Establishing a spatially explicit monitoring system for phosphorus use efficiency for crop fields

We evaluated a geospatially explicit phosphorus (P) use efficiency (PUE) monitoring method in crop fields using proximal sensing, field observations, and machine learning. Corn (Zea mays L.) yield and grain protein content were measured using an Ag Leader yield monitor and a CropScan sensor near Riesel, Texas. Topsoil P (0–15 cm) and grain P levels were analyzed for samples collected at strategic field locations. A random forest model was trained to predict PUE using soil electrical conductivity (EC_a) from a Veris instrument and topographic variables as predictors (R² = 0.78, root mean squared error = 0.01). CropScan sensor effectively estimated grain P content, supporting field-wide PUE upscaling. EC_a and elevation were the primary drivers of PUE variation. The resulting maps are valuable for monitoring PUE in crop fields and guiding variable-rate fertilizer applications. This scalable approach provides a robust framework for monitoring nutrient dynamics and efficiency, informing precision management strategies to enhance yield and sustainability in crop production systems.