Date of Award
Master of Science
Lloyd F. Seatz
A. G. Sharp
With such a wide diversity of soils and with the different crop plants varying in their fertility requirements, the problem of when to fertilize, what fertilizer to use, and at what rate apply it, has long been the concern of both the agricultural scientist and the farmer. Many field fertilizer experiments have been inducted to determine the fertility status of the soil and the response of a particular crop to applied fertilizers. Since soils vary widely in their fertility levels, the complete cataloging of soil-crop-fertIlizer relationships Is a monumental task. The problem Is made more complex by other soil factors such as pH, texture, structure, and drainage as well as by certain environmental conditions such as temperature and rainfall. Moreover, field fertilizer experiments are quite expensive; a full growing season must elapse before the crop is harvested and the results are known. The best that can be done in field investigations is to conduct work on as many different types of soil as possible and to make predictions as to what might be expected on soils somewhat similar to those tested.
Many other tests have been devised to increase the scope of this experimentation and to obtain quicker results. These tests have ranged all the way from bringing the different soils to a central location and cropping them in field cylinders or in pots in the greenhouse, to subjecting small samples of the soil to certain chemical or biological tests in the laboratory.
Many laboratory tests, both chemical and biological, have been devised in an effort to rapidly determine the fertility of a large number of soils. While total chemical analyses are of -value in characterizing a soil, they have the serious weakness of failing to indicate what amounts of the various nutrients will become available to the plant over a rather short period of time. The unavailability to plants of an element in the soil is just as serious as an absolute lack of that element. Total chemical analyses have largely been supplanted by what are commonly called rapid tests. These rapid tests usually employ weak acid solutions or buffered salt solutions of those acids. Many extracting agents such as solutions of acetic, citric, hydrochloric, sulfuric, and perchloric acid have been used.
There is some disagreement as to what a chemical extraction of the soil is supposed to measure. Peech (13, p. 3) says that
…before results of chemical tests can predict accurately the fertilizer needs of the soil, the extracting solution employed in the chemical tests must thus simulate the ability of plant roots to obtain from the soil sufficient amounts of this different nutrient elements to meet requirements for normal plant growth.
Bray (5, p. 58), on the other hand, contends that the soil test must measure the total amounts of the available force of nutrients in the soil. He further adds;
The old idea that a soil test must simulate plant feeding and remove the available nutrients in amounts proportional to the removed by crops is fundamentally unsound and has led to no progress in soil fertility.
In either event, the reliability of any extracting agent depends upon how well it can be correlated with the degree of plant response to fertilizers when applied to soils of various nutrient levels. Even with the discovery of a satisfactory extracting solution, there still remains the problem of the amount of fertilizer that should be applied, if the need for fertilizer is indicated. Here again, the field experiments, if properly set up and conducted, can serve as a guide.
Many states, including Tennessee, are now making fertilizer recommendations to farmers using chemical rapid tests as the basis for such recommendations. The advantages of soil testing have been given much favorable publicity and this service has been generally well received by farmers. Its limitations are frequently overlooked and confidence in this routine procedure can be misplaced.
It is the purpose of the present paper to present yield data from fertilizer experiments on a number of field crops in Tennessee and to correlate these results with soil tests for available phosphorus and potassium, pointing out examples where both satisfactory as well as unsatisfactory correlations are obtained.
Long, Oddest H., "Correlations of soil tests for available phosphorus and potassium with crop yield responses to fertilization. " Master's Thesis, University of Tennessee, 1951.