Leaf Hyperspectral Reflectance Research Articles

Dynamics of leaf chlorophyll fluorescence parameters can well be tracked by coupling VIS-NIR-SWIR hyperspectral reflectance and light drivers in partial least-squares regression

Chlorophyll α fluorescence (ChlF)-derived parameters provide deep insights into the photosynthetic functions of vegetation, and their rapid tracking is of great interest to plant studies. Unfortunately, traditional approaches such as direct measurements or regressive derivation from leaf traits are impossible to capture their dynamics, owing to their high variability under changing light conditions. Recent advances in reflectance spectroscopy address provide rapid and non-destructive estimates of ChlF parameters but lack thorough investigations. In this study, partial least-squares regression (PLSR), coupled with wavelength point and interval selection methods, was attempted to estimate ChlF parameters quickly and non-destructively using the coupled information from leaf spectroscopy, leaf traits, and light conditions recorded in a synchronous dataset covering ChlF parameters, leaf traits, and leaf hyperspectral reflectance under various light conditions. The results showed that PLSR models using only reflectance spectroscopy information predicted ChlF parameters well and maintained effectively across diverse light conditions, with the one coupled with stepwise regression variable selection performing best. Unexpectedly, the coupling of reflectance spectra and leaf traits achieved only marginal improvements. In comparison, the combination of spectral information with light conditions has further improved the estimation of ChlF parameters (R2: 0.92–0.96, RPIQ: 4.59–7.53) pronouncedly. The results of the present study highlight that the leaf hyperspectral reflectance together with light drivers can efficiently monitor different types of ChlF parameters, strengthening the use of hyperspectral reflectance in plant functions and thus providing valuable information in functional ecology.

Utilizing Hyperspectral Reflectance and Machine Learning Algorithms for Non-Destructive Estimation of Chlorophyll Content in Citrus Leaves

To address the demands of precision agriculture and the measurement of plant photosynthetic response and nitrogen status, it is necessary to employ advanced methods for estimating chlorophyll content quickly and non-destructively at a large scale. Therefore, we explored the utilization of both linear regression and machine learning methodology to improve the prediction of leaf chlorophyll content (LCC) in citrus trees through the analysis of hyperspectral reflectance data in a field experiment. And the relationship between phenology and LCC estimation was also tested in this study. The LCC of citrus tree leaves at five growth seasons (May, June, August, October, and December) were measured alongside measurements of leaf hyperspectral reflectance. The measured LCC data and spectral parameters were used for evaluating LCC using univariate linear regression (ULR), multivariate linear regression (MLR), random forest regression (RFR), K-nearest neighbor regression (KNNR), and support vector regression (SVR). The results revealed the following: the MLR and machine learning models (RFR, KNNR, SVR), in both October and December, performed well in LCC estimation with a coefficient of determination (R2) greater than 0.70. In August, the ULR model performed the best, achieving an R2 of 0.69 and root mean square error (RMSE) of 8.92. However, the RFR model demonstrated the highest predictive power for estimating LCC in May, June, October, and December. Furthermore, the prediction accuracy was the best with the RFR model with parameters VOG2 and Carte4 in October, achieving an R2 of 0.83 and RMSE of 6.67. Our findings revealed that using just a few spectral parameters can efficiently estimate LCC in citrus trees, showing substantial promise for implementation in large-scale orchards.

Coupling effective variable selection with machine learning techniques for better estimating leaf photosynthetic capacity in a tree species (Fagus crenata Blume) from hyperspectral reflectance

Numerous studies have attempted reflectance-based estimations of leaf photosynthetic capacity parameters using different statistical approaches. Although increasing attention has been paid to selecting effective variables for data-driven methods to assess vegetation parameters, there has been less attention to the estimation of leaf photosynthetic capacity. The primary objective of this study is to examine the potential of selecting effective variables for machine learning techniques to estimate leaf photosynthetic capacity in a typical temperate deciduous species (Fagus crenata Blume) from leaf hyperspectral reflectance or its spectral transformations, with or without additional leaf traits. The least absolute shrinkage and selection operator (LASSO) method was coupled to two machine learning models to extract the effective variables for assessing two key leaf photosynthetic parameters (Vcmax and Jmax). The results showed that two support vector machine (SVM) models, successfully extracted the effective bands by utilizing the LASSO method from the Der-Log-VIs-Traits and Der-Traits, exhibited the best performance for Vcmax (R2 = 0.66, RMSE = 9.73 μmol m−2 s−1, RPD = 1.72, and AIC = 524.86) and Jmax (R2 = 0.65, RMSE = 18.33 μmol m−2 s−1, RPD = 1.68, and AIC = 567.39), respectively, suggesting that the LASSO method can effectively locate important photochemistry variables from hyperspectral data. The models also performed better when based on the first-order derivative spectra, rather than from original or apparent absorption spectra. Furthermore, the results also revealed that the NIR and SWIR spectral regions are important for estimating leaf photosynthetic capacity. This study provides a useful reference for estimating leaf photosynthetic capacity from full-spectrum (visible through shortwave infrared) hyperspectral reflectance data.

AraDiv: a dataset of functional traits and leaf hyperspectral reflectance of Arabidopsis thaliana

Data from functional trait databases have been increasingly used to address questions related to plant diversity and trait-environment relationships. However, such databases provide intraspecific data that combine individual records obtained from distinct populations at different sites and, hence, environmental conditions. This prevents distinguishing sources of variation (e.g., genetic-based variation vs. phenotypic plasticity), a necessary condition to test for adaptive processes and other determinants of plant phenotypic diversity. Consequently, individual traits measured under common growing conditions and encompassing within-species variation across the occupied geographic range have the potential to leverage trait databases with valuable data for functional and evolutionary ecology. Here, we recorded 16 functional traits and leaf hyperspectral reflectance (NIRS) data for 721 widely distributed Arabidopsis thaliana natural accessions grown in a common garden experiment. These data records, together with meteorological variables obtained during the experiment, were assembled to create the AraDiv dataset. AraDiv is a comprehensive dataset of A. thaliana’s intraspecific variability that can be explored to address questions at the interface of genetics and ecology.

Wavelet Vegetation Index to Improve the Inversion Accuracy of Leaf V25cmax of Bamboo Forests

Maximum carboxylation rate (Vcmax) is a key parameter to characterize the forest carbon cycle process. Hence, monitoring the Vcmax of different forest types is a hot research topic at home and abroad, and hyperspectral remote sensing is an important technique for Vcmax inversion. Moso bamboo is a unique forest type with a high carbon sequestration capacity in subtropical regions, but the lack of Vcmax data is a major limitation to accurately modeling carbon cycling processes in moso bamboo forests. Our study area was selected in the moso bamboo forest carbon sink research base in Shanchuan Township, Anji County, Zhejiang Province, China, which has a typical subtropical climate and is widely distributed with moso bamboo forests. In this study, the hyperspectral reflectance and V25cmax (Vcmax converted to 25 °C) of leaves of newborn moso bamboo (I du bamboo) and 2- to 3-year-old moso bamboo (II du bamboo) were measured at different canopy positions, i.e., the top, middle and bottom, in the moso bamboo forest. Then, we applied a discrete wavelet transform (DWT) to the obtained leaf hyperspectral reflectance to construct the wavelet vegetation index (WVI), analyzed the relationship between the WVI and moso bamboo leaf V25cmax, and compared the WVI to the existing hyperspectral vegetation index (HVI). The ability of the WVI to characterize the moso bamboo V25cmax was interpreted. Finally, the partial least squares regression (PLSR) method was used to construct a model to invert the V25cmax of moso bamboo leaves. We showed the following: (1) There are significant leaf V25cmax differences between I du and II du bamboo, and there are also significant leaf V25cmax differences between the top, middle and bottom canopy positions of I du bamboo. (2) Compared to that with the HVI, the detection wavelength of the V25cmax of the WVI expanded to the shortwave infrared region, and thus the WVI had a higher correlation with the V25cmax. The absolute value of the correlation coefficient between the V25cmax of I du bamboo and SR2148,2188 constructed by cD1 was 0.75, and the absolute value of the correlation coefficient between the V25cmax of II du bamboo and DVI2069,407 constructed by cD2 was 0.67. The highest absolute value of the correlation coefficient between V25cmax and WVI at the three different canopy positions was also 13–21% higher than that with the HVI. The longest wavelength used by the WVI was 2441 nm. (3) The validation accuracies of the V25cmax inversion models constructed with the WVI as a variable were all higher than those of the models constructed with the HVI as a variable for all ages and positions, with the highest R2 value of 0.97 and a reduction of 20–60% in the root mean square error (RMSE) value. After the wavelet decomposition of the hyperspectral reflectance of moso bamboo leaves, the low-frequency components contained no noise, and the high-frequency components highlighted the original spectral detail features. The WVI constructed by these components increases the wavelength range of V25cmax interpretation. Therefore, the V25cmax retrieval model based on the WVI encompasses different resolutions and levels of spectral characteristics, which can better reflect the changes in bamboo leaves and can provide a new method for the inversion of the V25cmax of moso bamboo forests based on hyperspectral remote sensing.

Waterlogging Effects on Soybean Physiology and Hyperspectral Reflectance during the Reproductive Stage

In humid climates, waterlogging from excessive rainfalls can be a major limiting factor for soybean production, particularly during the reproductive stage. However, there is a limited understanding of how soybean growth and physiology respond to waterlogging during this critical stage. Here, we investigated the effects of waterlogging and subsequent reoxygenation on the growth, physiology, yields, and leaf hyperspectral reflectance traits of the soybean cultivar ‘Asgrow AG5332’. The crop was grown to stage R1 (initial flowering) in outdoor pot culture conditions, and then waterlogged for 16 days. The flooded pots were drained and continuously monitored for recovery for an additional 16 days. The results showed that soil oxygen levels declined rapidly to zero in about 5 days after waterlogging, and slowly recovered in about 5–16 days. However, it did not reach the same level as the control plants, which maintained an oxygen concentration of 18%. Increasing waterlogging duration negatively affected leaf chlorophyll index, water potential, and stomatal conductance, with a consequent decline in the photosynthetic rate. Further, decreased photosynthetic rate, leaf area, and mineral nutrients resulted in lower biomass and seed yield. Pod dry weight and leaf number were the most and least sensitive parameters, respectively, decreasing by 81% and 15% after 16 days of waterlogging. Waterlogged plants also had higher reflectance in the PAR, blue, green, and red regions, and lower reflectance in the near-infrared, tissue, and water band regions, indicating changes in chemistry and pigment content. The current study reveals that the soybean crop is susceptible to waterlogging during the reproductive stage, due to poor recovery of soil oxygen levels and physiological parameters. Understanding and integrating the growth, physiology, and hyperspectral reflectance data from this study could be used to develop improved cultivars to ensure the stability of soybean production in waterlogging-prone areas.

Discrimination of weeds from sugarcane in Louisiana using hyperspectral leaf reflectance data and pigment analysis

AbstractControlling weeds is a critically important task in sugarcane production systems. Weeds compete for light, nutrients, and water, and if they are not managed properly can negatively impact sugarcane yields. Accurate detection of weeds versus desired plants was assessed using hyperspectral and pigment analyses. Leaf samples were collected from four commercial Louisiana sugarcane varieties, and nine weed species commonly found in sugarcane fields. Hyperspectral leaf reflectance data (350 to 850 nm) were collected from all samples. Plant pigment (chlorophylls and carotenoids) levels were also determined using high-performance liquid chromatography, and concentrations were determined using authentic standards and leaf area. In all cases, leaf reflectance data successfully differentiated sugarcane from weeds using canonical discrimination analysis. Linear discriminant analysis showed that the accuracy of the classification varied from 67% to 100% for individual sugarcane varieties and weed species. In all cases, sugarcane was not misclassified as a weed. Plant pigment levels exhibited marked differences between sugarcane varieties and weed species with differences in chlorophyll and carotenoid explaining much of the observed variation in reflectance. The ratio of chlorophyll a to chlorophyll b showed significant differences between sugarcane and all weed species. The successful implementation of this technology as either an airborne system to scout and map weeds or a tractor-based system to identify and spray weeds in real-time would offer sugarcane growers a valuable tool for managing their crops. By accurately targeting weeds in sugarcane fields that are emerged and growing, the total amount of herbicide applied could be decreased, resulting in cost savings for the grower and reduced environmental impacts.

Detecção de estresse nutricional em cana-de-açúcar por espectroscopia de reflectância VIS-NIR-SWIR

ABSTRACT: This study applied spectroradiometry techniques with hyperspectral data to identify the correlations between sugarcane leaf reflectance and the contents of Nitrogen (N), phosphorus (P), Potassium (K), Sulfur (S), Calcium (Ca) and Magnesium (Mg). During the harvests 2019/20 and 2020/21, sugarcane was introduced to nutritional stress by the application of limestone doses. Liming was applied in a fractional way and, at the end of five years, the amounts corresponded to 0, 9, 15 and 21 t ha-1 of dolomitic limestone. The leaf hyperspectral reflectance data and the state of nutrients in the exponential growth phase of the culture were registered. The wavelengths correlated with N, P, K, S, Ca and Mg were identified using the Spearman’s correlation analysis. The test of similarity (ANOSIM) and the Principal Component Analysis (PCA) were applied to evaluate data variability, as well as the Partial Least Squares Regression (PLSR) for the prediction of the nutritional contents. The order of the degree of correlation in the region of visible was: P > K > N > Ca > S > Mg and for the region of the near infrared: P > K > Ca > N > S > Mg. P presented peaks with high correlations in the wavelengths 706-717 nm (-0.78) and 522-543 nm (-0.76). The values of the PLSR registered the best spectral responses in the region of VIS and red-edge, regions that are more sensitive to the deficiency of sulfur, potassium and phosphorus.

Tree Species Classification Using Plant Functional Traits and Leaf Spectral Properties along the Vertical Canopy Position

Plant functional traits are rarely used in tree species classification, and the impact of vertical canopy positions on collecting samples for classification also remains unclear. We aim to explore the feasibility and effectiveness of leaf traits in classification, as well as to detect the effect of vertical position on classification accuracy. This work will deepen our understanding of the ecological mechanism of natural forest structure and succession from new perspectives. In this study, we collected foliar samples from three canopy layers (upper, middle and lower) and measured their spectra, as well as eight well-known leaf traits. We used a leaf hyperspectral reflectance (LHR) dataset, leaf functional traits (LFT) dataset and LFT + LHR dataset to classify six dominant tree species in a subtropical evergreen broad-leaved forest. Our results showed that the LFT + LHR dataset achieved the highest classification results (overall accuracy (OA) = 77.65% and Kappa = 0.73), followed by the LFT dataset (OA = 74.26% and Kappa = 0.69) and the LHR dataset (OA = 69.06% and Kappa = 0.63). Along the vertical canopy, the OA and Kappa increased from the lower to the upper layers, and the combination data of the three canopy layers achieved the highest accuracy. For the individual tree species, the shade-tolerant species (including Machilus chinensis, Cryptocarya chinensis and Cryptocarya concinna) produced higher accuracies than the light-demanding species (including Schima superba and Castanopsis chinensis). Our results provide an approach for enhancing tree species recognition from the plant physiology and biochemistry perspective and emphasize the importance of vertical direction in forest community research.

A non-destructive testing method for early detection of ginseng root diseases using machine learning technologies based on leaf hyperspectral reflectance.

Ginseng is an important medicinal plant benefiting human health for thousands of years. Root disease is the main cause of ginseng yield loss. It is difficult to detect ginseng root disease by manual observation on the changes of leaves, as it takes a long time until symptoms appear on leaves after the infection on roots. In order to detect root diseases at early stages and limit their further spread, an efficient and non-destructive testing (NDT) method is urgently needed. Hyperspectral remote sensing technology was performed in this study to discern whether ginseng roots were diseased. Hyperspectral reflectance of leaves at 325-1,075 nm were collected from the ginsengs with no symptoms on leaves at visual. These spectra were divided into healthy and diseased groups according to the symptoms on roots after harvest. The hyperspectral data were used to construct machine learning classification models including random forest, extreme random tree (ET), adaptive boosting and gradient boosting decision tree respectively to identify diseased ginsengs, while calculating the vegetation indices and analyzing the region of specific spectral bands. The precision rates of the ET model preprocessed by savitzky golay method for the identification of healthy and diseased ginsengs reached 99% and 98%, respectively. Combined with the preliminary analysis of band importance, vegetation indices and physiological characteristics, 690-726 nm was screened out as a specific band for early detection of ginseng root diseases. Therefore, underground root diseases can be effectively detected at an early stage by leaf hyperspectral reflectance. The NDT method for early detection of ginsengs root diseases is proposed in this study. The method is helpful in the prevention and control of root diseases of ginsengs to prevent the reduction of ginseng yield.

Temporal instability of partial least squares regressions for estimating leaf photosynthetic traits from hyperspectral information

Partial least squares regression (PLSR) is applied increasingly often to predict plant photosynthesis from reflectance spectra. While its applicability across different areas has been examined in previous studies, its stability across time has yet to be evaluated. In this study, we assessed a series of PLSR models built upon three different band selection approaches (iterative stepwise, genetic algorithm, and uninformative variable elimination), in combination with different spectral transforms (original and first-order derivative spectra), for their stabilities in predicting the maximum carboxylation rate (Vcmax) and maximum electron transport rate (Jmax) from hyperspectral reflectance spectra at different temporal scales (seasonal and interannual). The results showed that both photosynthetic parameters can be estimated from leaf hyperspectral reflectance with moderate to good accuracy across different growing stages (R2 = 0.45–0.84) and years (R2 = 0.37–0.97). We further found that the iterative stepwise selection of informative bands when building PLSR models could greatly improve its predictive capacity compared with that of other PLSR models, especially those based on first-order derivative spectra. However, the selected bands of the models for both photosynthetic parameters were, unfortunately not consistent. Furthermore, we could not have identified any model with fixed spectra performed consistently across different seasonal stages and across different years. However, the blue spectral regions were popularly selected throughout the growing stages and in different years. The results demonstrate that leaf spectra-trait estimation using PLSR models varies with time and thus cast doubt over the use of a specific PLSR model to infer leaf traits across different temporal-spatial contexts. The development of a general applicable PLSR model is still in the works.

High-throughput characterization, correlation, and mapping of leaf photosynthetic and functional traits in the soybean (Glycine max) nested association mapping population.

Photosynthesis is a key target to improve crop production in many species including soybean [Glycine max (L.) Merr.]. A challenge is that phenotyping photosynthetic traits by traditional approaches is slow and destructive. There is proof-of-concept for leaf hyperspectral reflectance as a rapid method to model photosynthetic traits. However, the crucial step of demonstrating that hyperspectral approaches can be used to advance understanding of the genetic architecture of photosynthetic traits is untested. To address this challenge, we used full-range (500-2,400 nm) leaf reflectance spectroscopy to build partial least squares regression models to estimate leaf traits, including the rate-limiting processes of photosynthesis, maximum Rubisco carboxylation rate, and maximum electron transport. In total, 11 models were produced from a diverse population of soybean sampled over multiple field seasons to estimate photosynthetic parameters, chlorophyll content, leaf carbon and leaf nitrogen percentage, and specific leaf area (with R2 from 0.56 to 0.96 and root mean square error approximately <10% of the range of calibration data). We explore the utility of these models by applying them to the soybean nested association mapping population, which showed variability in photosynthetic and leaf traits. Genetic mapping provided insights into the underlying genetic architecture of photosynthetic traits and potential improvement in soybean. Notably, the maximum Rubisco carboxylation rate mapped to a region of chromosome 19 containing genes encoding multiple small subunits of Rubisco. We also mapped the maximum electron transport rate to a region of chromosome 10 containing a fructose 1,6-bisphosphatase gene, encoding an important enzyme in the regeneration of ribulose 1,5-bisphosphate and the sucrose biosynthetic pathway. The estimated rate-limiting steps of photosynthesis were low or negatively correlated with yield suggesting that these traits are not influenced by the same genetic mechanisms and are not limiting yield in the soybean NAM population. Leaf carbon percentage, leaf nitrogen percentage, and specific leaf area showed strong correlations with yield and may be of interest in breeding programs as a proxy for yield. This work is among the first to use hyperspectral reflectance to model and map the genetic architecture of the rate-limiting steps of photosynthesis.

Using hyperspectral leaf reflectance to estimate photosynthetic capacity and nitrogen content across eastern cottonwood and hybrid poplar taxa.

Eastern cottonwood (Populus deltoides W. Bartram ex Marshall) and hybrid poplars are well-known bioenergy crops. With advances in tree breeding, it is increasingly necessary to find economical ways to identify high-performing Populus genotypes that can be planted under different environmental conditions. Photosynthesis and leaf nitrogen content are critical parameters for plant growth, however, measuring them is an expensive and time-consuming process. Instead, these parameters can be quickly estimated from hyperspectral leaf reflectance if robust statistical models can be developed. To this end, we measured photosynthetic capacity parameters (Rubisco-limited carboxylation rate (Vcmax), electron transport-limited carboxylation rate (Jmax), and triose phosphate utilization-limited carboxylation rate (TPU)), nitrogen per unit leaf area (Narea), and leaf reflectance of seven taxa and 62 genotypes of Populus from two study plantations in Mississippi. For statistical modeling, we used least absolute shrinkage and selection operator (LASSO) and principal component analysis (PCA). Our results showed that the predictive ability of LASSO and PCA models was comparable, except for Narea in which LASSO was superior. In terms of model interpretability, LASSO outperformed PCA because the LASSO models needed 2 to 4 spectral reflectance wavelengths to estimate parameters. The LASSO models used reflectance values at 758 and 935 nm for estimating Vcmax (R2 = 0.51 and RMSPE = 31%) and Jmax (R2 = 0.54 and RMSPE = 32%); 687, 746, and 757 nm for estimating TPU (R2 = 0.56 and RMSPE = 31%); and 304, 712, 921, and 1021 nm for estimating Narea (R2 = 0.29 and RMSPE = 21%). The PCA model also identified 935 nm as a significant wavelength for estimating Vcmax and Jmax. Therefore, our results suggest that hyperspectral leaf reflectance modeling can be used as a cost-effective means for field phenotyping and rapid screening of Populus genotypes because of its capacity to estimate these physicochemical parameters.

Prediction of South American Leaf Blight and Disease-Induced Photosynthetic Changes in Rubber Tree, Using Machine Learning Techniques on Leaf Hyperspectral Reflectance.

The efficiency of visible and near-infrared (VIS/NIR) sensors and predictive modeling for detecting and classifying South American Leaf Blight (SALB) (Pseudocercospora ulei) in rubber trees (Hevea brasiliensis) has been poorly explored. Furthermore, the performance of VIS/NIR analysis combined with machine learning (ML) algorithms for predicting photosynthetic alterations caused by SALB is unknown. Therefore, this study aimed to detect and classify the SALB levels, as well as to predict, for the first time, disease-induced photosynthetic changes in rubber trees. Leaf hyperspectral reflectance combined with five ML techniques (random forest (RF), boosted regression tree (BRT), bagged classification and regression trees (BCART), artificial neural network (ANN), and support vector machine (SVM)) were used. The RF, ANN, and BCART models achieved the best performance for classifying the SALB levels on the training dataset (accuracies of 98.0 to 99.8%), with 10-fold cross-validation repeated five times, and test dataset (accuracies of 97.1 to 100%). The ANN and RF models were better at predicting leaf gas exchange-related traits such as net CO2 assimilation rate (A) and extrinsic water use efficiency (WUEe) in the training (R2 ranged from 0.97 to 0.99) and testing (R2 ranged from 0.96 to 0.99) phases. In comparison, lower performances (R2 ranged from 0.24 to 0.52) were evidenced for the photochemical traits. This research provides a basis for future designs of a remote monitoring system based on early detection and accurate diagnosis of biotic stress caused by SALB, which is fundamental for more effective rubber crop protection.

Wheat physiology predictor: predicting physiological traits in wheat from hyperspectral reflectance measurements using deep learning

BackgroundThe need for rapid in-field measurement of key traits contributing to yield over many thousands of genotypes is a major roadblock in crop breeding. Recently, leaf hyperspectral reflectance data has been used to train machine learning models using partial least squares regression (PLSR) to rapidly predict genetic variation in photosynthetic and leaf traits across wheat populations, among other species. However, the application of published PLSR spectral models is limited by a fixed spectral wavelength range as input and the requirement of separate custom-built models for each trait and wavelength range. In addition, the use of reflectance spectra from the short-wave infrared region requires expensive multiple detector spectrometers. The ability to train a model that can accommodate input from different spectral ranges would potentially make such models extensible to more affordable sensors. Here we compare the accuracy of prediction of PLSR with various deep learning approaches and an ensemble model, each trained and tested using previously published data sets.ResultsWe demonstrate that the accuracy of PLSR to predict photosynthetic and related leaf traits in wheat can be improved with deep learning-based and ensemble models without overfitting. Additionally, these models can be flexibly applied across spectral ranges without significantly compromising accuracy.ConclusionThe method reported provides an improved prediction of wheat leaf and photosynthetic traits from leaf hyperspectral reflectance and do not require a full range, high cost leaf spectrometer. We provide a web service for deploying these algorithms to predict physiological traits in wheat from a variety of spectral data sets, with important implications for wheat yield prediction and crop breeding.

Classification of soybean frogeye leaf spot disease using leaf hyperspectral reflectance.

In this study, the feasibility of classifying soybean frogeye leaf spot (FLS) is investigated. Leaf images and hyperspectral reflectance data of healthy and FLS diseased soybean leaves were acquired. First, image processing was used to classify FLS to create a reference for subsequent analysis of hyperspectral data. Then, dimensionality reduction methods of hyperspectral data were used to obtain the relevant information pertaining to FLS. Three single methods, namely spectral index (SI), principal component analysis (PCA), and competitive adaptive reweighted sampling (CARS), along with a PCA and SI combined method, were included. PCA was used to select the effective principal components (PCs), and evaluate SIs. Characteristic wavelengths (CWs) were selected using CARS. Finally, the full wavelengths, CWs, effective PCs, SIs, and significant SIs were divided into 14 datasets (DS1–DS14) and used as inputs to build the classification models. Models’ performances were evaluated based on the classification accuracy for both the overall and individual classes. Our results suggest that the FLS comprised of five classes based on the proportion of total leaf surface covered with FLS. In the PCA and SI combination model, 5 PCs and 20 SIs with higher weight coefficient of each PC were extracted. For hyperspectral data, 20 CWs and 26 effective PCs were also selected. Out of the 14 datasets, the model input variables provided by five datasets (DS2, DS3, DS4, DS10, and DS11) were more superior than those of full wavelengths (DS1) both in support vector machine (SVM) and least squares support vector machine (LS-SVM) classifiers. The models developed using these five datasets achieved overall accuracies ranging from 91.8% to 94.5% in SVM, and 94.5% to 97.3% in LS-SVM. In addition, they improved the classification accuracies by 0.9% to 3.6% (SVM) and 0.9% to 3.7% (LS-SVM).

Evaluating different methods for retrieving intraspecific leaf trait variation from hyperspectral leaf reflectance

Leaf mass per area (LMA), leaf dry matter content (LDMC) and leaf water content/ equivalent water thickness (EWT) are commonly used functional plant traits in ecology. Whereas spectroscopy has recently proven to be a powerful tool to collect such functional trait information across large scales, it remains unclear whether these reflectance-based trait predictions are accurate enough to reliably model trait variation at the intraspecific level (i.e. across individuals of one species). We explored the potential of hyperspectral leaf reflectance-based methods to predict LMA, LDMC and EWT at the intraspecific level for two herbs (Hieracium umbellatum and Jacobaea vulgaris) and two shrubs (Rosa rugosa and Rubus caesius), based on 2400 leaf samples. More specifically we tested i) inversion of the PROSPECT-D radiative transfer model, ii) a generic PLSR approach using the multibiome LMA PLSR model and iii) a data-specific PLSR approach at the species level. For the latter approach we furthermore assessed both model transferability across species and the trade-off between sample size and model accuracy. Although the PROSPECT-D model inversion and the multibiome LMA PLSR model were relatively accurate for intraspecific LMA predictions of shrubs (R2 > 71 and 76%, respectively, however NRMSE = 33–47%), their performance was lower for herbs (R2 < 61%, NRMSE = 28–50%). PROSPECT-D was furthermore slightly less successful in retrieving EWT at the intraspecific level (R2 < 70%, NRMSE = 16–43%), and unsuccessful in retrieving LDMC through combining LMA and EWT inversion results (R2 < 10%, NRMSE = 9–192%). The highest correlation accuracy was obtained for all three traits with the species-specific PLSR models (R2 > 70%, NRMSE < 10%). If high predictive accuracy is needed, we thus suggest the use of species-specific PLSR models. The training of species-specific PLSR models comes at the cost of a needed sample size of 100–160 leaves however, depending on the trait. Although transferability of species-specific PLSR models seems limited overall, our results suggest potentially high transferability across herbaceous species.

Early infestations by arthropod pests induce unique changes in plant compositional traits and leaf reflectance.

BACKGROUNDWith steadily growing interest in the use of remote‐sensing technologies to detect and diagnose pest infestations in crops, it is important to investigate and characterize possible associations between crop leaf reflectance and unique pest‐induced changes in plant compositional traits. Accordingly, we compiled plant compositional traits from chrysanthemum and gerbera plants in four treatments: non‐infested, or infested with mites, thrips or whiteflies, and we acquired hyperspectral leaf reflectance data from the same plants over time (0–14 days).RESULTSPlant compositional traits changed significantly in response to arthropod infestations, and individual chrysanthemum and gerbera plants were classified with 78% and 80% accuracy, respectively. Based on leaf reflectance, individual plants from the four treatments were classified with moderate accuracy levels of 76% (gerbera) and 73% (chrysanthemum) but with a clear distinction between non‐infested and infested plants. Accurate and consistent diagnosis of biotic stressors was not achieved.CONCLUSIONTo our knowledge, this is the first study in which infestations by multiple economically important arthropod pests are directly compared and associated with leaf reflectance responses and changes in plant compositional traits. It is important to highlight that imposed stress levels were low, period of infestation was short, and hyperspectral remote‐sensing data were acquired at four time points with analyses based on large data sets (3826 leaf reflectance profiles for chrysanthemum and 4041 for gerbera). This study provides novel insight into crop responses to different biotic stressors and into possible associations between plant compositional traits and hyperspectral leaf reflectance data acquired from crop leaves.

Predicting copper content in chicory leaves using hyperspectral data with continuous wavelet transforms and partial least squares

Aiming to accurate and rapid prediction of leaf Cu content in Cichorium intybus L., a hydroponic experiment with three cultivars was performed from May to August 2020. Eight different Cu concentration treatments (0, 5, 10, 25, 50, 100, 200, and 300 µmol/L) were established with four replications per treatment, and the in situ leaf hyperspectra (350–2500 nm) were taken on at four, six, eight and ten-leaf stages. Moreover, data from an independent experiment under 0, 25, 100, and 200 µmol/L Cu stress conditions from July to August 2020 were also collected to test the transferability of the established optimal monitoring model for leaf Cu content prediction. Continuous wavelet transform (CWT) technique was applied to process the collected reflectance hyperspectra; partial least square (PLS) and lambda-lambda r2 (LL r2) models were used to analyze the relationships between leaf Cu content and the hyperspectral reflectance. Results showed that CWT transformation technique can significantly improve the prediction accuracy of chicory leaf Cu content, and the best decomposition scales are CWT-3, CWT-4, and CWT-5. The PLS model for leaf Cu content prediction in the three decomposition scales yielded a relatively higher accuracy compared to the leaf raw spectra based on the full range hyperspectra, the residual prediction deviation (RPDval) was 5.386 for the CWT-3, 6.016 for the CWT-4, and 5.830 for the CWT-5, respectively. Finally, seven bands (735, 910, 1010, 1105, 1194, 1380, and 1590 nm) were identified as effective wavelengths for predicting the leaf Cu content of chicory. The newly-developed PLS model using the effective wavelengths also performed well for CWT-3 (r2val = 0.830; RPDval = 3.433), CWT-4 (r2val = 0.848; RPDval = 3.583) and CWT-5 (r2val = 0.804; RPDval = 3.136) leaf Cu content prediction. The overall results demonstrated the applicability and feasibility of the CWT and PLS algorithms for estimating the Cu status of chicory using in situ leaf hyperspectral reflectance data.

Evaluating the Performance of Hyperspectral Leaf Reflectance to Detect Water Stress and Estimation of Photosynthetic Capacities

Advanced techniques capable of early, rapid, and nondestructive detection of the impacts of drought on fruit tree and the measurement of the underlying photosynthetic traits on a large scale are necessary to meet the challenges of precision farming and full prediction of yield increases. We tested the application of hyperspectral reflectance as a high-throughput phenotyping approach for early identification of water stress and rapid assessment of leaf photosynthetic traits in citrus trees by conducting a greenhouse experiment. To this end, photosynthetic CO2 assimilation rate (Pn), stomatal conductance (Cond) and transpiration rate (Trmmol) were measured with gas-exchange approaches alongside measurements of leaf hyperspectral reflectance from citrus grown across a gradient of soil drought levels six times, during 20 days of stress induction and 13 days of rewatering. Water stress caused Pn, Cond, and Trmmol rapid and continuous decline throughout the entire drought period. The upper layer was more sensitive to drought than middle and lower layers. Water stress could also bring continuous and dynamic changes of the mean spectral reflectance and absorptance over time. After trees were rewatered, these differences were not obvious. The original reflectance spectra of the four water stresses were surprisingly of low diversity and could not track drought responses, whereas specific hyperspectral spectral vegetation indices (SVIs) and absorption features or wavelength position variables presented great potential. The following machine-learning algorithms: random forest (RF), support vector machine (SVM), gradient boost (GDboost), and adaptive boosting (Adaboost) were used to develop a measure of photosynthesis from leaf reflectance spectra. The performance of four machine-learning algorithms were assessed, and RF algorithm yielded the highest predictive power for predicting photosynthetic parameters (R2 was 0.92, 0.89, and 0.88 for Pn, Cond, and Trmmol, respectively). Our results indicated that leaf hyperspectral reflectance is a reliable and stable method for monitoring water stress and yield increase, with great potential to be applied in large-scale orchards.