Fertilizer Recommendation for Date Palm Using Multivariable Model.

Correct determination of the amount of fertilizer to be applied to a particular location in the field is central to the environmental and economic benefits of "variable-rate fertilizer application technology”. The objective of this study was to develop a code (FRDP) able to employ the results of soil testing, leaf analysis, and nominated site variable-value pairs to calculate fertilizer requirements for date palm (Phoenix dactylifera). At the core of FRDP was a multi-dimensional experimental data matrix that resulted from 67 fertilizer experiments, during two growing seasons that represented the range of soil fertility in a date growing region. The matrix was a collection of site vectors and yield vectors; and used to derive foliar and soil DRIS norms, a general yield regression model Y(N,K), and maintenance levels. The case base in FRDP was a combination of case space, feature space, and index space and included algorithms for assessing outliers and similarity between cases. If the foliar DRIS diagnosis (N,P,K,Fe,Mn, Zn, Cu) indicated that a micronutrient X was deficient then a remedial X fertilizer recommendation was suggested using a mass balance equation and maintenance level of X. FRDP employed site variable-value pairs (location, soil test values: N,P,K,Fe,Mn, Zn, Cu, salinity, organic matter, clay, silt, depth to saturation zone, and quantity and salinity of irrigation water), to compute the yield variables, regression coefficients, predicted yield vector Y(N,K), and subsequently the NK remedial recommendations. FRDP employed foliar DRIS diagnosis and P soil testing to make the P fertilizer recommendation. These FRDP generated fertilizer recommendations were experimentally validated during a third growing season. The date growers might, on the average, double their annual profits upon using the service offered by FRDP.