Suitability of Utilization of Compositional Nutrient Diagnosis on Diagnosis of Foliar Nutrient in Litchi and Longan

Abstract

【Objective】Determination of the high-yield production level is the premise of establishing foliar nutrient diagnosis indicators for crop. The problems of compositional nutrient diagnosis (CND) approach in determining the high-yield levels of litchi and longan were examined, with the aim to identify whether CND is a suitable approach for diagnosis of foliar nutrient in litchi and longan.【Method】A total of 511 leaf samples of litchi (Litchi chinensis Sonn) (c.v. Feizixiao) and 262 leaf samples of longan (Dimocarpus longan Lour) (c.v. Chuliang) were collected at terminal shoot maturing stage and fruit bulking stage in litchi and longan orchards in South China during 2016-2019. The fruit yields of all sample trees were recorded at each harvest period. Foliar nutrient (N, P, K, Ca, Mg, S, Fe, Mn, Cu, Zn and B) contents in all samples were determined. Based on the principle of CND approach, the models for the accumulative variance function of foliar nutrient content and fruit yield of litchi and longan were fitted. The appearance of anomalous extremes was examined, and their effects on the models were further analyzed. Then, the variations of the cutoff values before and after the elimination of anomalous extremes were compared, and the influences of different high-yield cutoff value selection principles on high-yield population were investigated.【Result】Abnormal values appeared in the models for fruit yield of both litchi and longan at different growth periods. The removal of the abnormal values had significant influence on the yield models and nutrient cutoff values, and the influence was irregular. No matter what kind of principle is used to determine the high-yield threshold, there were problems such as the small size of the high-yield population, the low proportion of high-yield population or the great discrepancy between the calculated yield level and the actual yield level. Moreover, the calculated high-yield threshold considerably varied with growth stage and year, indicating poor stability. In addition, there was a strong subjectivity in the treatment of abnormal values and the selection of high-yield threshold.【Conclusion】There are obvious shortcomings when CND approach is adopted to determine the high-yield population. The obstacles are probably ascribed to the joint restrictions of the sample size, the foliar nutrient content, and the counting scale of the present study. Therefore, the suitability of CND approach for foliar nutrient diagnosis in perennial woody fruit trees with alternative fruit bearing needs to be carefully evaluated.

Keywords

• litchi
• longan
• foliar nutrient diagnosis
• perennial woody fruit tree
• compositional nutrient diagnosis
• high-yield population