Grid Sampling Procedures

Consensus has not been reached on exactly how to sample for maximum benefit from grid sampling. The primary concerns being researched include:

— size of grid cell

— number of samples taken at each grid cell

— sampling scheme at each grid cell

— procedure for making application maps

— how many levels of a nutrient to apply

— how to determine boundaries between levels

— how to interpret soil test values

— how often to sample

A sample grid map is laid out on the field with equal distant spacing between the intersection points. The size of the grid cell is the distance between the intersection areas. Spacing distances as short as 100 feet have been used, but many fields are sampled at distances up to 400 feet. It is probably better to use as short a distance as possible and sample less often. For most non-mobile nutrients such as phosphorus, potassium, zinc and pH, sampling once every five years is sufficient. The grid sample size can be adjusted for areas of high variability. Sampling in grids is a little different than sampling on a field basis. Within each cell, a sample can either be a random point or a composite of several samples. Be sure there is enough soil for analysis. The samples should not be taken on a transect in order to avoid application patterns or other repetitive events that might have occurred.

The nutrient distribution maps are based on various techniques that are constantly being invented and refined. Specific map-making techniques might have specific sampling methods. The methods used now are kreiging, inverse distance, splines and tinning. Check with the firm making the maps or software company to know how to sample and what the maps represent.

Site-specific management is changing daily, so these suggestions may become quickly outdated. The solutions to the problems discussed above will be a combination of economic calculations and agronomic data. It is beyond the scope of this lesson to address these concerns in depth. However, you need to know the underlying assumptions used when constructing any map.