Spectral characterization and nutrient prediction of litchi canopy leaves during flower bud differentiation

In order to achieve efficient monitoring of litchi nutrient and elevate the litchi management level in precision fertilizing, the research on the litchi nutrient and the spectral data of litchi canopy leaves during flower bud differentiation was performed serving to high yield and safety production of litchi planting in South China. The four forms of spectral data were analyzed. The results showed that the spectrum of litchi canopy leaves had common spectral characteristics of a plant. Based on the laboratory analysis on total nitrogen, total phosphorus, chlorophyll, and organic C, the relationships between spectral data and nutrients were investigated. Those spectral bands with the most significant correlation were chosen to create regression forecasting equations. The results showed that the first-order derivative of reflectance had the capability to estimate chlorophyll and organic C, and the sensitive diagnostic bands were at 1 562 nm (r=0.8944) and 1 720 nm (r=0.7827), respectively. The most-sensitive diagnostic spectral bands of total nitrogen and total phosphorus were the first-order derivative of reflectance spectrum at 2 059 nm (r=0.8073), 1 347 nm (r=0.7575), respectively. The exponential function was best for predicting the total nitrogen (RMSE=0.002730), chlorophyll (RMSE=0.008138) and organic C (RMSE=0.000416). Linear function fitted the prediction of total phosphorus best (RMSE=0.000336). The results have great significance on implementation of precision fruit agriculture and the environmental protection of litchi agriculture in South China.