Multivariate diagnosis of nutrient imbalance in banana

Abstract

Abstract A data bank was established through field experimentation and survey of cropping enterprises for Robusta and Ney Poovan cultivars of banana. The cultivar differences were assessed by discriminant analysis. The two cultivars differed only for potassium (K) and calcium (Ca). Common multivariate compositional nutrient diagnosis (CND) norms were derived for routine diagnostic purposes using the data bank. Further, the principle component analysis (PCA) was used to extract the correlation structure among the nutrient elements in the high yielding population. The PCA conducted on log-transformed data provided three significant PCs and accounted for about 60% of the variance. The inverse relationship between nitrogen (N), phosphorus (P), and K on one hand and Ca and magnesium (Mg) on the other was highlighted in PC (N+P+Ca−Mg−), which explained about 23.6% of the variance. Elucidean distance ("d") computed from PCs provided nutrient element imbalance index for the crop. The diagnosis of overall nutrient imbalance in banana was made comparing the leaf nutrient concentrations with the CND norms. The imbalance of nutrients as reflected in "d" value was more with seven split applications of N and K compared to four split applications. The CND indices were instrumental in diagnosis treatment dependent P and zinc (Zn) imbalance in banana. Acknowledgments