In order to better explain spatial variations within fields, soil physical properties need to be studied in more depth. Relationships between soil physical parameters and yield, especially in the subsoil, are seldom studied since the characterization of soil variability at field or subfield scale using conventional methods is a labor intensive, very expensive, and time-consuming procedure, particularly when high-resolution data is required. However, soil physical properties may be able to explain a significant proportion of the variations in yield.

 

One property that can be measured fairly easy is soil penetration resistance. Maps of soil penetration resistance could reveal areas potentially limiting root growth and thus yield since restricted root growth may lead to temporal drought stress which limits the potential yield especially under rain fed crop production. Since the mechanical properties of the soil around the penetrating cone are complicated and depend on several factors, the result from a penetrometer can be difficult to interpret. The low specificity should however not be seen as a disadvantage rather as a way to detect any change in soil physical conditions. The influence of soil moisture on the penetration resistance can be minimized by restricting measurements to conditions of constant soil moisture conditions, e.g. at field capacity.

 

To measure the penetration resistance across entire fields efficiently a tractor pulled penetrometer was used. A modified parallel-plough support three horizontally mounted soil-penetrating cones. The cone-angle was 30° with a base diameter of 6.3 cm. The device was able to simultaneously measure at 10, 30 and 50 cm depth. In addition speed and position were recorded with a GPS and depth was measured with a separate system. The construction can be used in the field at a speed of 1.5 m/s. The 1000 Hz signals from the draught sensors are filtered and the result is a 1 Hz signal. This horizontal, on-line penetrometer can measure a large area considerably faster than a traditional penetrometer, which makes it easier to incorporate soil physical parameters into precision agriculture. The horizontal penetrometer shows values that are in the same range as a standard vertical penetrometer.

 

Four different fields in Sweden were used in this study. Yield was compared with penetration resistance as well as EM38-measurements at field scale. Two of the fields were also deep loosened in strips across the field, about half the fields were deep loosened. The loosening was done at autumn and no positive effects on yield were found the year after, the crops were autumn and spring rape seed. Penetration resistance varies considerably across all fields. There are however strong indications that penetration resistance can be used to find areas where the physical state of the soil has limited yield. Further years, as well as the EM38-measurements, will be analysed during this winter.