Rapid Prediction of Nutrient Concentration in Citrus Leaves Using Vis-NIR Spectroscopy

Area of Science:

• Agricultural Science
• Plant Physiology
• Spectroscopy

Background:

• Traditional foliar ionomic analysis for crop nutrition is laboratory-intensive and destructive.
• Spectroscopy presents a viable alternative for non-destructive plant monitoring.

Purpose of the Study:

• To develop a calibration model for predicting citrus leaf macro- and micronutrient concentrations using spectral measurements.
• To establish a rapid, non-destructive method for assessing citrus nutritional status.

Main Methods:

• Collected 592 'Clementina de Nules' citrus leaves over several months.
• Measured spectral absorbance (430-1040 nm) using a portable spectrometer.
• Determined foliar ionomics via ICP-OES and nitrogen via Kjeldahl method; developed Partial Least Squares Regression (PLS-R) models.

Main Results:

• PLS-R models achieved determination coefficients between 0.31 and 0.69.
• Highest prediction accuracy was observed for Phosphorus (P), Potassium (K), and Boron (B) (0.60, 0.63, 0.69).
• Key predictive wavelengths for P, K, and B were identified within the visible spectrum (430-750 nm), correlating with foliage pigments.

Conclusions:

• Spectroscopy is a promising technique for rapid, non-destructive prediction of foliar macro- and micronutrients in citrus.
• The identified spectral regions offer insights into nutrient-related pigment changes in citrus leaves.
• This approach could significantly reduce the cost and time associated with crop nutritional diagnosis.