| | The Alberta Environmentally Sustainable Agriculture Program (AESA) Soil Quality Resource Monitoring Program has provided a data set containing soil organic matter (OM), nitrogen (N), phosphorus (P), potassium (K), electrical conductivity (EC), and pH for three AESA sites in central Alberta. The data set has been used to assess the spatial variability of soil properties and evaluate the current sampling scheme. At each site, soil sampling locations along a transect are marked with global positioning systems (GPS) at the upper, mid and lower slope positions. Two transects at each slope are sampled to evaluate the spatial dependence of the selected soil properties. Spatial analysis of the soil properties will quantify the distance of spatial correlation with a view to expanding the current sampling scheme of 2 m.
.
Current AESA soil sampling protocol involves sampling from the 0 to 2 m area surrounding a central marker. A preliminary study by Cathcart and Haung, (2005), indicated that this area could be expanded to 4 m to provide additional sample locations. To confirm this finding, three AESA sites, Dapp (681), Carvel (684), and Tomahawk (692) were sampled in October, 2005. Soil from 162 locations at the 0-15 cm depth is analyzed for OM, N, P, K, EC and pH.
The variability of sampled soil nutrients is highest for K and lowest for EC based on the variance. The coefficient of variation (CV) confirmed this observation. All soil nutrients are positively skewed and most of the soil nutrients have a negative kurtosis. Spatial trends are evident for all soil variables except N which shows no clear trend. Histograms for each of the soil nutrients visually confirmed these descriptive statistics.
The range of spatial correlation determined by modeling the experimental, isotropic variogram varies among the soil properties from 3.5 m for EC to 21 m for P. Spatial dependence generally appears to be longest for phosphorus followed by organic matter, potassium and pH, EC and nitrogen. The N experimental, omni directional variogram is uncorrelated when calculated over all 3 sites but spatial correlation occurred at the mid slope and Carvel (684) site. Mobility of these elements in the soil, parent material, climate and topography may explain why N is uncorrelated at short distances while the other soil properties are spatially dependent at greater distances.
Directional variograms of the soil properties indicated that the short range spatial correlation varied from 1 to 20 m while long range correlation varied from 10 to 40 m. The strongest direction of correlation is in an east-west direction except for pH which is north-south. There are no directional trends for nitrogen. Consequently, the sampling area may be expanded up to 6 m in the direction of the slope and 4 m in the opposite direction from the central marker except EC which can only be expanded 1 m in the north-south direction. These directional trends are not evident at the individual AESA sites and additional transects at each site are needed to verify this conclusion.
Spatial analysis of selected soil properties sampled at three AESA sites indicated that the sampling area could be expanded except for N. Further soil nutrient sampling in different ecoregions will validate this initial conclusion. |
|