Near infrared (NIR) spectroscopy as a rapid and cost-effective method for nutrient analysis of plant leaf tissues

Prananto, J.A.; Minasny, B.; Weaver, T.

Advances in Agronomy, Vol 164 164: 1-49

2020


ISSN/ISBN: 0065-2113
DOI: 10.1016/bs.agron.2020.06.001
Accession: 071095907

Download citation:  
Text
  |  
BibTeX
  |  
RIS

Article/Abstract emailed within 0-6 h
Payments are secure & encrypted
Powered by Stripe
Powered by PayPal

Abstract
The efficient use of nutrients by plants can significantly improve the economic profitability and environmental sustainability of agricultural enterprises. Near infrared spectroscopy (NIRS) can provide a real-time, rapid, and non-destructive crop nutrient monitoring method to improve the timing of plant nutrient delivery. This review provides an analysis of the use of NIR (700-2500nm) spectrometer for measuring macro- and micronutrients in leaf tissues. It firstly discusses the history of NIRS in agriculture and the principle of NIRS in estimating leaf nutrient content. The review then discusses factors that influence the effectiveness of NIR spectra in estimating leaf macro- and micronutrient contents: sample preparation, the spectral range used, pre-processing and multivariate analysis methods, and conditions of application. Key findings are: (1) macronutrients (N, P, K, S, Ca, Mg) and micronutrients (Fe, Zn, Mn, Cu) can be predicted accurately using NIRS, (2) macronutrients are better predicted compared to micronutrients, (3) NIRS can detect macronutrients such as N, P, and S directly because they are major constituents of NIR-sensitive organic compounds, whereas micronutrients and macronutrients that exist mostly in inorganic forms such as Ca, Mg, and K are detected through association with organic compounds and indirect correlations with organic compounds, (4) dried and ground samples result in a better calibration compared to fresh leaf samples due to the standardization of moisture and particle size, (5) Partial least squares regression (PLSR) yields better model accuracy and robustness compared to other linear regression methods, (6) there are not many studies that use machine learning methods for calibration, and (7) NIRS performed better in a laboratory condition compared to field conditions due to interfering external factors such as moisture, temperature, solar radiation and leaf orientation. The overall review suggests that there is a great potential for the field application of NIRS to be used for in situ nutrient analysis of plant leaf tissue.