Onfield estimation of quality parameters in alfalfa through hyperspectral spectrometer data
Alfalfa is a forage of vast importance around the world. In the past, near-infrared spectroscopy (NIRS) technique have been explored in the lab to determine quality traits such as fibre content in dried and ground material. During the last decade, portable hyperspectral devices have emerged as a tools for in-field prediction, of not only crop yield but also a large range of quality and physiological markers. The objective of this study was to estimate quality parameters in an alfalfa crop using hyperspectral data acquired from a full-range (350–2500 nm) spectrometer under field conditions. Reflected spectra were measured in single leaves as well as at the canopy level, then reflectance was related to target parameters such as biomass, leaf pigments, sugars, protein, and mineral contents. Due to their large effect on crop quality parameters, meteorological conditions and phenological stages were included as predictors in the models. We found that meteorological and phenological variables improved the accuracies and percentage of variance explained (R2) for most of the parameters evaluated. Based on R2 values, the best prediction models were obtained for biomass (0.71), sucrose (0.65), flavonoids (Flav) (0.56) and nitrogen (0.70) with normalized root mean squared errors of 0.196, 0.32, 0.087 and 0.08, respectively. These parameters were associated mainly with visible (VIS) (approx. 350–700 nm) and near infrared (NIR) (700–1250 nm) regions of the spectrum. Regarding mineral composition, the best prediction models were developed for P (0.51), B (0.50) and Zn (0.44), associated with the short-wave infra-red (SWIR) region (1250–2500 nm). The results of this study demonstrated the potential of hyperspectral techniques to be used as a base for performing initial evaluations in the field of quality traits in alfalfa crops.
Main Authors: | , , , , |
---|---|
Other Authors: | |
Format: | artículo biblioteca |
Language: | English |
Published: |
Elsevier
2024-01-01
|
Subjects: | Alfalfa, Canopy, Hyperspectral technique, Quality parameters, Trait prediction, |
Online Access: | http://hdl.handle.net/10261/352149 http://dx.doi.org/10.13039/501100000780 https://api.elsevier.com/content/abstract/scopus_id/85179758704 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
dig-idab-es-10261-352149 |
---|---|
record_format |
koha |
spelling |
dig-idab-es-10261-3521492024-05-17T20:44:00Z Onfield estimation of quality parameters in alfalfa through hyperspectral spectrometer data Gámez, Angie L. Vatter, Thomas Santesteban, Luis G. Araus, Jose Luis Aranjuelo, Iker Diputación Foral de Navarra European Commission Alfalfa Canopy Hyperspectral technique Quality parameters Trait prediction Alfalfa is a forage of vast importance around the world. In the past, near-infrared spectroscopy (NIRS) technique have been explored in the lab to determine quality traits such as fibre content in dried and ground material. During the last decade, portable hyperspectral devices have emerged as a tools for in-field prediction, of not only crop yield but also a large range of quality and physiological markers. The objective of this study was to estimate quality parameters in an alfalfa crop using hyperspectral data acquired from a full-range (350–2500 nm) spectrometer under field conditions. Reflected spectra were measured in single leaves as well as at the canopy level, then reflectance was related to target parameters such as biomass, leaf pigments, sugars, protein, and mineral contents. Due to their large effect on crop quality parameters, meteorological conditions and phenological stages were included as predictors in the models. We found that meteorological and phenological variables improved the accuracies and percentage of variance explained (R2) for most of the parameters evaluated. Based on R2 values, the best prediction models were obtained for biomass (0.71), sucrose (0.65), flavonoids (Flav) (0.56) and nitrogen (0.70) with normalized root mean squared errors of 0.196, 0.32, 0.087 and 0.08, respectively. These parameters were associated mainly with visible (VIS) (approx. 350–700 nm) and near infrared (NIR) (700–1250 nm) regions of the spectrum. Regarding mineral composition, the best prediction models were developed for P (0.51), B (0.50) and Zn (0.44), associated with the short-wave infra-red (SWIR) region (1250–2500 nm). The results of this study demonstrated the potential of hyperspectral techniques to be used as a base for performing initial evaluations in the field of quality traits in alfalfa crops. Angie L. Gámez is the recipient of a PhD grant (reference 0011-1408-2020-000005) funded by the Government of Navarre and Nafosa S.L. This manuscript has been conducted within the context of the CropEqualT-CEC project funded by the European Union’s Horizon 2020, Belgium Marie Curie Rise research and innovation programme. Garazi Ezpeleta for technical support provided during the sample analyses in the lab and José María Oses and Bernardo Monreal for the being in charge of crop management. Peer reviewed 2024-03-27T16:36:36Z 2024-03-27T16:36:36Z 2024-01-01 artículo http://purl.org/coar/resource_type/c_6501 Computers and Electronics in Agriculture 216: 108463 (2024) 0168-1699 http://hdl.handle.net/10261/352149 10.1016/j.compag.2023.108463 1872-7107 http://dx.doi.org/10.13039/501100000780 2-s2.0-85179758704 https://api.elsevier.com/content/abstract/scopus_id/85179758704 en Publisher's version https://doi.org/10.1016/j.compag.2023.108463 Sí open application/pdf Elsevier |
institution |
IDAB ES |
collection |
DSpace |
country |
España |
countrycode |
ES |
component |
Bibliográfico |
access |
En linea |
databasecode |
dig-idab-es |
tag |
biblioteca |
region |
Europa del Sur |
libraryname |
Biblioteca del IDAB España |
language |
English |
topic |
Alfalfa Canopy Hyperspectral technique Quality parameters Trait prediction Alfalfa Canopy Hyperspectral technique Quality parameters Trait prediction |
spellingShingle |
Alfalfa Canopy Hyperspectral technique Quality parameters Trait prediction Alfalfa Canopy Hyperspectral technique Quality parameters Trait prediction Gámez, Angie L. Vatter, Thomas Santesteban, Luis G. Araus, Jose Luis Aranjuelo, Iker Onfield estimation of quality parameters in alfalfa through hyperspectral spectrometer data |
description |
Alfalfa is a forage of vast importance around the world. In the past, near-infrared spectroscopy (NIRS) technique have been explored in the lab to determine quality traits such as fibre content in dried and ground material. During the last decade, portable hyperspectral devices have emerged as a tools for in-field prediction, of not only crop yield but also a large range of quality and physiological markers. The objective of this study was to estimate quality parameters in an alfalfa crop using hyperspectral data acquired from a full-range (350–2500 nm) spectrometer under field conditions. Reflected spectra were measured in single leaves as well as at the canopy level, then reflectance was related to target parameters such as biomass, leaf pigments, sugars, protein, and mineral contents. Due to their large effect on crop quality parameters, meteorological conditions and phenological stages were included as predictors in the models. We found that meteorological and phenological variables improved the accuracies and percentage of variance explained (R2) for most of the parameters evaluated. Based on R2 values, the best prediction models were obtained for biomass (0.71), sucrose (0.65), flavonoids (Flav) (0.56) and nitrogen (0.70) with normalized root mean squared errors of 0.196, 0.32, 0.087 and 0.08, respectively. These parameters were associated mainly with visible (VIS) (approx. 350–700 nm) and near infrared (NIR) (700–1250 nm) regions of the spectrum. Regarding mineral composition, the best prediction models were developed for P (0.51), B (0.50) and Zn (0.44), associated with the short-wave infra-red (SWIR) region (1250–2500 nm). The results of this study demonstrated the potential of hyperspectral techniques to be used as a base for performing initial evaluations in the field of quality traits in alfalfa crops. |
author2 |
Diputación Foral de Navarra |
author_facet |
Diputación Foral de Navarra Gámez, Angie L. Vatter, Thomas Santesteban, Luis G. Araus, Jose Luis Aranjuelo, Iker |
format |
artículo |
topic_facet |
Alfalfa Canopy Hyperspectral technique Quality parameters Trait prediction |
author |
Gámez, Angie L. Vatter, Thomas Santesteban, Luis G. Araus, Jose Luis Aranjuelo, Iker |
author_sort |
Gámez, Angie L. |
title |
Onfield estimation of quality parameters in alfalfa through hyperspectral spectrometer data |
title_short |
Onfield estimation of quality parameters in alfalfa through hyperspectral spectrometer data |
title_full |
Onfield estimation of quality parameters in alfalfa through hyperspectral spectrometer data |
title_fullStr |
Onfield estimation of quality parameters in alfalfa through hyperspectral spectrometer data |
title_full_unstemmed |
Onfield estimation of quality parameters in alfalfa through hyperspectral spectrometer data |
title_sort |
onfield estimation of quality parameters in alfalfa through hyperspectral spectrometer data |
publisher |
Elsevier |
publishDate |
2024-01-01 |
url |
http://hdl.handle.net/10261/352149 http://dx.doi.org/10.13039/501100000780 https://api.elsevier.com/content/abstract/scopus_id/85179758704 |
work_keys_str_mv |
AT gamezangiel onfieldestimationofqualityparametersinalfalfathroughhyperspectralspectrometerdata AT vatterthomas onfieldestimationofqualityparametersinalfalfathroughhyperspectralspectrometerdata AT santestebanluisg onfieldestimationofqualityparametersinalfalfathroughhyperspectralspectrometerdata AT arausjoseluis onfieldestimationofqualityparametersinalfalfathroughhyperspectralspectrometerdata AT aranjueloiker onfieldestimationofqualityparametersinalfalfathroughhyperspectralspectrometerdata |
_version_ |
1802820047060598784 |