The objectives of this research were to develop and evaluate a field method for in situ measurement of soil properties using visible near-infrared reflectance spectroscopy (Vis-NIRS). A probe with an independent light source for acquiring soil reflectance spectra from soil cores was developed around an existing portable field spectrometer (ASD FieldSpecPro, Boulder, CO, USA; 350-2500 nm). Initial experiments tested the ability of the acquired spectra to predict plant root density, an important property in soil carbon dynamics. Reflectance spectra were acquired from soil containing ryegrass roots (Lolium multiflorum) grown in Allophanic and Fluvial Recent soils in a glasshouse pot trial. Differences in root density were created by differential nitrogen and phosphorus fertilization. Partial least squares regression (PLSR) was used to calibrate spectral data (pre-processed by smoothing and transforming spectra to the first derivative) against laboratory-measured root density data (wet-sieve technique). The calibration model successfully predicted root densities (r2 = 0.85, RPD = 2.63, RMSECV = 0.47 mg cm-3) observed in the pots to a moderate level of accuracy. This soil reflectance probe was then tested using a soil coring system to acquire reflectance spectra from two soils under pasture (0-60 mm soil depths) that had contrasting root densities. The PLSR calibration models for predicting root density were more accurate when soil samples from the two soils were separated rather than grouped. A more accurate prediction was found in Allophanic soils (r2 = 0.83, RPD = 2.44, RMSECV = 1.96 mg g-1) than in Fluvial Recent soils (r2 = 0.75, RPD = 1.98, RMSECV = 5.11 mg g-1). The Vis-NIRS technique was then modified slightly to work on a soil corer that could be used to measure root contents from deeper soil profiles (15- 600 mm depth) in arable land (90-day-old maize crop grown in Fluvial Recent soils). PLSR calibration models were constructed to predict the full range of maize root densities (r2 = 0.83, RPD = 2.42, RMSECV = 1.21 mg cm-3) and also soil carbon (C) and nitrogen (N) concentrations that had been determined in the laboratory (LECO FP- 2000 CNS Analyser; Leco Corp., St Joseph, MI, USA). Further studies concentrated on improving the Vis-NIRS technique for prediction of total C and N concentrations in differing soil types within different soil orders in the field. The soil coring method used in the maize studies was evaluated in permanent and recent pastoral soils (Pumice, Allophanic and Tephric Recent in the Taupo-Rotorua Volcanic Zone, North Island) with a wide range of soil organic matter contents resulting from different times (1-5 years) since conversion from forest soils. Without any sample preparation, other than the soil surface left after coring, it was possible to predict soil C and N concentrations with moderate success (C prediction r2 = 0.75, RMSEP = 1.23%, RPD = 1.97; N prediction r2 = 0.80, RMSEP = 0.10%, RPD = 2.15) using a technique of acquiring soil reflectance spectra from the horizontal cross-section of a soil core (H method). The soil probe was then modified to acquire spectra from the curved vertical wall of a soil core (V method), allowing the spectrometer’s field of view to increase to record the reflectance features of the whole soil sample taken for laboratory analysis. Improved predictions of soil C and N concentrations were achieved with the V method of spectral acquisition (C prediction r2 = 0.97, RMSECV = 0.21%, RPD = 5.80; N prediction r2 = 0.96, RMSECV = 0.02%, RPD = 5.17) compared to the H method (C prediction r2 = 0.95, RMSECV = 0.27%, RPD = 4.45; N prediction r2 = 0.94, RMSECV = 0.03%, RPD = 4.25). The V method was tested for temporal robustness by assessing its ability to predict soil C and N concentrations of Fluvial Recent soils under permanent pasture in different seasons. When principal component analysis (PCA) was used to ensure that the spectral dimensions (which were responsive to water content) of the data set used for developing the PLSR calibration model embraced those of the “unknown” soil samples, it was possible to predict soil C and N concentrations in “unknown” samples of widely different water contents (in May and November), with a high level of accuracy (C prediction r2 = 0.97, RMSEP = 0.36%, RPD = 3.43; N prediction r2 = 0.95, RMSEP = 0.03%, RPD = 3.44). This study indicates that Vis-NIRS has considerable potential for rapid in situ assessment of soil C, N and root density. The results demonstrate that field root densities in pastoral and arable soil can be predicted independently from total soil C, which will allow researchers to predict C sequestration from root production. The recommended “V” technique can be used to assess spatial and temporal variability of soil carbon and nitrogen within soil profiles and across the landscape. It can also be used to assess the rate of C sequestration and organic matter synthesis via root density prediction. It reduces the time, labour and cost of conventional soil analysis and root density measurement.

The objectives of this research were to develop and evaluate a field method for in situ measurement of soil properties using visible near-infrared reflectance spectroscopy (Vis-NIRS). A probe with an independent light source for acquiring soil reflectance spectra from soil cores was developed around an existing portable field spectrometer (ASD FieldSpecPro, Boulder, CO, USA; 350-2500 nm). Initial experiments tested the ability of the acquired spectra to predict plant root density, an important property in soil carbon dynamics. Reflectance spectra were acquired from soil containing ryegrass roots (Lolium multiflorum) grown in Allophanic and Fluvial Recent soils in a glasshouse pot trial. Differences in root density were created by differential nitrogen and phosphorus fertilization. Partial least squares regression (PLSR) was used to calibrate spectral data (pre-processed by smoothing and transforming spectra to the first derivative) against laboratory-measured root density data (wet-sieve technique). The calibration model successfully predicted root densities (r2 = 0.85, RPD = 2.63, RMSECV = 0.47 mg cm-3) observed in the pots to a moderate level of accuracy. This soil reflectance probe was then tested using a soil coring system to acquire reflectance spectra from two soils under pasture (0-60 mm soil depths) that had contrasting root densities. The PLSR calibration models for predicting root density were more accurate when soil samples from the two soils were separated rather than grouped. A more accurate prediction was found in Allophanic soils (r2 = 0.83, RPD = 2.44, RMSECV = 1.96 mg g-1) than in Fluvial Recent soils (r2 = 0.75, RPD = 1.98, RMSECV = 5.11 mg g-1). The Vis-NIRS technique was then modified slightly to work on a soil corer that could be used to measure root contents from deeper soil profiles (15- 600 mm depth) in arable land (90-day-old maize crop grown in Fluvial Recent soils). PLSR calibration models were constructed to predict the full range of maize root densities (r2 = 0.83, RPD = 2.42, RMSECV = 1.21 mg cm-3) and also soil carbon (C) and nitrogen (N) concentrations that had been determined in the laboratory (LECO FP- 2000 CNS Analyser; Leco Corp., St Joseph, MI, USA). Further studies concentrated on improving the Vis-NIRS technique for prediction of total C and N concentrations in differing soil types within different soil orders in the field. The soil coring method used in the maize studies was evaluated in permanent and recent pastoral soils (Pumice, Allophanic and Tephric Recent in the Taupo-Rotorua Volcanic Zone, North Island) with a wide range of soil organic matter contents resulting from different times (1-5 years) since conversion from forest soils. Without any sample preparation, other than the soil surface left after coring, it was possible to predict soil C and N concentrations with moderate success (C prediction r2 = 0.75, RMSEP = 1.23%, RPD = 1.97; N prediction r2 = 0.80, RMSEP = 0.10%, RPD = 2.15) using a technique of acquiring soil reflectance spectra from the horizontal cross-section of a soil core (H method). The soil probe was then modified to acquire spectra from the curved vertical wall of a soil core (V method), allowing the spectrometer’s field of view to increase to record the reflectance features of the whole soil sample taken for laboratory analysis. Improved predictions of soil C and N concentrations were achieved with the V method of spectral acquisition (C prediction r2 = 0.97, RMSECV = 0.21%, RPD = 5.80; N prediction r2 = 0.96, RMSECV = 0.02%, RPD = 5.17) compared to the H method (C prediction r2 = 0.95, RMSECV = 0.27%, RPD = 4.45; N prediction r2 = 0.94, RMSECV = 0.03%, RPD = 4.25). The V method was tested for temporal robustness by assessing its ability to predict soil C and N concentrations of Fluvial Recent soils under permanent pasture in different seasons. When principal component analysis (PCA) was used to ensure that the spectral dimensions (which were responsive to water content) of the data set used for developing the PLSR calibration model embraced those of the “unknown” soil samples, it was possible to predict soil C and N concentrations in “unknown” samples of widely different water contents (in May and November), with a high level of accuracy (C prediction r2 = 0.97, RMSEP = 0.36%, RPD = 3.43; N prediction r2 = 0.95, RMSEP = 0.03%, RPD = 3.44). This study indicates that Vis-NIRS has considerable potential for rapid in situ assessment of soil C, N and root density. The results demonstrate that field root densities in pastoral and arable soil can be predicted independently from total soil C, which will allow researchers to predict C sequestration from root production. The recommended “V” technique can be used to assess spatial and temporal variability of soil carbon and nitrogen within soil profiles and across the landscape. It can also be used to assess the rate of C sequestration and organic matter synthesis via root density prediction. It reduces the time, labour and cost of conventional soil analysis and root density measurement.

Production of Pinus radiata is a major contributor to New Zealand's economy and new plantings are a valuable carbon sink. Phosphorus (P) deficiency and high P fixing capacity of some volcanic ash soils (e.g. Allophanic Soil) may constrain radiata productivity. This thesis investigates the role of ectomycorrhizal (ECM) root processes in the acquisition of P by P. radiata fiom native soil and soil fertilised with two reactive phosphate rock (RPR) fertilisers. The application of finely-divided RPRs to a P deficient Allophanic Soil significantly increased P. radiata seedling growth and P uptake in 10 month pot trials. RPR dissolution was high in this soil, and it was further enhanced by the radiata rhizosphere processes. The development and formation of ECM in radiata seedlings was stimulated by low rates of RPR application but was hindered in unfertilised soils and high rates of RPR application. The P. radiata ECM roots induced acidification and increased oxalate concentration and phosphatase activities in the rhizosphere soil. These changes in rhizosphere biochemical properties were associated with enhanced solubilisation of fertiliser and soil inorganic P and increased mineralisation of organic P, leading to increased P bioavailability in the rhizosphere. ECM inoculation of P. radiata roots with Rhizopogen rubescens and Suillus luteus stimulated production of phosphatase enzymes and oxalate and induced acidification in the rhizosphere. The extent of root-induced changes in the rhizosphere soils was associated with ECM hyphae length density. A technique using pulse labelling of radiata shoots with 14CO2 showed promise in estimating the active ECM hyphae density. The 14C activity was highly correlated with ECM hyphae density measured by an agar film technique. Overall, observations made in this thesis indicate that sparingly soluble forms of organic and inorganic P in soils low in plant-available P are readily solubilised and utilised for P. radiata growth through ECM rhizosphere processes.

The soils used by the East Coast Organic Producers Trust (ECOPT; the grower group that this study is targeted towards) have exceptionally low soil Olsen P concentrations (ca. 6 mg/L). These and other limitations (e.g. poor weed and pest and disease control) result in many ECOPT growers being unable to produce economic yields on anything other than small scale gardens. Fertilisers and manures are seldom used by these growers, which exacerbates the problem. Thus, the object of this research was to provide information to ECOPT on which fertilisers and application strategies would provide the best returns on their phosphorus (P) fertiliser investment. The experimental work was carried out in two parts. A laboratory study tested a range of phosphate rock (PR) based fertilisers and application rates; Ben Guerir reactive phosphate rock (RPR; 67, 133, 267, 533 and 1,333 mg P/kg soil), BioPhos and BioSuper (267 and 1,333 mg P/kg soil) and a no fertiliser Control. Soil fertiliser mixtures were incubated for 155 days and periodic measurements of PR dissolution, soil pH and Bic-P (analogous to Olsen P but expressed in µg/g) were undertaken. The field study used fewer application rates and two application methods; banded and broadcast. Broadcast plots were applied at 678 mg P/kg soil (488 kg P/ha); banded RPR was applied at 236, 678 and 1475 mg P/kg soil (40, 115 and 250 kg P/ha respectively) and banded BioPhos and BioSuper at 678 mg P/kg soil (115 kg P/ha). A Control was also included. Fertilisers were applied in October 2004 and changes in soil pH and Bic-P were measured in the broadcast plots only over a 344 day period. Potato (Solanum tuberosum L. cv. Desiree) was the test crop. Regression analysis was used to generate exponential equations to describe the changes in Bic-P over time (∆Bic-P). Differences between fertilisers in the amount of P dissolved and pH fluxes were used to explain the differences in ∆Bic-P. BioSuper dissolved quicker and generated greater ∆Bic-P than RPR and BioPhos, which were similar. Higher application rates produced greater increases in Bic-P than lower rates but decreased the % of P applied that dissolved. The increase in Bic-P over time from fertiliser application was much slower in the field compared with the laboratory. This was put down to differences in experimental conditions; mainly soil pH and soil aggregate surface area. Potato tuber yield (mean = 35 t/ha) did not respond to any of the fertiliser treatments despite a significant increase in P concentration of the shoots mid-way through the season in all broadcast treatments (shoot P concentration was not analysed in the banded plots). Water and N availability were the main limiting factors in this season as the crop was not irrigated and soil N supply was insufficient to produce a full canopy. Phosphorus response curves generated using the fertiliser response model PARJIB (Reid, 2002), and an economic analysis, indicated that for RPR and BioPhos the optimum economic application rate was 200 kg P/ha and for BioSuper it was 100 kg P/ha (applied every third and second year respectively).

Sulphur is an important nutrient to plants, and reports of its deficiency have been increasing worldwide. Sulphur starvation causes losses in both yield and quality, and it reduces nitrogen use efficiency of plants. As the timing for fertilisation can be decisive for avoiding deleterious effects, improvements in the description of the sulphur balance in fields are a valuable contribution for assisting fertiliser management. Sulphate is the most important inorganic form of sulphur in soils. Being the mobile form, sulphate is readily available for plants, and also prone to be leached. Therefore the description of the movement of sulphate is the key component of the sulphur balance. Leaching of sulphate from the soil can be significantly delayed by its adsorption onto the soil particles. Soil type and pH are the main factors defining the sulphate adsorption capacity; although the presence of other ions in the soil solution can have a considerable effect. It has been reported that in some soils, typically volcanic and tropical soils with variable-charge characteristics, the co-presence of sulphate and calcium can substantially enhance their retention via ion-pair adsorption (IPA). To determine the influence of cations on the movement of sulphate, series of batch and miscible displacement experiments were conducted using two New Zealand soils, of contrasting ion adsorption capacities: the Taupo sandy and Egmont loam soils. These experiments demonstrated the occurrence of cooperative adsorption of sulphate and calcium in the Egmont soil, but not in the Taupo soil. Batch experiments were conducted to examine the IPA adsorption process in the Egmont soil in more detail. Based on the analyses of the results from these two series of experiments, plus the review of published data, three different mathematical approaches for evaluating the amount of solute adsorbed as ion-pairs are proposed. A computer program was built for solving an adsorption model using these three approaches, and was used to compare the model's predictions and the observed adsorption data. An extension of this program, coupling the adsorption model with a solute transport description, was used to simulate the movement of sulphate and calcium. Comparisons between the data from the miscible displacements and the results from this model are used to demonstrate the applicability of the proposed IPA description for modelling the transport of these ions in the soil. Finally, results from a pot trial with Egmont soil are used to examine the relevance of IPA for the movement of sulphate under non-equilibrium conditions, and with active plant growth. Although the results from this experiment regarding IPA were statistically non-significant, some insights could be obtained and are discussed. More studies involving IPA under non-equilibrium experiments are needed for a better understanding of the relevance of IPA in field conditions.

Modelos matemáticos de predição de erosão e de produção de sedimentos apresentam um grande potencial para serem utilizados no planejamento dos recursos naturais, pois através destes é possível estimar impactos causados pelas atividades agrícolas. Com base nessa premissa, utilizou-se o modelo SWAT (Soil and Water Assessment Tool), em uma pequena bacia rural (1,19 km²) localizada no município de Arvorezinha-RS, para avaliar a sua aplicabilidade e simular a produção de sedimentos em três diferentes cenários de uso e manejo de solo (floresta, cultivo do fumo em sistema convencional e mínimo). A aplicabilidade do modelo foi avaliada através da comparação dos resultados hidrológicos e sedimentológicos simulados pelo modelo, sem utilização da calibração, com valores observados obtidos para uma série de dados de cinco anos de monitoramento hidrossedimentológico (2002-2006). Resultados referentes à aplicabilidade do modelo através da simulação do escoamento superficial para a escala de tempo diário não foram satisfatórios, enquanto que valores mensais e anuais foram mais adequados. Para a vazão líquida, da mesma forma que para o escoamento superficial, as melhores simulações foram verificadas também para vazões mensais e anuais. Já a produção de sedimentos os resultados não foram satisfatórios, tanto para simulações diárias como mensais e anuais. Possíveis razões que explicam os resultados da produção de sedimentos estão relacionados com limitações tanto da equação que simula o aporte de sedimentos (MUSLE), como do equacionamento da propagação dos sedimentos no canal. Os resultados das simulações hidrológicas para diferentes cenários mostraram não haver diferença para o escoamento superficial entre os sistemas convencional e o cultivo mínimo. Em relação à erosão, verificou-se que o cultivo mínimo foi capaz de reduzir em 30% a produção de sedimentos quando comparado com o sistema convencional. / Mathematical models to predict erosion and sediment yield have great potential for use in natural resource planning, as they enable the user to estimate impacts caused by agricultural activities. Based on this premise, the SWAT model (Soil and Water Assessment Tool) was applied in a small rural catchment (1,19 km²) in Arvorezinha - RS, in Southern Brazil, to evaluate its applicability and to estimate sediment yield under three different soil management scenarios: forest, conventional tillage with tobacco, and minimum tillage with tobacco. The applicability of the model was evaluated by comparing hydrologic and sediment results simulated by the uncalibrated model with observed values obtained over the course of five years (2002-2006) monitoring program. Applicability results for the model based on simulating surface runoff were unacceptable for the daily time scale, although results for monthly and annual time scales were better. This was also true for flow rate, with better simulations on the monthly and annual time scales. However for sediment yield simulations, the SWAT model did not present satisfactory results over any time frame. This may be related to limitations in the equation that simulates sediment load (MUSLE) or to the equivalency of sediment propagation in the canal. Hydrologic simulations for the different soil management scenarios did not show differences in surface runoff between conventional and minimum tillage, although the minimum tillage was shown to reduce sediment yield by 30% when compared with conventional tillage.

O entendimento da dinâmica hidrossedimentológica em uma bacia hidrográfica pode ser realizado pelo monitoramento das variáveis hidrossedimentológicas e pela modelagem desses processos. Nesse contexto, essa pesquisa analisou a eficiência do modelo Soil and Water Assessment Tool (SWAT) na previsão dos processos hidrossedimentológicos na bacia do Taboão (Pejuçara, RS), considerando as saídas (vazão e produção de sedimentos) em um passo de tempo mensal e diário. Para tal, foram utilizados dados de chuva horária dos anos 2008 a 2016, monitorada em quatro pluviógrafos instalados na bacia (PVGs 34, 40, 43 e 51), e dados climáticos da estação meteorológica de Cruz Alta. As informações de vazão, para os anos de 2011 a 2016, foram obtidas por meio da conversão dos dados de nível de água monitorados no exutório da bacia, utilizando uma curva-chave cota x vazão. A concentração de sedimentos suspensos (CSS), para os anos de 2013 a 2015, foi estimada por meio de modelos de redes neurais artificias (RNAs), empregando como entrada dados de turbidez e de nível de água, monitorados no exutório da bacia. O preenchimento das falhas dos registros de precipitação horária foi executado por meio de modelos de Combinações de RNAs (CRNAs) associados à média simples (MS) ou à média ponderada pelo inverso da distância (MP), utilizando como entrada dados pluviométricos dos postos vizinhos. As falhas nos dados de nível de água foram preenchidas por modelos de RNAs, que usaram como entrada níveis de água monitorados em sub-bacias embutidas ou adjacente à bacia do Taboão (bacias do Donato, Turcato, Alemão e Andorinhas), e dados de precipitação média dos quatro pluviógrafos utilizados nessa pesquisa Foram determinadas as defasagens temporais entre os níveis de água das diferentes bacias, e testados o uso da precipitação média com aplicação de filtro temporal linear e/ou exponencial. Os registros falhos nos dados de turbidez foram preenchidos por modelos de RNAs, que empregaram como entrada informações de nível de água monitoradas, de 10 em 10 minutos, no exutório da bacia. A calibração do modelo SWAT para a previsão dos processos hidrológicos foi realizada usando dados de vazão, diários e mensais, para os anos de 2013, 2014 e 2016 e; a etapa de verificação foi executada para os anos de 2011 e 2015. Considerou-se o Método de Green & Ampt para determinação da infiltração de água no solo e 2 anos (2008-2009) para período de aquecimento do modelo SWAT. A calibração do modelo para a produção de sedimentos foi realizada para os anos de 2013 e 2015 e o processo de verificação foi efetuado para o ano de 2014. A calibração e a análise de sensibilidade dos parâmetros foram realizadas com auxílio do SWAT-CUP, utilizando o algoritmo SUFI-2. O coeficiente de Nash–Sutcliffe (NS) das RNAs para preenchimento das falhas de precipitação variou entre 0,35, classificado como “Insatisfatório”, e 0,86, avaliado como “Muito Bom”, considerando critérios propostos por Moriasi et al. (2007). Das 13 RNAs desenvolvidas para preenchimento das falhas nos níveis de água, apenas uma delas foi classificada como de desempenho “Satisfatório” durante o treinamento e; as demais enquadraram-se como de desempenho “Muito Bom”. Na etapa de verificação, sete RNAs foram consideradas com desempenho “Muito Bom” e cinco com “Bom” desempenho No preenchimento das falhas de turbidez, das cinco RNAs desenvolvidas, quatro mostraram “Bom” desempenho durante o treinamento, e uma rede teve desempenho “Muito Bom”; enquanto que, no processo de verificação, duas RNAs tiveram desempenho “Muito Bom”, uma delas foi classificada com desempenho “Bom” e; duas RNAs foram consideradas com desempenho “Satisfatório”. As estatísticas de desempenho dos modelos de RNAs desenvolvidos para o preenchimento das falhas de nível de água, de turbidez e de precipitação também demonstraram que tais redes representam uma alternativa interessante para a obtenção de séries contínuas desses dados, possibilitando o uso posterior dos registros para a modelagem hidrossedimentológica. A calibração do modelo SWAT para estimativa da vazão mensal mostrou desempenho “Muito Bom” (NS=0,78), e para a determinação da vazão diária foi considerado “Bom” (NS=0,72). Na etapa de verificação, o modelo manteve o “Bom” desempenho (NS=0,68) para estimativa da vazão diária, decaindo para desempenho “Satisfatório” (NS=0,64) para a simulação em escala mensal. Para a estimativa da produção de sedimentos mensal, o desempenho do modelo foi considerado “Bom” tanto na calibração (NS=0,66) quanto na verificação (NS=0,70). Na escala diária o desempenho foi “Satisfatório” para a calibração (NS=0,64) e “Insatisfatório” para a verificação (NS=0,38) Tais resultados indicam que o modelo SWAT é uma ferramenta promissora para aplicações na previsão hidrossedimentológica na bacia do Taboão, especialmente em termos de simulações dos processos hidrológicos. No entanto, existem limitações para aplicações na estimativa da produção de sedimentos, sobretudo quando considerados os processos em escala diária. Essas limitações são consequência da presença de processos erosivos na bacia (voçorocas), que não são simulados pelas rotinas presentes no modelo SWAT, bem como pelo escoamento dominante ser do tipo subsuperficial, com ocorrência de pipping; indicando-se, portanto, adequações nas rotinas do modelo para melhor representatividade desses processos. / The understanding of hydrosedimentological dynamics in a watershed can be obtained by monitoring the hydrossedimentological variables and by modeling these processes. In this context, this research analyzed the efficiency of the Soil and Water Assessment Tool (SWAT) in predicting the hydrosedimentological processes in the Taboão basin (Pejuçara, RS), considering the outputs (flow and sediment production) in a monthly and daily time step. For that, hourly rainfall data from 2008 to 2016 were monitored at four pluviographs installed in the basin (PVGs 34, 40, 43 and 51), and climate data were obtained from the Cruz Alta meteorological station. The flow information for the years 2011 to 2016 was obtained by converting the monitored water level data into flow by using a rating curve. The suspended sediment concentration (SSC), from 2013 to 2015, was estimated using artificial neural network (ANN) models, using as input turbidity and water level data, monitored in the basin. The filling of the hourly rainfall records was performed by models of Combinations of RNAs (CRNAs) associated with the simple mean (MS) or weighted mean to the inverse distance (MP), using as input rainfall data from the neighboring stations. Failures in the water-level data were filled by RNA models, which used as input water levels monitored in sub-basins adjacent or embedded to the Taboão basin (Donato, Turcato, Alemão and Andorinha basins), and mean precipitation data of the four pluviographs used in this research. The temporal lags between the water levels of the different basins were determined and the use of the average precipitation with linear and exponential temporal filters was tested The turbidity data records were filled by RNA models, using water level information monitored at every 10 minutes. The SWAT model calibration for predicting the hydrological processes was performed using daily and monthly flow data for the years 2013, 2014 and 2016 and the verification step was performed for the years 2011 and 2015; considering Green & Ampt Method for infiltration estimation and 2 years of warm-up period (2008-2009). The calibration of the model for sediment yield was performed for the years 2013 and 2015 and the verification process was carried out for the year 2014. The calibration and sensitivity analysis of the parameters were performed with the assistance of SWAT-CUP, using the SUFI-2 algorithm. The Nash-Sutcliffe Coefficient (NS) of the RNAs used to fill precipitation faults varied between 0.35, classified as "Unsatisfactory", and 0.86, evaluated as "Very Good", considering criteria proposed by Moriasi et al. (2007). Of the 13 RNAs developed to fill water level failures, only one of them was classified as a "Satisfactory" performance during training and; the others have been classified as "Very Good" performance. In the verification step, seven RNAs were considered to have "Very Good" performance and five had "Good” performance. In the fulfillment of the turbidity faults, of the five RNAs developed, four showed "Good" performance during the training, and one network had "Very Good" performance; while in the verification process two ANNs performed "Very Good", one of them was classified as "Good" and; two ANNs were considered to have "Satisfactory" performance The performance statistics of the ANN models developed to fill the water level, turbidity and precipitation failures also demonstrated that such networks represent an interesting alternative to obtain continuous series of these data, allowing the later use of the records for hydrossedimentological modeling. In the verification processes, the model maintained a “Good” performance (NS=0.68) to estimate the daily flow, decreasing to "Satisfactory" performance (NS=0.64) for the monthly scale simulation. For the estimation of sediment yield the model performance was considered "Good" for monthly calibration period (NS=0.66) and also for the verification (NS=0.70). In daily scale the performance was "Satisfactory" for calibration (NS=0.64) and “Unsatisfactory” in the verification (NS=0.38). These results indicate that the SWAT model is a promising tool for applications in the hydrosedimentological forecasting in the Taboão basin, especially in terms of hydrological processes simulations. However, there are limitations to applications in the estimation of sediment production, especially when considering daily scale processes. These limitations are due to the presence of erosive processes in the basin (gully erosion), which are not simulated by the routines present in the SWAT model, as well as by the existence of the lateral flow with occurrence of pipping; indicating, therefore, the need for adjustments in the routines of the model to better represent these processes.

Modelos matemáticos de predição de erosão e de produção de sedimentos apresentam um grande potencial para serem utilizados no planejamento dos recursos naturais, pois através destes é possível estimar impactos causados pelas atividades agrícolas. Com base nessa premissa, utilizou-se o modelo SWAT (Soil and Water Assessment Tool), em uma pequena bacia rural (1,19 km²) localizada no município de Arvorezinha-RS, para avaliar a sua aplicabilidade e simular a produção de sedimentos em três diferentes cenários de uso e manejo de solo (floresta, cultivo do fumo em sistema convencional e mínimo). A aplicabilidade do modelo foi avaliada através da comparação dos resultados hidrológicos e sedimentológicos simulados pelo modelo, sem utilização da calibração, com valores observados obtidos para uma série de dados de cinco anos de monitoramento hidrossedimentológico (2002-2006). Resultados referentes à aplicabilidade do modelo através da simulação do escoamento superficial para a escala de tempo diário não foram satisfatórios, enquanto que valores mensais e anuais foram mais adequados. Para a vazão líquida, da mesma forma que para o escoamento superficial, as melhores simulações foram verificadas também para vazões mensais e anuais. Já a produção de sedimentos os resultados não foram satisfatórios, tanto para simulações diárias como mensais e anuais. Possíveis razões que explicam os resultados da produção de sedimentos estão relacionados com limitações tanto da equação que simula o aporte de sedimentos (MUSLE), como do equacionamento da propagação dos sedimentos no canal. Os resultados das simulações hidrológicas para diferentes cenários mostraram não haver diferença para o escoamento superficial entre os sistemas convencional e o cultivo mínimo. Em relação à erosão, verificou-se que o cultivo mínimo foi capaz de reduzir em 30% a produção de sedimentos quando comparado com o sistema convencional. / Mathematical models to predict erosion and sediment yield have great potential for use in natural resource planning, as they enable the user to estimate impacts caused by agricultural activities. Based on this premise, the SWAT model (Soil and Water Assessment Tool) was applied in a small rural catchment (1,19 km²) in Arvorezinha - RS, in Southern Brazil, to evaluate its applicability and to estimate sediment yield under three different soil management scenarios: forest, conventional tillage with tobacco, and minimum tillage with tobacco. The applicability of the model was evaluated by comparing hydrologic and sediment results simulated by the uncalibrated model with observed values obtained over the course of five years (2002-2006) monitoring program. Applicability results for the model based on simulating surface runoff were unacceptable for the daily time scale, although results for monthly and annual time scales were better. This was also true for flow rate, with better simulations on the monthly and annual time scales. However for sediment yield simulations, the SWAT model did not present satisfactory results over any time frame. This may be related to limitations in the equation that simulates sediment load (MUSLE) or to the equivalency of sediment propagation in the canal. Hydrologic simulations for the different soil management scenarios did not show differences in surface runoff between conventional and minimum tillage, although the minimum tillage was shown to reduce sediment yield by 30% when compared with conventional tillage.

Made available in DSpace on 2014-06-11T19:24:11Z (GMT). No. of bitstreams: 0 Previous issue date: 2006-12-05Bitstream added on 2014-06-13T18:48:39Z : No. of bitstreams: 1 golla_ar_me_prud.pdf: 1858401 bytes, checksum: 19a2eedb3dd7b43c07d8803318256180 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / O manejo adequado de solos e água é imprescindível para a manutenção da qualidade ambiental e de vida da população. Entretanto, isso não vem ocorrendo na microbacia hidrográfica do córrego Palmitalzinho -Regente Feijó/São Paulo. Nessa área, são encontrados diversos problemas ambientais, tais como, baixa fertilidade e compactação do solo, erosão e assoreamento dos cursos d'água. Em função disto, neste trabalho teve-se como objetivos: levantar as atividades agropecuárias e sua disposição espacial na microbacia hidrográfica do córrego Palmitalzinho; identificar aspectos negativos associados ao manejo inadequado dos recursos naturais; propor práticas de conservação e manejo dos recursos solo e água e levantar as atividades desenvolvidas pelo Programa Estadual de Microbacias Hidrográficas. Para atingir os objetivos foram realizados levantamentos de campo em todas as propriedades, entrevistas com os proprietários/funcionários das mesmas, assim como coletas de dados e informações em órgãos públicos envolvidos com o Programa Estadual de Microbacias Hidrográficas (PEMBH) (Casa da Agricultura de Regente Feijó e Escritório de Desenvolvimento Rural de Presidente Prudente). Quanto à ocupação da área, foram encontradas culturas de café, milho, feijão, cana-de-açúcar, pêra, havendo o predomínio de pastagens. Os problemas associados ao manejo inadequado dos recursos naturais identificados foram: desestruturação do solo através da utilização do sistema convencional de plantio/semeadura, correção e adubação deficientes, compactação dos solos, pastagem degradada, presença de plantas daninhas, falta de manutenção ou ausência de práticas mecânicas de conservação, estradas mal locadas, erosões, área de preservação permanente com mata ciliar incompleta e reserva legal não averbada.... / The appropriate management of the soils and water is indispensable for the maintenance of the environmental quality and of life of the population. However, that is not happening in the Palmitalzinho river hydrography microbasin- Regente Feijó county- São Paulo state. The local presents several environmental problems, such as compacted soil and with low fertility, erosion and sedimentation of the courses of water. By the reason of this, this research has as objectives: to present the agricultural activities and their spacial disposition in the Palmitalzinho river microbasin; to identify negative aspects related to the inadequate handling of natural resources; to propose conservation practices and management of the resources soil and water; and to present the activities developed by Programa Estadual de Microbacias Hidrográficas (PEMBH). Properties were traveled, as well as information were collected in public organs involved with PEMBH. The Palmitazinho river microbasin presents culture of coffee, corn, bean, sugar-cane, pear, having the predominance of pasture. The problems related to the inadequate management of the identified natural resources were: degradation of the soil through the use of the conventional system of tillage, low investment in fertilizers and limestone, compactation of the soils, pasture with low production of biomass, presence of harmful plants, maintenance lack or absence of terraces, dirt road with problems, erosions, area of permanent preservation with incomplete riparian forest and not protocoled legal reserves. The suggestions proposed include the use of the areas according to their use capacity, accomplishment of chemical analysis of the soil for correct use of fertilizers and limestone, descompactation of the soil, terracing and its maintenance, adaptation of dirt road,...(Complete abstract, click electronic address below).

Submitted by Claudia Rocha (claudia.rocha@udesc.br) on 2018-03-05T14:35:00Z No. of bitstreams: 1 PGCS16MA130.pdf: 1221495 bytes, checksum: 3a9e92a3d45df8b56495a6f9754adb04 (MD5) / Made available in DSpace on 2018-03-05T14:35:00Z (GMT). No. of bitstreams: 1 PGCS16MA130.pdf: 1221495 bytes, checksum: 3a9e92a3d45df8b56495a6f9754adb04 (MD5) Previous issue date: 2016-02-05 / Capes / The water erosion degraded the soil and contaminates the environment, influenced, among other factors, the slope length. This work aimed to quantify the losses of soil (PS) and water (PA) in different slope lengths in Cambissolo Húmico soil, in a field experiment in natural rainfall condition in Lages, SC and estimate losses soil to the conditions of the experiment by USLE and RUSLE model. The treatments with two replications, were constituted in slope length of 11 m, 22 m, 33 m and 44 m, at an average gradient of 8%, in soil cultivated with corn prior to the survey. At the end of the evaluation period, the shoot residue was removed from the surface, leaving the crop roots in bare soil and discovered. We evaluated the erosion between 04/07/2014 and 06/18/2015 in individual events of 41 erosive rainfall. The PS were adjusted to the slope of 0.09 m m-1 (standard for the Universal Soil Loss Equation - USLE), through the factor S. Erosive rainfall totaled 1,349 mm, with Ei30 of 6,066 MJ mm ha-1-h-1. Soil losses increased with increasing slope length and were greater in the summer and spring than in autumn and winter; water losses showed the same trend, with lower among slope lengths and these seasons. Soil and water losses by erosion related to linear and positively with the erosivity and the rain height, respectively. Soil losses were satisfactorily estimated by USLE and RUSLE, with lower than those observed in the field values, keeping a linear and positive relationship between the variables. The RUSLE estimated soil losses better than the USLE, compared to those observed in the field, especially in longer lengths slopes. Soil losses estimated by RUSLE related to linear and positively with these estimated losses USLE / A erosão hídrica degrada o solo e contamina o ambiente, influenciada, dentre outros fatores, pelo comprimento de declive. Com este trabalho objetivou-se quantificar as perdas de solo (PS) e de água (PA) em diferentes comprimentos de declive num CAMBISSOLO HÚMICO ALUMÍNICO léptico, em experimento de campo, em condição de chuva natural, em Lages, SC e estimar as perdas de solo para as condições do experimento pelo modelo USLE e RUSLE. Os tratamentos, com duas repetições, constituíram-se nos comprimentos de declive de 11 m, 22 m, 33 m e 44 m, em uma declividade média de 8%, em solo cultivado com milho anteriormente à pesquisa. Ao final do ciclo do milho, o resíduo da parte aérea foi removido da superfície, permanecendo as raízes da cultura em solo sem cultivo e descoberto. Avaliou-se a erosão entre 04/07/2014 e 18/06/2015, em eventos individuais de 41 chuvas erosivas. As PS foram ajustadas para a declividade de 0,09 m m-1 (padrão para a Equação Universal de Perda de Solo - USLE), por meio do fator S. As chuvas erosivas totalizaram 1.349 mm, com EI30 de 6.066 MJ mm ha-1 h-1. As perdas de solo aumentaram com o aumento do comprimento de declive e foram maiores no verão e na primavera do que o no outono e inverno; as perdas de água apresentaram a mesma tendência, com menores variações entre os comprimentos de declive e as referidas estações do ano. As perdas de solo e água por erosão hídrica relacionaram-se linear e positivamente com a erosividade e com a altura de chuva, respectivamente. As perdas de solo foram satisfatoriamente estimadas pela USLE e RUSLE, com valores menores do que os observados em campo, mantendo uma linear e positiva relação entre as variáveis. A RUSLE estimou melhor as perdas de solo do que a USLE, em comparação aos valores observados em campo, especialmente em comprimentos de declives mais longos. As perdas de solo estimadas pela RUSLE relacionaram-se linear e positivamente com as referidas perdas estimadas pela USLE