Spatial econometric analysis of a field-scale site-specific nitrogen fertilizer experiment on wheat (Triticum aestuvum L.) yield and quality

Knowledge of site-specific response may help farmers to tailor their management decisions with the help of precision farming technologies. However, farmers often have only a vague idea of the economic potential for site-specific management of their fields, which is important for investment decisions on precision farming technologies. This study presents an on-farm experimental approach to identify the economic potential of site-specific fertilization strategies at low costs. A strip trial with winter wheat (Triticum aestivum, L.) was established with precision farming technologies. Twelve different nitrogen fertilizer rates split in two applications were applied to 30 plots over a total strip length of 1.5 km. Geo-referenced yield was recorded with the harvester. Furthermore, electrical conductivity of the soil was measured and grain quality was surveyed with hand selected samples. With the help of advanced spatial statistical methods, within-field site-specific response was modeled with sufficient accuracy at comparably low costs. Electrical conductivity of the soil, elevation above sea level, and derivates of a digital elevation model were used as covariates to identify a possible economic potential for site-specific fertilization. Yield and protein response was best predicted with spatially adjusted regression models with site-characteristics or their interaction with management variables. Protein content was essential for achieving best economic results. The economic potential for site-specific fertilization strategies for the analyzed field was below 2 €/ha. However, the approach to identify the potential may be transferred to other locations with greater potential for site-specific farming.

Research highlights▶ Precision agriculture technologies can be used for on-farm experiments. ▶ Specific errors due to experimental design can be addressed with statistical models. ▶ Knowledge of spatial patterns of site-characteristics improve response models.