Is Soil-Test Potassium Meaningful for Crop Production and Soil Fertility Management in Arkansas?.

The utility of soil-testing for estimating soil-K availability and making fertilizer recommendations has been questioned requiring that relevant research be summarized to identify its successes and failures. This presentation will summarize correlation-calibration relationships between Mehlich-3 soil K and crop yield and nutrient uptake research conducted in Arkansas. Research with rice (Oryza sativa) and irrigated soybean (Glycine max) show Mehlich-3 K in the 0-10 cm depth is a good predictor of tissue K concentration (r² = 0.73-0.97) at key growth stages and relative crop yield (r² = 0.63-0.76). Despite these strong correlations results for a single, long-term experiment (1:1 rice:soybean rotation) indicate problems with K availability measurements. The long-term trial showed the annual decline in crop yields and soil-test K was 1.9% (n=14 and r² = 0.42) and 3 mg kg^-1 (r² = 0.37), respectively, for a Calhoun silt loam when no K is applied. The relationship between relative crop yield and soil-test K was not significant (p=0.1865; r² = 0.14), but calculation of a relative soil-test K improved the relationship describing the annual soil-test K decline (r² =0.56) and the correlation with relative yield (r² = 0.43). In a contrasting cropping system, preliminary correlations for bermudagrass (Cynodon dactylon) season-total yield or K uptake with soil-test K were strong (r² = 0.81-0.86).  For one 5-yr trial, the rate of soil-test K depletion across years was strong and non-linear (r² = 0.87) with reductions of 22.1 mg kg^-1 yr^-1 the first four years.  Across all forage trials, soil-test K changed linearly by 0.56 mg kg^-1 for each 1 kg K ha^-1 added or removed (r² = 0.52). Soil-K-availability indices may be better correlated for cropping systems with greater K demands as greater K removal may minimize the effect of annual fluctuations in K availability indices.