Potassium Recommendations for Corn in North Dakota

Comprehensive K fertilizer calibration work in North Dakota has never been conducted, since K soil concentrations in most of the state, until the last 20 years, have indicated very high levels. Since 1992, increased rainfall and well-adapted varieties of soybean and corn have resulted in a shift in production from small grains and sugar beets, to corn and soybean. This has resulted in greater K uptake and removal, so that soils that used to have soil test levels around 600 ppm K, have now dropped to levels near 100 to 200 ppm.

The objectives of the study are to: 1) determine the critical K level where corn yield increase with K fertilization is likely; 2) investigate whether the use of the moist K soil test is more predictive of crop response than the dry soil test; and 3) observe and assess in-season changes in soil test K. Ten sites in eastern North Dakota were examined, with most testing 100 ppm K or below when sampled in mid-April. The experimental design was a randomized complete block with six K treatments (0, 30, 60, 90, 120, and 150 lbs K₂O/A). Soil samples were obtained in-season from check plots twice each month at the 0 to 6 in and 6 to 12 in depth.

Results of the first year were largely unexpected. Three sites of the 10 sites were responsive to K at the 5% probability level and one site at the 10% probability level. Corn yields were generally high, with some low-K sites yielding up to 200 bu/A with no response to K. Two of the responsive sites, had K soil test levels of about 100 ppm in April, and the other two responsive K sites had levels near or above 150 ppm. Several sites had initial K tests of about 100 ppm, but did not respond to K. A moist, refrigerated and archived sample from each site was sent to a laboratory for mineral speciation and clay mineral quantitative analysis. There were unexpected variations in the levels of illite and smectite clay fractions. Certainly much more needs to be learned about K availability in North Dakota in order to better predict crop response.