The primary objective of this research was to develop a landscape-oriented, process-based approach that can enhance understanding and prediction of SOC fluxes in IMLs by incorporating the key mechanisms impacting soil carbon dynamics when moving from the soilscape to the landscape. The mechanisms that are considered to be the focus of this study are redistribution of SOC due to erosion and deposition without neglecting the importance of litter incorporation into the soil column, decomposition due to microbial activity, and physical and chemical stabilization of carbon. To accomplish this objective, field experiments were performed to examine how selective entrainment of different soil size fractions, quantified through the enrichment ratio (ER), varies with management and hillslope position. Differential modes in soil mobilization between rill and interrill areas were either elevated or dampened depending on the prevalent management practice, the gradient of the site and landscape position. Sites where sediment and runoff fluxes were highest were found to have lower ER values (around unity) due to the mobilization of all size classes making static and dynamic samples almost identical. The size fractions analyzed in these experiments were found to have varying levels of carbon associated with them, especially the larger aggregates, which encapsulate organic material. Neglecting them in transport estimates could lead to large errors in predicted fluxes of SOC. For this reason, a careful attention was placed on identifying how aggregate stability varies with respect to management and hillslope position, through controlled experiments looking size distributions to reflect tillage disturbance and aggregate stability to assess resistance to rainsplash. Lastly, a landscape-oriented modeling framework was developed that captures not only the SOC spatial heterogeneity in IMLs but also determines the impacts that redistribution has on this heterogeneity and ultimately on SOC dynamics. The integrative modeling framework considers the collective effects of both rainsplash/rainfall- and tillage-induced erosion on SOC redistribution in IMLs through an ER-module developed and woven within this framework to connect an upland erosion model with a soil biogeochemical model. It provides not only size fraction updates to the active layer and ER values, but also explicitly considers the effects of splash-driven interrill erosion on those ER estimates. The model was applied to twentieth-century changes in SOC across a representative agricultural hillslope in the study watershed and compared to recent SOC data. The chronosequence in SOC storage within the erosional zone revealed that soils were continually depleted of the rich organic matter long after the 1930’s “Dust bowl” due to enhanced erosion that accompanied agricultural practices. However, conservation tillage and enhanced crop production that began in the late 1980’s reversed the downward trend in SOC losses, causing nearly 26% of the lost SOC to be regained. Results from this study can be used to aid policy and decision makers in developing a food-system that accounts for the co-evolution of human and natural activity, to develop sustainable agro-ecosystems through the use of data supported recommended best management practices.

Submitted by VERA LÚCIA DA SILVA FARIAS null (verlucbio@yahoo.com.br) on 2017-01-12T20:11:31Z No. of bitstreams: 1 Tese_Vera_Lucia_da_Silva_Farias.pdf: 2109156 bytes, checksum: a352f68e4cb2df0cff49ade4c7573779 (MD5) / Approved for entry into archive by LUIZA DE MENEZES ROMANETTO (luizamenezes@reitoria.unesp.br) on 2017-01-16T17:36:32Z (GMT) No. of bitstreams: 1 farias_vls_dr_jabo.pdf: 2109156 bytes, checksum: a352f68e4cb2df0cff49ade4c7573779 (MD5) / Made available in DSpace on 2017-01-16T17:36:32Z (GMT). No. of bitstreams: 1 farias_vls_dr_jabo.pdf: 2109156 bytes, checksum: a352f68e4cb2df0cff49ade4c7573779 (MD5) Previous issue date: 2016-12-16 / Com a crescente preocupação com as perdas de fósforo na água de enxurrada e sedimentos erodidos enriquecidos com P, que podem aumentar o risco de eutrofização dos corpos d´água superficiais, o interesse no uso de modelos que avaliam os impactos do uso do solo, tornou-se importante. O modelo WEPP, tem uma ampla gama de aplicabilidade, uma vez que pode ser usado para simular processos de erosão, escoamento e transporte de elementos orgânicos e químicos. Entretanto, no Brasil, não há casos em que um algoritmo tenha sido associado ao WEPP para modelar o transporte de P. Assim, o objetivo deste trabalho foi avaliar o desempenho de modelos para estimar as perdas de fósforo disponível, no sedimento erodido predito pelo WEPP, em uma pequena bacia hidrográfica. O presente trabalho foi realizado em área localizada no município de Tabapuã, noroeste do Estado de São Paulo. Os dados para alimentar o banco de dados do WEPP basearam-se nos componentes: climático, hidrológico e desenvolvimento vegetal. Os sedimentos erodidos foram classificados em: frações primárias, areia, silte, agregados pequenos e grandes. Com tais frações foi calculada a taxa de enriquecimento por fósforo com o WEPP (WER). Outros dois algoritmos foram utilizados para estimar o fósforo transportado com o sedimento erodido: ln(ER) = 2,682 – 0,278 ln(Sed) e) ln(ER) = 2 - 0,2 ln(Sed). Para a validação das estimativas de P nos sedimentos, utilizaram-se os critérios: o coeficiente de Nash-Sutcliffe (NS), o erro quadrático médio (RMSE), o coeficiente de massa residual (CRM) e o índice de concordância (d). Os modelos testados para estimar as taxas de enriquecimento do sedimento erodido e enxurrada por P são eficientes, quando em uso com as predições do Water Erosion Prediction Project (WEPP). As taxas de enriquecimento obtidas com o WEPP (WER), equação ln(ER) = 2,682 – 0,278 ln(Sed) e ln(ER) = 2 - 0,2 ln(Sed), respectivamente, apresentaram coeficiente de Nash-Sutcliffe (NS) próximo de 1. As perdas de P com o sedimento e enxurrada apresentaram moderado grau de dependência especial (GDE), enquanto que a erosão estimada pelo WEPP apresentou alto GDE. Perdas de fósforo solúvel com a enxurrada acima de 0,02 mg L-1, valor crítico para eutrofização, podem ser obtidas em 81% da área da bacia com uma probabilidade superior a 75%. / With the growing concern over phosphorus losses with the runoff water and eroded sediment enriched with P that can increase the risk of eutrophication of surface water bodies, interest in the use of models that assess the impacts of land use, has if important. The WEPP model has a wide range of applicability, since it can be used to simulate erosion, drainage and transport and organic chemicals. However, in Brazil, there aren’t cases in which an algorithm has been associated with the WEPP to model the transport P. Thus, the objective with this study was to evaluate the performance of mode loss of phosphorus with eroded sediment in agricultural environment to estimate phosphorus losses available in the sediment eroded predicted the WEPP in a small catchment. This work was carried out in an area located in the municipality of Tabapuã, northwest of São Paulo. The sediment losses were estimated with the WEPP model. Two approaches (P-empirical) estimated ER according to an empirical relationship, and the other approach used WER calculated by WEPP (P-WEPP). Models to estimate enrichment rates of the sediment (Sed) and runoff with the Water Erosion Prediction Project (WEPP), WER - WEPP, an equation ln(ER) = 2.682 - 0.278 ln(Sed) and ln(ER) = 2 - 0.2 ln (Sed) present coefficient of Nash-Sutcliffe (NS) next 1.

Raindrop kinetic energy and sheet flow can disintegrate aggregates during interrill erosion, a process responsible for non point source pollution. Also, the dissolution process during aggregate wetting can affect interrill erosion. These factors can be responsible for changes in particle size distribution in the sediment, especially when different tillage systems are compared. The effect of soil tillage and management on soil properties is not uniform, which determine a wide range of runoff and sediment delivery rate. Variety in these rates can be associated with pore functions and their interactions with aggregate stability. One of the objectives of this study was to analyze the wetting behavior of soil aggregates from soils under conventional tillage compared with soils under no tillage. It was expected that the wetting rate is a function of pore system and that different tillage systems would affect the soil wetting behavior based on their impact on soil structure and shape. The second objective was to analyze the relationships among soil wetting rate, particle movement, organic carbon (OC) and iron release with the sediment produced via interrill erosion. A rainfall simulation experiment was performed in the field to determine the effect of low and fast soil wetting on total soil loss through high and low kinetic rainfall energy, sediment particle size distribution and OC loss. Two soils that differed in soil textural composition and that were under conventional and no tillage were investigated. Soil loss depended largely on soil characteristics and wetting rate. Particle size distribution of sediment was changed by treatment and the proportion of particles smaller than 0.053 mm increased over time, at any kinetic energy wetting level. Temporal OC and iron release were constant, which required a continuous source principally due to aggregate slaking. An empirical model was proposed to improve an interrill erosion equation by using a bond-dissolution mechanism that identified soil as a regulator of particle release.