Near-infrared spectroscopy for rapid and cost-effective nutritional diagnosis of oil palm: An original study
Abstract
Oil palm (Elaeis guineensis Jacq.) is one of the world's most profitable tropical crops. Profitability of this crop is mainly linked to correct management of fertilization generally based on foliar analysis. Conventional chemical analyses remain costly, slow and logistically demanding. Near-infrared (NIR) spectroscopy offers a rapid and cost-effective alternative, yet broad deployment is limited by the difficulty of building calibration models that remain accurate across diverse conditions. This study evaluated the feasibility of developing a generic NIR calibration for predicting major foliar nutrients in oil palm and assessed the usefulness of a spiking strategy to improve model transferability. A dataset of 436 leaf samples collected from 11 sites on three continents was assembled to maximize variability. Partial least squares models were built using raw spectra in the 1250–2530 nm range and evaluated through cross-validation, leave-one-group-out validation and spiking. High calibration performances were obtained for most macronutrients (0.80 < r² < 0.98). However, prediction accuracy dropped substantially when entire fields were excluded from calibration, particularly for nutrients with weaker NIR signatures such as P, K and Ca. Spiking consistently improved prediction ability compared to leave-one-group-out validation, and correcting field-specific biases, while requiring only limited additional laboratory analyses. These results demonstrate that combining a diverse global calibration dataset with targeted spiking provides an efficient and scalable framework for NIR-based foliar nutrient diagnosis in oil palm. With the growing availability of portable NIR devices, this approach has strong potential for on-site, real-time nutrient monitoring in both industrial plantations and smallholder contexts.
Related
- Near-infrared spectroscopy for rapid and non-destructive assessment of PFAS effects in blue mussels (Mytilus spp.): a proof-of-concept study
- Monitoring a Polyurethane Synthesis by Fiber-Coupled Attenuated Total Reflection Fourier Transform Infrared Spectroscopy and Multivariate Curve Resolution-Alternating Least Squares
- Solvent-Dependent Self-Association of Chiral 2-Acetoxy-2-Phenylacetic Acid Revealed by Infrared and Vibrational Circular Dichroism Spectroscopy
- Gas-Phase Far-Infrared and Rotational Spectroscopy of 1-and 2-Cyanonaphthalene: Experiment and Theory
- Comprehensive assessment of prediction of pH in red meat based on near-infrared spectroscopy (NIRS)
BibTeX
@article{vrignonbrenas2026nearinfrared,
title = {{Near-infrared spectroscopy for rapid and cost-effective nutritional diagnosis of oil palm: An original study}},
author = {Sylvain Vrignon-Brenas and Valentin Avit and Jean Ollivier and Vivien Sarazin and Albert Flori and Gilles Chaix},
journal = {Smart Agricultural Technology},
year = {2026},
month = nov,
doi = {10.1016/j.atech.2026.102626},
url = {https://hal.science/hal-05849319v1},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - Near-infrared spectroscopy for rapid and cost-effective nutritional diagnosis of oil palm: An original study
AU - Sylvain Vrignon-Brenas
AU - Valentin Avit
AU - Jean Ollivier
AU - Vivien Sarazin
AU - Albert Flori
AU - Gilles Chaix
PY - 2026
DA - 2026/11/03
JO - Smart Agricultural Technology
DO - 10.1016/j.atech.2026.102626
UR - https://hal.science/hal-05849319v1
AB - Oil palm (Elaeis guineensis Jacq.) is one of the world's most profitable tropical crops. Profitability of this crop is mainly linked to correct management of fertilization generally based on foliar analysis. Conventional chemical analyses remain costly, slow and logistically demanding. Near-infrared (NIR) spectroscopy offers a rapid and cost-effective alternative, yet broad deployment is limited by the difficulty of building calibration models that remain accurate across diverse conditions. This study evaluated the feasibility of developing a generic NIR calibration for predicting major foliar nutrients in oil palm and assessed the usefulness of a spiking strategy to improve model transferability. A dataset of 436 leaf samples collected from 11 sites on three continents was assembled to maximize variability. Partial least squares models were built using raw spectra in the 1250–2530 nm range and evaluated through cross-validation, leave-one-group-out validation and spiking. High calibration performances were obtained for most macronutrients (0.80 < r² < 0.98). However, prediction accuracy dropped substantially when entire fields were excluded from calibration, particularly for nutrients with weaker NIR signatures such as P, K and Ca. Spiking consistently improved prediction ability compared to leave-one-group-out validation, and correcting field-specific biases, while requiring only limited additional laboratory analyses. These results demonstrate that combining a diverse global calibration dataset with targeted spiking provides an efficient and scalable framework for NIR-based foliar nutrient diagnosis in oil palm. With the growing availability of portable NIR devices, this approach has strong potential for on-site, real-time nutrient monitoring in both industrial plantations and smallholder contexts.
SN - 2772-3755
ER -