Jose, A IKrishna Kumar, NKAU2018-12-032018-12-031983http://krishikosh.egranth.ac.in/handle/1/5810085090A study was undertaken to standardize the foliar diagnostic technique in coconut palm and to work out regression models for predicting the yield based on foliar nutrient contents. Palm were selected from three different zones of Kerala State, namely the Coconut Research Station Balaramapuram, the Agricultural Research Station, Mannuthy. And the regional agricultural research station Pilicode. Leaf samples drawn from the leaf positions 2, 10 and 14 separately from each palm were analysed for nitrogen, phosphorus, potassium, calcium, magnesium and sodium. Attempts were made to standardize the leaf position, the nutrient status of which will best reflect the yield and to establish the critical levels of the nutrients in the index leaf. Regression models were also worked out to predict the yield based on tissue nutrient contents and the number of leaves retained by the palm. Observations revealed that application of nitrogen, phosphorus and potassium resulted in an increase in the content of these nutrients in the 2nd, 10th and 14th leaves. The number of leaves retained by the palm was mainly a function of potassium applied. The leaf number was highly correlated with the potassium per cent of the leaf lamina of the three leaf positions the highest correlation of 0.710** was registered for the leaf position 10. The number of leaves retained was also significantly correlated with yield (r = 0.7335**). The optimum number of leaves to be retained for maximum production was worked out to be 46.62. Yield of the palms was significantly correlated with the nitrogen per cent of leaf lamina of 2nd, 10th and 14th leaves, the highest coefficient of partial correlation being registered by the 10th leaf (r= 0.499**). The partial correlation coefficients between yield and the phosphorus per cent of leaf lamina of the three leaf positions were not significant. The coefficient of partial correlation between yield and potassium per cent of leaf lamina of leaf position 2 and 10 were significant, the highest value of 0.432** being recorded by the 10th leaf. On the other hand, the contents of calcium, magnesium and sodium in the leaf lamina showed significant correlation with yield only in the case of the leaf position 14. The optimum contents of nitrogen and potassium in the 10th leaf for maximum yield was 2.9 and 1.8 per cent respectively. Yield prediction models worked out using the percentage of nitrogen, phosphorus, potassium, calcium, magnesium and sodium, and the leaf number indicated that the model worked out for the 10th leaf had the maximum accuracy of prediction. Models worked out eliminating calcium, magnesium and sodium also confirmed the supremacy of the 10th leaf for the prediction of yield. Thus the leaf lamina of the leaf position 10 can be recommended as the best tissue for foliar diagnosis in coconut. Yield can be predicted with an accuracy of 85.3 per cent by the regression model, Y = -92.924 + 44.682 N – 0.0004 P + 49.397 K + 6.292 L – 6.970 NxP + 30.729 NxK – 2.218 LxN + 17.449 PxK – 0.205 LxK Utilizing nitrogen (N), phosphorus (P) and potassium (K) contents of the leaf lamina of 10th leaf and the number of leaves retained. Yield can also be predicted with an accuracy of 86.2 per cent based on the following regression model worked out for the leaf position 10. Y = -34.619 + 29.594 N – 33.827 P + 51.279 K + 6.547 L +23.646 N2 – 0.932 NxP + 10.044 NxK – 2.493 LxN +20.294 PxK – 54.768 K2 + 0.378 LxK.ennullThesis