Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Plant, Enviromental and Soil Sciences

First Advisor

Donald L. Robinson


Environmental concerns emphasize the requirement for diagnostic tools compatible with more intensively managed agricultural systems. Two recent approaches, the Barber-Cushman mechanistic model and the Diagnosis and Recommendation Integrated System (DRIS), appear to offer more accurate predictions of soil nutrient supply and more flexible means of assessing plant nutrient status, respectively, than conventional methods. The objectives of this research were to verify the aforementioned models for predicting K uptake and assessing P and K status of LA S-1 white clover. In growth chamber experiments, the Barber-Cushman model correctly predicted K uptake by plants for up to 21 days from Rita muck, Norwood and Providence silt loam and Ruston fine sandy loam soils at zero and 530 mg kg$\sp{-1}$ levels of added K. Over-prediction beyond 21 days may be partly due to an incorrect estimation of mean half-distances between root axes. In the Rita and Norwood soils where initial exchangeable K levels were high, prediction of K uptake was correct for up to 42 days. Under-prediction of K uptake from Providence and Ruston soils at the higher K level may have resulted from an underestimation of root parameters. The DRIS norms were developed from first year data of P and K rate studies on Dexter and Providence soils of adequate and relatively low P and K levels, respectively. Two subsequent years of data were used to verify diagnoses of plant P and K status using increasingly wider (zero to 16/3 $\sigma$) norm ranges in the calculation of DRIS diagnostic indices. The use of a norm range increased P and K overall-diagnostic accuracies as much as 23 and 27%, respectively, on the Dexter soil. On the Providence soil these values were 5 and 23%. Widest norm ranges resulted in a decrease in P and K response-prediction accuracies of 10 and 23% respectively, on the Providence soil. A norm range of 8/3 $\sigma$ appeared appropriate for use where the nutrients of interest were in relatively low supply. Wider norm ranges of 12/3 or 16/3 $\sigma$ were indicated for sites with adequate fertility.