tillförsel av TOC till Mälaren. Det finns däremot inget signifikant samband mellan TOC-transport per ytenhet och årsmedeltemperatur. Arealflödesberäkningar tyder på att den största tillförseln av TOC- transport per ytenhet kommer från den nordöstra delen av Mälaren. Fyrisån står för den största tillförseln av TOC. Hydrologiska, kemiska och meteorologiska data inkluderades i modeller för att kunna skatta TOC-halten i Mälaren. Temperatur-, evapotranspirations- och nederbördsdata användes i en hydrologisk modell, HBV- modellen, för att simulera vattenföringen från avrinningsområdet. Sedan användes en processbaserad modell, INCA- C, som drivs av hydrologisk data och beräknade grundvattenbildning och markfuktighet för att simulera tidsmässiga mönster i TOC. Invariablerna till INCA-modellen, markfuktigheten och HER (grundvattenbildning), simulerades med hjälp av HBV- modellen. Dessa modeller tillämpades i Kolbäcksån (ett av Mälarens största avrinningsområden). Modelleringen av Kolbäcksåns TOC- halt resulterade i en modell som anpassade dynamiken mellan 1996 och 2009, men missar den mellan 2009 och juni 2010, med bäst anpassning mellan 2006 och 2008. R2- och NS värden som erhölls för modellen var 0,086 och -0,059.<br>In this report, it has been investigated how the amount of organic carbon, TOC, varies in time and space in the basin of Mälaren, and what controls the TOC content in the lake. It is important to understand the dynamics of the TOC in the lake and its catchment because increased TOC in the water affects water quality and causes problems in the preparation of drinking water. Particularly, it can react with chlorine / UV- light and form carcinogenic substances. It can also increase the number of microbes in water distribution systems. In addition the work includes analysis of the relation between water chemistry variables, annual fluxes calculations (g/m2/year) of element flows to the lake and a modeling approach to a watershed. Annual fluxes calculations (g/m2/year) indicate that the largest supply of TOC to the lake comes from the northeast of the lake. Fyrisån accounts for the largest input of TOC to the lake. The high TOC-flux is due to a small proportion of open water in the catchment. Hydrological, chemical and meteorological data have been included in models to estimate the TOC content in the Mälaren. Input data processing, especially precipitation data, has been an important part of the work as it affects the whole model. Temperature, evapotranspiration and precipitation data were used in a hydrological model, HBV model, to simulate the flow from the catchment area. Then a process-based model, INCA-C, operated by the hydrological data and soil moisture, has been used to simulate the temporal patterns in TOC. The input variables to INCA-C- model, soil moisture and HER (Hydrological effective rainfall), have been simulated using the HBV- model. Those models were applied in Kolbäcksån, one of the lake's largest catchments. The modeling of Kolbäcksån resulted in a model that captured the dynamics of a few periods of the whole time series. The modeling of Kolbäcksån TOC-concentration resulted in a model that captured the dynamics between 1996 and 2009, but misses it between 2009 and June 2010. R2 and NS values obtained for the model were 0.086 and -0.059, respectively.

Submitted by Cláudia Bueno (claudiamoura18@gmail.com) on 2016-04-04T20:27:23Z No. of bitstreams: 2 Tese - Janaína de Moura Oliveira - 2015.pdf: 1868666 bytes, checksum: 5c342df236d2e6fc6f34b5b3a1245073 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5)<br>Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2016-04-05T10:47:36Z (GMT) No. of bitstreams: 2 Tese - Janaína de Moura Oliveira - 2015.pdf: 1868666 bytes, checksum: 5c342df236d2e6fc6f34b5b3a1245073 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5)<br>Made available in DSpace on 2016-04-05T10:47:36Z (GMT). No. of bitstreams: 2 Tese - Janaína de Moura Oliveira - 2015.pdf: 1868666 bytes, checksum: 5c342df236d2e6fc6f34b5b3a1245073 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2015-06-26<br>Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES<br>Integrated crop-livestock (iCL) and integrated crop-livestock-forest (iCLF) systems are pointed out as potential soil carbon sinks. However, there are few scientific studies that evaluated the real contribution of these production systems. This work included two studies. The first was to evaluate soil carbon accumulation and its origin in iCLF in the transition zone of the Cerrado-Amazon biomes; the second aimed to calibrate and validate the CQESTR model for the Cerrado ecosystem and to evaluate the effect of soil management practices, including iCL and various scenarios on soil organic carbon (SOC) over time. For the first study two areas under iCLF (iCLF1 and iCLF3, with one and three rows of Eucalyptus urograndis by hedgerow, respectively) were selected. They were cultivated in this system since 2009 in Nova Canaã do Norte, MT. A continuous pasture was used as reference. Soil samples were taken from eight layers (0.0 to 1.0 m) for the evaluation of the bulk density, texture, total C and N and δ¹³C. The second study was conducted in the Cerrado biome. The evaluated areas (Paddock 4 - P4 and Paddock 5 - P5 has been being managed in iCL since 2000. Bulk density and the organic matter content were determined for the 0.0-0.1 and 0.1-0.3 m layers. The CQESTR is a process based model which simulates the effect of climate, crop rotation and tillage management practices on SOC. The model was calibrated with P5 data and validated with P4 data. Its performance was evaluated using statistical regression analysis and the root mean square deviation (MSD). For the first study, the soil C stocks and isotopic composition were affected by the implementation of the iCLF system. The forest component was an important factor for soil C accumulation for both areas under iCLF. The N can be a limiting factor for C accumulation. We concluded that iCLF affected soil C and N stocks in the short term, however, longer iCLF deployment time would be necessary to elucidate the impact of iCLF in the long-term. In the second study model calibration was performed by adjusting the basic decomposition rate coefficient. The measured and simulated values were significantly correlated with an MSD of 2.11, indicating that the model captured spatial-temporal dynamics of SOC in the topsoil. However, CQESTR underestimated SOC for the 0,1-0,3 m layer, probably due to lack of site specific grass or crop root biomass and distribution data under tropical conditions. Additional calibration is required to improve prediction of SOC stabilization process in the subsoil layers of tropical soils. In the long term (20 years), for the superficial (0,0-0,1 m) soil layer, the model simulated C accumulation in iCL and C loss in soybean/corn grain production system independently of the use of zero-tillage or conventional tillage in either of these systems under Cerrado conditions.<br>Os sistemas de integração lavoura-pecuária (iLP) e integração lavourapecuária- floresta (iLPF) são apontados como potenciais acumuladores de carbono no solo. Entretanto, ainda há poucos estudos científicos que avaliaram a real contribuição desses sistemas de produção. O presente trabalho incluiu dois estudos para avaliação desses sistemas. O primeiro teve por objetivo avaliar a acumulação e a origem do carbono do solo em iLPF na região de transição dos biomas Cerrado-Amazônia; e o segundo estudo teve por objetivo calibrar e validar o modelo CQESTR para o ecossistema Cerrado bem como avaliar o efeito de práticas de manejo do solo, incluindo iLP e vários cenários no carbono orgânico do solo (COS) ao longo do tempo. Para o primeiro estudo foram selecionadas duas áreas sob iLPF (iLPF1 e iLPF3, sistemas com uma linha e três linhas de Eucalyptus urograndis por renque, respectivamente) cultivadas nesse sistema desde 2009 e uma pastagem no município de Nova Canaã do Norte, MT. Amostras de oito camadas (0,0-1,0 m) foram tomadas para avaliação da densidade, textura, teor de C e N total e δ¹³C. O segundo estudo foi conduzido no bioma Cerrado, em área que vem sendo manejada em iLP desde 2000. Foram avaliadas duas áreas, os Piquete 4 (P4) e Piquete 5 (P5). A densidade do solo e o teor de matéria orgânica foram determinados para as camadas 0,0-0,1 e 0,1-0,3 m. O CQESTR é um modelo de simulação de C baseado em processos que simula o efeito do clima, rotações de cultura e práticas de manejo no COS. O modelo foi calibrado com dados do P5 e validado com P4. Seu desempenho foi avaliado usando análise estatística de regressão e o desvio médio quadrático (MSD). No primeiro estudo, a composição isotópica do solo e os estoques de C foram afetados pela implantação do sistema iLPF. O componente florestal foi importante fator na acumulação de C em ambas as áreas sob iLPF. O N pode ser um fator limitante para a acumulação de C. Conclui-se que o iLPF afeta os estoques de C e N do solo no curto prazo, entretanto, novas avaliações com maior tempo de implantação do iLPF poderiam auxiliar na elucidação do comportamento desses elementos no sistema em longo prazo. No segundo estudo, a calibração do modelo foi realizada pelo ajuste do coeficiente da taxa de decomposição básica. Os valores simulados e medidos foram significativamente correlacionados com um MSD de 2,11, indicando que o modelo capturou satisfatoriamente a dinâmica temporal do COS na camada superficial. Entretanto, o CQESTR subestimou o COS para a camada subsequente 0,1-0,3 m, provavelmente devido às diferenças na biomassa e distribuição de raízes de gramíneas de clima tropical e temperado. Calibração adicional é requerida para melhorar a predição do COS e processos de estabilização nas camadas subsuperficiais de solos tropicais. Para a camada 0,0-0,1 m, em longo prazo (20 anos), o modelo simulou acumulação de C em iLP e decréscimo de C em sistema de produção com sucessão soja/milho, tanto sob plantio direto quanto preparo convencional em condições do Cerrado.

The decline of soil fertility is recognized by the European Union (EU) as the cause of yields reduction in many arable lands. The Soil Thematic Strategy proposed by the European Commission in 2006, identified the decline of organic matter as one of the main soil threats in EU. Organic carbon content is a recognised indicator of soil quality. Several studies have investigated this relationship through long-term field level experiments. This thesis presents a different approach: starting from data and information at EU level, a regional case study is investigated. The general objective of this thesis is to evaluate and quantify the impact of specific management practices in preserving or sequestering soil organic carbon in EU and regionally. The thesis is structured in five chapters: the first is a general introduction on the need for preserving soil organic carbon in the agricultural land and a review on the relevant legislation at international and European level. The second is a scoping chapter that presents a comparison on the available data on organic carbon content at EU level. The third chapter is a meta-analysis on soil organic carbon sequestration data available in scientific literature and reflection the management practices applied at EU scale. In the fourth chapter, the CENTURY model is applied at regional level in order to estimate the actual values of soil organic carbon stock and to model the implementation of the most promising management practices in two different climatic scenarios. The last chapter outlines the general conclusions and recommendations.<br>La ridotta fertilitá dei suoli è riconosciuta dall’Unione Europea (UE) come preludio di una minore produttivitá delle aree agricole. La Strategia tematica del suolo, prodotta dalla Commissione Europea nel 2006, aveva identificato il declino della sostanza organica come una delle otto principali minacce dei suoli in UE, in quanto il contenuto di carbonio organico è un indicatore della qualitá dei suoli. Molti studi si sono concentrati su esperimenti a lungo termine a taglio locale. Questo lavoro ha un approccio diverso: a partire da dati ed informazioni a livello UE viene indagato un caso studio a taglio regionale. L’obiettivo generale di questo lavoro è valutare e quantificare quali sono le pratiche agricole piú promettenti nel preservare o sequestrare carbonio organico nei suoli dell’UE. La tesi è strutturata in cinque capitoli: il primo è un’introduzione generale sulla necessitá di preservare il carbonio organico presente nei suoli agricoli e una review della legislazione disponibile a livello internazionale ed Europeo. Il secondo capitolo indaga e confronta i dati disponibili sui livelli di carbonio nel suolo a livello UE. Il terzo è una meta-analisi su dati in letteratura sulla capacitá di sequestrare carbonio da parte delle pratiche agricole utilizzate dei suoli dell’UE. Nel quarto capitolo viene applicato il modello CENTURY a livello regionale per ricostruire i valori di stock di carbonio organico attuali e modellare l’applicazione di pratiche agricole promettenti in due diversi scenari climatici. Infine, l’ultimo capitolo riporta le conclusioni generali del lavoro e alcune linee guida.

This thesis presents the first modeling study on long-term carbon dynamics for the University of Guelph Elora Agricultural Research Station and the Environment Canada Borden Forest Research Station at the daily and half-hourly time-step. The daily version of the CENTURY (DayCENT) model and the Carbon- and Nitrogen-coupled Canadian Land Surface Scheme (CN-CLASS) model were validated for quantifying the effects of agricultural management and component respiration on the carbon budget. DayCENT indicated that conventional tillage (CT) enhanced the annual heterotrophic respiration relative to no-till (NT) by 38.4, 93.7 and 64.2 g C m-2 yr-1 for corn, soybean and winter wheat, respectively. The seasonal variation of total soil organic carbon (SOC) pool was greater in CT than NT due to tillage effects on carbon transfer from the active surface SOC pool to the active soil SOC pool at a rate of 50-100 g C m-2 yr-1. NT accounted for a 10.7 g C m-2 yr-1 increase in the slow SOC pool (20-year turnover time) at a site in Elora, Ontario, Canada. I found that the plant phenology algorithms used in CN-CLASS were not constructed and validated for crop growth, resulting in a high degree of uncertainty in the simulations. Therefore, I designed and tested a new agricultural module for CN-CLASS. The regression analysis indicated that the new crop module improved the net ecosystem productivity (NEP) simulation for a cornfield, with the coefficient of determination (r2) of annual NEP increasing from 0.51 in the original CN-CLASS to 0.78 in the modified version of the model. I verified CN-CLASS to simulate the dynamics of component respiration for tracing the contributions from litterfall, SOC and root respiration in a deciduous mixedwood forest in Borden, Ontario, Canada. The model estimated that the annual ecosystem CO2 respiration was 1366 g C m-2 yr-1, contributed by heterotrophic respiration (57%), maintenance respiration (37%) and growth respiration (6%). The annual accumulated soil respiration was estimated at 782 g C m-2 yr-1, which was dominated by CO2 emissions from soil organic matter (60%). The base respiration rates required further verification based on field measurements. Based on the verified modeling approach in this thesis, the modeling core of DayCENT can be constructed as an integral platform for Agriculture and Agri-Food Canada National Carbon and Greenhouse Gas Accounting and Verification System. The crop phenological module in CN-CLASS allows us to conduct further agricultural studies concerning global carbon budget and environmental change. The validated respiration algorithms in CN-CLASS would be helpful in developing global biological CO2 transport model for tracing emission sources.<br>Natural Science and Engineering Research Council of Canada

Magister Scientiae (Biodiversity and Conservation Biology)<br>This study examined the magnitude of soil CO2 efflux in an arid South African ecosystem, the flux responses as well as those of key limiting nutrients to soil temperature increases and moisture reductions consistent with a future climate change scenario, and compared measured soil respiration rates with those predicted with empirically and theoretically-based soil respiration models. Measurements of soil respiration rate, temperature, moisture, N and P contents were conducted monthly over a 12-month period in natural environments and those artificially manipulated with replicated open-top warming chambers (average 4.1oC increase) and precipitation exclusion chambers (average 30.1% decrease in rainfall, 26.2% decrease in fog and dewfall) distributed in five different soil-vegetation units.Measured soil respiration rates were over 3-fold less than those reported for temperate and tropical forest ecosystems with 61.5% of the total soil CO2 efflux contributed by root respiration (derived from the differences between moderately vegetated and sparsely vegetated areas) in moderately vegetated soils. Massive increases (up to 15 times) in soil CO2 efflux occurred during wet phases, but even these large CO2 pulses were only comparable in magnitude with soil CO2 effluxes reported for temperate semi-arid grasslands. There was considerable intra-annual and inter-site variability in the magnitude and direction of soil respiration and N and P responses to elevated temperatures and reduced precipitation levels with poor correspondence evident between soil CO2 efflux and soil organic matter content. Soil CO2 effluxes declined in response to precipitation exclusion by 7.1% over all sites and increased in response to warming by 42.1% over all sites. The large increase in response to warming was assisted by a 7.5% enhancement in soil moisture content due to precipitation interception by the chamber walls and its channelling to the soil surface.Relatively smaller respiration increases in response to warming occurred in moderately vegetated soils, these attributed to soil thermal insulation by the plant canopy cover. Soil P and N contents increased in response to warming by 11.3% and 13.3% respectively over all sites, with soil P declining in response to precipitation exclusion by 5.8% over all sites and soil N increasing in response to precipitation exclusion over all sites by 5.8%. Standard least squares regressions quantified the relationships between soil respiration rate and measured soil physical and chemical properties, and their interactions for each of the 5 soil-vegetation units. These relationships were incorporated in an empiricallybased soil respiration (EMR) model which was compared with a theoretically based generalized soil respiration model (GRESP). GRESP model functions included measured Q10 coefficients at soil moisture contents above field capacity, these assumed reduced by half for dry conditions, and maximum retentive and field capacities of soils. EMR modelled soil respiration rates displayed slightly better correspondence with measured soil respiration rates than GRESP modelled soil respiration rates. This apparent from the higher regression coefficients and lower sums of squared residuals, with EMR model residuals also more closely approximating normal distributions. However, despite the EMR model’s slight superiority, it was concluded that more precise laboratory-based measurements of soil retentive and field capacities and their Q10 coefficients at different soil moisture contents could improve the GRESP model’s accuracy thereby providing a more convenient and uncomplicated means of predicting respiration responses to current and future climates over a wide range of arid soil types