Soil analytical quality control by traditional and spectroscopy techniques: Constructing the future of a hybrid laboratory for low environmental impact

Author(s)

Dematte, Jose Alexandre M. ^[1] ; Dotto, Andre Carnieletto ^[1] ; Bedin, Luis Gustavo ^[1] ; Sayao, Veridiana Maria ^[1] ; Barros e Souza, Arnaldo ^[1]

Affiliation

^[1] Univ Sao Paulo, Coll Agr Luiz de Queiroz ESALQ, Dept Soil Sci, Av Padua Dias 11, BR-13418260 Piracicaba, SP - Brazil

• Document type: Journal article
• Source: Geoderma; v. 337, p. 111-121, MAR 1 2019.
• Web of Science Citations: 5

Abstract

Soil analysis is an important information in agriculture and environmental monitoring. It is usually performed by wet chemical analysis with high cost and chemical products consumption. In the world, it is estimated that 1.5 billion ha is used as agricultural area. If every 5 ha 2 samples (2 depths) were collected, we would have 600 million soil samples for chemical and granulometric analysis. Considering just the analysis of organic matter (OM) by wet combustion method in the laboratory as an example, we would be utilizing about 840 thousand kg of dichromate and ammonium ferrous sulfate and 3 million L of sulfuric acid. The use of these reagents can have a huge ecological consequence if they do not have an adequate final disposal. An alternative methodology such as proximal sensing can be utilized with low environmental impact. Therefore, the objective of this study was to: i) evaluate the analytical quality of soil attributes via different traditional laboratories and sensors, ii) evaluate the prediction of the models using sensors, iii) assess the uncertainties of lime recommendation analyzed by the laboratories. We applied 96 soil samples at two depths collected in Sao Paulo State, Brazil. The determination of 15 soil attributes was performed by four different routine laboratories, and they were predicted by 4 sensors (400-2500 nm). Results indicate that the determination of attributes via chemical analysis with low quality led to high error in spectral models. The great predictive performances of clay, OM, cation exchange capacity (CEC), and pH enable the use of sensors in the evaluation of these attributes. Overall, the criteria for classification of analytical results showed that sand, silt, clay, pH, OM, CEC, and base saturation were the attributes that can be determined by the spectroscopy technique with high-quality outcome. The lime recommendation derived from proximal sensor analysis can be used as an efficient method, since it presented a high correlation with the laboratory result. In this sense, a hybrid laboratory analysis can be developed to optimize analysis with better quality control, which is indicated as a great opportunity in the near future. (AU)

• FAPESP's process: 14/22262-0 - Geotechnologies on a detailed digital soil mapping and the Brazilian soil spectral library: development and applications
• Grantee: José Alexandre Melo Demattê
• Support Opportunities: Research Projects - Thematic Grants
• FAPESP's process: 17/03207-6 - Geotechnologies in detailed pedological digital mapping and Brazilian spectral soil library: development and applications
• Grantee: Andre Carnieletto Dotto
• Support Opportunities: Scholarships in Brazil - Post-Doctoral