Strategy for Innovation in Soil Tests Illustrated for P Tests

Abstract

Soil phosphorus (P) tests are used for P fertilization recommendations, environmental evaluations, and occasionally for legislation purposes. The basis of fertilization recommendation as function of soil P status was established in the 1950s–1960s. Since then the agroeconomic environment has altered: Environmental protection became increasingly important and P rock resources for fertilizers appeared exhaustible. Also, new insights in soil testing and fertilization recommendations reflecting more efficient use of P became available. However, these new insights seem hard to implement into agricultural practice, to a large extent because replacing existing soil tests and recommendations would imply a very significant effort with respect to introducing new tests and recommendations by fertilization trials in practice. The same would apply for environmental evaluations. Here, a novel, three-step schedule for introducing new soil tests is proposed: (1) establishing new promising soil tests, (2) creating regression models between the old and new soil tests, and (3) implementing the new soil test stepwise by fertilization trials. In this way, the knowledge based on the old soil tests can be used until the new soil tests and their subsequent crop responses are validated sufficiently. As a novel P test we considered combining soil P intensity [as reflected by P–calcium chloride (CaCl2)] with P capacity [as reflected by P-ammonium lactate (Al)] and P-buffering capacity (as reflected by P-Al/P-CaCl2 ratio) characteristics. Researchers tested whether this novel soil test can predict P water (Pw), P–calcium lactate / acetate (CAL), and P-Olsen values. To test the hypothesis, four datasets were used (two with Pw, one with P-CAL, and one with P-Olsen). In all datasets additional soil characteristics were available including soil type. Regression models with Radj 2 from 0.80 to 0.93 were obtained by using P-Al, P-CaCl2, and soil type. It can be concluded that these regressions can be used as a helpful intermediate instrument when introducing fertilization recommendations based on new soil tests. Predicting one soil P test out of other soil characteristics, analogous to the predicted Pw, P-CAL, and P-Olsen, could also be helpful in comparing P statuses of agricultural land in different nations.