Previous Article in Journal
A Novel Graph-Based Innovative Trend Analysis Technique for Studying the Crop Trends in Kerala, India
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
AgriEngineering
Volume 7
Issue 2
10.3390/agriengineering7020037
agriengineering-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Detection of Fusarium Head Blight in Wheat Using NDVI from Multispectral UAS Measurements and Its Correlation with DON Contamination

by
Igor Petrović
1,
Filip Vučajnk
2 and
Valentina Spanic
3,*
1
Faculty of Mechanical Engineering, University of Ljubljana, 1000 Ljubljana, Slovenia
2
Department of Agronomy, University of Ljubljana, 1000 Ljubljana, Slovenia
3
Agricultural Institute Osijek, Department for Breeding & Genetics of Small Cereal Crops, 31000 Osijek, Croatia
*
Author to whom correspondence should be addressed.
AgriEngineering 2025, 7(2), 37; https://doi.org/10.3390/agriengineering7020037
Submission received: 3 December 2024 / Revised: 29 January 2025 / Accepted: 31 January 2025 / Published: 3 February 2025
Browse Figures
Versions Notes

Abstract

:
Fusarium head blight (FHB) is a serious fungal disease of wheat and other small cereal grains, significantly reducing grain yield and producing mycotoxins that affect food safety. There is a need for disease detection technologies to determine the right time to apply fungicides, as FHB infection begins before visible symptoms appear. Using multispectral remote sensing by an unmanned aircraft system (UAS), wheat plants were observed under field conditions infested with FHB and simultaneously protected with fungicides sprayed with four different types of nozzles, as well as corresponding control plots infested with FHB only. The results showed that the levels of deoxynivalenol (DON) differed significantly between the five treatments, indicating that the control had the highest DON concentration as no fungicide treatment was applied. This study revealed that the assessment of the normalized difference vegetation index (NDVI) after FHB infection could be useful for predicting DON accumulation in wheat, as a significant negative correlation between DON and NDVI values was measured 24 days after anthesis. The decreasing NDVI values at the end of the growth cycle were expected due to senescence and yellowing of the wheat spikes and leaves. Therefore, significant differences in the NDVI were observed between three measurement points on the 13th, 24th, and 45th day after anthesis. Additionally, the green normalized difference vegetation index (GNDVI) and normalized difference red-edge index (NDRE) were in significant positive correlation with the NDVI at 24th day after anthesis. The use of appropriate measurement points for the vegetation indices can offer the decisive advantage of enabling the evaluation of very large breeding trials or farmers’ fields where the timing of fungicide application is particularly important.

1. Introduction

Fusarium head blight (FHB), primarily caused by Fusarium graminearum and F. culmorum, is a major threat to wheat (Triticum aestivum L.) production, significantly affecting grain yield [1]. The consumption of contaminated food can cause health problems due to the ability of some Fusarium spp. to produce mycotoxins. The most important Fusarium mycotoxins are type A and B trichothecenes and zearalenone (ZEN) [2]. The type A group encompasses T-2 and HT-2 toxins, while the B group consists of deoxynivalenol (DON) and its derivatives (3-acetyldeoxynivalenol (3Ac-DON) and 15-acetyldeoxynivalenol (15Ac-DON)), nivalenol (NIV), and fusarenone X [3]. However, DON is the most frequently detected Fusarium mycotoxin, with lower concentrations of 15Ac-DON, 3Ac-DON, NIV, HT-2 toxin, T-2 toxin, and ZEN occasionally found [2,4]. Mycotoxins are usually mutagenic, teratogenic, and estrogenic, provoking poisoning, head and abdominal pain, and diarrhea in humans as well as thinness in animals [5,6]. Some mycotoxins provoke acute toxicity and chronic illness depending on the concentrations in food and feed. To preserve human and animal health, maximal permitted concentrations of some mycotoxins are regulated by European Union (EU) legislation [7,8,9].
FHB severity is increasing due to the climate changes caused by weather fluctuations (temperature and humidity). The most favorable conditions for FHB infection are longer periods of moisture (>80% humidity) and warm temperatures (20–30 °C) at anthesis [10] when anther extrusion occurs. The initial symptoms appear a few days after flowering when spikelets become dark brown on the glumes or spikes start to bleach above the point of infection. Further, wilting and blight will spread over the spike which can lead to premature senescence of infected spikes, especially if the entire spike is affected [11]. At full maturity, grains could be shriveled with a chalky white or pink coating. The disease’s progression is accompanied by the production of secondary metabolites. The concentration of these metabolites in wheat grains was in positive correlation with FHB severity [12]. However, it was previously reported that the DON concentration was not always highly correlated with FHB severity [13]. The contamination of grains with FHB metabolites depends on many factors, such as the climatic conditions, abundance of inoculum, agro-technique applied, harvesting methods and time, as well as varietal resistance to FHB. It is known that FHB-resistant wheat varieties are the most effective way to combat FHB [14]. To reduce Fusarium contamination, integrated disease management is advised through a combination of fungicides with resistant wheat varieties and rotation of crops [15]. The application of fungicides is performed within a short time window that coincides with anthesis, when the disease starts to develop. The application of fungicide for efficient control can be conducted up to eleven days after anthesis to avoid grain contamination with DON, as was reported in the research of Freije et al. [16]. Fungicides will reduce disease severity and mycotoxin contamination; however, the percentage of reduction was highly variable due to the timing of fungicide application [17].
Till now, FHB symptoms were predominantly monitored based on individual spikes or small plots, while the entire field could not have been observed. Technological innovations are strongly needed to solve these problems. Previously, spatial resolution and observation time have been the most crucial in applications for smart farming, but recent advances in high-resolution cameras located on drones improved these limitations [18]. Unmanned aerial vehicles (UAVs) together with the sensors on board and the necessary equipment for a safe and efficient flight form an unmanned aerial system (UAS). A UAS can serve as a powerful tool for many applications in the mapping of crop conditions, yield prediction, disease detection, weed control, etc. [19]. For example, multispectral remote sensing from a UAS facilitated the rapid detection of FHB outbreaks in wheat on a large-scale, providing a solution for wheat disease detection in larger areas [20]. Efficient information from hyperspectral measurement in FHB control by early detection was reported in the research of Mustafa et al. [21]. The characteristics derived from remote sensing measured on different dates and under varied weather conditions around the anthesis stage were combined and could efficiently predict FHB severity [22]. Phenotyping for FHB resistance in winter wheat needs to be more accurate for effective breeding, as was shown in the research of McConachie et al. [23]. The same authors concluded that a spectral index threshold can be used to distinguish FHB-infected from healthy tissue and to quantify infection. It is very important to note that under FHB infections of wheat, optical sensing methods were able to map locations of severe infections and to separate the harvest of heavily diseased areas of fields to control the entrance of toxins into the food [24]. Therefore, the need for accurate early detection of FHB for large-scale, real-time, and non-destructive testing during the high disease likelihood stage is more and more prominent [25]. For example, the normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) are frequently used vegetation indexes employed to derive phenological metrics, which tend to characterize the greenness of plants’ tissues [26]. It was also shown in the research of Mirasi et al. [27] that a crucial time to estimate wheat grain yield using NDVI values was during the grain milky and later maturity stage. On the other hand, NDVI values can be used to evaluate the health condition of plants based on how a plant reflects different light waves.
Here, we investigate the relationships between NDVI values measured three times after anthesis with an unmanned aircraft system (UAS) and DON concentration in grains after harvest. Additionally, other remote sensing indices will be calculated and correlations with the NDVI, and the DON content will be obtained. This study aims to prove the feasibility of FHB detection using multispectral remote sensing with a UAS after the anthesis stage and the possible correlation with DON contamination.

2. Materials and Methods

2.1. Field Trial

The field trial was carried out during the 2023/2024 vegetation season at the Biotechnical Faculty in Ljubljana, Slovenia (46°03′ N 14°30′ E) (Figure 1). The soil texture at this site is silty loam with 30% clay. Plots were completely randomized in a block design with three replications (Figure 2). Basic processing of the soil consisted of ploughing in September 2023 using a reversible plough, the Lemken VariOpal. On 10 October 2023, the rotary harrow Lemken Zirkon 7 was used for soil preparation two times. The mineral fertilizer NPK 8-20-30 was applied at a rate of 500 kg/ha before the second ploughing with the rotary harrow. The trial was planted using the mechanical seeder Amazone D9 Special. The winter wheat variety used in the trial was Aurelius, a high-quality wheat from Saatbau, Austria, with a sowing rate of 220 kg/ha. This variety was observed to have good resistance to FHB [28]. During spring, the plants were fertilized three times with calcium ammonium nitrate fertilizer (CAN 27%) in an amount of 200 kg/ha at each time of application. The first fertilization was conducted in the tillering growth stage (BBCH 25), the second was conducted in the stem elongation growth stage (BBCH 32), and the last one was conducted in the heading growth stage (BBCH 51). The herbicide Hussar OD, at a dose of 0.1 l/ha (Iodosulfuron-methyl-sodium 100 g/L), was applied on 6th April 2024 to protect plants against weeds. On 12 April 2024, the fungicide Ascra X pro (65 g/L fluopyram, 65 g/L bixafen, 130 g/L prothioconazole, and N,N-Dimethyl decanamide) was sprayed at a dose of 1.5 L/ha together with the plant growth regulator Moddus evo (250 g/L trinexapac-ethyl) at a dose of 0.5 L/ha. The average temperatures during the vegetative season were 11.96 °C with a rainfall amount of 1439.9 mm.

2.2. Fungicide Application

The spraying with fungicide was applied on 25 May 2024 in the morning in the anthesis growing phase. The fungicide used was Prosaro at a dose of 1 l/ha (125 g/L prothioconazole and 125 g/L tebuconazole). For fungicide application, the mounted sprayer Agromehanika 600 EN was used, with a nominal volume of 600 l and a spray boom of 12 m, which is hydraulically foldable. The tractor was a Fendt 208 S with a nominal power of 60 kW. On the day of spraying, the average air temperature was 16 °C, the average wind speed was 1.1 m/s, and the relative moisture was 70%. The treatments were four different flow rates for the same nozzle type. The type of nozzle was an asymmetrical twin flat spray air-injector nozzle, namely IDTA. The front angle of spraying was 120°, and the back angle of spraying was 90°. In the direction of travel, the front spray angle was 30° to the vertical line, and the back spray angle was 50° to the vertical line. Moreover, 60% of the spray volume comes from the front spray jet, and 40% of the spray volume comes from the rear spray jet [29]. The first nozzle treatment was conducted with the IDTA 120-02 nozzle at 120 L/ha (T1 in Figure 2), the second nozzle treatment was conducted with the IDTA 120-03 nozzle at 180 L/ha (T2 in Figure 2), the third nozzle treatment was conducted with the IDTA 120-04 nozzle at 240 L/ha (T3 in Figure 2), and the fourth nozzle treatment was conducted with the IDTA 120-08 nozzle at 480 L/ha (T4 in Figure 2). The fifth treatment was an unsprayed control (TC in Figure 2), where the fungicide was not applied. The spraying speed was 8 km/h with a spraying pressure of 3.0 bar.

2.3. Fusarium Inoculum and Inoculations

The Fusarium graminearum (originated from Croatia) and F. culmorum (originated from Austria) species were grown on synthetic nutrient-deficient agar (SNA) for 14 days at 24 °C. The agar pieces were transferred to a mixture of wheat and oat grains (3:1 by volume) previously soaked in water overnight. After six weeks of spore multiplication, the concentration of Fusarium spores was calibrated using a hemocytometer (Bürker-Türk, Hecht Assistent, Sondheim vor der Rhön, Germany) to reach 1 × 105 spores per ml. To inoculate plants, wheat spikes were sprayed with a conidial suspension (F. graminearum plus F. culmorum) of 100 mL on an area of 6 m2. The inoculum was applied when 50% of the plants per plot were at the stage of anthesis [30] in the late afternoon on 25 May 2024 by usage of the electric back-sprayer Solo 416 and AVI Twin 110 03 nozzle for all treatments. The spraying pressure with the electric back-sprayer was 2.5 bar. After inoculation, water was applied on a couple occasions during the day to maintain moisture for successful infection. The inoculation was repeated two days later.

2.4. Unmanned Aerial Multispectral Measurement System

The unmanned aircraft system (UAS) used for the multispectral image acquisition was a DJI Mavic 3M (SZ DJI Technology Co., Ltd., City Shenzhen, China). The UAS uses its own GPS and GLONASS receiver for positioning, which is enhanced by the real-time kinematic module RTK connected to the Slovenian national network of fixed Global Navigation Satellite System (GNSS) stations, SIGNAL. The absolute positioning accuracy of the UAS is expected to be in the range of 10 cm. The bands of the multispectral camera are green (G) (560 ± 16 nm), red (R): 650 ± 16 nm, red edge (RE): 730 ± 16 nm, and near infrared (NIR): 860 ± 26 nm, with a 1/2.8-inch CMOS image sensor and 5 million pixels. The UAS also has an RGB camera with a 4/3 CMOS image sensor and 20 million pixels.
The flights with the UAS were performed on 7 June, 18 June, and 9 July 2024. The flight mission was planned with the DJI Pilot 2 application (SZ DJI Technology Co., Ltd., Shenzhen, China), with a nadir mapping flight at 12 m above the ground with 80% frontal and 80% lateral overlap. Based on the camera resolution and the height above the ground, the average ground sampling distance for RGB images is 0.33 cm/pixel and for multispectral images 0.52 cm/pixel.
All acquired images were imported into the photogrammetric software Pix4Dmapper (Pix4D SA, City Prilly, Switzerland) to create the 3D map of the wheat field. The RGB images were used in a 3D maps project template that generated an orthophoto image of the field. The multispectral images were processed using the agriculture multispectral template, which generated a map with specific spectral values for each pixel that were used to calculate a vegetation index.
Once the 3D model of the field was created, 2 wooden stakes marking the center of each of the test blocks were identified and marked on multiple images, creating a 3D point in the point cloud. This was necessary to align the UAS images and minimize the relative error between the flight surveys at different dates. Therefore, the process was repeated for each flight date. Based on these points, the 6 m × 1 m blocks (as shown in Figure 2) were marked, georeferenced, and transferred to the index map. The average index values of the blocks were calculated and used in the statistical analysis.
The spectrometric data from the red and near infrared bands are extracted from the georeferenced images at each pixel and the normalized difference vegetation index [20] is calculated using the following equation:
NDVI = (NIR − R)/(NIR + R),
where NIR stands for the near-infrared spectral band, and R stands for the visible red spectral band. Additionally, the red-edge spectral band (RE) was used for the evaluation of the normalized difference red-edge index (NDRE):
NDRE = (NIR − RE)/(NIR + RE),
And the green spectral band (G) was used for the evaluation of the green normalized difference vegetation index (GNDVI):
GNDVI = (NIR − G)/(NIR + G).

2.5. Deoxynivalenol Measurement

At full maturity of the wheat plants, 70 spike samples for DON analysis were taken manually immediately after the last measurement with the UAS. Firstly, the wheat spikes were picked from inoculated trial plots, then the spikes were threshed, and further grain samples were additionally cleaned with air. Further, 50 g of grains from each sample were taken to the Agricultural and Forestry Institute of Ptuj (Slovenia) for DON analysis. The quantitative test (lateral flow) was based on immunochromatographic principles [31].

2.6. Statistical Analysis

To minimize the effect on the environment, a randomized complete block design was applied. Samples were evaluated and collected from each plot, whereas evaluations were performed in three biological replications. Collected data were statistically analyzed using the Statistica software (version 14, TIBCO Software Inc., Palo Alto, CA, USA). Fisher’s LSD test at a 5% probability level was calculated to test significant differences among mean values. The analyzed parameters were expressed as the mean value of three replicates ± standard deviation (SD).

3. Results

3.1. DON Concentration and NDVI Values in Different Treatments

Analysis of variance for DON content and three NDVI values measured at three different measurement points showed a significant effect of treatment (different nozzle types for fungicide protection against FHB) for deoxynivalenol (DON) concentration only (Table 1).
DF: degree of freedom; MS: mean square; DON: deoxynivalenol concentration; NDVI 1: normalized difference vegetation index (NDVI) value measured on 7th of June 2024; NDVI 2: NDVI value measured on 18th of June 2024; NDVI 3: NDVI value measured on 9th of July 2024
The control treatment exhibited the highest recorded DON concentration (10,451.0 µg/kg) in the grains, followed by the fungicide treatment sprayed with nozzle type 110-02 (4434.0 µg/kg) (Figure 3). The DON concentrations for treatments with nozzle types 110-03 and 110-04 were significantly different from the DON concentration for the first two treatments, while the application of fungicide with nozzle type 110-08 resulted in the significantly lowest DON accumulation in the grains (810.7 µg/kg).

3.2. Vegetative Indices in Different Treatments and Measurement Points

The analysis of variance revealed significant differences between the normalized difference vegetation index (NDVI), normalized difference red-edge index (NDRE), and green normalized difference vegetation index (GNDVI) values at three measurement points after anthesis (13th, 24th, and 45th day after anthesis) (Table 2).
Figure 4 shows that the NDVI values differed significantly between all treatments at the different measurement times measured with the UAS. The NDVI value measured on the 24th day after anthesis was significantly lower compared to the NDVI on the 13th day after anthesis, while the NDVI measured at the last measurement point (on the 45th day after anthesis) was significantly lower compared to the first and second measurement points for all treatments.

3.3. Relations Between Investigated Traits

A significant negative correlation was obtained between DON accumulation and the NDVI, NDRE, and GNDVI values measured on 24th day after anthesis (Table 3). On the 13th day after anthesis, disease symptoms were still very low, without a change in the color of spike tissue, while symptoms on the 45th day after anthesis were not visible anymore, as plants were already in senescence. There is also a significant positive correlation between the NDVI and GNDVI values measured on the 13th day after anthesis and between the NDVI, NDR, and GNDVI values measured on the 24th day after anthesis, while the NDVI and GNDVI measured on the 45th day after anthesis exhibited a significant negative correlation with the GNDVI value measured on the 13th day after anthesis.
DON: deoxynivalenol concentration; NDVI1: normalized difference vegetation index (NDVI) value measured on the 13th day after anthesis; NDVI2: NDVI value measured on the 24th day after anthesis; NDVI3: NDVI value measured on the 45th day after anthesis; NDRE1: normalized difference red-edge index (NDRE) value measured on the 13th day after anthesis; NDRE2: NDRE value measured on the 24th day after anthesis; NDRE3: NDRE value measured on the 45th day after anthesis; GNDVI1: green normalized difference vegetation index (GNDVI) value measured on the 13th day after anthesis; GNDVI22: GNDVI value measured on the 24th day after anthesis; GNDVI3: GNDVI value measured on the 45th day after anthesis.
Through a general regression model (Pareto chart), it was evident that measurements at different points after anthesis were significant for the NDVI and GNDVI but not for the NDRE (Figure 5).

4. Discussion

FHB disease has always been a major obstacle to wheat productivity, as it decreases the grain yield and quality. Also, FHB endangers the health of humans and animals through the accumulation of mycotoxins. FHB monitoring in wheat fields has mainly relied on symptom assessment by professional plant protection personnel who are unable to meet the demands of monitoring this devastating disease on a large scale. Detection of Fusarium fungi in the early stage of the infection process may have a strong role in efficient fungicide treatment. Only a few days after infection (up to 14 days), but depending on weather conditions and disease pressure, the healthy spikes will remain green. The diseased spikelets will lose color, and the infection will gradually spread over the spike [32]. Physiological and biochemical changes in plants alter spectral reflectance, forming the foundation of optical technologies for diagnosing FHB severity [33]. The remote sensing method has the potential for detecting changes before infections are visible on the plants [34]. To detect spectral variations in plants, vegetation indices (VIs) are a simple and effective tool [35] because of the nature of plant tissue damage from diseases. In the current research, the multispectral images of the wheat plots were made using a UAS, and vegetation indices were extracted from the multispectral images. To define the greenness availability exposure, the normalized difference vegetation index (NDVI) was calculated to show the relative abundance and spatial distribution of vegetation. Previously, it was shown that the NDVI measured using a UAV high-throughput phenotyping platform can increase the precision of selection of breeding material [36]. Also, there is a strong relation of the NDVI with grain yield for some physiological characteristics that influence grain yield under different stresses. Using spectral data, water accumulation and nitrogen and protein content in the grains were estimated through the NDVI in the research of Tan et al. [37]. The added value of NDVI application is found in developing new strategies for fertilization that can adjust nitrogen fertilization to the needs of plants [38]. For example, Cabrera-Bosquet et al. [39] observed a strong correlations between NDVI values and total green area, dry aboveground biomass and nitrogen content, and green area without spikes. Additionally, the normalized difference red-edge index (NDRE) and green normalized difference vegetation index (GNDVI) values were calculated to check for correlations with NDVI values. Similar to the NDVI, the NDRE is used for the measurement of the amount of chlorophyll in the plants. In the previous research, the possibility of using the NDVI and NDRE at later stages of plant development was indicated [40]. The same authors stressed that correction of the NDVI index can be accomplished by using the NDRE index, allowing for a better understanding of the obtained results. Another remote index is the GNDVI which shows plants’ “greenness” or photosynthetic activity. It was reported that the GNDVI measures chlorophyll content more accurately than the NDVI, as the GNDVI has a higher saturation threshold [41].
Therefore, the present study, involving a comprehensive range of fungicide treatments with different nozzle types together with Fusarium inoculation and NDVI measurement after treatments, was used to confirm the importance of these vegetative indices on DON production. The measurement of FHB severity by multispectral imaging is possible as FHB mainly can be found on wheat spikes, resulting in bleaching and the loss of chlorophyll, which can be seen in the red-edge band [42]. However, it is not known if there is a relationship between NDVI values of diseased spikes and DON accumulation from grains obtained from these spikes. Usually, FHB severity correlates with mycotoxin accumulation, but DON content is not always in significant correlation with FHB severity [2].

4.1. DON in Different Treatments

DON was selected for measurement as it is one of the most abundant mycotoxins. Previously, DON was highly correlated with other detected mycotoxins in wheat grains [14]. From the current research, it was evident that fungicide treatment with any type of nozzle may reduce DON accumulation at rates from 58% to 92%, compared to controls without fungicide treatment. Similar results were obtained in durum wheat when the Avi Twin 04 nozzle caused a lower but significant reduction in DON content by 45% [31]. The explanation for observing the lowest DON content in grains when the nozzle type 110-08 was used for fungicide application in the current research lies in the nozzle’s better covering of spikes with droplets of fungicide. The DON content in samples with FHB inoculation without fungicide application was comparable with DON content from two research studies of Spanic et al. [14,43], where maximal values of DON reached more than 20,000 µg/kg. It is important to highlight how the timing of fungicide applications is crucial to prevent or shorten fungal growth and spread. This is especially important as most diseases can spread many days before FHB symptoms become noticeable. When visible symptoms can be seen, it has likely already caused significant damage to plant tissue, and it is often too late for effective remediation [44]. When there are warm and humid conditions, the first FHB symptoms appear in less than 14 days, although the initial infection started even earlier [45]. Thus, to improve the effectiveness of the fungicide and prevent additional sprays, the right timing may be assessed through remote sensing by detecting infection before the symptoms of the disease show up on the plants. This is especially significant when there are many plots to evaluate, as it is very well known that programs for wheat improvement require the rapid assessment of large numbers of plots across multiple locations and years. The classical approach to plant disease evaluation in field conditions is labor intensive and tends to be subjective with lower efficiency; remote sensing techniques can be a good supplement for evaluating plant diseases [46].

4.2. Relations of Investigated Traits

It is important to note that the NDVI values measured by multispectral UAS analysis may comprise traits that can be used in FHB symptom evaluation to predict DON accumulation; in the current research, a significant negative correlation was observed between DON accumulation and the NDVI values on the 24th day after flowering. This means that the healthier plants in the plots exhibited lower accumulation of DON. Previous reports revealed the fact that the NDVI shows plant light absorption and the reflection of light by plants, indicating the status of plant health [47]. Developing multispectral data analysis methods based on time series can be used to monitor the trend of development and how the FHB is transmitted [20]. The results of remote sensing measurements taken on different dates and under different weather conditions around anthesis could predict FHB severity [22]. Another study showed that the high sensitivity of hyperspectral imaging and infrared thermography could enable early detection of FHB before the symptoms become visible. Moreover, a 78% precision for differentiation between control and FHB-infected spikelets at 3 days after inoculation was observed [48]. From the results of the current study, it can be concluded that more measurements time points between the 13th and 24th day after wheat anthesis are needed to ensure precise determination of time points for fungicide treatments.
However, the current research did not reveal significant differences between NDVI values at one measurement point between the control and different types of nozzles. This is most likely due to the UAS’s relative height of 12 m above wheat plants which provides a larger ground sampling distance, along with the natural spike illumination and nadir flight. Contrary to this, the research of Almoujahed et al. [49] showed significant differences in spectral reflectance according to different varietal resistance levels. The reason for this could be a special platform setup and a hyperspectral camera with a spectral range of 400–1000 nm and lamps for extra illumination. Another study also showed that extracted wavelet features could be used in the detection of FHB severity, showing better sensitivity to FHB than traditional spectral features [50]. Also, in many studies, observation data were pooled at different stages of the entire growth stage to explore characteristic bands and construct spectral indices [33]. However, the NDVI value significantly changed between different measurement points, gradually decreasing from the first to third measurement. This was expected as NDVI values should increase continuously until wheat maturity, after which the values will decrease, indicating senescence. The significant effect of senescence on NDVI values has already been concluded [51]. Therefore, NDVI measurements have significant potential to rapidly detect variations in senescence among wheat genotypes [52]. It is suggested that powdery mildew occurrence in wheat may be predicted very well by integrating meteorological and remote sensing data [53]. Also, another research study suggested that the NDVI can be used to predict the tolerance of wheat varieties to the root-lesion nematode Pratylenchus thornei [54]. The research of Du et al. confirmed the effectiveness of remote sensing technologies for the detection of winter wheat disease/pest infection [55]. There is also evidence that plant health status or the level of stress in plants may be measured by remote sensing [56].
Furthermore, there were significant correlations obtained between the NDVI and NDRE and the NDVI and GNDVI on the 24th day after anthesis, which, moreover, exhibited a significant negative correlation with DON content. This confirmed previous reports about the strong correlations between NDVI and NDRE values [57] and between NDVI and GNDVI values [58]. However, our results demonstrated that the NDVI had the highest correlation with DON accumulation. Previously, it was also reported that the NDVI exhibited better performance than the NDRE and GNDVI in the health estimation of trees [59]. It was reported previously that NDVI values had a high degree of relation to grain yield at the final stages of wheat development in different environments [60]. The current research showed that the effectiveness of the fungicide can be improved and additional sprays prevented; the right timing for application may be realized through remote sensing that can detect infection before the symptoms of disease become visible. Practically, it is possible to use NDVI measurements in the evaluation of wheat senescence and to predict DON accumulation, but different spectral features still need to be explored, and the UAS relative height needs to be adjusted. While the NDVI is a prominent choice in senescence detection and the assessment of DON accumulation, other VIs such as the NDRE and GNDVI could also be employed in wheat research due to their significant negative correlations with DON content. However, the Pareto chart showed that there were no significant differences for the NDRE at the three measurement points. Due to a lower number of plot replicates, we would repeat the experiments with more plots over the treatment and would add more genotypes.

5. Conclusions

This investigation, given the potential of FHB epidemics in wheat, expands the scope of traditional agronomic research by employing the NDVI to evaluate health status and photosynthesis efficiency in plants after strong Fusarium infection with or without fungicide protection and shows the relation of the NDVI with DON accumulation in grains. This research showed that it is important to select the right nozzle to adequately cover a spike (the lowest DON concentration was obtained by usage of nozzle type 110-08 for fungicide application). The potential of the NDVI to differentiate wheat stages of maturity was also demonstrated. However, significant differences under the same treatment with the same nozzle type and control were obtained only for DON accumulation, which exhibited a significant correlation with the NDVI value measured on the 25th day after anthesis. Therefore, this approach enables early, non-destructive detection of FHB stress, facilitating optimal fungicide application timing. For future research, however, more multispectral measurements at different times and different flight parameters with different UASs, as well as the inclusion of different wheat varieties and precise fungicide treatments, are needed to obtain comprehensive conclusions.

Author Contributions

Conceptualization, F.V. and I.P.; methodology, F.V.; software, I.P: and V.S.; validation, F.V., I.P. and V.S.; formal analysis, F.V.; investigation, F.V. and I.P.; resources, V.S. and I.P.; data curation, F.V.; writing—original draft preparation, F.V., I.P. and V.S.; writing—review and editing, F.V.; visualization, I.P.; supervision, V.S.; project administration, F.V.; funding acquisition, V.S. All authors have read and agreed to the published version of the manuscript.

Funding

The authors would like to thank the Slovenian research agency ARIS for financial contribution to this work. This work was partly financed by the research programs Animal health, environment and food safety P4-0092 and Mechanics in Engineering P2-0263.

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Khan, M.K.; Pandey, A.; Athar, T.; Choudhary, S.; Deval, R.; Gezgin, S.; Hamurcu, M.; Topal, A.; Atmaca, E.; Santos, P.A.; et al. Fusarium head blight in wheat: Contemporary status and molecular approaches. 3 Biotech. 2020, 10, 1–17. [Google Scholar] [CrossRef] [PubMed]
  2. Spanic, V.; Zdunic, Z.; Drezner, G.; Sarkanj, B. The Pressure of Fusarium Disease and Its Relation with Mycotoxins in the Wheat Grain and Malt. Toxins 2019, 11, 198. [Google Scholar] [CrossRef] [PubMed]
  3. Yu, J.; Pedroso, I.R. Mycotoxins in Cereal-Based Products and Their Impacts on the Health of Humans, Livestock Animals and Pets. Toxins 2023, 15, 480. [Google Scholar] [CrossRef] [PubMed]
  4. Scudamore, K.; Patel, S. The Fate of Deoxynivalenol and Fumonisins in Wheat and Maize during Commercial Breakfast Cereal Production. World Mycotoxin J. 2008, 1, 437–448. [Google Scholar] [CrossRef]
  5. Bottalico, A.; Perrone, G. Toxigenic Fusarium Species and Mycotoxins Associated with Head Blight in Small-Grain Cereals in Europe. Eur. J. Plant Pathol. 2002, 108, 611–624. [Google Scholar] [CrossRef]
  6. Maresca, M. From the Gut to the Brain: Journey and Pathophysiological Effects of the Food-Associated Trichothecene Mycotoxin Deoxynivalenol. Toxins 2013, 5, 784–820. [Google Scholar] [CrossRef] [PubMed]
  7. European Commission Commission Regulation (EC) No 1881/2006 of 19 December 2006 Setting Maximum Levels for Certain Contaminants in Foodstuffs. Off. J. Eur. Union 2006, 364, 5–24.
  8. European Commission Commission Recommendation of 17 August 2006 on the Presence of Deoxynivalenol, Zearalenone, Ochratoxin A, T-2 and HT-2 and Fumonisins in Products Intended for Animal Feeding. Off. J. Eur. Union 2006, 229, 7–9.
  9. European Commission Commission Recommendation of 27 March 2013 on the Presence of T-2 and HT-2 Toxin in Cereals and Cereal Products. Off. J. Eur. Union 2013, 91, 12–15.
  10. Shude, S.P.; Yobo, K.S.; Mbili, N.C. Progress in the management of Fusarium head blight of wheat: An overview. S. Afr. J. Sci. 2020, 116, 60–66. [Google Scholar] [CrossRef] [PubMed]
  11. Bushnell, W.R.; Hazen, B.E.; Pritsch, C. Histology and Physiology of Fusarium Head Blight. In Fusarium Head Blight of Wheat and Barley; The American Phytopathological Society: St. Paul, MN, USA, 2003; pp. 44–83. [Google Scholar]
  12. Moretti, A.; Pascale, M.; Logrieco, A.F. Mycotoxin Risks under a Climate Change Scenario in Europe. Trends Food Sci. Technol. 2019, 84, 38–40. [Google Scholar] [CrossRef]
  13. Somers, D.J.; Fedak, G.; Savard, M. Molecular Mapping of Novel Genes Controlling Fusarium Head Blight Resistance and Deoxynivalenol Accumulation in Spring Wheat. Genome 2003, 46, 555–564. [Google Scholar] [CrossRef] [PubMed]
  14. Spanic, V.; Maricevic, M.; Ikic, I.; Sulyok, M.; Sarcevic, H. Three-Year Survey of Fusarium Multi-Metabolites/Mycotoxins Contamination in Wheat Samples in Potentially Epidemic FHB Conditions. Agronomy 2023, 13, 805. [Google Scholar] [CrossRef]
  15. McMullen, M.; Bergstrom, G.; De Wolf, E.; Dill-Macky, R.; Hershman, D.; Shaner, G.; Van Sanford, D. A Unified Effort to Fight an Enemy of Wheat and Barley: Fusarium Head Blight. Plant Dis. 2012, 96, 1712–1728. [Google Scholar] [CrossRef] [PubMed]
  16. Freije, A.N.; Wise, K.A. Impact of Fusarium Graminearum Inoculum Availability and Fungicide Application Timing on Fusarium Head Blight in Wheat. Crop Prot. 2015, 77, 137–147. [Google Scholar] [CrossRef]
  17. Bolanos-Carriel, C.; Wegulo, S.N.; Baenziger, P.S.; Funnell-Harris, D.; Hallen-Adams, H.E.; Eskridge, K.M. Effects of Fungicide Chemical Class, Fungicide Application Timing, and Environment on Fusarium Head Blight in Winter Wheat. Eur. J. Plant Pathol. 2020, 158, 667–679. [Google Scholar] [CrossRef]
  18. Inoue, Y. Satellite- and Drone-Based Remote Sensing of Crops and Soils for Smart Farming–a Review. Soil. Sci. Plant Nutr. 2020, 66, 798–810. [Google Scholar] [CrossRef]
  19. Liu, J.; Xiang, J.; Jin, Y.; Liu, R.; Yan, J.; Wang, L. Boost Precision Agriculture with Unmanned Aerial Vehicle Remote Sensing and Edge Intelligence: A Survey. Remote Sens. 2021, 13, 4387. [Google Scholar] [CrossRef]
  20. Dong, P.; Wang, M.; Li, K.; Qiao, H.; Zhao, Y.; Bacao, F.; Shi, L.; Guo, W.; Si, H. Research on the Identification of Wheat Fusarium Head Blight Based on Multispectral Remote Sensing from UAVs. Drones 2024, 8, 445. [Google Scholar] [CrossRef]
  21. Mustafa, G.; Zheng, H.; Khan, I.H.; Zhu, J.; Yang, T.; Wang, A.; Xue, B.; He, C.; Jia, H.; Li, G.; et al. Enhancing Fusarium Head Blight Detection in Wheat Crops Using Hyperspectral Indices and Machine Learning Classifiers. Comput. Electron. Agric. 2024, 218, 108663. [Google Scholar] [CrossRef]
  22. Xiao, Y.; Dong, Y.; Huang, W.; Liu, L.; Ma, H.; Ye, H.; Wang, K. Dynamic Remote Sensing Prediction for Wheat Fusarium Head Blight by Combining Host and Habitat Conditions. Remote Sens. 2020, 12, 3046. [Google Scholar] [CrossRef]
  23. McConachie, R.; Belot, C.; Serajazari, M.; Booker, H.; Sulik, J. Estimating Fusarium Head Blight Severity in Winter Wheat Using Deep Learning and a Spectral Index. Plant Phenome J. 2024, 7, e20103. [Google Scholar] [CrossRef]
  24. West, J.S.; Canning, G.G.M.; Perryman, S.A.; King, K. Novel Technologies for the Detection of Fusarium Head Blight Disease and Airborne Inoculum. Trop. Plant Pathol. 2017, 42, 203–209. [Google Scholar] [CrossRef]
  25. Feng, G.; Gu, Y.; Wang, C.; Zhou, Y.; Huang, S.; Luo, B. Wheat Fusarium Head Blight Automatic Non-Destructive Detection Based on Multi-Scale Imaging: A Technical Perspective. Plants 2024, 13, 1722. [Google Scholar] [CrossRef] [PubMed]
  26. Huete, A.; Didan, K.; Miura, T.; Rodriguez, E.P.; Gao, X.; Ferreira, L.G. Overview of the Radiometric and Biophysical Performance of the MODIS Vegetation Indices. Remote Sens. Env. Environ. 2002, 83, 195–213. [Google Scholar] [CrossRef]
  27. Mirasi, A.; Mahmoudi, A.; Navid, H.; Valizadeh Kamran, K.; Asoodar, M.A. Evaluation of Sum-NDVI Values to Estimate Wheat Grain Yields Using Multi-Temporal Landsat OLI Data. Geocarto Int. 2021, 36, 1309–1324. [Google Scholar] [CrossRef]
  28. AgroChema. Available online: https://www.agrochema.lt/en/products-for-agriculture/winter-wheat-aurelius/ (accessed on 29 November 2024).
  29. Lechler GmbH IDTA. Available online: https://www.lechler.com/de-en/products-nozzles-spray-technology-systems/product-range/agriculture/nozzles-for-broadcast-spraying/idta (accessed on 29 November 2024).
  30. Zadoks, J.C.; Chang, T.T.; Konzak, C.F. A Decimal Code for the Growth Stages of Cereals. Weed Res. 1974, 14, 415–421. [Google Scholar] [CrossRef]
  31. Vučajnk, F.; Španić, V.; Trdan., S.; Košir, I.J.; Ocvirk, M.; Vidrih, M. Performance of Symmetric Double Flat Fan Nozzles against Fusarium Head Blight in Durum Wheat. Agriculture 2024, 14, 2–16. [Google Scholar] [CrossRef]
  32. Spanic, V.; Sarcevic, H. Evaluation of Effective System for Tracing FHB Resistance in Wheat: An Editorial Commentary. Agronomy 2023, 13, 2116. [Google Scholar] [CrossRef]
  33. Zhang, D.; Wang, Q.; Lin, F.; Yin, X.; Gu, C.; Qiao, H. Development and Evaluation of a New Spectral Disease Index to Detect Wheat Fusarium Head Blight Using Hyperspectral Imaging. Sensors 2020, 20, 2260. [Google Scholar] [CrossRef] [PubMed]
  34. Eiko, T.; Heege, H.J. Site-Specific Sensing for Fungicide Spraying. In Precision in Crop Farming; Springer: Dordrecht, The Netherlands, 2013; pp. 295–311. [Google Scholar] [CrossRef]
  35. Ren, Y.; Huang, W.; Ye, H.; Zhou, X.; Ma, H.; Dong, Y.; Shi, Y.; Geng, Y.; Huang, Y.; Jiao, Q.; et al. Quantitative Identification of Yellow Rust in Winter Wheat with a New Spectral Index: Development and Validation Using Simulated and Experimental Data. Int. J. Appl. Earth Obs. Geoinf. 2021, 102, 102384. [Google Scholar] [CrossRef]
  36. Hassan, M.A.; Yang, M.; Rasheed, A.; Yang, G.; Reynolds, M.; Xia, X.; Xiao, Y.; He, Z. A Rapid Monitoring of NDVI across the Wheat Growth Cycle for Grain Yield Prediction Using a Multi-Spectral UAV Platform. Plant Sci. 2019, 282, 95–103. [Google Scholar] [CrossRef] [PubMed]
  37. Tan, C.W.; Zhang, P.P.; Zhou, X.X.; Wang, Z.X.; Xu, Z.Q.; Mao, W.; Li, W.X.; Huo, Z.Y.; Guo, W.S.; Yun, F. Quantitative Monitoring of Leaf Area Index in Wheat of Different Plant Types by Integrating NDVI and Beer-Lambert Law. Sci. Rep. 2020, 10, 1–10. [Google Scholar] [CrossRef]
  38. Aranguren, M.; Castellón, A.; Aizpurua, A. Wheat Yield Estimation with NDVI Values Using a Proximal Sensing Tool. Remote Sens. 2020, 12, 2749. [Google Scholar] [CrossRef]
  39. Cabrera-Bosquet, L.; Molero, G.; Stellacci, A.; Bort, J.; Nogués, S.; Araus, J. NDVI as a Potential Tool for Predicting Biomass, Plant Nitrogen Content and Growth in Wheat Genotypes Subjected to Different Water and Nitrogen Conditions. Cereal Res. Commun. 2011, 39, 147–159. [Google Scholar] [CrossRef]
  40. Voitik, A.; Vasyl, K.; Pushka, O.; Kutkovetska, T.; Shchur, T.; Kocira, S. Comparison of NDVI, NDRE, MSAVI and NDSI Indices for Early Diagnosis of Crop Problems. Agric. Eng. 2023, 27, 47–57. [Google Scholar] [CrossRef]
  41. Daliman, S.; Michael, M.J.A.; Rendra, P.P.R.; Sukiyah, E.; Hadian, M.S.D.; Sulaksana, N. Dual Vegetation Index Analysis and Spatial Assessment in Kota Bharu, Kelantan using GIS and Remote Sensing. BIO Web Conf. 2024, 131, 05009. [Google Scholar] [CrossRef]
  42. Gao, C.; Ji, X.; He, Q.; Gong, Z.; Sun, H.; Wen, T.; Guo, W. Monitoring of Wheat Fusarium Head Blight on Spectral and Textural Analysis of UAV Multispectral Imagery. Agriculture 2023, 13, 293. [Google Scholar] [CrossRef]
  43. Spanic, V.; Katanic, Z.; Sulyok, M.; Krska, R.; Puskas, K.; Vida, G.; Drezner, G.; Šarkanj, B. Multiple Fungal Metabolites Including Mycotoxins in Naturally Infected and Fusarium-Inoculated Wheat Samples. Microorganisms 2020, 8, 578. [Google Scholar] [CrossRef]
  44. Gold, K.M. Plant Disease Sensing: Studying Plant-Pathogen Interactions at Scale. mSystems 2021, 6, e0122821. [Google Scholar] [CrossRef]
  45. Peršić, V.; Božinović, I.; Varnica, I.; Babić, J.; Španić, V. Impact of Fusarium Head Blight on Wheat Flour Quality: Examination of Protease Activity, Technological Quality and Rheological Properties. Agronomy 2023, 13, 662. [Google Scholar] [CrossRef]
  46. Zhang, J.; Huang, Y.; Pu, R.; Gonzalez-Moreno, P.; Yuan, L.; Wu, K.; Huang, W. Monitoring Plant Diseases and Pests through Remote Sensing Technology: A Review. Comput. Electron. Agric. 2019, 165, 104943. [Google Scholar] [CrossRef]
  47. Hnizil, O.; Baidani, A.; Khlila, I.; Nsarellah, N.; Laamari, A.; Amamou, A. Integrating NDVI, SPAD, and Canopy Temperature for Strategic Nitrogen and Seeding Rate Management to Enhance Yield, Quality, and Sustainability in Wheat Cultivation. Plants 2024, 13, 1574. [Google Scholar] [CrossRef]
  48. Mahlein, A.K.; Alisaac, E.; Al Masri, A.; Behmann, J.; Dehne, H.W.; Oerke, E.C. Comparison and Combination of Thermal, Fluorescence, and Hyperspectral Imaging for Monitoring Fusarium Head Blight of Wheat on Spikelet Scale. Sensors 2019, 19, 2281. [Google Scholar] [CrossRef] [PubMed]
  49. Almoujahed, M.B.; Rangarajan, A.K.; Whetton, R.L.; Vincke, D.; Eylenbosch, D.; Vermeulen, P.; Mouazen, A.M. Detection of Fusarium Head Blight in Wheat under Field Conditions Using a Hyperspectral Camera and Machine Learning. Comput. Electron. Agric. 2022, 203, 107456. [Google Scholar] [CrossRef]
  50. Huang, L.; Wu, K.; Huang, W.; Dong, Y.; Ma, H.; Liu, Y.; Liu, L. Detection of Fusarium Head Blight in Wheat Ears Using Continuous Wavelet Analysis and PSO-SVM. Agriculture 2021, 11, 998. [Google Scholar] [CrossRef]
  51. Di Bella, C.M.; Paruelo, J.M.; Becerra, J.E.; Bacour, C.; Baret, F. Effect of Senescent Leaves on NDVI-Based Estimates of FAPAR: Experimental and Modelling Evidences. Int. J. Remote Sens. 2004, 25, 5415–5427. [Google Scholar] [CrossRef]
  52. Anderegg, J.; Yu, K.; Aasen, H.; Walter, A.; Liebisch, F.; Hund, A. Spectral Vegetation Indices to Track Senescence Dynamics in Diverse Wheat Germplasm. Front. Plant Sci. 2020, 10, 1749. [Google Scholar] [CrossRef]
  53. Zhang, J.; Pu, R.; Yuan, L.; Huang, W.; Nie, C.; Yang, G. Integrating Remotely Sensed and Meteorological Observations to Forecast Wheat Powdery Mildew at a Regional Scale. IEEE J. Sel. Top. Appl. Earth Obs. 2014, 7, 4328–4339. [Google Scholar] [CrossRef]
  54. Robinson, N.A. Use of Normalised Difference Vegetation Index (NDVI) to assess tolerance of wheat cultivars to root-lesion nematodes (Pratylenchus thornei). Master’s Thesis, University of Southern Queensland, Darling Heights, QLD, Australia, 2019. [Google Scholar] [CrossRef]
  55. Du, X.; Li, Q.; Shang, J.; Liu, J.; Qian, B.; Jing, Q.; Dong, T.; Fan, D.; Wang, H.; Zhao, L.; et al. Detecting advanced stages of winter wheat yellow rust and aphid infection using RapidEye data in North China Plain. GIScience Remote Sens. 2019, 56, 1093–1113. [Google Scholar] [CrossRef]
  56. Oladejo, O.S.; Blesing, L.O.; Rotimi, A.D.; Oguntola, E. Assessment of plant health status using remote sensing and GIS techniques. Adv. Plants Agric. Res. 2018, 8, 517–525. [Google Scholar] [CrossRef]
  57. Dunn, B.W.; Dunn, T.S.; Hume, I.; Orchard, B.A.; Dehaan, R.; Robson, A. Remote Sensing PI Nitrogen Uptake in Rice. IREC Newsl. 2016, 195, 48–50. [Google Scholar]
  58. Suplito, M.R.; David, F.B.; Olalia, L.C. Relationship of vegetation indices and SPAD meter readings with sugarcane leaf nitrogen under Pampanga Mill District, Philippines condition. IOP Conf. Ser. Earth Environ. Sci. 2020, 540, 012016. [Google Scholar] [CrossRef]
  59. Lay, H.; Shao, Z.; Hor, S. Application of UAV-Based multispectral images for accessing oil palm trees health using online AI Platform. In Proceedings of the 2nd Intercontinental Geoinformation Days (IGD), Mersin, Turkey, 5–6 May 2021; pp. 155–158. [Google Scholar]
  60. Lemma, A.Z.; Hailemariam, F.M.; Abebe, K.A.; Bishaw, Z. Normalized difference vegetation index as screening trait to complement visual selections of durum wheat drought tolerant genotypes. Afr. J. Plant Sci. 2022, 16, 1–7. [Google Scholar] [CrossRef]
Agriengineering 07 00037 g001
Figure 1. Field location and orientation at the Biotechnical Faculty in Ljubljana, Slovenia, with the orthophoto made from the UAS-captured images from a 12 m altitude.
Figure 1. Field location and orientation at the Biotechnical Faculty in Ljubljana, Slovenia, with the orthophoto made from the UAS-captured images from a 12 m altitude.
Agriengineering 07 00037 g001
Agriengineering 07 00037 g002
Figure 2. Randomized complete block design of the testing field. Each bracket represents a 6 m × 1 m plot, on which five different treatments were replicated three times. The R stands for the replication number and the T for the treatment number. T1, T2, T3, and T4 stand for 110-02, 110-03, 110-04, and 110-08 nozzle types for fungicide application, and TC stands for the control treatment, respectively.
Figure 2. Randomized complete block design of the testing field. Each bracket represents a 6 m × 1 m plot, on which five different treatments were replicated three times. The R stands for the replication number and the T for the treatment number. T1, T2, T3, and T4 stand for 110-02, 110-03, 110-04, and 110-08 nozzle types for fungicide application, and TC stands for the control treatment, respectively.
Agriengineering 07 00037 g002
Agriengineering 07 00037 g003
Figure 3. Deoxynivalenol (DON) concentration in µg/kg in grains of winter wheat for four treatments with usage of different nozzle types for fungicide protection and control without usage of fungicide. Different letters show significant differences.
Figure 3. Deoxynivalenol (DON) concentration in µg/kg in grains of winter wheat for four treatments with usage of different nozzle types for fungicide protection and control without usage of fungicide. Different letters show significant differences.
Agriengineering 07 00037 g003
Agriengineering 07 00037 g004
Figure 4. Normalized difference vegetation index (NDVI) value of winter wheat in four treatments with usage of different nozzle types (T1-110-02, T2-110-03, T3-110-04, and T4-110-08) for fungicide protection and control without usage of fungicide (TC) at three measurement points (13th (blue column), 24th (red column), and 45th (grey column) day after anthesis). Different letters show significant differences.
Figure 4. Normalized difference vegetation index (NDVI) value of winter wheat in four treatments with usage of different nozzle types (T1-110-02, T2-110-03, T3-110-04, and T4-110-08) for fungicide protection and control without usage of fungicide (TC) at three measurement points (13th (blue column), 24th (red column), and 45th (grey column) day after anthesis). Different letters show significant differences.
Agriengineering 07 00037 g004
Agriengineering 07 00037 g005
Figure 5. Pareto chart showing the significant effect of three measurement points after anthesis on investigated traits.
Figure 5. Pareto chart showing the significant effect of three measurement points after anthesis on investigated traits.
Agriengineering 07 00037 g005
Table 1. Analysis of variance for deoxynivalenol (DON) concentration and normalized difference vegetation index (NDVI) values measured on 13th (NDVI 1), 24th (NDVI 2) and 45th (NDVI 3) day after anthesis of winter wheat.
Table 1. Analysis of variance for deoxynivalenol (DON) concentration and normalized difference vegetation index (NDVI) values measured on 13th (NDVI 1), 24th (NDVI 2) and 45th (NDVI 3) day after anthesis of winter wheat.
Source of Variation DF MS
DON NDVI 1 NDVI 2 NDVI 3
Treatment 4 41,740,892.8 *** 0.00072 0.00368 0.000173
Replication 3 7,271,865.8 *** 0.00049 0.00789 * 0.00006
Error 12 443501 0.00073 0.00147 0.000093
*, ***-significant at 0.5 and 0.001.
Table 2. Analysis of variance for normalized difference vegetation index (NDVI), normalized difference red-edge index (NDRE), and green normalized difference vegetation index (GNDVI) values across different treatments and measurement points.
Table 2. Analysis of variance for normalized difference vegetation index (NDVI), normalized difference red-edge index (NDRE), and green normalized difference vegetation index (GNDVI) values across different treatments and measurement points.
Source of Variation DF MS
NDVI NDREGNDVI
Measurement point 2 0.6780 *** 0.1953 ***0.2530 ***
Replication 2 0.0044 * 0.00520.0017
Treatment 4 0.0018 0.00090.0014
Error 36 0.0011 0.00310.0009
*, ***-significant at 0.5 and 0.001.
Table 3. Correlation analysis between four investigated traits.
Table 3. Correlation analysis between four investigated traits.
DONNDVI1NDVI2NDVI3NDRE1NDRE2NDRE3GNDVI1GNDVI2GNDVI3
DON1.00−0.17−0.65 **−0.19−0.10−0.60 *0.15−0.14−0.58 *0.19
NDVI1−0.171.000.35−0.360.170.31−0.310.91 **0.33−0.50
NDVI2−0.65 **0.351.000.10−0.140.93 **−0.080.140.99 **−0.13
NDVI3−0.19−0.360.101.00−0.130.190.22−0.470.100.34
NDRE1−0.100.17−0.14−0.131.00−0.24−0.380.36−0.20−0.24
NDRE2−0.60 *0.310.93 **0.19−0.241.000.100.080.90 **0.04
NDRE30.15−0.31−0.080.22−0.380.101.00−0.40−0.080.90 **
GNDVI1−0.140.91 **0.14−0.47 *0.360.08−0.401.000.11−0.55 *
GNDVI2−0.58 *0.340.99 **0.10−0.200.90 **−0.080.111.00−0.13
GNDVI30.19−0.50−0.130.34−0.240.040.90 **−0.55 *−0.131.00
*,**-significant at 0.05 and 0.01, respectively.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Petrović, I.; Vučajnk, F.; Spanic, V. Detection of Fusarium Head Blight in Wheat Using NDVI from Multispectral UAS Measurements and Its Correlation with DON Contamination. AgriEngineering 2025, 7, 37. https://doi.org/10.3390/agriengineering7020037

AMA Style

Petrović I, Vučajnk F, Spanic V. Detection of Fusarium Head Blight in Wheat Using NDVI from Multispectral UAS Measurements and Its Correlation with DON Contamination. AgriEngineering. 2025; 7(2):37. https://doi.org/10.3390/agriengineering7020037

Chicago/Turabian Style

Petrović, Igor, Filip Vučajnk, and Valentina Spanic. 2025. "Detection of Fusarium Head Blight in Wheat Using NDVI from Multispectral UAS Measurements and Its Correlation with DON Contamination" AgriEngineering 7, no. 2: 37. https://doi.org/10.3390/agriengineering7020037

APA Style

Petrović, I., Vučajnk, F., & Spanic, V. (2025). Detection of Fusarium Head Blight in Wheat Using NDVI from Multispectral UAS Measurements and Its Correlation with DON Contamination. AgriEngineering, 7(2), 37. https://doi.org/10.3390/agriengineering7020037

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.
Back to TopTop