Investigating soil physical properties and yield response in a grassland field using a dual-sensor penetrometer and EM38

Precision-farming applications are mainly based on site-specific information of soil properties at the field scale. For this purpose, a number of novel sensor techniques have been developed but not intensively tested under different field conditions. This study presents a combined application of a self-developed dual-sensor vertical penetrometer (DVP) for measuring volumetric soil water content (VSWC) and cone index (CI), and an EM38 for soil apparent electrical conductivity (ECa) in a pasture (1.4 ha). To verify the feasibility of the DVP for interpreting the depth-specific information in the field, not only the soil physical properties and their geographical coordinates were measured, but also geo-referenced yield data were collected. We found that the yield pattern was quite similar to the soil water-content pattern of each layer (layer-1: 515cm; layer-2: 1525cm, layer-3: 2535cm) and ECa pattern. Using the map-based comparisons in conjunction with the statistical analyses, the effect of each measured soil physical property (VSWC, CI, and ECa) on the yield was investigated. The regression between the yield and VSWC at each layer fitted a quadratic equation (R2 = 0.515 at 515cm; R2 = 0.623, at 1525cm; R2 = 0.406 at 2535cm). The negative correlation between yield and CI at each layer fitted a linear model with R2 0.510.