Sulfur Deficiency and its Effect on Cotton Production on Coastal Plain Soils

GROWTH response to application of sulfur has been reported for various regions and various crops of the North American continent. Sulfur deficiencies have been shown to exist in the podzols of Alberta and Minnesota (i, 5), the Pacific Northwest, and especially in the basaltic region of Oregon (2, 6). Deficiencies have also been indicated for scattered areas of the Middle West and South (3, 4, 7). In the northern regions the deficiencies have been discovered usually as a result of declining crops of legumes. In the southern areas the test crops usually have been legumes, cotton, or tobacco. In practically all areas the sulfur deficiency is accompanied by deficiencies of one or more of the primary elements, and in most cases has been masked by these. Sulfur deficiencies in Arkansas were first demonstrated under greenhouse conditions in 1937 on Coastal plain soils of the Ruston series with sudan grass (9). The yield response to sulfur amounted to about 20% of the comparable heavy treatment without sulfur. Greenhouse and field experiments conducted since that time with cotton have shown that several other series are deficient in sulfur. The soils responding occur both on upland and terrace and range all the way from clay to sand. In general, however, the exploited and depleted and the sandy soils are most likely to be deficient in sulfur, as well as in other nutrient elements. The treatments used in exploring sulfur deficiencies were obtained as follows: The treatments including sulfur consisted of ammonium sulfate, 20% and sodium nitrate, 16%, each supplying one-half the required nitrogen; potassium chloride, 50%, and superphosphate, 20%. For the sulfur-free treatment, the materials were sodium nitrate, potassium chloride, and superphosphate, 45%. Chemical tests for sulfur gave negative results for this mixture. All materials were of fertilizer grade, the nitrate and sulfate salts being synthetic products. These mixtures were distributed by hand in a furrow about 4 inches deep under the cotton rows. In this report the treatment unit is defined as one formula unit of 4-10-4 or the equivalent of 4 pounds nitrogen, 10 pounds PsOs, and 4 pounds K2O per acre. For the mixture with combined sulfur, the sulfur content is estimated at 12 pounds per treatment unit. Some idea of the extent to which sulfur deficiencies exist has been secured through a series of fertilizer experiments under cotton. The phases relevant to the sulfur problem are reported for 10 sites in Table i and Fig. i. At five other sites, not shown in the table, no significant yield difference was noted between the two test treatments, and consequently a sulfur deficiency was not established. It will be observed that of the 10 sites tabulated only 6 show a statistically significant response to sulfur treatment, 4 sites for the 1940 experiments indicating no significant sulfur deficiency at the treatment rates used. It is probable, however, that with continued heavy fertilization that some of these soils would soon register sulfur deficiencies. Because of the high statistical efficiency of the 1940 experiments and the broader experience provided, it appeared desirable to include all site data in this study. The response to sulfur varies with seasonal factors as shown for the three sites operating in 1938 and 1939. For the Chidester experiment both seasons were below the normal in rainfall, the second season being the most favorable., However, greater experimental variability increased the level of yield required for significance in 1939 and the increase due to sulfur failed to reach significance. For the Okolona site the 1938 August rainfall was about 0.25 inch, as shown by the weather records for nearby Boughton (8), and the lack of moisture severely retarded growth and reduced the probable yields especially for treatments with heavier growth, the sulfur-free treatments losing least thereby and eventually outyielding the sulfur treatment at the higher rate. At the Nashville site the 1938 season was favorable, although the rainfall was below normal each month from May to September. For the 1939 season the rainfall was again below normal and the AugustSeptember drouth was more severe, less than 2.0 inches of total rainfall being recorded for this period at the Highland weather station nearby. As in the case of the other sites quoted, the lack of moisture at