Submitted by Cássia Santos (cassia.bcufg@gmail.com) on 2016-10-19T11:04:18Z No. of bitstreams: 2 Dissertação - Alex Leonardo Tosta - 2014.pdf: 2466765 bytes, checksum: 244fe49c8fdc9852a7fbf54d5cee5710 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Jaqueline Silva (jtas29@gmail.com) on 2016-10-19T16:10:36Z (GMT) No. of bitstreams: 2 Dissertação - Alex Leonardo Tosta - 2014.pdf: 2466765 bytes, checksum: 244fe49c8fdc9852a7fbf54d5cee5710 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2016-10-19T16:10:36Z (GMT). No. of bitstreams: 2 Dissertação - Alex Leonardo Tosta - 2014.pdf: 2466765 bytes, checksum: 244fe49c8fdc9852a7fbf54d5cee5710 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2014-06-06 / The development of more adapted genotypes to water stress conditions and/ or more efficient water use is a priority that begins to be sought in different programs of genetic improvement of plants. Thus, knowledge of methodologies to be used in breeding programs to produce genotypes more tolerant to water stress conditions and more efficient water use is of paramount importance to this alternative to be viable. Based on the above, the objective was to study the response of onion genotypes to different levels of soil water that help the selection of plants with tolerance to drought and/ or which are more efficient in water use. We evaluated the effect of five levels of water deficit (100, 80, 60, 40 and 20% of ETc ) in eight genotypes of onion (“TX 08” , “Vale Ouro IPA – 11”, “BRS 367” , “Primavera”, “Optima F1”, “Franciscana IPA – 10” , “CNPH 6179org” and “Alfa Tropical”). The irrigations were performed whenever the strain of the soil water in treatments with 100% of ETc reached the critical soil water tension of 10 kPa. The design was used in a randomized block design (RBD) with five replications in a factorial 5 x 8 (blades x genotypes ) . Each experimental unit corresponded to a 10-liter vessel with 6 onion plants. The TX08 and IPA-10 genotype showed potential for use in breeding programs to drought tolerance, because their moderate sensitivity to drought, high water use efficiency and high productivity / O desenvolvimento de genótipos mais adaptadas às condições de estresse hídrico e/ou mais eficientes no uso de água é uma prioridade que começa a ser buscada em diferentes programas de melhoramento genético de plantas. Sendo assim, o conhecimento de metodologias para serem utilizadas em programas de melhoramento destinados a produção de genótipos mais tolerantes a condições de estresse hídrico e mais eficientes no uso de água é de suma importância para que esta alternativa se viabilize. Com base no exposto, objetivou-se estudar a resposta de genótipos de cebola a diferentes níveis de estresse hídrico que auxilie a seleção de genótipos com maior tolerância ao déficit hídrico e/ou que sejam mais eficientes no uso de água. Avaliou-se o efeito de cinco níveis de déficit hídrico (100, 80, 60, 40 e 20% da ETc) em oito genótipos de cebola (“TX 08”, “Vale Ouro IPA-11”, “BRS 367”, “Primavera”, “Optima F1”, “Franciscana IPA-10”, CNPH 6179org e “Alfa Tropical”). Foi empregado o delineamento em blocos casualizados (DBC), com cinco repetições em esquema fatorial 5 x 8 (lâminas x genótipos). Cada unidade experimental correspondeu a um vaso de 10 litros com 6 plantas de cebola. As irrigações eram realizadas sempre que a tensão de água no solo, avaliada por tensiômetros, atingia de 10 kPa nos tratamentos com 100% da ETc. O genótipo TX08 e IPA-10 demonstraram potencial para serem utilizados em programas de melhoramento visando tolerância a seca, pois apresentam moderada sensibilidade ao déficit hídrico, alta eficiência no uso de água e alta produtividade.

Submitted by Claudia Rocha (claudia.rocha@udesc.br) on 2018-03-06T16:10:50Z No. of bitstreams: 1 PGCS14MA140.pdf: 1574830 bytes, checksum: 4217f8ffb682e61890fd939d7f9ca3e2 (MD5) / Made available in DSpace on 2018-03-06T16:10:50Z (GMT). No. of bitstreams: 1 PGCS14MA140.pdf: 1574830 bytes, checksum: 4217f8ffb682e61890fd939d7f9ca3e2 (MD5) Previous issue date: 2014-07-25 / Capes / Several techniques have been used to obtain soil water retention curve. Direct methods are time consuming and costly. Mathematical techniques have been used to replace direct methods through pedotransfer functions (PTFs), using predictors easily obtained and inexpensive to obtain parameters that expend more time and resources. PTFs generated from multiple regressions have shown reliable results but requires great knowledge of statistics. A new approach to obtaining PTFs, artificial neural networks (ANN) that does not require prior knowledge of statistical and database minimum, have been used in various areas of knowledge, and have been shown to be accurate and reliable. The study aimed to compare the efficiency of ANNs and multiple regressions to generate PTFs for estimating the retention and availability of water in soils of Santa Catarina state, located in a subtropical zone (region) of Brazil. The PTFs were generated for 940 horizons of representative soils of St. Catarina having as input data the particle size distribution (seven classes), the organic matter content, bulk density, particle density and output data water retention between from zero to 1500 kPa tension. The ANNs were generated using the software WEKA, with six neurons in the hidden layer using the error backpropagation algorithm (Multilayer Perceptron) and the cross-validation to stratify sample. Results showed that PTFs obtained from ANNs are more efficient than multiple regression. The PTFs that include specific texture data (five classes of sand), soil organic matter, soil bulk density, particle density, total porosity and their interactions better estimate the retention and availability of water in these soils. From these results, it is possible to develop a program that uses these PTFs generated to provide these values of SWRC, and is therefore an extremely useful tool for studies about soil water retention and avaibility / Várias técnicas vêm sendo utilizadas para obtenção da curva de retenção de água no solo. Como métodos diretos demandam tempo e são de alto custo, tem sido estudadas funções de pedotransferência (FPTs), as quais utilizam preditores de fácil obtenção e baixo custo para obter os parâmetros. FPTs geradas a partir de regressões múltiplas têm mostrado resultados confiáveis, mas exigem grande conhecimento estatístico para sua determinação. Uma nova abordagem para obtenção de FPTs é o uso de redes neurais artificiais (RNAs), que não exigem conhecimento prévio estatístico e podem ser obtidas utilizando-se um banco de dados mínimos. As RNAs vêm sendo utilizadas em diversas áreas do conhecimento, e têm se mostrado precisas e confiáveis. Este trabalho teve por objetivo comparar a eficiência de RNAs e regressões múltiplas na geração de FPTs para estimar a retenção e a disponibilidade de água em solos do estado de Santa Catarina. As FPTs foram geradas a partir de 940 horizontes de solos representativos de Santa Catarina, tendo como dados de entrada a granulometria (sete classes de tamanhos de partículas), o teor de matéria orgânica, a densidade do solo e de partículas e, como dados de saída, a retenção de água nas tensões de 0 até 1.500 kPa. As RNAs foram geradas utilizando o software WEKA, tendo como arquitetura a rede Multilayer Perceptron, com seis neurônios na camada escondida, utilizando-se a função de retropropagação do erro (Backpropagation) e validação cruzada para estratificar a amostra. As FPTs obtidas a partir de RNAs são mais eficientes que as regressões múltiplas. As FPTs pontuais, que incluem dados de granulometria (cinco divisões de areia), matéria orgânica do solo, densidade do solo, densidade de partículas, porosidade total e suas interações estimam melhor a retenção e a disponibilidade de água nestes solos. A partir destes resultados, é possível desenvolver um programa que utiliza as FPTs geradas para estimar a curva de retenção de água, sendo, portanto uma ferramenta extremamente útil para estudos que envolvam a retenção e a disponibilidade de água no solo

Iron deficiency chlorosis (IDC) is a frequent problem throughout many areas of the United States where soils are high in calcium carbonate (CaCO3), including the Blackland Prairie regions of Mississippi. The main objectives of this study were to 1) determine the effects of seven different cropping systems on IDC visual symptomology and grain yield in rainfed soybeans grown in calcareous soils, and 2) evaluate the effects of soil water tension (SWT) on IDC. Rotating soybeans with corn produced significantly higher yields for both tolerant and susceptible soybean varieties. IDC symptomology was worse, and yields were lower in cropping systems with lower average SWT, or wetter soils, throughout the growing season. Last, this study evaluated the use of multispectral imagery and apparent soil electrical conductivity (ECa) to identify IDC-prone areas of a field for the site-specific implementation of management strategies that produced higher yields in the plot-scale study.

Ziel der Arbeit ist es, ein genaueres Prozessverständnis zur Abflussbildung an Hängen auf der Basis der räumlichen Verbreitung periglazialer Deckschichten zu erhalten. Das Untersuchungsgebiet ist ein 6 ha großes, forstlich bewirtschaftetes Quelleinzugsgebiet im Osterzgebirge. Das anstehende Gestein ist Gneis. Der oberflächennahe Untergrund ist aus zwei- und dreigliedrigen Deckschichten zusammengesetzt. Auf der Punkt-, Hang- und Kleineinzugsgebietsskala werden hydrometrische, hydrochemische und geoelektrische Methoden sowie Färbeversuche eingesetzt, um auf die dominierenden Abflussbildungsprozesse schließen zu können. Aus der Synthese der Teilergebnisse werden drei typische Prozessabläufe in Abhängigkeit von der Gebietsvorfeuchte abgeleitet. Diese verdeutlichen, dass bei geringer Vorfeuchte Sättigungsoberflächenabfluss im Quellsumpf vorherrscht, bei mittlerer bis hoher Vorfeuchte dagegen Zwischenabfluss der dominierende Abflussprozess ist. Die Abflusswirksamkeit der Niederschläge steigt zudem mit zunehmender Vorfeuchte nichtlinear an. Es wird herausgestellt, dass die hydraulisch anisotropen Eigenschaften der Basislage entscheidend die oberflächennahen Fließwege des Wassers beeinflussen. Sie besitzt durch ihre hohe Lagerungsdichte einerseits vertikal wasserstauende Eigenschaften. Andererseits kann Wasser, begünstigt durch das dominant sandige Substrat und das hangparallel eingeregelte Bodenskelett innerhalb der Schicht bevorzugt lateral geleitet werden. Die gewonnenen Erkenntnisse verdeutlichen die Bedeutung der Eigenschaften der periglazialen Deckschichten für die Abflussbildung an Mittelgebirgshängen. / The aim of this study is to contribute to the understanding of runoff processes on slopes based on the spatial distribution of periglacial cover beds. The study area is a 6 ha large forested spring catchment in the Eastern Ore Mountains, Saxony. Bedrock is gneiss overlain by periglacial cover beds comprising two or three layers. On plot, hillslope and small-catchment scales hydrometrical, hydrochemical and geoelectrical methods as well as tracer experiments are used to determine the constitutive runoff processes. From the synthesis of partial results, three pre-moisture controlled process cycles are derived. With low pre-moisture, saturation overland flow dominates in the spring bog. In contrast, with medium or high pre-moisture interflow occurs. Besides, with rising pre-moisture runoff coefficients increase in a non-linear manner. It is emphasised that the hydraulic anisotropic structure of the Basal Layer is the major control factor for subsurface water-flow paths. On the one hand, this layer acts as an aquitard for seeping water because of its high bulk density. On the other hand, water within the layer is able to flow laterally because of the sandy texture and the coarse clasts oriented parallel to the slope. These findings highlight the importance of relic periglacial cover beds for runoff generation in subdued mountains.

Ziel der Arbeit ist es, ein genaueres Prozessverständnis zur Abflussbildung an Hängen auf der Basis der räumlichen Verbreitung periglazialer Deckschichten zu erhalten. Das Untersuchungsgebiet ist ein 6 ha großes, forstlich bewirtschaftetes Quelleinzugsgebiet im Osterzgebirge. Das anstehende Gestein ist Gneis. Der oberflächennahe Untergrund ist aus zwei- und dreigliedrigen Deckschichten zusammengesetzt. Auf der Punkt-, Hang- und Kleineinzugsgebietsskala werden hydrometrische, hydrochemische und geoelektrische Methoden sowie Färbeversuche eingesetzt, um auf die dominierenden Abflussbildungsprozesse schließen zu können. Aus der Synthese der Teilergebnisse werden drei typische Prozessabläufe in Abhängigkeit von der Gebietsvorfeuchte abgeleitet. Diese verdeutlichen, dass bei geringer Vorfeuchte Sättigungsoberflächenabfluss im Quellsumpf vorherrscht, bei mittlerer bis hoher Vorfeuchte dagegen Zwischenabfluss der dominierende Abflussprozess ist. Die Abflusswirksamkeit der Niederschläge steigt zudem mit zunehmender Vorfeuchte nichtlinear an. Es wird herausgestellt, dass die hydraulisch anisotropen Eigenschaften der Basislage entscheidend die oberflächennahen Fließwege des Wassers beeinflussen. Sie besitzt durch ihre hohe Lagerungsdichte einerseits vertikal wasserstauende Eigenschaften. Andererseits kann Wasser, begünstigt durch das dominant sandige Substrat und das hangparallel eingeregelte Bodenskelett innerhalb der Schicht bevorzugt lateral geleitet werden. Die gewonnenen Erkenntnisse verdeutlichen die Bedeutung der Eigenschaften der periglazialen Deckschichten für die Abflussbildung an Mittelgebirgshängen. / The aim of this study is to contribute to the understanding of runoff processes on slopes based on the spatial distribution of periglacial cover beds. The study area is a 6 ha large forested spring catchment in the Eastern Ore Mountains, Saxony. Bedrock is gneiss overlain by periglacial cover beds comprising two or three layers. On plot, hillslope and small-catchment scales hydrometrical, hydrochemical and geoelectrical methods as well as tracer experiments are used to determine the constitutive runoff processes. From the synthesis of partial results, three pre-moisture controlled process cycles are derived. With low pre-moisture, saturation overland flow dominates in the spring bog. In contrast, with medium or high pre-moisture interflow occurs. Besides, with rising pre-moisture runoff coefficients increase in a non-linear manner. It is emphasised that the hydraulic anisotropic structure of the Basal Layer is the major control factor for subsurface water-flow paths. On the one hand, this layer acts as an aquitard for seeping water because of its high bulk density. On the other hand, water within the layer is able to flow laterally because of the sandy texture and the coarse clasts oriented parallel to the slope. These findings highlight the importance of relic periglacial cover beds for runoff generation in subdued mountains.