Variabilidade espacial de atributos químicos do solo em sistemas plantio direto e convencional

• Ano de defesa: 2017
• Autor(a) principal: Miziara Neto, Salvador Jorge
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Agronomia
• Palavras-chave em Português: Agronomia; Solos - Correção; Geologia- Métodos estatísticos; Solos - Amostragem; Amostragem de solo; Dependência espacial; Geoestatística; Krigagem; Soil sampling; Spatial dependence; Geoestatistics; Kriging; CNPQ::CIENCIAS AGRARIAS::AGRONOMIA
• Link de acesso: https://repositorio.ufu.br/handle/123456789/21121; http://doi.org/10.14393/ufu.di.2017.371

The understanding of the spatial variability of soil chemical attributes is crucial when thinking about a more profitable and environmentally sustainable agriculture. Soil heterogeneity begins with the pedogenetic processes, which have direct interaction with the climatic elements and the biotic factors happening over time. The usual soil sampling contemplates the collection of small number of samples, taken at depths of 0 to 0,20 m. These samplings produce insufficient results to provide soil correction, according to the spatial variability of its characteristics. This study aimed to evaluate the variability of the chemical attributes pH, cation exchange capacity (CEC), sum of bases (SB) and base saturation percentage (BS%), in two areas, under conventional and notillage planting systems, in Brazilian Cerrado biome, located in Uberaba, MG. The area in conventional tillage is irrigated by center pivot, 500 m away from the no-till area, cultivated under rainfed conditions. A sampling grid of 200 x 400 m was established for each system, with 45 georeferenced points, 50 m apart each other. Five subsamples were collected at each point, at depths 0 - 0,05, 0,05 - 0,10, 0,10 - 0,15 and 0,15 - 0,20 m. The samples were analyzed for pH, SB, CEC and BS% values. Data were submitted to descriptive statistics, determining mean, median, mode, standard deviation, coefficient of variation, asymmetry, kurtosis, minimum and maximum values and skewness. The classical statistics allowed to evaluate possible variations among the attributes, without, however, consider the spatial position of the sampling points. With the use of variograms and data interpolation, by kriging, it is possible to characterize the spatial variability of the soil attributes and define the domain amplitude of each sampling. By GS + software, the spatial dependence was evaluated by adjusting semivariograms, prevailing the exponential model, followed by spherical. Results indicated spatial variability of the evaluated attributes, with levels varying from low to high classes. In the no-till area lower vertical variation of the attributes was observed. Horizontal variability was significant, with spatial dependence occurring for all variables. In the no tillage area, layer 0 - 0,10 m presented higher levels of all attributes. Except for the attribute BS%, from layer 0,10 - 0,15 m in no-tillage system, all other variables showed spatial dependence, thus enabling the creation of kriging maps. The spatial dependence range varied from 51 m to 220 m in conventional system and from 55 m to 155 m in no tillage system. The use of geostatistical techniques allowed a better understanding of soil attributes variability dynamics in the studied systems.