Nitrogen‑phosphorus responses and Vis/NIR prediction in fresh tea leaves

Nitrogen and phosphorus are essential nutrients for the growth and development of tea plants.However, the nitrogen content (NC) and phosphorus content (PC) in different parts of fresh tea has not been paid attention. In this study, the NC and PC responses different nitrogen stress were analyzed, and a quantitative regression model for predicting NC and PC was established by using Vis/NIR spectroscopy and a variety of intelligent algorithms. Among them, NC and PC of different parts had significant difference. The selection of preprocessing algorithms has a significant impact on the predictive performance of the model. The VMDSG-D1-VCPA-IRIV-SVR prediction model for NC and the VMDSG-CARS-Stacking prediction model for PC have better prediction effects, and the correlation coefficients of the test set are more than 0.85, and the RPD is greater than 1.8. In conclusion, this study is helpful to guide the precise fertilization and in-situ detection of fresh tea leaves.