Perhaps the greatest challenge in soil testing is calibration of the tests. It is essential that the results of soil tests be calibrated against crop responses from applications of the plant nutrients in question. This information is obtained from field and greenhouse fertility experiments conducted over a wide range of soils. Yield responses from rates of applied nutrients can then be related to the quantity of available nutrients in the soil.
The results of long-term soil test calibration studies on different soil types are then utilized to establish recommended amounts of plant nutrients to apply to a particular crop at a given soil test level. For instance, if the soil test P level is in the range of 0–10 ppm (which is low), the P recommendation for a 150 bu/acre corn crop may be 100 lbs/acre of P2O5; whereas, if the soil test P level is above 40 ppm (very high), the recommendation may be 0 to 20 lbs/acre.
In this example to the right, more than 85 percent of the fields testing very low in a particular plant nutrient may give a profitable yield response to the added nutrient. At a very high level of the plant nutrient, there is only a 15 percent probability of a profitable yield increase to the added nutrient. These values are arbitrary, but they illustrate the idea of expectation of response.
The tools of site-specific precision management now allow growers to manage more homogenous areas within fields. Some of those areas have much higher yield potential than the database with which most of today’s soil tests were calibrated indicates. This lack of calibration for high-yielding areas is one of the factors driving interest in using yield monitors and global positioning satellites to conduct strip trials to determine the adequacy of existing soil fertility programs. New precision ag tools have the ability to develop algorithms that allow for management of multiple site-specific zones within individual fields. This means a balanced crop nutrition prescription can be delivered to each square foot of every field.