Skip to main content
ARS Home » Southeast Area » Stoneville, Mississippi » Crop Production Systems Research » Research » Publications at this Location » Publication #299960

Title: Early detection of crop injury from glyphosate on soybean and cotton using plant leaf hyperspectral data

item ZHAO, FENG - Beihang University
item Huang, Yanbo
item GAO, YIQING - Beihang University
item Reddy, Krishna
item Lee, Matthew
item Fletcher, Reginald
item Thomson, Steven

Submitted to: Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/2/2014
Publication Date: 2/20/2014
Citation: Zhao, F., Huang, Y., Gao, Y., Reddy, K.N., Lee, M.A., Fletcher, R.S., Thomson, S.J. 2014. Early detection of crop injury from glyphosate on soybean and cotton using plant leaf hyperspectral data. Remote Sensing. 6:1538-1563.

Interpretive Summary: Plant leaf optical spectrum could provide useful information for early detection of crop injury from off-target glyphosate drift. The scientists at the School of Instrumentation Science and Opto-Electronics Engineering, Beihang University in Beijing, China and USDA-ARS Crop production Systems Research Unit, Stoneville, MS worked collaboratively to study the spectrum portions of the visible and near-infrared reflectance spectra to reveal the stress-induced changes of leaf interior structure and growth status to differentiate the injured leaves from the healthy leaves of soybean and cotton. The results indicated that the conventional data analysis method could not perform satisfactorily while the features generated from the spectral data through a new data processing method could provide consistent results for early detection of glyphosate injury for soybean and cotton leaves. This study shows the potential to use crop leaf hyperspectral reflectance for the early detection of glyphosate injury, especially from off-target drift.

Technical Abstract: In this paper, we describe early detection of crop injury from glyphosate using traditionally used spectral indices and newly extracted features from leaf hyperspectral reflectance data in non-glyphosate-resistant (non-GR) soybean and non-GR cotton. Spectral bands used in the new features are selected based on the sensitivity analysis results of a physically based leaf radiation transfer model (leaf optical PROperty SPECTra model, PROSPECT), which helps extend the effectiveness of these features to a wide range of leaf structures and growing conditions. The new features are extracted by canonical analysis technique, which could provide the largest separability to distinguish the injured leaves from the healthy ones. The approaches have been validated with greenhouse measured data acquired in glyphosate treatment experiments. Results indicate that the glyphosate injury could be detected by NDVI (Normalized Difference Vegetation Index) and RVI (Ratio Vegetation Index) in 48 hours after treatment (HAT) in soybean and in 72 HAT in cotton. The other spectral indices either showed little use for separation or did not show consistent separation for soybean and cotton. Compared with the traditionally used spectral indices, the new features are more reliable for the early detection of glyphosate injury in non-GR soybean and non-GR cotton leaves, with a consistent trend of higher spraying rate having larger values. This trend became more and more obvious with time. The experimental leaves sprayed with different glyphosate rates showed some separability at 24 HAT by the new features, and could be totally distinguished at and beyond 48 HAT for both soybean and cotton. These findings demonstrate the feasibility of using leaf hyperspectral reflectance measurements for the early detection of glyphosate injury through these newly proposed features.