Soil spatial variability in high-yield Peruvian Amazon coffee: a geostatistical approach for precision fertilization

• Autores: Mejía Maita, Sharon Yahaira, Quispe Matos, Kenyi Rolando, Díaz Chuquizuta, Henry, Rengifo Sánchez, Raihil Rabindranath, Mercado Chinchay, Ruth Lizbeth, Cuevas Gimenez, Juan Pablo, Solórzano Acosta, Richard Andi
• Formato: artículo
• Fecha de Publicación: 2025
• Institución: Instituto Nacional de Innovación Agraria
• Repositorio: INIA-Institucional
• Lenguaje: inglés
• OAI Identifier: oai:repositorio.inia.gob.pe:20.500.12955/3123
• Enlace del recurso: http://hdl.handle.net/20.500.12955/3123; https://doi.org/10.3389/fsoil.2025.1701602
• Nivel de acceso: acceso abierto
• Materia: Precision agriculture; Agricultura de precisión; Soil zoning; Zonificación de suelos; Coffee yield; Rendimiento de café; Applied geostatistics; Geoestadística aplicada; Soil fertility; Fertilidad del suelo; https://purl.org/pe-repo/ocde/ford#4.01.00; Coffea; Soil; Suelo; Fertilizers; Abono; Geostatistics; Geoestadística; Yield increases; Aumento del rendimiento

Fertilization practices in coffee plantations often overlook the spatial variability of soils, particularly in mountainous regions with acidic conditions. Although geostatistics has been used to map nutrient distributions, its integration with multivariate analysis to identify differentiated fertilization zones in coffee systems remains limited. This study evaluated the influence of soil properties, altitude, and crop age on coffee yield by combining principal component analysis (PCA) and ordinary kriging to design site-specific fertilization strategies. A total of 70 soil samples were collected from three districts of the Peruvian high jungle (San Martín and Amazonas), measuring physical and chemical properties, altitude, and crop age. The following analyses were applied: (1) Spearman correlations to assess associations with yield, (2) PCA to identify fertility gradients, and (3) geostatistical models with cross-validation. The PCA identified two main gradients: PC1 (32.41% of variance) associated with cation exchange capacity (CEC) and organic matter, and PC2 (17.88%) associated with the availability of K and P and crop age. Cross-validation confirmed high accuracy in the spatial prediction of available P and K across the three study areas. Kriging maps revealed zones with high available K (>150 mg kg⁻¹) and P (>20 mg kg⁻¹) associated with yields >1.5 t ha⁻¹. The integration of PCA and geostatistics enabled the delineation of management zones with differentiated nutrient requirements, reducing fertilization needs by up to 30% in areas with high fertility potential (e.g., Alto Saposoa). Overall, the results provide a solid methodological basis for implementing precision fertilization strategies in tropical coffee systems, promoting more efficient nutrient use and greater production sustainability.