Abstract: Background and Objective: Gray mold is a serious problem in strawberry production worldwide. Early diagnostic of diseased strawberry fruits allows export decision. The aim of the study was to test the capability of spectral measurement for detecting the asymptomatic Botrytis infection. Materials and Methods: Asymptomatic and symptomatic samples of 25 strawberry fruits were measured for gray mold using two reliable systems, qPCR and spectroradiometer, to detect the causal pathogen, Botrytis cinerea. Results: Molecular results showed the presence of gray mold in many asymptomatic fruits. Moreover, spectral analysis demonstrated a higher reflectance in healthy fruit than that of infected ones throughout the visible near infra red (VNIR) spectral range, while the short wave infra red (SWIR) spectral zone showed different degrees of gray mold infection. Conclusion: The VNIR found to be the best spectral zone that could differentiate between healthy and infected strawberry fruits due to infection impact on the cellular pigments of the fruit, while SWIR was the best spectral zone to classify infection degrees because of the changes in cellular structure and water content due to infection.
INTRODUCTION
Gray mold threatens all strawberry production areas: The field, storage, transit, marketing and export of strawberry fruits, due to the severe rot when the fruits begin to ripen. Botrytis cinerea Pers., is a wide host range necrotrophic fungus causing rot symptoms1 and can be difficult to be combated. Early detection of diseased strawberry fruits, before observation of symptoms, allow accurate diagnosis2-4, especially before fruit export. Molecular techniques have previously been used for pathogenic characterization4-7. Quantitative real time PCR (qPCR) application has been developed and offered the ability of simultaneous detection and quantification of pathogenic DNA based on its sequence and concentration8,9. The qPCR system has many advantages: high sensitivity, quantitative properties and specificity, which make this method suitable for routine diagnostic and disease management decisions, but it is highly expensive and time consuming.
The measurements of spectral reflectance of vegetation contain information on plant pigment concentration, cellular structure, plant anomalies10 and moisture content11. Spectral reflectance was previously reported to measure powdery mildew infection in wheat under greenhouse condition. In addition, VIS-NIR reflectance spectroscopy (325-1075 nm) had been used for early detection of many diseases like B. cinerea infection in eggplant leaves at symptomless stages10, powdery mildew disease in wheat12, Sclerotinia rot infection on celery13 and Huanglongbing (HLB) bacterial infection on citrus14,15. Delalieux et al.16 had also investigated the detection of apple scab using hyperspectral spectroscopy and found that the spectral domain between 1300-1700 nm is the most effective region to separate stressed from healthy plant leaves. This is the first study using both diagnostic techniques, spectroradiometer and qPCR, to distinguish Botrytis infected strawberries from healthy ones. The aim of this study was to investigate an alternative inexpensive methodology for early detection of gray mold using spectroradiometer.
MATERIALS AND METHODS
Sampling: Asymptomatic and symptomatic samples were collected, during 2013-2014 at Faculty of Agriculture, Ain Shams University, from two varieties of strawberry fruits (Fragaria ananassa Duchesne var. Sweet Charlie and Festival), which have been grown under natural condition. Fifteen samples from sweet charlie (samples: A, B, C, D, E, F, G, H, I, J, K, L, M, N, O) and ten samples from Festival (1, 2, 3, 4, 5, 15, 16, 17, 18, 19) were tested microscopically for their infection with gray mold and then kept at 4°C for further analysis.
Spectral reflectance measurements: An indoor black house laboratory was set up in the National Authority for Remote Sensing and Space Sciences (NARSS) Station in order to measure the spectral reflectance of healthy and infected samples using the ASD FieldSpec® 3 JR spectroradiometer. According to Zhou et al. 17 and Zhang et al.18, measurements have been carried out in the whole spectral range: Visible, near-infrared and short wave infrared, starting from 350-2500 nm (Table 1) with a spectral resolution of 3 nm from 350-1000 and 30 nm from 1000-2500 nm. A halogen light (50 W) was used as the light source for such indoor measurement. A white reference (spectralon) was used every 2 min in order to calibrate the spectral measurements. For each measurement, the number recorded was based on an average of five individual spectral measurements.
DNA extraction and molecular analysis preparation: Samples from all fruits were lyophilized and stored at -20°C. The extraction of DNA was performed by Plant Genomic DNA extraction KIT® (GeneJET, Thermo Scientific), according to the manufacturer’s manual. DNAs were quantified and adjusted to 10 ng μL–1 before molecular amplification.
Real-time quantitative PCR (qPCR): Two specific primers, Bc424f: 5'-GCT TCC CCC GTA TCG AAG A-3'; Bc491r: 5'-CGA ACG GCC AGG TCA TCT-3', were used according to Saito et al.19. The qPCR assay based on SYBR® green was performed on all DNA samples. Five microliter from each DNA sample (10 ng μL–1) was mixed in a final volume of 25 μL with 2X Fermentas SYBR green master-mix.
| Table 1: | Spectral regions of the electromagnetic radiation under study |
Amplification program was performed on a PicoRealTM system as the following: 95°C for 10 min, followed by 45 cycles for 30 sec at 95°C, 55°C for 45 sec and 72°C for 45 sec. Standard curve samples for qPCR were obtained by isolating DNA from B. cinerea mycelia. Negative control (NTC, no B. cinerea DNA) was included in each run. This amplification test was performed twice.
Dissociation curve analysis: The specificity of the qPCR reaction was confirmed using melting curve analysis in order to establish that differences observed in Ct values between reactions are valid and not due to the presence of non-specific products or contamination.
DNA quantification: Absolute quantification was performed by plotting unknown concentration of fruit samples on a standard curve of a serial dilution of known concentration. The standard curve was typically a plot of the cycle quantification (Cq, Ct) against the logarithm of the amount of pathogen’s DNA. A linear regression analysis of the standard plot was used to calculate the amount of pathogen’s DNA in unknown fruit samples. The slope was related to the PCR efficiency of the samples which was determined by performing serial dilutions of such samples. For a graph where log (DNA copy) was represented on the y-axis and Ct on the x-axis: PCR efficiency = ((10–1/slope) 1)×100%. (©2013 Thermo Fisher Scientific, Inc.) (http://www. thermoscientific.com/onebio).
RESULTS
Screening of gray mold degrees in strawberries: Nine fruits (samples: A, B, C, D, E, F, G, H and I) from sweet charlie and only one from festival (sample 1) appeared healthy. Whereas, six samples (J, K, L, M, N and O) from sweet charlie and nine (2, 3, 4, 5, 15, 16, 17, 18 and 19) from festival showed Botrytis infection. Some of asymptomatic fruits upon collection were revealed gray mold within 2 days in the laboratory. The symptomatic fruits were grouped under the two infection degrees of gray mold: Low infection without rot and high infection associated with rot (Fig. 1).
Spectral detection of gray mold in strawberry fruit: All strawberry fruits, taken from two varieties, demonstrated the same trend of spectral reflectance for healthy and infected ones despite the slightly difference of spectral pattern due to plant variety (Fig. 2a, b). Results showed a higher spectral reflectance for the healthy fruit at NIR spectral zone: 700-1200 and 700-900 nm in festival and sweet charlie, respectively. However, such healthy fruit has always a clear distinguished spectral pattern located between low and severely infected samples at SWIR spectral zone: 1200-2500 nm. At SWIR spectral zone, the lowest spectral reflectance was observed for the fruit with a low infection level, while severely infected fruit demonstrated the highest spectral reflectance at such spectral zone for both strawberry varieties (Fig. 2a, b).
| Fig. 1(a-b): | Infection degrees of gray mold in two strawberry varieties (natural infection), (a) Sweet charlie and (b) Festival. From left to right: healthy, low infection, high infection |
| Fig. 2(a-b): | Spectral reflectance pattern of healthy, low and severely infected strawberry fruits in two varieties (a) Festival and (b) Sweet charlie |
Calibration of qPCR curves: Genomic DNA obtained from B. cinerea was used as a positive template (Cont+) for qPCR with primers Bc424f and Bc491r. The PCR product melting temperature was 77±1.00°C and no contaminating products nor primer dimers shown in this reaction. The standard curve for B. cinerea was generated by plotting the log of DNA quantity against the Ct value determined by qPCR (Fig. 3). The slope of the standard curve was -0.297, corresponds to DNA amplification efficiency of 98%. The detection limit was defined as the lowest population of B. cinerea that could be detected using SYBR Green qPCR method.
Molecular detection of gray mold in strawberry fruits: Amplification of genomic DNAs of samples using qPCR, taken from all symptomatic and asymptomatic strawberry fruits (Var. Festival and Sweet Charlie), was performed by SYBR green. All symptomatic samples (2-3-4-5-15-16-17-18-19-J-K-L-M-N-O) showed positive amplification curves at different Ct values depending on Botrytis quantity within each infected fruit sample (Table 2). The asymptomatic sample, A, did not show any amplification curve, while the other 9 asymptomatic samples (1, B, C, D, E, F, G, H, I) showed amplification curves at different Ct values, indicating that they have a quiescent infection (Table 2). The maximum Ct values: 31.2 and 33.6 were observed in samples: D and 18, respectively and defined as the lowest pathogen quantities (0.4 and 0.1pg, respectively) that were detected using SYBR green real time PCR method (Table 2).
| Fig. 3: | Standard curve revealed a linear relationship (R2 = 0.98) between the DNA quantity of serial dilutions of target genomic DNA of B. cinerea and the cycle threshold (Ct) |
| Table 2: | Quantification of DNAs of B. cinerea in symptomatic and asymptomatic strawberry fruits using qPCR |
| Ct: Cycle threshold, Q (pg): Quantity of Pathogen’s DNA (picogram), Cont+: Positive control (Pathogen’s DNA) | |
DISCUSSION
The present study tests the capability of using spectral reflectance of strawberry fruits to detect asymptomatic Botrytis infection. The presence of B. cinerea in asymptomatic plant varieties had been previously reported4,5,9, concluded that alternative reliable and accurate techniques should be tested in parallel for early diagnostic of gray mold. This study used qPCR as a calibrated molecular system for early detection of B. cinerea to evaluate the accuracy of spectral reflectance for gray mold detection. Such alternative model could be potentially used to provide inexpensive synergic verification for early diagnostic. In the current study, post harvested fruits taken from naturally infected strawberry have been tested for gray mold infection. The measurements were in good agreements with existing data published in literature for asymptomatic and symptomatic plants, even if the hyperspectral reflectance levels differ. Generally, results showed that the healthy fruit having a higher spectral reflectance than that of infected ones throughout VNIR spectral range (700-900 nm) and showed a slightly difference of the spectral pattern between two strawberry varieties, may due to a different plant behavior according to the amount of water content20. Likewise, the typical spectral signature for healthy plant had been investigated throughout VNIR spectral range (VIS700-NIR1200 nm). This was in accordance with the previous studies on infected beans with Botrytis fabae which produced a decrease in NIR reflectance, at 800 nm due to infection21. In fact, VIS-NIR spectroscopy was previously applied to detect B. cinerea disease at spectral range 325-1075 nm in eggplant leaves at asymptomatic stages10 and also investigated Sclerotinia rot disease on celery13. However, the current study investigated the potential for the use of spectral information for gray mold detection at red bands (700 nm) in the visible zone and NIR bands which revealed a good potential to discriminate Botrytis infection. In contrast, spectral reflectance pattern of infected fruits was inversed at SWIR spectral zone (1200-2500 nm) which refers to the effect on fruit composition due to Botrytis infection (negative relation of reflectance pattern was found between VNIR and SWIR spectral zones). Accordingly, hyperspectral spectroscopy was found to be useful in classifying infected trees by Venturia inaequalis at the spectral regions, 1350-1750 and 2200-2500 nm16. Indeed, it was confirmed that VNIR (700-900 nm) was the best spectral zone for differentiation between healthy and infected strawberry fruits due to infection effect on the cellular pigments of the fruit, while SWIR (1200-2500 nm) was the best spectral zone to classify infection degrees due to change in cellular structure and water content. Moreover, results showed that SWIR zone could distinguish between spectral patterns of Botrytis populations within the infected fruits at SWIR2 spectral zone (2055-2315 nm) according to Aboelghar and Abdel Wahab22. In fact, vegetation stress was a result of complex physiological processes, such changes affect the spectral behavior of plants through spectral reflectance, producing changes in the normal spectral reflectance patterns of plants23,24. Therefore, using remote sensing for detection of plant diseases assumes that stress induced by the infection interferes with physical plant structure and affects the light energy absorption, therefore, altering the characteristics of spectral reflectance of the plants. Previous studies had shown that changes in cell composition created a qualitative or quantitative variance in the spectral reflectance between healthy and infected plants21. Furthermore, it was shown that Botrytis removes water from the grape, leaving behind solids with a higher percent, such as fruit acids, sugars and minerals, thus the shortwave infrared (1400-2500 nm) was strongly influenced by the water in plant tissue, particularly at 1450 and 1940 nm.
Visible and near infrared (VIS-NIR) spectroscopy had been used extensively to detect plant anomalies10,13,16, while other works had highlighted the importance of the red edge (maximum slope of plant reflectance at 680-720 nm) in order to predict plant stress. In addition, several studies had shown that spectral reflectance at 600-710 and 750-900 nm, the red and near infra-red wavelengths, respectively, were correlated with disease severity estimates. In the visible range, reflectance was considered to be influenced by plant pigments and in the near-infrared range reflectance is affected by changes in the anatomical structure of plant25.
The differences in the reflectance measurements at specific wavelength intervals between Botrytis-infected and healthy strawberries and their correlation with qPCR results for the presence of B. cinerea suggested spectral reflectance method as a promising application for non-cost-effective and non-destructive diagnostic method for gray mold indoors as well as outdoors.
CONCLUSION
The two diagnostic applications, spectroradiometer and qPCR, were evaluated to compare their reliability to distinguish Botrytis infected fruits from healthy ones. Both different systems discriminated between the healthy and infected strawberry fruits and demonstrated their accordance in measurement results. Generally, the qPCR cycle and the spectral reflectance values of healthy fruit were higher than those of infected ones along with the whole sample collection. The results demonstrated that VNIR is the best spectral zone that would discriminate between infected and healthy fruits, while SWIR-2 is the best spectral zone to distinguish between population patterns of Botrytis within the infected fruits.
SIGNIFICANCE STATEMENTS
This study carried out the possible synergistic usage of an economic technique (remote sensing) and will help the researcher to early detect gray mold without sample preparation prior to measurements. Thus, a new methodology on gray mold diagnostic and possibly other combinations, would be used as a routine work.
ACKNOWLEDGMENT
This study was supported financially by the Science and Technology Development Fund (STDF), Egypt with Grant No. 2131.