DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Hyperspectral Leaf Reflectance as Proxy for Photosynthetic Capacities: An Ensemble Approach Based on Multiple Machine Learning Algorithms

Abstract

Global agriculture production is challenged by increasing demands from rising population and a changing climate, which may be alleviated through development of genetically improved crop cultivars. Research into increasing photosynthetic energy conversion efficiency has proposed many strategies to improve production but have yet to yield real-world solutions, largely because of a phenotyping bottleneck. Partial least squares regression (PLSR) is a statistical technique that is increasingly used to relate hyperspectral reflectance to key photosynthetic capacities associated with carbon uptake (maximum carboxylation rate of Rubisco, Vc,max) and conversion of light energy (maximum electron transport rate supporting RuBP regeneration, Jmax) to alleviate this bottleneck. However, its performance varies significantly across different plant species, regions, and growth environments. Thus, to cope with the heterogeneous performances of PLSR, this study aims to develop a new approach to estimate photosynthetic capacities. A framework was developed that combines six machine learning algorithms, including artificial neural network (ANN), support vector machine (SVM), least absolute shrinkage and selection operator (LASSO), random forest (RF), Gaussian process (GP), and PLSR to optimize high-throughput analysis of the two photosynthetic variables. Six tobacco genotypes, including both transgenic and wild-type lines, with a range of photosynthetic capacities were used to test the framework. Leafmore » reflectance spectra were measured from 400 to 2500 nm using a high-spectral-resolution spectroradiometer. Corresponding photosynthesis vs. intercellular CO2 concentration response curves were measured for each leaf using a leaf gas-exchange system. Results suggested that the mean R2 value of the six regression techniques for predicting Vc,max (Jmax) ranged from 0.60 (0.45) to 0.65 (0.56) with the mean RMSE value varying from 47.1 (40.1) to 54.0 (44.7) μmol m-2 s-1. Regression stacking for Vc,max (Jmax) performed better than the individual regression techniques with increases in R2 of 0.1 (0.08) and decreases in RMSE by 4.1 (6.6) μmol m-2 s-1, equal to 8% (15%) reduction in RMSE. Better predictive performance of the regression stacking is likely attributed to the varying coefficients (or weights) in the level-2 model (the LASSO model) and the diverse ability of each individual regression technique to utilize spectral information for the best modeling performance. Further refinements can be made to apply this stacked regression technique to other plant phenotypic traits.« less

Authors:
; ; ;
Publication Date:
Research Org.:
Univ. of Illinois at Urbana-Champaign, IL (United States)
Sponsoring Org.:
USDOE Advanced Research Projects Agency - Energy (ARPA-E)
OSTI Identifier:
1524099
Alternate Identifier(s):
OSTI ID: 1613660
Grant/Contract Number:  
AR0000598
Resource Type:
Published Article
Journal Name:
Frontiers in Plant Science
Additional Journal Information:
Journal Name: Frontiers in Plant Science Journal Volume: 10; Journal ID: ISSN 1664-462X
Publisher:
Frontiers Media SA
Country of Publication:
Switzerland
Language:
English
Subject:
59 BASIC BIOLOGICAL SCIENCES; Plant Sciences

Citation Formats

Fu, Peng, Meacham-Hensold, Katherine, Guan, Kaiyu, and Bernacchi, Carl J. Hyperspectral Leaf Reflectance as Proxy for Photosynthetic Capacities: An Ensemble Approach Based on Multiple Machine Learning Algorithms. Switzerland: N. p., 2019. Web. doi:10.3389/fpls.2019.00730.
Fu, Peng, Meacham-Hensold, Katherine, Guan, Kaiyu, & Bernacchi, Carl J. Hyperspectral Leaf Reflectance as Proxy for Photosynthetic Capacities: An Ensemble Approach Based on Multiple Machine Learning Algorithms. Switzerland. https://doi.org/10.3389/fpls.2019.00730
Fu, Peng, Meacham-Hensold, Katherine, Guan, Kaiyu, and Bernacchi, Carl J. Mon . "Hyperspectral Leaf Reflectance as Proxy for Photosynthetic Capacities: An Ensemble Approach Based on Multiple Machine Learning Algorithms". Switzerland. https://doi.org/10.3389/fpls.2019.00730.
@article{osti_1524099,
title = {Hyperspectral Leaf Reflectance as Proxy for Photosynthetic Capacities: An Ensemble Approach Based on Multiple Machine Learning Algorithms},
author = {Fu, Peng and Meacham-Hensold, Katherine and Guan, Kaiyu and Bernacchi, Carl J.},
abstractNote = {Global agriculture production is challenged by increasing demands from rising population and a changing climate, which may be alleviated through development of genetically improved crop cultivars. Research into increasing photosynthetic energy conversion efficiency has proposed many strategies to improve production but have yet to yield real-world solutions, largely because of a phenotyping bottleneck. Partial least squares regression (PLSR) is a statistical technique that is increasingly used to relate hyperspectral reflectance to key photosynthetic capacities associated with carbon uptake (maximum carboxylation rate of Rubisco, Vc,max) and conversion of light energy (maximum electron transport rate supporting RuBP regeneration, Jmax) to alleviate this bottleneck. However, its performance varies significantly across different plant species, regions, and growth environments. Thus, to cope with the heterogeneous performances of PLSR, this study aims to develop a new approach to estimate photosynthetic capacities. A framework was developed that combines six machine learning algorithms, including artificial neural network (ANN), support vector machine (SVM), least absolute shrinkage and selection operator (LASSO), random forest (RF), Gaussian process (GP), and PLSR to optimize high-throughput analysis of the two photosynthetic variables. Six tobacco genotypes, including both transgenic and wild-type lines, with a range of photosynthetic capacities were used to test the framework. Leaf reflectance spectra were measured from 400 to 2500 nm using a high-spectral-resolution spectroradiometer. Corresponding photosynthesis vs. intercellular CO2 concentration response curves were measured for each leaf using a leaf gas-exchange system. Results suggested that the mean R2 value of the six regression techniques for predicting Vc,max (Jmax) ranged from 0.60 (0.45) to 0.65 (0.56) with the mean RMSE value varying from 47.1 (40.1) to 54.0 (44.7) μmol m-2 s-1. Regression stacking for Vc,max (Jmax) performed better than the individual regression techniques with increases in R2 of 0.1 (0.08) and decreases in RMSE by 4.1 (6.6) μmol m-2 s-1, equal to 8% (15%) reduction in RMSE. Better predictive performance of the regression stacking is likely attributed to the varying coefficients (or weights) in the level-2 model (the LASSO model) and the diverse ability of each individual regression technique to utilize spectral information for the best modeling performance. Further refinements can be made to apply this stacked regression technique to other plant phenotypic traits.},
doi = {10.3389/fpls.2019.00730},
journal = {Frontiers in Plant Science},
number = ,
volume = 10,
place = {Switzerland},
year = {2019},
month = {6}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record
https://doi.org/10.3389/fpls.2019.00730

Citation Metrics:
Cited by: 5 works
Citation information provided by
Web of Science

Save / Share:

Works referenced in this record:

Improving yield by exploiting mechanisms underlying natural variation of photosynthesis
journal, April 2012

  • Lawson, Tracy; Kramer, David M.; Raines, Christine A.
  • Current Opinion in Biotechnology, Vol. 23, Issue 2
  • DOI: 10.1016/j.copbio.2011.12.012

Review article: Attributes of neural networks for extracting continuous vegetation variables from optical and radar measurements
journal, January 1998

  • Kimes, D. S.; Nelson, R. F.; Manry, M. T.
  • International Journal of Remote Sensing, Vol. 19, Issue 14
  • DOI: 10.1080/014311698214433

PhotoSpec: A new instrument to measure spatially distributed red and far-red Solar-Induced Chlorophyll Fluorescence
journal, October 2018

  • Grossmann, Katja; Frankenberg, Christian; Magney, Troy S.
  • Remote Sensing of Environment, Vol. 216
  • DOI: 10.1016/j.rse.2018.07.002

Precipitation Estimation from Remotely Sensed Imagery Using an Artificial Neural Network Cloud Classification System
journal, December 2004

  • Hong, Yang; Hsu, Kuo-Lin; Sorooshian, Soroosh
  • Journal of Applied Meteorology, Vol. 43, Issue 12
  • DOI: 10.1175/JAM2173.1

Group Lasso-Based Band Selection for Hyperspectral Image Classification
journal, December 2017


Improving carbon fixation pathways
journal, August 2012


Imaging plants dynamics in heterogenic environments
journal, April 2012


Detection of early plant stress responses in hyperspectral images
journal, July 2014


Least angle regression
journal, April 2004


The Spatial LASSO With Applications to Unmixing Hyperspectral Biomedical Images
journal, October 2015


Stacked generalization
journal, January 1992


Support vector machines in remote sensing: A review
journal, May 2011


Partial least-squares regression: a tutorial
journal, January 1986


Can improvement in photosynthesis increase crop yields?
journal, March 2006


Support Vector Machines for classification and regression
journal, January 2010

  • Brereton, Richard G.; Lloyd, Gavin R.
  • The Analyst, Vol. 135, Issue 2
  • DOI: 10.1039/B918972F

Fitting photosynthetic carbon dioxide response curves for C 3 leaves
journal, September 2007


Temperature response of carbon isotope discrimination and mesophyll conductance in tobacco: Temperature response of mesophyll conductance
journal, September 2012


Hyperspectral image analysis for water stress detection of apple trees
journal, July 2011


Recent advances in sensing plant diseases for precision crop protection
journal, March 2012

  • Mahlein, Anne-Katrin; Oerke, Erich-Christian; Steiner, Ulrike
  • European Journal of Plant Pathology, Vol. 133, Issue 1
  • DOI: 10.1007/s10658-011-9878-z

Remotely estimating photosynthetic capacity, and its response to temperature, in vegetation canopies using imaging spectroscopy
journal, September 2015


Plant phenomics and the need for physiological phenotyping across scales to narrow the genotype-to-phenotype knowledge gap
journal, July 2015

  • Großkinsky, Dominik K.; Svensgaard, Jesper; Christensen, Svend
  • Journal of Experimental Botany, Vol. 66, Issue 18
  • DOI: 10.1093/jxb/erv345

Stacked regressions
journal, July 1996


High-throughput field phenotyping using hyperspectral reflectance and partial least squares regression (PLSR) reveals genetic modifications to photosynthetic capacity
journal, September 2019

  • Meacham-Hensold, Katherine; Montes, Christopher M.; Wu, Jin
  • Remote Sensing of Environment, Vol. 231
  • DOI: 10.1016/j.rse.2019.04.029

Proximal Remote Sensing Buggies and Potential Applications for Field-Based Phenotyping
journal, July 2014


A general regression neural network
journal, January 1991

  • Specht, D. F.
  • IEEE Transactions on Neural Networks, Vol. 2, Issue 6
  • DOI: 10.1109/72.97934

What is the maximum efficiency with which photosynthesis can convert solar energy into biomass?
journal, April 2008


Redesigning photosynthesis to sustainably meet global food and bioenergy demand
journal, June 2015

  • Ort, Donald R.; Merchant, Sabeeha S.; Alric, Jean
  • Proceedings of the National Academy of Sciences, Vol. 112, Issue 28
  • DOI: 10.1073/pnas.1424031112

Hyperspectral remote sensing of foliar nitrogen content
journal, December 2012

  • Knyazikhin, Y.; Schull, M. A.; Stenberg, P.
  • Proceedings of the National Academy of Sciences, Vol. 110, Issue 3
  • DOI: 10.1073/pnas.1210196109

Remote sensing of canopy chemistry
journal, January 2013


Satellite remote sensing of grasslands: from observation to management
journal, February 2016

  • Ali, Iftikhar; Cawkwell, Fiona; Dwyer, Edward
  • Journal of Plant Ecology, Vol. 9, Issue 6
  • DOI: 10.1093/jpe/rtw005

Hyperspectral Imaging Techniques for Rapid Identification of Arabidopsis Mutants with Altered Leaf Pigment Status
journal, April 2012

  • Matsuda, Osamu; Tanaka, Ayako; Fujita, Takao
  • Plant and Cell Physiology, Vol. 53, Issue 6
  • DOI: 10.1093/pcp/pcs043

Regression Shrinkage and Selection Via the Lasso
journal, January 1996


With ‘Phenomics,’ Plant Scientists Hope to Shift Breeding Into Overdrive
journal, July 2009


High-Throughput SNP Genotyping to Accelerate Crop Improvement
journal, September 2014


Ten principles for a landscape approach to reconciling agriculture, conservation, and other competing land uses
journal, May 2013

  • Sayer, J.; Sunderland, T.; Ghazoul, J.
  • Proceedings of the National Academy of Sciences, Vol. 110, Issue 21
  • DOI: 10.1073/pnas.1210595110

Breeding Technologies to Increase Crop Production in a Changing World
journal, February 2010


Raising yield potential of wheat. II. Increasing photosynthetic capacity and efficiency
journal, October 2010

  • Parry, M. A. J.; Reynolds, M.; Salvucci, M. E.
  • Journal of Experimental Botany, Vol. 62, Issue 2
  • DOI: 10.1093/jxb/erq304

Machine learning regression algorithms for biophysical parameter retrieval: Opportunities for Sentinel-2 and -3
journal, March 2012


Consistent land surface temperature data generation from irregularly spaced Landsat imagery
journal, October 2016


Mapping forest change using stacked generalization: An ensemble approach
journal, January 2018


FluoSpec 2—An Automated Field Spectroscopy System to Monitor Canopy Solar-Induced Fluorescence
journal, June 2018

  • Yang, Xi; Shi, Hanyu; Stovall, Atticus
  • Sensors, Vol. 18, Issue 7
  • DOI: 10.3390/s18072063

Novel throughput phenotyping platforms in plant genetic studies
journal, October 2007


Hyperspectral reflectance as a tool to measure biochemical and physiological traits in wheat
journal, December 2017

  • Silva-Perez, Viridiana; Molero, Gemma; Serbin, Shawn P.
  • Journal of Experimental Botany, Vol. 69, Issue 3
  • DOI: 10.1093/jxb/erx421

Gaussian processes uncertainty estimates in experimental Sentinel-2 LAI and leaf chlorophyll content retrieval
journal, December 2013


Applying hyperspectral imaging to explore natural plant diversity towards improving salt stress tolerance
journal, February 2017


What gas exchange data can tell us about photosynthesis: Gas exchange data and photosynthesis
journal, December 2015

  • Sharkey, Thomas D.
  • Plant, Cell & Environment, Vol. 39, Issue 6
  • DOI: 10.1111/pce.12641

Reduction of Ribulose-1,5-Bisphosphate Carboxylase/Oxygenase Content by Antisense RNA Reduces Photosynthesis in Transgenic Tobacco Plants
journal, January 1992

  • Hudson, Graham S.; Evans, John R.; von Caemmerer, Susanne
  • Plant Physiology, Vol. 98, Issue 1
  • DOI: 10.1104/pp.98.1.294

Image-based phenotyping of plant disease symptoms
journal, January 2015


For most large underdetermined systems of linear equations the minimal 𝓁1‐norm solution is also the sparsest solution
journal, January 2006

  • Donoho, David L.
  • Communications on Pure and Applied Mathematics, Vol. 59, Issue 6, p. 797-829
  • DOI: 10.1002/cpa.20132

Support-vector networks
journal, September 1995

  • Cortes, Corinna; Vapnik, Vladimir
  • Machine Learning, Vol. 20, Issue 3
  • DOI: 10.1007/BF00994018

High-Throughput Phenotyping of Maize Leaf Physiological and Biochemical Traits Using Hyperspectral Reflectance
journal, November 2016

  • Yendrek, Craig R.; Tomaz, Tiago; Montes, Christopher M.
  • Plant Physiology, Vol. 173, Issue 1
  • DOI: 10.1104/pp.16.01447

Multigene manipulation of photosynthetic carbon assimilation increases CO2 fixation and biomass yield in tobacco
journal, May 2015

  • Simkin, Andrew J.; McAusland, Lorna; Headland, Lauren R.
  • Journal of Experimental Botany, Vol. 66, Issue 13
  • DOI: 10.1093/jxb/erv204

High-throughput Phenotyping and Genomic Selection: The Frontiers of Crop Breeding ConvergeF: High-throughput Phenotyping and Genomic Selection
journal, May 2012

  • Cabrera-Bosquet, Llorenç; Crossa, José; von Zitzewitz, Jarislav
  • Journal of Integrative Plant Biology, Vol. 54, Issue 5
  • DOI: 10.1111/j.1744-7909.2012.01116.x

Artificial Neural Network for Classification and Analysis of Degraded Soils
journal, March 2017

  • Bonini Neto, Alfredo; dos Santos Batista Bonini, Carolina; Santos Bisi, Beatriz
  • IEEE Latin America Transactions, Vol. 15, Issue 3
  • DOI: 10.1109/TLA.2017.7867601

A Review of Imaging Techniques for Plant Phenotyping
journal, October 2014


Using leaf optical properties to detect ozone effects on foliar biochemistry
journal, May 2013

  • Ainsworth, Elizabeth A.; Serbin, Shawn P.; Skoneczka, Jeffrey A.
  • Photosynthesis Research, Vol. 119, Issue 1-2
  • DOI: 10.1007/s11120-013-9837-y

Natural genetic variation in plant photosynthesis
journal, June 2011


Machine Learning Techniques for Predicting Crop Photosynthetic Capacity from Leaf Reflectance Spectra
journal, June 2017


Regularization Paths for Generalized Linear Models via Coordinate Descent
journal, January 2010

  • Friedman, Jerome; Hastie, Trevor; Tibshirani, Robert
  • Journal of Statistical Software, Vol. 33, Issue 1
  • DOI: 10.18637/jss.v033.i01

In vivo temperature response functions of parameters required to model RuBP-limited photosynthesis
journal, September 2003


Phenomics – technologies to relieve the phenotyping bottleneck
journal, December 2011


Estimation of thermophysical properties of dimethyl ether as a commercial refrigerant based on artificial neural networks
journal, December 2010

  • Moghadassi, A. R.; Nikkholgh, M. R.; Parvizian, F.
  • Expert Systems with Applications, Vol. 37, Issue 12
  • DOI: 10.1016/j.eswa.2010.04.065

A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species
journal, June 1980

  • Farquhar, G. D.; von Caemmerer, S.; Berry, J. A.
  • Planta, Vol. 149, Issue 1
  • DOI: 10.1007/BF00386231

Relationships between hyperspectral data and components of vegetation biomass in Low Arctic tundra communities at Ivotuk, Alaska
journal, January 2017

  • Bratsch, Sara; Epstein, Howard; Buchhorn, Marcel
  • Environmental Research Letters, Vol. 12, Issue 2
  • DOI: 10.1088/1748-9326/aa572e

Geographic stacking: Decision fusion to increase global land cover map accuracy
journal, May 2015


Leaf optical properties reflect variation in photosynthetic metabolism and its sensitivity to temperature
journal, October 2011

  • Serbin, Shawn P.; Dillaway, Dylan N.; Kruger, Eric L.
  • Journal of Experimental Botany, Vol. 63, Issue 1
  • DOI: 10.1093/jxb/err294

Random Forests
journal, January 2001


Generating ensembles of heterogeneous classifiers using Stacked Generalization: Generating ensembles of heterogeneous classifiers
journal, January 2015

  • Sesmero, M. Paz; Ledezma, Agapito I.; Sanchis, Araceli
  • Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, Vol. 5, Issue 1
  • DOI: 10.1002/widm.1143

Regression and ANN models for estimating minimum value of machining performance
journal, April 2012

  • Zain, Azlan Mohd; Haron, Habibollah; Qasem, Sultan Noman
  • Applied Mathematical Modelling, Vol. 36, Issue 4
  • DOI: 10.1016/j.apm.2011.09.035

Improved temperature response functions for models of Rubisco-limited photosynthesis
journal, February 2001


Early detection and classification of plant diseases with Support Vector Machines based on hyperspectral reflectance
journal, October 2010


PLS-regression: a basic tool of chemometrics
journal, October 2001

  • Wold, Svante; Sjöström, Michael; Eriksson, Lennart
  • Chemometrics and Intelligent Laboratory Systems, Vol. 58, Issue 2
  • DOI: 10.1016/s0169-7439(01)00155-1

Works referencing / citing this record:

Modeling Hyperspectral Response of Water-Stress Induced Lettuce Plants Using Artificial Neural Networks
journal, November 2019

  • Osco, Lucas Prado; Ramos, Ana Paula Marques; Moriya, Érika Akemi Saito
  • Remote Sensing, Vol. 11, Issue 23
  • DOI: 10.3390/rs11232797