How soil sampling reflects DTPA-extractable micronutrients and their spatial variability in a cultivated field

This study investigated the spatial variability of DTPA-extractable Fe, Mn, Zn, and Cu in a cultivated field, their relationships with selected soil properties, and the optimum number of samples required to estimate mean micronutrient concentrations precisely. Forty-nine soil samples were collected at 5-m intervals from a square grid within a 30 × 30 m cultivated plot. Mean DTPA-extractable concentrations were 71.80 mg kg⁻¹ for Fe, 350.26 mg kg⁻¹ for Mn, 2.58 mg kg⁻¹ for Zn, and 4.11 mg kg⁻¹ for Cu. The coefficients of variation were higher for Fe (20.6%) and Zn (19.2%) than for Mn (11.8%) and Cu (8.4%). Spatial variability was best described by spherical models for Fe and Mn and exponential models for Zn and Cu. The spatial correlation ranges followed the order Zn (67 m) > Cu (17 m) > Fe (15 m) > Mn (8 m). Nugget-to-sill ratios indicated strong spatial dependence for Fe (0.039%), Mn (8.03%), and Cu (11.32%) and moderate spatial dependence for Zn (42.86%). In principal component analysis, Fe, Cu, and soil organic matter (OM) had high negative loadings on PC1, whereas Zn contributed most strongly to PC2. Fe was positively correlated with soil OM and negatively correlated with Mn, Zn, pH, Ca, and Na. Mn was positively correlated with Zn and Cu, whereas Zn was positively correlated with silt and pH. Cu was negatively correlated with pH, Ca, Mg, and K. The numbers of samples required to estimate mean concentrations within allowable errors of 5%, 10%, and 20% followed the order Fe (69, 17, and 4) > Zn (59, 15, and 4) > Mn (23, 6, and 2) > Cu (11, 3, and 1), respectively. Thus, greater spatial variability in a micronutrient requires a larger sample size to estimate its field mean precisely.