Abstract
Measures of soil electrical conductivity (EC) and elevation are relatively inexpensive to collect and result in dense data sets which allow for mapping with limited interpolation. Conversely, soil fertility information is expensive to collect so that relatively few samples are taken and mapping requires extensive interpolation with large estimation errors, resulting in limited usefulness for site-specific applications in precision agriculture. Principal component (PC) analysis and cokriging can be applied to create meaningful field scale summaries of groups of attributes and to decrease the estimation error of maps of the summarized attributes. Deep (0–90 cm) and shallow (0–30 cm) EC, elevation, and soil fertility attributes were measured in fields under corn (Zea mays L.) and soybean (Glycine max L.) rotations, at two sites in Illinois (IL) and two sites in Missouri (MO). Soil fertility and topography attributes were summarized by PC analysis. The first topography PC (TopoPC1) contrasted flow accumulation against elevation and curvature, to describe the main topographic pattern of the fields. The first soil fertility PC (SoilPC1) consistently grouped together cation exchange capacity (CEC), Ca, Mg, and organic matter (OM). SoilPC1 was well correlated to soil EC for all sites and cokriging with EC had higher r 2 in the crossvariogram models compared to ordinary kriging. The second and third soil fertility PCs (SoilPC2 and SoilPC3) were concerned with soil pH and P, and reflected historic land use patterns. Maps of SoilPC2 and SoilPC3 had little relationship to soil EC or topography and so could not be improved by cokriging. Abbreviations: CEC – cation exchange capacity; EC – soil bulk electrical conductivity; OM – soil organic matter; PC – principal component; SoilPC – soil fertility principal component; Sph – spherical variogram function; TopoPC – topography principal component.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have