The Korean Journal of Crop Science. 1 December 2024. 209-215
https://doi.org/10.7740/kjcs.2024.69.4.209

ABSTRACT


MAIN

  • INTRODUCTION

  • MATERIALS AND METHODS

  •   Experimental field

  •   Data acquisition by UAV

  •   Image processing and data analysis

  • RESULTS

  • DISCUSSION

INTRODUCTION

Rice (Oryza sativa L.) is cultivated across a wide range of environmental conditions (Shrestha et al., 2020) and serves as one of the most important crops supporting the world’s growing population. However, lodging, characterized by the breaking or bending of stems, often limits grain yield and quality in rice production (Shrestha et al., 2008). Lodging, as seen in other crops, can be influenced by various abiotic factors such as soil temperature, CO2 levels, precipitation, deep water, and resource complementarities (Ishimaru et al., 2008; Zhang et al., 2014), as well as biotic factors like sheath blight and brown planthoppers (Wu et al., 2012; Elanchezhyan et al., 2020). Abiotic lodging typically occurs during the panicle initiation stage, and since the plant is not in a dead state, it continues its physiological activities such as photosynthesis. (Wu et al., 2022). In contrast, biotic lodging results in the destruction of the chlorophyll and a loss of photosynthetic capability, leading to reduced vigour and the occurrence of lodging (Ramli et al., 2018). Severe lodging typically leads to significant losses in both grain yield and quality, while also complicating harvest operations (Chauhan et al., 2019; Dahiya et al., 2018).

In recent years, remote sensing technology has emerged as a valuable tool for assessing crop lodging (Li et al., 2021). While satellite imagery and radar systems provide useful data, they have limitations in terms of resolution and sensitivity to confounding factors like soil conditions and crop stress (Khanal et al., 2020). Unmanned aerial vehicles (UAVs), on the other hand, offer a more flexible and cost-effective solution for smaller areas, capturing high-resolution imagery that provides insights into crop color and 3D structure (Jang et al., 2020). The detection of lodging regions using UAVs was conducted based on morphological traits resulting from relative height differences among the plants (Yang et al., 2020; Tan et al., 2021). However, a key challenge in current research is the inability to differentiate between abiotic and biotic causes of lodging, such as pest or disease-related factors. Overcoming this limitation would greatly improve the practical applications of remote sensing for crop management.

Smart agronomic practices, such as the use of lodging-resistant cultivars, appropriate fertilization, application of plant growth regulators, and chemical treatments, can reduce the risk of crop lodging (Liao et al., 2023). However, the combined effects of these strategies on both grain yield and lodging risk in rice paddies remain poorly understood. Identifying the specific causes of lodging could enhance the efficiency of these management approaches. Therefore, the present study aims to establish parameters for detecting lodging caused by both abiotic and biotic factors, specifically focusing on brown hopper-induced lodging, using UAV-acquired images.

MATERIALS AND METHODS

Experimental field

This experiment was conducted in two commercial rice fields in Gyeongsangnam-do, South Korea. One field is located in Pyeong-ri, Namhae-eup, and the other in Yongso-ri, Idong-myeon, about 10 km apart (Fig. 1). The rice field in Pyeong-ri has abiotic lodging partially distributed due to rainfall and wind during the monsoon season, while the rice field in Yongso-ri has biotic lodging partially distributed due to the infestation by brown grasshopper (Nilaparvata lugens). The areas of Pyeong-ri and Yongso-ri are 1,784 m2 and 1,253 m2, respectively, and the soil types are loamy sand and loam (Heuktoram, https://soil.rda.go.kr). The experimental region, Namhae experiences an average annual temperature of 16°C, and an annual precipitation of 1,600 mm.

https://cdn.apub.kr/journalsite/sites/kjcs/2024-069-04/N0840690402/images/kjcs_2024_694_209_F1.jpg
Fig. 1.

Experimental sites located in Pyeong-ri and Yongso-ri, Namhae-gun (a, b), Sampling area from abiotic, and biotic lodging areas in each field (c, d).

Data acquisition by UAV

The experiment in the rice field utilized the DJI P4 multispectral UAV (Fig. 2). This UAV weighs 1,487 g at take-off and has an average flight duration of 30 minutes. Its imaging system consists of six 1/2.9-inch CMOS sensors, enabling simultaneous capture of RGB images via the visible camera and spectral data using a multispectral system. The multispectral system covers five spectral bands with the following ranges: Blue (B): 450 nm ± 16 nm, Green (G): 560 nm ± 16 nm, Red (R): 650 nm ± 16 nm, Red Edge (RE): 730 nm ± 16 nm, and Near-Infrared (NIR): 840 nm ± 26 nm.

https://cdn.apub.kr/journalsite/sites/kjcs/2024-069-04/N0840690402/images/kjcs_2024_694_209_F2.jpg
Fig. 2.

Unmanned aerial vehicle (UAV) used in this study (a), calibrated radiometric panel (CRP) (b), standard reflectance spectrum of the CRP as per the manufacturer (c), spectral response of the blue, green, red, red-edge, and near-infrared (NIR) bands (d).

The UAV mission was executed using the GS Pro app (DJI, China) at an altitude of 20 m, with 80% overlap and side lap to ensure thorough coverage of the target field. Waypoints for the mission were established, and location data was recorded at each waypoint through camera triggers. The DJI P4 multispectral system, featuring real-time kinematic-global positioning system (RTK-GPS), provided positional accuracy within 5 cm, allowing for precise navigation along the predetermined waypoints. The flights were conducted at each rice field between 11:00 am and 1:00 pm. on September 2, 2024. Immediately after the each flight, a Calibrated Radiometric Panel (CRP, Micasense, USA) was captured in the ground for radiometric calibration to convert the sensor’s digital numbers into reflectance values (Fig. 3).

Image processing and data analysis

The mean reflectance values for each band of the multispectral sensor were calculated using the spectral response data from the UAV’s multispectral sensor and the reflectance data from the CRP supplied by the manufacturer, following the Equation 1 (Table 1). Pix4Dmapper Pro 3.0.17 (Pix4D SA, Lausanne, Switzerland) was employed to process the UAV’s overlapping images into reflectance maps for each spectral band. The “Radiometric Processing and Calibration” setting was configured to “Camera and Sun Irradiance” to ensure accurate correction using CRP images captured in ground. This radiometric calibration accounted for irradiance variations during the UAV flight, converting the multispectral digital numbers into accurate reflectance values (Wang, 2021).

(1)
rk400900RCRP(λ)Ckλd400900Ck(λ)d

Where rk denotes the calculated mean reflectance values of the CRP, RX(λ) refers to the standard reflectance spectrum of the CRP provided by the manufacturer, Ck(λ) represents the spectral response of the image sensor, and 𝑘 indicates to one of the bands: Blue (B), Green (G), Red (R), Red Edge (RE), or Near-Infrared (NIR).

Table 1.

Average reflectance value of each spectral band on the calibrated radiometric panel used in this study.

Spectral band Mean reflectance value
Blue
(450 nm ± 16 nm)
0.518
Green
(560 nm ± 16 nm)
0.522
Red
(650 nm ± 16 nm)
0.523
Red edge
(730 nm ± 16 nm)
0.523
Near-infrared
(840 nm ± 26 nm)
0.521

From the reflectance maps generated for the two fields, polygons were manually created for no lodging areas (H), areas affected by abiotic lodging due to natural disasters (AL), and areas affected by biotic lodging from brown grasshoppers (BL), as shown in Fig. 1. These polygons were then used as sample regions for H, AL, and BL.

Vegetation indices related to crop vigor and growth status, as shown in Table 2, were generated using the reflectance maps. Data was extracted from each generated vegetation index map using the sample polygons, and statistical comparisons were made among H, AL, and BL. The means and skewness of each dataset (H, AL, BL) was calculated and one way analysis of variance (ANOVA) was performed with Tukey HSD (P < 0.05) post hoc test for significant mean comparisons. All of statistical analysis was performed using scipy.stats module in Python.

Table 2.

Vegetation indices used in this study.

Vegetation index Formulation Features Reference
Enhanced vegetation
index
(EVI)
2.5×(NIR-RED)(NIR+6RED-7.5BLUE)+1 Crop growth status
(Leaf area index, biomass)
(Qiao et al., 2019)
Excess green
(EXG)
2×GREEN-RED-BLUE Crop growth status
(Leaf area index, biomass)
(Rossi, 2019)
Green normalized
difference vegetation
index
(GNDVI)
NIR-GREENNIR+GREEN Chlorophyll content (SPAD) (Gitelson et al., 1996)
Green ratio vegetation
index
(GRVI)
NIRGREEN Crop growth status
(biomass)
(Mandal et al., 2021)
Normalized difference
red-edge index
(NDRE)
NIR-RENIR+RE Crop vigour (Gitelson et al., 1994)
Normalized difference
vegetation index
(NDVI)
NIR-REDNIR+RED Crop growth status
and crop vigour
(Rouse et al., 1974)
Normalized green minus
red difference index
(NGRDI)
GREEN-REDGREEN+RED Nitrogen contents (Choudhary et al., 2021)
Optic soil-adjusted
vegetation index
(OSAVI)
1.5×NIR-REDNIR+RED+0.16 Chlorophyll content, water status
prediction, and plant stress
identification
(Rondeaux et al., 1996)
Red-edge re-normalized
difference
vegetation index
(RDVI)
NIR-REDNIR+RE Crop grwoth status
(Leaf area index, biomass)
(Roujean et al., 1995)
Visual atmospheric
resistance index
(VARI)
GREEN-REDGREEN+RED-BLUE Crop growth status
(Leaf area index, plant height)
(Feng et al., 2022)

RESULTS

The comparative analysis of vegetation indices is detailed in Table 3. A clear distinction was observed between lodged areas (AL, BL) and non-lodged areas for all vegetation indices. Although significant differences between AL and BL were found in most indices, except for EXG, GNDVI, and NDRE, the average values showed that the distinction between AL and BL was not as pronounced as that between H.

Interestingly, in terms of skewness, the vegetation indices in AL were found to be greater than 1, except for GNDVI and NDRE, indicating that the distribution is skewed to one side. All skewness values were positive, suggesting that the data overlaps with values from the H region (Fig. 3). This implies that the abiotic lodging areas include regions of rice where chlorophyll is not severely damaged, which could serve as a distinguishing feature from biotic lodging areas.

Table 3.

Results of statistical analyses.

Vegetation
Index
Mean Skewness p-value
H BL AL H BL AL
EVI 0.527ay 0.163b 0.21c -0.253a 0.689b 1.76c ***
EXG 0.172a 0.05b 0.06b 0.137a 1.210b 1.284b ***
GNDVI 0.586a 0.406b 0.40b -0.544a 0.341b 0.408b ***
GRVI 0.185a -0.171b -0.11c -0.22a 0.531b 1.300c ***
NDRE 0.224a 0.12b 0.11b -0.164a 0.376b 0.377b ***
NDVI 0.636a 0.253b 0.30c -0.729a 0.700b 1.123c ***
NGRDI 0.088a -0.171b -0.106c 0.215a 0.531b 1.295c ***
OSAVI 0.50a 0.188b 0.250c -0.677a 0.832b 1.13c ***
RDVI 0.400a 0.176b 0.270c -0.455a 0.707b 1.090c ***
VARI 0.214a -0.222b -0.15c -0.424a 0.742b 1.273c ***

yLetters represent significant differences among the cultivars by the non-parametric statistical analysis, one-way analysis of variance (ANOVA) with Tukey’s HSD (P < 0.05) post-hoc test.

*, **, and *** Significant at 0.05, 0.01, and 0.001 probability levels, respectively.

https://cdn.apub.kr/journalsite/sites/kjcs/2024-069-04/N0840690402/images/kjcs_2024_694_209_F3.jpg
Fig. 3.

Histograms for GNDVI and NDVI. Skewness of GNDVI in the AL data is less than 1, indicating that the distribution is not skewed to one side and does not distinctly separate from BL data. NDVI has a skewness greater than 1, observed as an overlap with H data.

DISCUSSION

Vegetative indices (VIs), which are mathematical combinations of spectral bands from remote sensing data, offer a valuable tool for estimating chlorophyll levels in plants (Xue & Su, 2017). These indices, such as the Normalized Difference Vegetation Index (NDVI), Green Normalized Difference Vegetation Index (GNDVI), and Chlorophyll Index (CI), utilize red, green, and near-infrared bands to assess plant health. While these indices provide valuable insights, factors like soil background, atmospheric conditions, and other plant pigments can influence their accuracy. To ensure reliable chlorophyll estimation, it is essential to use vegetative indices in conjunction with other information and methods.

Chlorophyll, the pigment responsible for the green color of plants, is crucial for photosynthesis (Pareek et al., 2017). When a plant is diseased, its metabolic processes can be disrupted, leading to decreased chlorophyll production or breakdown (Christ et al., 2014). This can result from reduced photosynthesis, nutrient deficiencies, tissue damage, or hormonal imbalances. While not all diseases directly affect chlorophyll levels, many cause symptoms like yellowing or browning of leaves due to chlorophyll degradation (Hörtensteiner et al., 2006). The extent of chlorophyll reduction can vary depending on the severity of the disease and the plant species involved.

Abiotic factors such as wind and excessive nitrogen can also contribute to a gradual decline in chlorophyll levels (Sharma et al., 2020). Wind lodging causes physical damage to plant tissues, including leaves, leading to chlorophyll breakdown and reduced photosynthesis. Additionally, damage to the root system can impair nutrient uptake, including nitrogen, which is essential for chlorophyll production (Karthika et al., 2018). Excessive nitrogen can accelerate plant senescence and disrupt nutrient balance, leading to chlorophyll degradation. The rate of chlorophyll decline varies depending on the severity of these factors, plant species, and other environmental conditions.

Biotic factors such as diseases and pests can cause a more rapid and severe depletion of chlorophyll compared to abiotic factors like wind lodging or excess nitrogen (Ogle, 1997). These biotic stressors often directly damage plant tissues, leading to chlorophyll breakdown and necrosis. Additionally, they can deplete essential nutrients and inhibit photosynthesis, further accelerating chlorophyll decline. Addressing these issues promptly is crucial to minimize damage and preserve the plant’s photosynthetic capacity.

The differences in chlorophyll reduction observed in the current study led to distinct shapes in the peaks of VIs, despite similar average peak values between abiotic and biotic-induced lodging. Using these finding, it is possible to determine whether the lodging is biotic or abiotic by analyzing the skewness of the VI value distribution in lodging areas identified morphologically. This suggests the importance of selecting appropriate parameters to distinguish target traits. The results of this study are expected to aid in parameter decision-making for future research.

Acknowledgements

This work was supported by Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, Forestry (IPET) through Open Field Smart Agriculture Technology Short-term Advancement Program, funded by Ministry of Agriculture, Food and Rural Affairs (MAFRA) (322032-3). Also, this research was supported by the BK21 FOUR, Global Smart Farm Educational Research Center, Seoul National University, Seoul, Korea and Advanced Institute of Convergence Technology, Suwon, 16229, Republic of Korea. This research also was carried out with the support of “Cooperative Research Program for Agriculture Science & Technology Development (Development of an approach to analyze timing and locality of migratory pests occurrence for rice using information derived from an automated surveillance system: RS-2021-RD009729)” Rural Development Administration, Republic of Korea.

References

1

Chauhan, S., R. Darvishzadeh, M. Boschetti, M. Pepe, and A. Nelson. 2019. Remote sensing-based crop lodging assessment: Current status and perspectives. ISPRS J. Photogramm. Remote Sens. 151 : 124-40.

10.1016/j.isprsjprs.2019.03.005
2

Choudhary, S. S., S. Biswal, R. Saha, and C. Chatterjee. 2021. A non-destructive approach for assessment of nitrogen status of wheat crop using unmanned aerial vehicle equipped with RGB camera. Arabian J. Geosci. 14(17) : 1739.

10.1007/s12517-021-08139-3
3

Christ, B. and S. Hörtensteiner. 2014. Mechanism and significance of chlorophyll breakdown. J. Plant Growth Regul. 33 : 4-20.

10.1007/s00344-013-9392-y
4

Dahiya, S., S. Kumar, C. Chaudhary, and C. Chaudhary. 2018. Lodging: Significance and preventive measures for increasing crop production. Int. J. Chem. Stud. 6(1) : 700-705.

5

Elanchezhyan, K., T. Sathyan, and K. R. Manikandan. 2020. Brown plant hopper (BPH) and their management in rice. Res. Today 2(4) : 90-2.

6

Feng, H., H. Tao, Z. Li, G. Yang, and C. Zhao. 2022. Comparison of UAV RGB imagery and hyperspectral remote-sensing data for monitoring winter wheat growth. Remote Sens. 14(15) : 3811.

10.3390/rs14153811
7

Gitelson, A. A., Y. J. Kaufman, and M. N. Merzlyak. 1996. Use of a green channel in remote sensing of global vegetation from EOS-MODIS. Remote Sens. Environ. 58(3) : 289-298.

10.1016/S0034-4257(96)00072-7
8

Gitelson, A., and M. N. Merzlyak. 1994. Quantitative estimation of chlorophyll-a using reflectance spectra: Experiments with autumn chestnut and maple leaves. J. Photochem. Photobiol., B 22(3) : 247-252.

10.1016/1011-1344(93)06963-4
9

Hörtensteiner, S. 2006. Chlorophyll degradation during senescence. Annu. Rev. Plant Biol. 57(1) : 55-77.

10.1146/annurev.arplant.57.032905.10521216669755
10

Ishimaru, K., E. Togawa T. Ookawa, T. Kashiwagi, Y. Madoka, and N. Hirotsu. 2008. New target for rice lodging resistance and its effect in a typhoon. Planta 227 : 601-609.

10.1007/s00425-007-0642-817960419
11

Jang, G., J. Kim, J. K. Yu, H. J. Kim, Y. Kim, D. W. Kim, K. H. Kim, C. W. Lee, and Y. S. Chung. 2020. Cost-effective unmanned aerial vehicle (UAV) platform for field plant breeding application. Remote Sens. 12(6) : 998.

10.3390/rs12060998
12

Karthika, K. S., I. Rashmi, and M. S. Parvathi. 2018 Biological functions, uptake and transport of essential nutrients in relation to plant growth. Plant Nutrients and Abiotic Stress Tolerance. pp. 1-49.

10.1007/978-981-10-9044-8_1
13

Khanal, S., K. Kc, J. P. Fulton, S. Shearer, and E. Ozkan. 2020. Remote sensing in agriculture-accomplishments, limitations, and opportunities. Remote Sens. 12(22) : 3783.

10.3390/rs12223783
14

Li, X., X. Li, W. Liu, B. Wei, X. Xu. 2021. A UAV-based framework for crop lodging assessment. Eur. J. Agron. 123 : 126201.

10.1016/j.eja.2020.126201
15

Liao, P., S. M. Bell, L. Chen, S. Huang, H. Wang, J. Miao, Y. Qi, Y. Sun, B. Liao, Y. Zeng, and H. Wei. 2023. Improving rice grain yield and reducing lodging risk simultaneously: A meta-analysis. Eur. J. Agron. 143 : 126709.

10.1016/j.eja.2022.126709
16

Mandal, D., A. Bhattacharya, Y. S. Rao, D. Mandal, A. Bhattacharya, and Y. S. Rao. 2021. Radar vegetation indices for crop growth monitoring. Radar Remote Sensing for Crop Biophysical Parameter Estimation. pp. 177-228.

10.1007/978-981-16-4424-5_7
17

Ogle, H. J. 1997. Abiotic diseases of plants. Plant Pathogens and Plant Diseases; Brown, JF, Ogle, HJ, Eds. pp. 156-71.

18

Pareek, S., N. A. Sagar, S. Sharma, V. Kumar, T. Agarwal, G. A. González‐Aguilar, and E. M. Yahia. 2017. Chlorophylls: Chemistry and biological functions. Fruit and Vegetable Phytochemicals: Chemistry and Human Health, 2nd Edition. pp. 269-284.

10.1002/9781119158042.ch14
19

Qiao, K., W. Zhu, Z. Xie, and P. Li, 2019. Estimating the seasonal dynamics of the leaf area index using piecewise LAI-VI relationships based on phenophases. Remote Sens. 11(6) : 689.

10.3390/rs11060689
20

Ramli, N., S. Yusup, B. W. B. Kueh, P. S. D. Kamarulzaman, N. Osman, M. A. Rahim, R. Aziz, S. Mokhtar, and A. B. Ahmad. 2018. Effectiveness of biopesticides against brown planthopper (Nilaparvata lugens) in paddy cultivation. Sustainable Chem. Pharm. 8 : 16-20.

10.1016/j.scp.2018.01.001
21

Rondeaux, G., M. Steven, and F. Baret. 1996. Optimization of soil-adjusted vegetation indices. Remote Sens. Environ. 55(2) : 95-107.

10.1016/0034-4257(95)00186-7
22

Rossi, G. 2019. Vegetative vigor of maize crop obtained through vegetation indexes in orbital and aerial sensors images. Rev. Bras. Eng. Biossistemas 13 : 195-206.

10.18011/bioeng2019v13n3p195-206
23

Roujean, J. L. and F. M. Breon. 1995. Estimating PAR absorbed by vegetation from bidirectional reflectance measurements. Remote Sens. Environ. 51(3) : 375-384.

10.1016/0034-4257(94)00114-3
24

Rouse Jr, J. W., R. H. Haas, D. W. Deering, J. A. Schell, and J. C. Harlan. 1974. Monitoring the vernal advancement and retrogradation (green wave effect) of natural vegetation (No. E75-10354).

25

Sharma, A., V. Kumar, B. Shahzad, M. Ramakrishnan, G. P. Singh Sidhu, A. S. Bali, N. Handa, D. Kapoor, P. Yadav, K. Khanna, and P. Bakshi. 2020. Photosynthetic response of plants under different abiotic stresses: a review. J. Plant Growth Regul. 39:509-31.

10.1007/s00344-019-10018-x
26

Shrestha, C., I. Ona, S. Muthukrishnan, and T. Mew. 2008. Chitinase levels in rice cultivars correlate with resistance to the sheath blight pathogen Rhizoctonia solani. Eur. J. Plant Pathol. 120 : 69-77.

10.1007/s10658-007-9199-4
27

Shrestha, J., M. Kandel, S. Subedi, and K. K. Shah. 2020. Role of nutrients in rice (Oryza sativa L.): A review. Agrica 9(1) : 53-62.

10.5958/2394-448X.2020.00008.5
28

Tan, S., A. K. Mortensen, X. Ma, B. Boelt, and R. Gislum. 2021. Assessment of grass lodging using texture and canopy height distribution features derived from UAV visual-band images. Agric. For. Meteorol. 308 : 108541.

10.1016/j.agrformet.2021.108541
29

Wang, C. 2021. At-sensor radiometric correction of a multispectral camera (RedEdge) for sUAS vegetation mapping. Sensors 21(24) : 8224.

10.3390/s2124822434960318PMC8704258
30

Wu, D. H., C. T. Chen, M. D. Yang, Y. C. Wu, C. Y. Lin, M. H. Lai, and C. Y. Yang. 2022. Controlling the lodging risk of rice based on a plant height dynamic model. Bot. Stud. 63(1) : 25.

10.1186/s40529-022-00356-736008613PMC9411474
31

Wu, W., J. Huang, K. Cui, L. Nie, Q. Wang, F. Yang, F. Shah, F. Yao, and S. Peng. 2012. Sheath blight reduces stem breaking resistance and increases lodging susceptibility of rice plants. Field Crops Res. 128 : 101-8.

10.1016/j.fcr.2012.01.002
32

Xue, J. and B. Su. 2017 Significant remote sensing vegetation indices: A review of developments and applications. J. Sens. 2017(1) : 1353691.

10.1155/2017/1353691
33

Yang, M. D., J. G. Boubin, H. P. Tsai, H. H. Tseng, Y. C. Hsu, and C. C. Stewart. 2020. Adaptive autonomous UAV scouting for rice lodging assessment using edge computing with deep learning EDANet. Comput. Electron. Agric. 179 : 105817.

10.1016/j.compag.2020.105817
34

Zhang, J, G. Li, Y. Song, Z. Liu, C. Yang, S. Tang, C. Zheng, S. Wang, and Y. Ding. 2014. Lodging resistance characteristics of high-yielding rice populations. Field Crops Res. 161 : 64-74.

10.1016/j.fcr.2014.01.012
페이지 상단으로 이동하기