CRITICAL LIMITS OF SOIL TESTS FOR K AS DERIVED FROM RELATIONSHIPS BETWEEN TOPSOIL K AND TISSUE WATER K OF PLANTS

A statistical procedure is proposed to simultaneously determine minimu m critical soil test values for K and critical plant K concentrations from relationship between topsoil K and the K concentration in the tis sue water of winter wheat plants. Soil K was extracted from both, fiel dmoist and airdried samples, using the conventional NH(4)acetate (exch angeable K), CAL and CaCl2 procedures. The closest soil-plant correlat ion (r(2)=0.602) was found for the exchangeable K of moist soils, the weakest for the K-CaCl2 of dried samples (r(2)=0.547) and KCAL was inb etween (r(2)=0.583). In all cases saturation curves were obtained appr oaching constant levels of 207 mmol K l(-1) tissue water (plant critic al value) when asymptotic models were fitted to the data. The amount o f soil K related to the onset of the plateau is defined as the soil K critical level. It is derived by fitting several asymptotic functions to the data among which the RICHARDS function (RICHARDS, 1959) perform ed best. Tentative critical limits for the routine NH4 extraction are 1.25 mg K g(-1) clay for moist and 1.12 mg K-1 clay for dried soils, r espectively, and for the CAL procedure 0.85 mg K g(-1) clay for moist and 0.93 mg K g(-1) clay for dried samples, respectively. For K-CaCl2, moist and dried soil samples yielded similar critical values of about 80 mg K g(-1) soil. Critical values like these appear to be useful li mits when reductions of soil K levels in previousely overfertilized so ils are in progress.