Doctoral Dissertations

Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Agricultural Economics

Major Professor

Joe A. Martin

Committee Members

Irving Dubov, Luther Keller, Keith Phillips


This study analyzed the effects of varying soil moisture conditions on per acre corn yields for selected counties in West Tennessee. The basic assumption underlying the study was that abnormal moisture conditions during the growing season adversely affected corn yield. It was hypothesized that too much as well as too little available soil moisture contributed to reductions in per acre corn yields.

To test the hypothesis, a 14 county area of West Tennessee was selected as a study area. These counties were selected because they produced a substantial portion of the total corn crop grown in Tennessee.

The objectives of the study were as follows:

1. To make quantitative estimates of corn yield response to soil moisture and precipitation variables during the critical period of the corn growing season in selected counties in West Tennessee.

2. To evaluate the incremental changes in moisture variables during the critical months of the corn growing season.

3. To construct precipitation probability distributions for counties included in the study, and to use the distributions to predict changes in monthly moisture conditions.

To achieve the objectives, two types of analyses were used. One involved the use of Palmer's Drought Index and the other included the use of two-week precipitation totals.

The regression equation containing the Index explained about 85 percent of the corn yield variation. The results indicated that for drier than normal conditions during the months of May, July, and August, increases in the Index generally brought about increased corn yield. Increases in the Index to slightly wetter than normal were associated with increased yield. Further increases in the Index, to extremely wet conditions, were associated with decreased yield. Results for June were irrational with respect to the signs and magnitudes of the coefficients.

The regression equation containing inches of precipitation by two-week period explained about 38 percent of the corn yield variation. Results of this analysis closely paralleled those of the equation containing Index values. However, coefficients for the first two-week periods in May, June, and August exhibited irrational signs.

As precipitation rose from 0.00 to 4.00 inches, corn yield first increased at a decreasing rate then declined. Marginal physical product became negative at 4.00 inches of precipitation in the second period of both May and June, and at 1.00 inch in the second period of July. In the first period in July and the second period in August, marginal physical product did not diminish even with 4.00 inches of precipitation.

A combination of data from Palmer's Drought Index and precipitation probability distributions provided a model which was used in predicting changes in soil moisture conditions. The model accurately predicted monthly changes during the May through August period about 61 percent of the time. This model was seen as a valuable tool for both policy makers and for individual producers.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."