In previous studies in hydroponically growing Romaine lettuce, it was shown that the levels of internalized TV RNA was higher than the levels of infectious virus that could be recovered from the same samples . In addition it was found that employing an RNAse pre-treatment to degrade viral RNA that was not found in an intact viral capsid reduced the level of TV RNA detection . Similarly, in studies investigating murine norovirus and hepatitis A virus internalization in green onions and spinach more samples were found to be positive for viral RNA compared to infectious virus . Two types of growth substrates, soil and hydroponic solution are commonly used in pathogen internalization studies conducted with plants. Pathogen internalization has been detected more often in hydroponically grown plants compared to plants grown in soil . In this study, both hydroponic and soil growing green onions were contaminated with TV at a level of 1 × 106 PFU/ml. Similar to previous studies, we found that internalization of TV via the root of hydroponically growing green onion occurred while no TV internalization occurred in soil grown green onions. The same phenomena was also observed with HAV and MNV, which were not detected in green onion plants grown in soil for up to 20 days . Compared to soil systems,black plastic plant pots a drastic increase in virus internalization was observed in hydroponic systems with both HAV and MNV internalized up to 4 log RT-qPCR units . Most investigators have suggested that the motility in hydroponic solution may provide more opportunity for uptake and internalization into leafy green plants compared to soil systems .
The virus capsid has an ionic charge and it may remain bound to charged particles in the soil matrix reducing viral uptake by the roots . It is also possible that root damage induced by transplanting of seedlings into the hydroponic system also increased internalization in this study and others . In this study, no virus internalization occurred in radishes while there was viral internalization detected in green onions and Romaine lettuce. Internalization of Salmonella Typhimurium was also observed in lettuce and radish but not cress or spinach seedlings . In the same study, radish seedlings were susceptible to Salmonella internalization while mature radishes were resistant, indicating the importance of plant age on pathogen internalization and dissemination . The radishes used in this study were mature . In our study we found no significant decrease in the level of virus detected in the feed water of hydroponically growing radishes. This indicates that no viral uptake occurred via radish root and it is possible that these mature radishes were not permissive to pathogen internalization. Previously it was shown TV was internalized and disseminated in growing strawberry plants while the virus was not detected in peppers of growing green pepper plants . It was found that the green peppers had a significant anti-viral effect on TV and that this may play a role in the lack of internalization in bell pepper plants . In this study, we found that the levels of TV detected in Romaine lettuce increased over the study period, which is indicative of internalization and dissemination with no inhibition. However, the internalization and dissemination kinetics was different in green onions with high levels of internalized virus detected on day 1 post inoculation and decreasing over the study period. This may indicate that an inhibitory compound in the green onion may be inactivating the virus.
Green onions contain a wide array of phenolic compounds which have the potential to act as antimicrobials . We found that sap from the green onion root lead to a 1-log reduction in infectious TV only after 14 days of exposure. Previously, it was found that a component of radishes, trans-4–methylthio-3-butenyl isothiocyanate , possessed antimicrobial activity . In addition, the pigment found in the skin of radishes, anthocyanins, have also been shown to have antiviral activity. Different cyanidin glycosides present in radishes have antiviral activity against influenza A and B viruses and herpes-1 virus . Our data indicates that radish roots are not permissive to virus penetration and internalization; therefore, we investigated whether radishes had antiviral properties. A maximum 1-log reduction in infectious TV titer was achieved after 14 days of exposure of TV to radish sap, indicating only minimal antiviral effects. In conclusion, we have shown that i) human NoV and TV can be internalized in hydroponically grown green onions but not in soil grown green onions, ii) the magnitude of TV internalization is influenced by the level of virus present, and iii) different types of plants have different susceptibility for viral internalization. Salinity stress is a major abiotic stress that has significant adverse effects on crop productivity and yield. These negative effects include interference of root function in absorbing water, as well as the prevention of physiological and biochemical processes such as nutrient uptake and assimilation . Unfortunately, many regions around the world are facing a rapid increase in soil salinity and sodicity. It is estimated that at least 0.3 million hectares of farmland is becoming unusable annually, and another 20–46 million ha are suffering decreases in production potential each year . Nevertheless, even with lower yield potential, these salt-affected farmlands must continue to produce crops so the increasing demand for food can be met and food security concerns mitigated.
The lack of new productive land threatens food security, thus the productivity of existing marginal lands must improve.There are numerous potential solutions for mitigating salt stress, including genetic engineering of plants with salt tolerance and application of exogenous compounds such as hormones, growth regulators, or nanoparticles . Among the potential solutions, selecting plant varieties with high tolerance to salt stress appears to be one of the most promising approaches in utilizing salt-affected soil for crop production . Although some progress has been made using measurement of photosynthetic parameters as a more sensitive method to screen for salt tolerance , the standard process of selecting either conventionally–bred or transgenic salt-tolerant crop lines relies on laborious phenotyping to assess tolerance. Despite the emergence of innovative platforms, precise instrumentation, sophisticated sensors, and rapid development of advanced machine learning and deep learning algorithms, phenotyping is still a barrier to variety development. While DNA sequencing and plant genotyping has rapidly evolved, phenotyping still depends on conventional methods which are not as accurate or efficient. In general, these techniques can be time-consuming, destructive, subjective, and costly. In recent years, non-contact sensing technology, in particular imaging, has been extensively deployed as a potential substitute for conventional methods for high-throughput phenotyping of plants. Thanks to the advances in developing sensors with high spatial and spectral resolution, different imaging sensors including visible, fluorescence, thermal, and spectral imaging are available, each tailored for specific applications. Each of these sensing technologies can vary in their application, as well as limitations, in the context of plant phenotyping . Among these techniques, hyperspectral imaging is uniquely suited to provide insights into the internal activities of plants, leaf tissue structure, leaf pigments, and water content . HSI also provides the ability to investigate physiological dynamics of plants caused by environmental variables , and consequently has drawn substantial attention for plant phenotyping . Few research studies have attempted to identify salt stress in plants using hyperspectral reflectance. In a previous study, three potential indicators including blue, yellow, and red edge positions of vegetation reflectance spectrum were calculated to detect four levels of salt stress imposed on Chinese castor bean . The authors claimed that blue and red edge positions shift to the shorter wavelength in response to salt stress and therefore could be used to detect salt stress. However, the pattern of shifting to the shorter wavelength was not consistent across all treatments and hence further research is required. In another paper,black plastic planting pots the application of HSI to identify plant tolerance to salt stress in a high throughput phenotyping system was reviewed . They concluded more efficient and fully automated methods are required to analyze complex hyperspectral images. To leverage the full potential of HSI, a large high-quality hyperspectral dataset and several preprocessing tasks are necessary .
However, there are two major challenges that hamper the application of HSI. The first major challenge is accounting for the variance caused by the complex interaction between incident light and leaf surfaces due to non-Lambertian reflectance properties. The direction of reflected light is a function of leaf geometry, including leaf angle, and curvature. Several researchers have focused on pre-processing techniques to address the problems related to leaf angle and curvature . One method to resolve this problem is to generate a high-resolution 3D representation of plants by upfront geometric calibration of the hyperspectral camera . However, this proposed method depends on highly intensive processing and is only suitable for close-range imaging. The second major challenge is analyzing the complex and high-dimensional hyperspectral images in order to extract meaningful features and recognize latent patterns associated with the desired phenotyping trait in a more interpretable manner. To address this issue, machine learning and deep learning algorithms can be leveraged. Recent reviews of various ML algorithms emphasize the potential of these methods in the context of agriculture and provide guidelines for plant scientists to deploy them . Singh et al. reported that ML algorithms are a promising approach to analyze large datasets generated by sophisticated imaging sensors mounted to platforms that can cover large areas. Despite several studies that focus on the application of HSI for plant phenotyping, research is limited in the context of handling, processing, and analyzing hyperspectral images. This research was motivated by the need to identify salt tolerant wheat lines to mitigate yield losses due to salinity, and to ultimately maintain or improve production on saline soils. The objectives of this study were to rank wheat lines based on their tolerance to salt stress, assess the difference between the salt tolerance of lines to attain a quantitative ranking rather than a qualitative ranking, and evaluate the feasibility of precise ranking of wheat lines as early as one day after applying salt treatment. We hypothesized that the spectral response of wheat leaves experiencing salt stress would deviate from the control leaves even one day after adding the stress, and this deviation would be larger for a susceptible line compared to a salt tolerant line. To the best of our knowledge, no previous study has investigated early detection of salt tolerant plant lines using advanced phenotyping tools and approaches. This research proposes a machine learning approach to analyze hyperspectral images of wheat lines to rank their salt stress tolerance in a quantitative, interpretable, and non-destructive manner while reducing cost, time, and labor input.To develop analytical methods for analysis of hyperspectral images, four bread wheat lines were selected with varying levels of salt tolerance. The cultivar Kharchia was included as it is historically known to maintain a stable harvest index and yield well in high salt conditions , and the salt-sensitive cultivar Chinese Spring was selected as well . Two additional “unknown” lines were selected for screening from a set of wheat alloplasmic lines developed in Japan . Alloplasmic lines are created by substitution backcrossing to replace the cytoplasmic genomes of one species with those of another while maintaining the original nuclear genome background, and have shown promise for improving stress tolerance and other developmental traits . The two alloplasmic lines selected were Aegilops columnaris KU11-2 [abbreviated co hereafter] and Ae. speltoides aucheri KU2201B [abbreviated sp hereafter] with the cytoplasmic genome type preceding the nuclear genome background, which in this case is Chinese Spring . Screening was performed in a hydroponic system in a Conviron growth chamber to ensure uniform conditions. Hydroponic systems are commonly used to screen plants for salt tolerance, including wheat. In all experiments, growth conditions in the Conviron were set at 22◦C during light conditions and 18◦C during the dark, 16 h photoperiod, 375 µmol m−2 s −1 light intensity, and 50% relative humidity. Three hydroponic tanks were used per treatment . Each hydroponic tank contained a grid of 16 Cone-tainers filled with perlite. Within each tank, there were four genotypes each with four individual replicates . For each treatment , there were three replicate tanks; hence, there were a total of 48 Cone-tainers for each treatment.