An optimized sample for assessing soil property variations across the field and within management zones efficiently.

• Autor(a) principal: VASQUES, G. de M.
• Data de Publicação: 2022
• Outros Autores: RODRIGUES, H. M., TAVARES, S. R. de L., HERNANI, L. C., OLIVEIRA, R. P. de
• Idioma: eng
• Título da fonte: Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice)
• Texto Completo: http://www.alice.cnptia.embrapa.br/alice/handle/doc/1140307
• topic: Proximal soil sensing; Spatial simulated annealing; Spatial trends; Sensoriamento Remoto; Agricultura de Precisão; Precision agriculture; Remote sensing

An optimized sampling design to assess soil property variation across the field and within management zones is proposed and validated in a 72-ha crop field in southeastern Brazil. An optimized sample (18 sites) was derived by spatial simulated annealing from proximal sensor covariates. Soil properties were measured at 0-10 cm and validated against those measured at 72 sites on a regular grid. The optimized and regular grid samples had equal global spatial trend models and means for soil clay, pH and exchangeable Ca, Mg and K, and different ones for organic C and available P. Within zones, equal means between sampling designs were found for all soil properties in the ?North? zone, and for most properties in the other two zones. Soil property correlations against proximal sensor variables were honored by the optimized samples in most cases, both globally and within zones. The optimized soil sample reduces costs while keeping most soil information for guiding management decisions.