USE OF HOT KCL-NH4-N TO ESTIMATE FERTILIZER IN REQUIREMENTS

We need an easy-to-use chemical index for estimating the amount of N t hat becomes available during the growing season, to improve N use effi ciency. This paper discusses how producers may, in future, use crop gr owth models that incorporate indices of soil N availability, to make m ore accurate, risk-sensitive estimates of fertilizer N requirements. I n a previous study, we developed an equation, using 42 diverse Saskatc hewan soils, that related potentially mineralizable N (N-0) to NH4-N e xtracted with hot 2 M KCl (X), (i.e., N-0 = 37.7 + 7.7X, r(2) = 0.78). We also established that the first order rate constant (k) for N mine ralization at 35 degrees C is indeed a constant for arable prairie soi ls (k = 0.067 wk(-1)). We modified the N submodel of CERES-wheat to in clude k and N-0 (values of N-0 were derived from the hot KCl test). Wi th long-term weather data (precipitation and temperature) as input, th is model was used to estimate probable N mineralization during a growi ng season and yield of wheat (grown on fallow or stubble), in response to fertilizer N rates at Swift Current. The model output indicated th at the amount of N mineralized in a growing season for wheat on fallow was similar to that for wheat on stubble, as we hypothesized. Further the model indicated that rate of fertilizer N had only minimal effect on N mineralized. We concluded that, despite the importance of knowin g the N-min capability of a soil, it is available water, initial level s of available N and rate of fertilizer N that are the main determinan ts of yield in this semiarid environment. The theoretical approach we have proposed must be validated under field conditions before it can b e adopted for use.