Optimization of Leaf Nutrient Mapping in Oil Palm A Geospatial Comparison for Site-Specific Management

Abstract

Accurate nutrient mapping is crucial to optimize fertilizer application and ensure sustainable oil palm production. In this study, four spatial interpolation methods, including Inverse Distance Weighting (IDW), Ordinary Kriging (OK), Universal Kriging (UK), and Radial Basis Function (RBF), were evaluated and compared to estimate leaf macronutrient content in oil palm plantations across North Sumatra, Indonesia. A total of 3,191 georeferenced leaf samples were collected from the 17th frond and analyzed for nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), and magnesium (Mg). Spatial interpolation was conducted using each method, and performance was evaluated based on the Mean Error (ME), Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE). The results showed that all methods achieved acceptable prediction accuracy, with MAPE values ranging from 3% to 10%. OK and IDW performed optimally across most nutrients, with excellent predictions for nitrogen, potassium, calcium, and magnesium. The resulting maps facilitated site-specific nutrient management, offering a cost-effective and environmentally friendly alternative to uniform fertilization practices. This study supports precision agriculture by enabling an accurate site-specific nutrient diagnosis across extensive oil palm plantations.