High-throughput phenotyping as an auxiliary tool in the selection of corn hybrids for agronomic traits¹

ABSTRACT

High-throughput phenotyping (HTP) using vegetation indices (VIs) is an important data source for managing plant breeding programs and can be a promising tool in indirect selection. This study hypothesized that VIs are correlated with agronomic traits in corn, and therefore, HTP can be an auxiliary tool for selecting superior genotypes. The objectives were: i) to analyze the association between agronomic traits and VIs, and ii) to identify superior corn hybrids for the evaluated traits using multivariate techniques. Ten corn hybrids (AGRI 330, AGRI 340, FS575PWU, KTZ006VP3, MG545AW, MG580PW, MG711PW, MZ1780, MZ1952, and TROPI 102) were evaluated for plant height (PH), stem diameter (SD), and grain yield (GY). The VIs studied were NDVI, NDRE, EVI, GNDVI, SAVI, and MSAVI. Pearson’s correlation network was constructed to analyze the relationship between the variables, and a canonical analysis was performed to verify the inter-relationship between the variables and hybrids. The VIs evaluated are strongly positively correlated with each other and with PH. The most productive hybrids are MG545AW, FS575PWU and KTZ006VP3. Hybrid MZ1952 has higher correlations with VIs and PH. The findings reveal that VIs can be excellent auxiliary variables for selecting agronomically superior genotypes, being a promising alternative to increase corn breeding efficiency.

Keywords
plant breeding; indirect selection; multivariate analysis; precision agriculture; vegetation indices

INTRODUCTION

Corn (Zea mays L.) is among the most important crops in the economic and social aspects and stands out for being the most grown and consumed grain globally. Given the gradual world population growth (ONU, 2019), the demand for food tends to increase even more. In this scenario, plant breeding can increase crop yield (Borém et al., 2017; Bailey-Serres et al., 2019; Heinz et al., 2020; Taveira et al., 2020). Since the beginning of agriculture, selecting plants for traits desirable to humans has been established. In the 1960s, the Green Revolution relied on genetic improvement in grains, among them corn, for global food security (Bailey-Serres et al., 2019).

In corn breeding programs, one of the ways to obtain high-yielding hybrids is through direct selection on the grain yield trait. However, selecting a particular trait can influence others, which makes crucial the knowledge of the degree of association between traits so that selection achieves the breeder’s goal (Cruz et al., 2012). Besides searching for the traits of interest, the optimization of the plant evaluation process is also essential in breeding programs to achieve greater efficiency. Thus, when there is a high correlation between a trait that is easy to evaluate and traits of interest that require time and labor with recurrent evaluations in the field, like agronomic traits, the indirect selection is an excellent alternative (Samecima Junior, 2018). Indirect selection aims to optimize costs and labor and advance the selection process by seeking to improve a trait of interest by selecting based on another trait that is more easily analyzed (Ramalho et al., 2012; Borém et al., 2017).

High-throughput phenotyping (HTP) is an important data source for managing breeding programs by providing spatial and temporal information (Shanahan et al., 2001; Da Silva et al., 2020). The use of sensors in agriculture has enabled non-contact and nondestructive evaluation and can be a helpful tool in indirect selection to minimize time, labor, and costs while providing more accurate information. One of the main HTP tools is the use of vegetation indices (VIs). VIs are ratios between the reflected radiation from two or more bands and are efficient algorithms for estimating vegetation canopy, growth, vigor, nutritional status, among other applications (Xue & Su, 2017; Silva Júnior et al., 2018; Prado Osco et al., 2020; Ramos et al., 2020; Santana et al., 2021, Santana et al., 2022).

Among the several VIs available, the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Red Edge (NDRE) are the most widely used in plant growth monitoring studies (Bonfil, 2017; Da Silva et al., 2020; Santana et al., 2022). NDVI is correlated with plant height and leaf mass, mainly in absorbing green radiation around 550 nm (Moriwaki et al., 2019). NDRE includes the red band and has higher sensitivity for estimating leaf chlorophyll content and N content than NDVI, especially in C4 plants (Gitelson & Merzlyak, 1994; Portz et al., 2012). Another widely used VI is the Soil Adjusted Vegetation Index (SAVI), which minimizes the soil background effects on spectral reflectance, which is especially important in situations where the soil surface is not entirely covered by vegetation (Zhao et al., 2018)

Spectral reflectance is related to plant yield (Chang et al., 2016). Plant leaf tissues can reflect, absorb, or transmit solar radiation, and the relationship between these factors is variable according to organ surface characteristics and their physiologies. Furthermore, light emission ranges according to wavelength (Silva Júnior et al., 2018). High-yielding phenotypes are related to their respective VIs, which, when observed precisely, are related to their gene constitutions and hence can be subject to selection and improvement (Rutkoski et al., 2016).

Despite being a promising technique, using VIs as auxiliary tool for selecting genotypes in corn breeding programs is still incipient, and studies evaluating the relationship between plant traits and VIs are crucial to improve the efficiency of genotype selection using this approach. Our hypothesis is that the VIs are correlated with agronomic traits in corn, and therefore, HTP can be an auxiliary tool for selecting superior genotypes. In this sense, the objectives of this study were: (i) to analyze the association between agronomic traits and VIs and (ii) to identify superior corn hybrids for the evaluated traits using multivariate techniques.

MATERIAL AND METHODS

Conducting the experiment

The experiment was carried out in the 2019/2020 crop season at the experimental field of the Universidade Federal de Mato Grosso do Sul, located in the municipality of Chapadão do Sul (18°41’33’’S, 52°40’45’’W, with an altitude of 810 m), Mato Grosso do Sul, Brazil. The region’s climate according to the Köppen classification is Aw type (Savanna Tropical). The soil of the experimental field was classified as Red Latosol clayey dystrophic with the following chemical properties in the 0-0.20 m layer: pH (H₂O) = 6.2; exchangeable Al (cmolc dm^-3) = 0.0; Ca+Mg (cmolc dm^-3) = 4.31; P (mg dm^-3) = 41.3; K (cmolc dm^-3) = 0.2; organic matter (g dm^-3) = 19.74; V (%) = 45.0; Al saturation (%) = 0.0; sum of bases (cmolc dm^-3) = 2.3; CEC (cmolc dm^-3) = 5.1. Weather conditions throughout the experiment are shown in Figure 1.

Figure 1
Weather conditions during the 2019/2020 crop season.

Ten treatments were evaluated in a randomized block design with three replicates. The treatments were composed of single hybrids acquired from Agricom Seeds: AGRI 330, AGRI 340, FS575PWU, KTZ006VP3, MG545AW, MG580PW, MG711PW, MZ1780, MZ1952, and TROPI 102. The plots consisted of five-meters rows, with a spacing of 0.45 m between rows and a density of four plants m^-1.

Sowing occurred in December 2019 using conventional soil tillage (plowing and leveling harrow). Fungicide (Pyraclotrobin + Methyl Thiophanate) and insecticide (Fipronil) were used in the treatment of seeds with at a dose of 200 mL of the commercial product for 100 kg seeds to ensure protection against the pests and soil fungi. Cultural practices were performed according to the needs of the crop.

Assessing agronomic traits and vegetation indices

When the grains were at the R2 stage, usually recommended for the beginning of harvesting (Harrison et al., 1996), the following agronomic traits were evaluated: plant height (PH, m) and stem diameter (SD, cm), sampled from five plants randomly in each plot using a tape measure; and grain yield (GY, kg ha^-1), sampled along two meters of the central rows of each plot and, after being weighed and converted to 13% humidity, it was extrapolated to kg ha^-1.

At 60 days after emergence (DAE), a Sensefly eBee RTK fixed-wing remotely piloted aircraft (RPA) was used as a HTP resource. The eBee has autonomous take-off, flight plan, and landing control and was equipped with the senseFly Parrot Sequoia® multispectral camera, which has a sunshine sensor and an additional 16 MP RGB camera for scouting. The overflight was performed at 100 m altitude, allowing a spatial image resolution of 0.10 m. The overflight was carried out near the zenith due to the minimization of the shadows on the plants at 11 a.m for six minutes. Radiometric calibration was made for the entire scene using a manufacturer-supplied calibrated reflective surface. Multispectral reflectance images were obtained for green (550 nm ± 40 nm), red (660 nm ± 40 nm), red (735 nm ± 10 nm), and near infrared (NIR, 790 nm ± 40 nm) spectral bands. Reflectance values at these wavelengths enabled the calculation of VIs studied. Further details on flight procedures and the acquisition of wavelengths for calculating VIs can be found in Prado Osco et al. (2020) and Santana et al. (2022).

The acquired images were mosaiced and orthorectified by the Pix4Dmapper software. The calculated VIs were: NDVI (Normalized Difference Vegetation Index), NDRE (Normalized Difference Red Edge Index), EVI (Enhanced Vegetation Index), GNDVI (Green Normalized Difference Vegetation Index), SAVI (Soil-Adjusted Vegetation Index), and MSAVI (Modified Soil-Adjusted Vegetation Index). The VIs evaluated here were chosen because of their higher correlation with plant biomass according to Raper &Varco (2015), and are detailed in Table 1.

Table 1
Description of the vegetation indices calculated using the Sequoia multispectral sensor

Abbreviation	Vegetation Index	Equation	References
NDVI	Normalized Difference Vegetation Index	$\frac{NIR-red}{NIR+red}$	(Rouse et al., 1974)
NDRE	Normalized Difference Red Edge Index	$\frac{NIR-\mathit{\text{rededge}}}{NIR+\mathit{\text{rededge}}}$	(Gitelson & Merzlyak, 1994)
EVI	Enhanced Vegetation Index	$\frac{NIR-red}{(NIR+6\mathit{\text{red}}-7,5\mathit{\text{green}})+1}$	(Justice et al., 1998)
GNDVI	Green Normalized Difference Vegetation Index	$\frac{NIR-\mathit{\text{green}}}{NIR+\mathit{\text{green}}}$	(Gitelson et al., 1996)
SAVI	Soil-Adjusted Vegetation Index	$(1+0,5)\frac{NIR-red}{NIR+red+0.5}$	(Huete, 1988)
MSAVI	Modified Soil-Adjusted Vegetation Index	$$\begin{gathered} 0,5[2NIR+1- \\ \frac{\sqrt{(2NIR+1{)}^2-8(NIR-red})]}{2} \end{gathered}$$	(Qi et al., 1994)

• NIR: near-infrared reflectance; green: green reflectance; red: red reflectance, rededge: red transition reflectance.

Statistical Analysis

To study the existing relationship between the traits and VIs, Pearson’s correlations were estimated between the agronomic traits and VIs. A correlation network was constructed to graphically express the functional relationship between the correlations, where the distance between the nodes (variables) is proportional to the absolute value of the correlation between them. The thickness of the lines was controlled by enforcing a cut-off value of 0.60, meaning that only correlations above 0.60 were highlighted. Lastly, positive correlations were plotted by green lines, while negative correlations were highlighted in red. A multivariate canonical analysis was performed to verify the inter-relationship between the variables (agronomic traits and VIs) and treatments (hybrids). All statistical analyses were performed using Rbio software (Bhering, 2017) and followed the procedures recommended by Cruz et al. (2012).

RESULTS AND DISCUSSION

Canonical analysis provides a multivariate analysis of variance (MANOVA), in which treatment effect was significant (p < 0.05) for all variables analyzed. Thus, it was possible to proceed to Pearson’s correlation and canonical analyses to verify the association between variables and the interrelationship between variables and treatments, respectively. Correlation network revealed the existence of high-magnitude positive correlations for most variables evaluated (Figure 2). There was a low-magnitude negative correlation only between PH and SD. Although the correlation between the traits PH and SD was negative of low magnitude, i.e., the taller plants had smaller stem diameters, such result is consistent with the literature. According to Taiz et al. (2017), the microfibrils present in most stem cells have an anisotropic behavior, which is growing in a preferential direction, providing a greater expansion in length than in width. Consequently, the cell structure tends to reorganize itself, narrowing on the sides and expanding vertically.

Figure 2
Correlation network between agronomic traits plant height (PH), stem diameter (SD) and grain yield (GY), and vegetation indices NDVI, NDRE, EVI, GNDVI, SAVI and MSAVI, evaluated in ten corn hybrids.

The VIs showed a strong positive correlation between them, which is expected because, although there are differences between them, their formulas are similar, favoring smaller differences between their values. This leads to the fact that plants with high NDVI values also obtained high values for the other indices. Moreover, PH also showed a high positive correlation with all the VIs. The correlation between VIs and PH supports that plant height can be estimated from vegetation indices, as reported in a previous study carried out by Prado Osco et al. (2020) with corn crop in Brazil. This is mainly due to the fact that VIs are able to measure plant attributes such as chlorophyll content and biomass, which in turn are highly related with plant height.

NDVI, created by Rouse et al. (1974), uses the wavelengths of the red (660 ± 12 nm) and near-infrared (770 ± 12 nm) bands. Plants that have higher reflectance in the green and near-infrared tend to absorb more red (Shiratsuchi, 2014). The chlorophyll molecule reflects the visible range, while the near-infrared is reflected by the leaf mesophyll (Oliveira, 2017), pointing to suitable conditions for photosynthesis. Therefore, NDVIs close to 1 (maximum value) refer to healthier and more yielding plants.

However, the red and near-infrared bands may be influenced by the biomass, impairing the accuracy of yields. An alternative is to employ VIs that use wavelengths in the green range (Oliveira, 2017). In this sense, we calculated NDRE, covering wavelengths in the rededge and near-infrared (NIR) bands; EVI, allowing the relation between green, red and NIR; GNDVI, relating green to NIR; SAVI and MSAVI, which, although they also use only NIR and red, are adjusted according to the parameters of the soil and the vegetation itself, respectively.

Regarding GY, the interaction with the other evaluated traits was moderate, allowing its indirect selection. Quantitative traits, such as yield, are difficult to select because they are highly influenced by the environment. One strategy in this situation is indirect selection, which also reduces time and labor (Samecima Junior, 2018). Thus, traits with high heritability and presenting a linear correlation with yield identify superior genotypes in this aspect (Ribeiro et al., 2010). The plants that presented higher heights and VIs were the highest yielding ones, and hence the linear correlation was positive.

Figure 3 shows the canonical analysis, which allows knowing the interrelationship between the treatments and the evaluated variables. The accumulated variance in the first two canonical variables was 99.6%, indicating high accuracy in interpreting the constructed biplot. Cruz et al. (2011) and Mingoti (2007) recommend employing this procedure when the first two canonical variables retain at least 80% of the total variation. This analysis made it possible to verify that hybrids MG545AW, FS575PWU, and KTZ006VP3 stood out for their grain yield since these hybrids are on the same axis of the GY trait.

Figure 3
Canonical analysis applied to the agronomic traits plant height (PH), stem diameter (SD) and grain yield (GY), and vegetation indices NDVI, NDRE, EVI, GNDVI, SAVI and MSAVI, evaluated in ten corn hybrids.

A high-magnitude positive relationship between VIs and PH provided by Pearson’s correlation analysis was also observed in the canonical analysis, as they fall on the same axis. The hybrid MZ1952 was the most correlated with such variables according to its graphical positioning, also occupying the same axis. Although GY is not inserted in this quadrant, its proximity is significant. Similar results were obtained by Taveira et al. (2020), who evaluated the association between spectral data and agronomic traits in soybean using a multivariate approach. The authors verified a strong association between the studied VIs (NDVI and NDRE) but found no association between VIs and agronomic traits.

The other hybrids showed no correlation with agronomic traits and VIs. AGRI330, AGRI340, MZ1780, and TROPI102, which occupy the upper right axis, are close to the VIs and hybrid MZ1952 but not close enough to be correlated, due to the separation by axis. The hybrids most distant from the evaluated variables and the other hybrids were MG711PW and MG580PW, which occupied the upper left axis. The trait SD did not fit on the axes and, consequently, did not obtain a relationship with the other variables and hybrids by the canonical analysis.

The findings reported in this study reveal that VIs can be excellent auxiliary variables for selecting superior genotypes in breeding programs. In a corn breeding program where hundreds of lines are assessed annually, the measurement of agronomic traits is costly, as it is a labor and time-consuming task. Furthermore, long periods of field evaluations cause exhaustion and human fatigue, which may result in lower accuracy of the assessments (Prado Osco et al., 2020; Ramos et al., 2020; Samecima Junior, 2018). In this sense, VIs can be used as auxiliary traits in the indirect selection, where genotypes with higher values for the VIs studied here could be selected for better agronomic performance, reducing the number of in-field measurements. Therefore, HTP using VIs arises as a promising alternative to increase the efficiency of crop breeding programs, providing accurate results in a shorter time.

CONCLUSIONS

The vegetation indices NDVI, NDRE, EVI, GNDVI, SAVI and MSAVI are strongly positively correlated with each other and with plant height, which, in turn, show moderate positive correlation with grain yield. The stem diameter has low negative correlation with plant height.

The hybrids with higher yields were MG545AW, FS575PWU and KTZ006VP3. Hybrid MZ1952 has higher correlations with vegetation indices and plant height compared to the other varieties. AGRI330, AGRI340, MZ1780 and TROPI102 are closer to each other but have no significant correlation with the variables evaluated by canonical analysis; the same occurred with the hybrids MG711PW and MG580PW.

ACKNOWLEDGEMENTS, FINANCIAL SUPPORT AND FULL DISCLOSURE

This study was financed in part by the Coordination for the Improvement of Higher Education Personnel - Brazil (CAPES) - Finance Code 001, National Council for Scientific and Technological Development (CNPq) and Federal University of Mato Grosso do Sul (UFMS). We declare that there is no conflict of interest.

• 1
  This work is part of the research project of the last author.

REFERENCES

• Bailey-Serres J, Parker JE, Ainsworth EA, Oldroyd GED & Schroeder JI (2019) Genetic strategies for improving crop yields. Nature, 575:109-118.
• Bhering LL (2017) Rbio: A Tool For Biometric And Statistical Analysis Using The R Platform. Crop Breeding and Applied Biotechnology, 17:187-190.
• Bonfil DJ (2017) Wheat phenomics in the field by RapidScan: NDVI vs. NDRE. Israel Journal of Plant Sciences, 64:41-54.
• Borém A, Miranda GV & Fritsche-Neto R (2017) Melhoramento de plantas. Viçosa, UFV. 523p.
• Chang LIU, Peng-Sen SUN & Shi-Rong L (2016) A review of plant spectral reflectance response to water physiological changes. Chinese Journal of Plant Ecology, 40:80-91.
• Cruz CD, Ferreira FM & Pessoni LA (2011) Biometria aplicada ao estudo da diversidade genética. Visconde do Rio Branco, Suprema. 620p.
• Cruz CD, Regazzi AJ & Carneiro PCS (2012) Modelos biométricos aplicados ao melhoramento genético. Viçosa, UFV. 732p.
• Da Silva EE, Baio FHR, Teodoro LPR, Da Silva Junior CA, Borges RS & Teodoro P (2020) UAV-multispectral and vegetation indices in soybean grain yield prediction based on in situ observation. Remote Sensing Applications: Society and Environment, 18:100318.
• Gitelson A, Kaufman YJ & Merzlyak MN (1996) Use of a green channel in remote sensing of global vegetation from EOS-MODIS. Remote Sensing of Environment, 58:289-298.
• Gitelson A & Merzlyak MN (1994) Spectral reflectance changes associated with autumn senescence of Aesculus hippocastanum L. and Acer platanoides L. leaves. Spectral features and relation to chlorophyll estimation. Journal of Plant Physiology, 143:286-292.
• Harrison JH (1996) Managing corn silage for maximum nutritive value. In: Cornell Nutrition Conference, New York. Proceedings, Cornell University. p.29-37.
• Heinz R, Teodoro LPR, Gonçalves MC, De Azevedo Peixoto L, Bhering LL & Teodoro PE (2020) Potential of maize (Zea mays L.) populations derived from commercial single-cross hybrids for extraction of partially inbred lines under different nitrogen availability. Revista de la Facultad de Ciencias Agrarias UNCuyo, 52:32-42.
• Huete AR (1988) A soil-adjusted vegetation index (SAVI). Remote Sensing of Environment, 25:295-309.
• Justice CO, Vermote E, Townshend JRG, Defries R, Roy DP, Hall DK, Salomonson VV, Privette JL, Riggs G, Strahler A, Lucht W, Myneni RB, Knyazikhin Y, Running SW, Nemani RR, Wan Z, Huete RR, Van Leeuwen W, Wolfe RE, Giglio L, Muller J, Lewis P & Barnsley MJ (1998) The Moderate Resolution Imaging Spectroradiometer (MODIS): Land remote sensing for global change research. IEEE Transactions on Geoscience and Remote Sensing, 36:1228-1249.
• Mingoti SA (2007) Análise de dados através de métodos de estatística multivariada: uma abordagem aplicada. Brasil, UFMG. 297p.
• Moriwaki T, Falcioni R, Tanaka FAO, Cardoso KAK, Souza LA, Benedito E, Nanni MR, Bonato CM & Antunes WC (2019) Nitrogen-improved photosynthesis quantum yield is driven by increased thylakoid density, enhancing green light absorption. Plant Science, 278:01-11.
• Oliveira MF (2017) Modelos de predição de produtividade da cultura do milho por meio de NDVI em arranjos espaciais. Master Thesis. Universidade Estadual Paulista ‘Júlio de Mesquita Filho’, Jaboticabal. 44p.
• Prado Osco L, Marcato Junior J, Ramos APM, Furuya DEG, Santana DC, Teodoro LPR, Gonçalves WN, Baio FHR, Pistori H, Silva Junior CA & Teodoro PE (2020) Leaf Nitrogen Concentration and Plant Height Prediction for Maize Using UAV-Based Multispectral Imagery and Machine Learning Techniques. Remote Sensing, 12:3237.
• Portz G, Molin JP & Jasper J (2012) Active crop sensor to detect variability of nitrogen supply and biomass on sugarcane fields. Precision Agriculture, 13:33-44.
• Qi J, Chehbouni A, Huete AR, Kerr YH & Sorooshian S (1994) A modified soil adjusted vegetation index. Remote Sensing and Environment, 48:119-126.
• Ramalho MAP, Santos JB, Pinto CB, Souza EA, Gonçalves FMA & Souza JC (2012) Genética na Agropecuária. Brasil, Editora UFLA. 565p.
• Ramos APM, Osco LP, Furuya DEG, Gonçalves WN, Santana DC, Teodoro LPR, Silva Junior CA, Capristo-Silva G, Li J, Baio FHR, Marcato Junior J, Teodoro PE & Pistori H (2020) A random forest ranking approach to predict yield in maize with uav-based vegetation spectral indices. Computers and Electronics in Agriculture, 178:105791.
• Raper TB & Varco JJ (2015) Canopy-scale wavelength and vegetative index sensitivities to cotton growth parameters and nitrogen status. Precision Agriculture, 16:62-76.
• Ribeiro ND, Cargnelutti Filho A & Poersch NL (2010) Critério de seleção indireta para a produtividade de grãos em feijão. Ciência Rural, 40:986-989.
• Rouse JW, Haas RH, Schell JA, Deering DW & Harlan JC (1974) Monitoring the vernal advancement and retrogradation (green wave effect) of natural vegetation. United States, NASA/GSFC. 8p. (Progress Report, number 7).
• Rutkoski J, Poland J, Mondal S, Autrique E, Pérez LG, Crossa J, Reinolds M & Singh R (2016) Canopy temperature and vegetation indices from high-throughput phenotyping improve accuracy of pedigree and genomic selection for grain yield in wheat. G3-Genes Genomes Genetics, 6:2799-2808.
• Samecima Junior EH (2018) índice de vegetação por diferença normalizada e caracteres Agronômicos em genótipos de milho. Master Thesis. Universidade Estadual Paulista ‘Júlio de Mesquita Filho’, Jaboticabal. 44p.
• Santana DC, Cotrim MF, Flores MS, Baio FHR, Shiratsuchi LS, Da Silva Junior CA, Teodoro LPR & Teodoro PE (2021) UAV-based multispectral sensor to measure variations in corn as a function of nitrogen topdressing. Remote Sensing Applications: Society and Environment, 23:100534.
• Santana DC, Dos Santos RG, Teodoro LPR, Da Silva Junior CA, Baio FHR, Coradi PC, & Teodoro PE (2022) Structural equation modelling and factor analysis of the relationship between agronomic traits and vegetation indices in corn. Euphytica, 218:01-08.
• Silva Júnior CA, Nanni MR, Shakir M, Teodoro PE, De Oliveira-Júnior JF, Cezar E & Shiratsuchi LS (2018) Soybean varieties discrimination using nonimaging hyperspectral sensor. Infrared Physics and Technology, 89:338-350.
• Shanahan JF, Schepers JS, Francis DD, Varvel GE, Wilhelm WW, Tringe JM, Schlemmer MR & Major DJ (2001) Use of remote-sensing imagery to estimate corn grain yield. Agronomy Journal, 93:583-589.
• Shiratsuchi LS, Brandão ZN, Vicente LE, Victoria DC, Ducati JR, Oliveira RP & Vilela MF (2014) Sensoriamento remoto: conceitos básicos e aplicações na agricultura de precisão. In: Bernardi ACC, Naime JM, Resende AV, Bassoi LH & Inamasu RY (Eds.) Agricultura de precisão: resultados de um novo olhar. Brasília, Embrapa. p.58-73.
• Taiz L, Zeiger E, Moller IM & Murphy A (2017) Fisiologia e desenvolvimento vegetal. 6ª ed. Brasil, Editora Artmed. 858p.
• Taveira A, Pantaleão A, Campos C, Baio F, Teodoro L & Teodoro P (2020) Selection of soybean F₃ populations for agronomic and physiological traits and vegetation indices using multivariate approaches. Revista de la Facultad de Ciencias Agrarias UNCuyo, 52:22-31.
• ONU - United Nations Organization: Department of Economic and Social Affairs, Population Division (2019) World Population Prospects 2019: Highlights. New York, United Nations/Department of Economic and Social Affairs. 46p.
• Xue J & Su B (2017) Significant remote sensing vegetation indices: A review of developments and applications. Journal of Sensors, 2017:01-17.
• Zhao B, Duan A, Ata-Ul-Karim ST, Liu Z, Chen Z, Gong Z, Zhang J, Xiao J, Liu Z, Qin A & Ning D (2018) Exploring new spectral bands and vegetation indices for estimating nitrogen nutrition index of summer maize. European Journal of Agronomy, 93:113-125.

Publication Dates

• Publication in this collection
  10 Mar 2023
• Date of issue
  Jan-Feb 2023

History

• Received
  30 Nov 2021
• Accepted
  17 May 2022