| Abstract: | Abstract Soil tests are used to predict fertilizer and liming needs of legumes being introduced into permanent pastures. However, the routine drying of soil samples may change the soil characteristics so much that recommendations based on test results may be poorly related to actual plant needs in the field. Test results from four soils, based on dried and moist samples, were related through stepwise multiple regression to percent nodulation and dry weight of white clover (Trifolium repens L. ) seedlings grown on soil cores in the greenhouse. Highly variable test results were obtained for Al, Mn, and P, depending on if a sample was dried or maintained moist. These discrepancies have implications for making P fertilizer and liming recommendations. Test results for other elements also varied widely. Extractable Mn and P accounted for most of the variation observed in percent nodulation, whether soil samples were dried or not. There was a significant interaction of P, Al, and Mn extracted from dried soils with plant weight. This contrasted with the significant interaction of extractable P, Ca, K, and Mg with plant weight observed when soil samples were not dried before analysis. Drying soil samples before analysis improved, rather than harmed, the ability of soil tests to predict white clover growtl responses on these soils. However, the soil tests explained only 40 percent of the variation in plant growth. |
