Abstract
Plant demand for K in rice is comparable to nitrogen. With yields and management practices changing, refining K fertility management and decision-making tools is necessary. Our objectives were to determine (1) how soil K balances affect soil K indices, (2) the best soil test indicator of plant K availability, and (3) the relationships between plant and soil K indices. We assessed soil (plow layer) and flag-leaf samples from 55 commercial rice fields in California. Growers provided historical information on K fertility practices, straw management, and yields to develop a soil K balance. A soil K balance average of − 15 kg K ha −1 yr −1 (range: − 72 to 47) suggested an increased likelihood of K infertility; however, K balance was not correlated with soil K indices or flag-leaf K. This, plus the potential for K losses and fixation of surplus K, suggests that attempting to build up soil K may not be an effective strategy. Soil K indices were poorly correlated with plant K, but 1 N NH 4 OAc-extractable K (Kext) was the best of the K indices tested. Soils that were low in Kext had low clay content, CEC, and K saturation, a high Ca + Mg:K ratio, and showed evidence of K fixation. Soil K varied regionally and may be related to irrigation water and soil parent material. The critical level for Kext (currently 60 mg K kg −1 ) may need to be revised based on our findings, as there was evidence of K deficiencies above this concentration.