Potassium dynamics and exchange equilibria in Loess-derived soils

A lack of cotton response to the recommended application of fertilizer potassium (K) has been well documented, prompting a renewed interest in the evaluation of soil and plant analysis. A study was conducted to evaluate varying rates and residual effects of fertilizer K on no-till cotton, in conjunction with the evaluation of cation exchange studies to establish the dynamics of K behavior in the loess-derived soils of west Tennessee. The soils were comprised of kaolinite, mica, and HIV-vermiculite. The effect of four rates of fertilizer K (0, 67, 134, and 269 kg K ha^-1) , on yield and plant and soil K levels was assessed. Total vs. acetic acid-extractable plant tissue content was studied as a function of tissue type (leaf and petiole) and sampling time (at bloom and 2 after bloom). Mehlich 1 (M1) vs. Mehlich 3 (M3) soil K was determined on samples collected biannually, prior to fertilization and post-harvest, as a function of soil sampling position (in-row (IR) and between-row (BR)) and soil depth (0-7.5 and 7.5-15 cm). In preparation for cation exchange, the soils were initially saturated with Ca²⁺, Mg²⁺, or K⁺ and equilibrated at times varying from 2 h to 4 w. Exchange isotherms were generated as a function of equilibration times and total normality (TN). The results from the field study showed that soil moisture had a pronounced impact on the plant and soil K content and subsequently on cotton yield, as precipitation in June 1997 exceeded that received in June 1998. The plant K digestion and extraction methods and the M1 and M3 soil tests were highly correlated suggesting either test be used for the determination of K levels. The plant tissue type and time of sampling were also significantly correlated, although petiole collected at bloom may prove to be a more sensitive K indicator. The benefit of residual fertilizer K increased with increasing fertilizer levels. The plant and soil test data were able to adequately evaluate crop response to fertilizer K and illustrated the effect of moisture stress. The results from the cation exchange study showed that K⁺ was preferred over Ca²⁺ and Mg²⁺ by the exchanger for heterovalent exchange and Ca²⁺ was preferred over Mg²⁺ by the exchanger for homovalent exchange. There were no effects of K fertilization on the exchange isotherms. However, a decrease in the CEC of the soils as a result of K fertilization indicated the collapse of the HIV-vermiculite layers. Exchange isotherms for the heterovalent systems varied as a function of equilibration time and as a function of initial cation saturation. The data suggested a slow equilibrium followed an initial rapid K⁺ fixation. In general, heterovalent cation exchange was hysteretic; while homovalent was reversible. The In K_v values of the heterovalent exchange systems decreased with an increased K⁺ loading, with the exception of the K⁺-saturated soils at 0.01 1N, suggesting decreasing selectivity for K⁺ by the exchanger as the amount of K⁺ adsorbed increased. The Ca²⁺ -Mg²⁺ exchange behaved ideally and reversibly as indicated by constant In K_v values. It can be concluded that the soils used in this study fixed relatively large quantities of fertilizer K, which became available for plant use over time.