Due to initially high soil test K values, K soil test correlation and calibration for corn in North Dakota has previously not been intensely investigated. Potassium fertilizer rate experiments were conducted on 25 sites from 2014 to 2016. The previously published soil test K critical value of 150 mg kg−1 predicted crop response correctly at 16 of the sites. Alternative soil test methods, including a resin-based extraction at two timings, sodium tetraphenylboron extractions at two timings, and 1 mol L−1 NH4OAc extraction using moist soil were conducted; however, the currently used 1 mol L−1 ammonium acetate extraction using dry soil was most predictive. Mineral analysis of soil from all sites was determined for potassium feldspar content of whole soil, and clay species, particularly smectite, illite, and kaolinite, were determined on the clay fraction. Cluster analysis revealed that a smectite/illite ratio of 3.5 separated the sites into two unique K response data sets. Sites with a smectite/illite ratio >3.5 had a K critical level of ~200 mg kg−1, whereas sites with a smectite/illite ratio <3.5 had a K critical level of ~130 mg kg–1. For soils with K soil tests between 130 and 200 mg kg−1, consideration of clay chemistry improves the predictability of crop yield response with K fertilization.
Bibliographical notePublisher Copyright:
© Soil Science Society of America.