Quantifying the Variability and Bias of NH4-N and NO3-N Soil Sampling Techniques.

Nitrogen (N) application rate recommendations and modelling the N cycle in soil depend on measurement of soil NH4-N and NO3-N. Due to soil sampling and lab measurement costs, NH4-N and NO3-N concentration for a sampling area is typically measured using a single composite soil sample. However, the variability and bias associated with single soil samples is poorly characterized, particularly due to lability of NH4-N and NO3-N. The objectives of this study were to quantify the variability and bias of NH4-N and NO3-N measurements associated with common soil sampling techniques under different environmental and management conditions. Three soil sample types were assessed: 1) Random samples - composites of 10 randomly-positioned soil cores from the sampling area; 2) Row samples - composites of 5 cores from in crop rows and 5 cores from between crop rows; and 3) Board samples - composites of 16 cores collected across crop rows under even spacing. Samples were collected from 23 fields across the US Midwest under combinations of N application method, tillage, soil texture and crop rotation. At each location, 64 cores were collected from 1.5 m × 1.5 m sampling areas before planting and before sidedress application. Soil sample NO3-N and NH4-N variability was quantified for each sampling area using the standard deviation (SD) of the individual cores’ measurements. The results indicate that the SD of NH4-N and NO3-N can be estimated from the mean concentration of these two N species (R2 = 0.8). For a 10-core row sampled measurement of 25 ppm NO3-N, the estimated standard error is ±3.9 or ±6.6 ppm for broadcast or knifed N application methods, respectively. This data suggests that improved soil sampling procedures to account for variability are needed to enable broad utility of soil sampling data in digital agriculture.