Dados do Trabalho

Título

SOIL SUBSAMPLING IMPROVEMENT USING PROXIMAL SENSING AND MANAGEMENT ZONE APPROACH

Resumo

The delimitation of management zones to facilitate soil sampling and soil fertility characterization has been widely used in precision agriculture. Thus, this work aimed to explore how reliable these zones can be for representing phosphorus availability (Pavail), exploring the issue of subsampling by soil class and dividing the field by homogeneous zones guided by soil proximal sensing. A grain production area of 52 hectares in the state of São Paulo, Brazil, was divided into two parts, according to the soil classification (LV2-3 and LV1-3), resulting in two areas of 29 and 23 ha. Using a previously existing sample, four subsampling schemes were simulated to collect 15 soil-subsamples at 0-20cm depth, obtaining four average Pavail values to represent the same area. As the average Pavail varied substantially among the simulations (50-61 mg /dm3 for one area and 38-52 mg /dm3 for another), three homogeneous soil zones were created to reduce de soil variability based on the magnetic susceptibility obtained by the EM38 sensor up to a depth of 75cm. Within each zone, three subsampling schemes (15 points) were simulated to assess whether this field subdivision would provide a more stable average Pavail for each zone than the assessment for the entire field. As a result, the variation in the Pavail between the subsampling simulations that achieved 24% and 37% for Areas 1 and 2, respectively, reduced to 4-10% for Area 1 and 1-4.5% for Area 2. Concerning the average Pavail of each zone, it was possible to observe that the average value of 55 mg/dm3 was reduced to three zones of 61, 48, and 53 mg/dm³ values for Area 1 and comes from 43 mg/dm3 for Area 2 to three zones with values of 48, 45 and 39 mg/dm3, better representing the Pavail variability present within the fields. Thus, subdividing the areas into homogeneous zones with proximal sensing is an option for better detailing the soil's nutrient content, aiming for the variable-rate fertilizer application.

Palavras-chave

management zones, soil sampling, precision agriculture

Instituição financiadora

Fundação de Amparo a Pesquisa do Estado de são Paulo - Fapesp

Agradecimentos

Fundação de Amparo a Pesquisa do Estado de são Paulo - Fapesp; Faculdade de Engenharia Agrícola - UNICAMP, e ao GITAP