Determinación de la materia orgánica del suelo mediante espectroscopia de reflectancia vis-NIR en áreas cultivadas con boniato (Ipomoea batatas (L.) Lam)
Fecha
2018-12-06
Autores
González Aguiar, Diana
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central "Marta Abreu" de Las Villas.Facultad de Ciencias Agropecuarias.Departamento de Agronomía
Resumen
Con el objetivo de determinar el contenido de materia orgánica del suelo con el uso de la técnica de espectroscopia de reflectancia en el visible e infrarrojo cercano (vis-NIR) en áreas cultivadas con boniato, se realizó un muestreo de suelos entre enero y marzo de 2016. Las muestras fueron colectadas en el paisaje edáfico y en dos campos específicos de la de la empresa agropecuaria “Valle del Yabú”. Se analizó el contenido de materia orgánica por un método convencional y mediante la técnica de espectroscopia de reflectancia vis-NIR en el rango de longitud de onda de 379 - 1700 nm. Se evaluó la variabilidad espacial y se desarrollaron modelos de calibración y validación mediante la técnica estadística multivariable de regresión por mínimos cuadrados parciales (PLSR). La precisión del modelo se evaluó mediante el coeficiente de determinación R2. La materia orgánica mostró una variabilidad espacial de baja a moderada en el paisaje edáfico y en los campos específicos de Conyedo y Albarrán. El modelo de calibración obtenido en el paisaje edáfico se evaluó de muy bien con un coeficiente R2 de 0,95. Se validó este modelo en los campos específicos y se obtuvo una precisión de bien y muy bien con R2 de 0,89 a 0,90. Los mapas desarrollados con los valores convencionales y estimados por espectroscopia vis-NIR mostraron similitudes. Al campo específico de Albarrán correspondió la menor variabilidad espacial y se obtuvieron los máximos valores del rendimiento agrícola real y estimado con 6,15 y 6,06 t ha-1 respectivamente.
In order to determine the soil organic matter content with the use of the visible and near infrared (vis-NIR) reflectance spectroscopy technique in areas cultivated with sweet potato, soil sampling was carried out between January and March 2016. The samples were collected in the edaphic landscape and in two specific fields of the agricultural company "Valle del Yabú". The soil organic matter content was analysed by means of a conventional method and vis-NIR reflectance spectroscopy technique in the wavelength range of 379 - 1700 nm. Soil spatial variability was evaluated and calibration and validation models were developed using the multivariate statistical technique of partial least squares regression (PLSR). The model accuracy was assessed by the coefficient of determination R2. The organic matter showed a low to moderate spatial variability in the landscape and in the specific fields of Conyedo and Albarrán. The calibration model obtained in the landscape was evaluated as Successful with R2 of 0.95. This model was validated in the specific fields and an accuracy of Moderately Successful and Successful with R2 from 0.89 to 0.90 was obtained. The maps developed with the conventional values and estimated by vis-NIR spectroscopy showed similarities. The smallest spatial variability and the maximum values for real yield (6.15 t ha-1) and estimated yield (6.06 t ha-1) corresponded to the specific field of Albarrán.
In order to determine the soil organic matter content with the use of the visible and near infrared (vis-NIR) reflectance spectroscopy technique in areas cultivated with sweet potato, soil sampling was carried out between January and March 2016. The samples were collected in the edaphic landscape and in two specific fields of the agricultural company "Valle del Yabú". The soil organic matter content was analysed by means of a conventional method and vis-NIR reflectance spectroscopy technique in the wavelength range of 379 - 1700 nm. Soil spatial variability was evaluated and calibration and validation models were developed using the multivariate statistical technique of partial least squares regression (PLSR). The model accuracy was assessed by the coefficient of determination R2. The organic matter showed a low to moderate spatial variability in the landscape and in the specific fields of Conyedo and Albarrán. The calibration model obtained in the landscape was evaluated as Successful with R2 of 0.95. This model was validated in the specific fields and an accuracy of Moderately Successful and Successful with R2 from 0.89 to 0.90 was obtained. The maps developed with the conventional values and estimated by vis-NIR spectroscopy showed similarities. The smallest spatial variability and the maximum values for real yield (6.15 t ha-1) and estimated yield (6.06 t ha-1) corresponded to the specific field of Albarrán.
Descripción
Palabras clave
Suelo, Materia Orgánica, Espectroscopia, Reflectancia Vis-NIR, Boniato, Ipomoea Batatas L., Empresa Agropecuaria Valle del Yabú