Using Deep Learning for Image-Based Potato Tuber Disease Detection.

Many plant diseases have distinct visual symptoms, which can be used to identify and classify them correctly. This article presents a potato disease classification algorithm that leverages these distinct appearances and advances in computer vision made possible by deep learning. The algorithm uses a deep convolutional neural network, training it to classify the tubers into five classes: namely, four disease classes and a healthy potato class. The database of images used in this study, containing potato tubers of different cultivars, sizes, and diseases, was acquired, classified, and labeled manually by experts. The models were trained over different train-test splits to better understand the amount of image data needed to apply deep learning for such classification tasks. The models were tested over a data set of images taken using standard low-cost RGB (red, green, and blue) sensors and were tagged by experts, demonstrating high classification accuracy. This is the first article to report the successful implementation of deep convolutional networks, popular in object identification, to the task of disease identification in potato tubers, showing the potential of deep learning techniques in agricultural tasks.

Dickeya solani sp. nov., a pectinolytic plant-pathogenic bacterium isolated from potato (Solanum tuberosum).

Pectinolytic bacteria have been recently isolated from diseased potato plants exhibiting blackleg and slow wilt symptoms found in a number of European countries and Israel. These Gram-reaction-negative, motile, rods were identified as belonging to the genus Dickeya, previously the Pectobacterium chrysanthemi complex (Erwinia chrysanthemi), on the basis of production of a PCR product with the pelADE primers, 16S rRNA gene sequence analysis, fatty acid methyl esterase analysis, the production of phosphatases and the ability to produce indole and acids from α-methylglucoside. Differential physiological assays used previously to differentiate between strains of E. chrysanthemi, showed that these isolates belonged to biovar 3. Eight of the isolates, seven from potato and one from hyacinth, were analysed together with 21 reference strains representing all currently recognized taxa within the genus Dickeya. The novel isolates formed a distinct genetic clade in multilocus sequence analysis (MLSA) using concatenated sequences of the intergenic spacer (IGS), as well as dnaX, recA, dnaN, fusA, gapA, purA, rplB, rpoS and gyrA. Characterization by whole-cell MALDI-TOF mass spectrometry, pulsed field gel electrophoresis after digestion of whole-genome DNA with rare-cutting restriction enzymes, average nucleotide identity analysis and DNA-DNA hybridization studies, showed that although related to Dickeya dadantii, these isolates represent a novel species within the genus Dickeya, for which the name Dickeya solani sp. nov. (type strain IPO 2222(T) = LMG25993(T) = NCPPB4479(T)) is proposed.

The role of onion-associated fungi in bulb mite infestation and damage to onion seedlings.

In Israel Rhizoglyphus robini is considered to be a pest in its own right, even though the mite is usually found in association with fungal pathogens. Plant protection recommendations are therefore to treat germinating onions seedlings, clearly a crucial phase in crop production, when mites are discovered. The aim of this study was to determine the role of fungi in bulb mite infestation and damage to germinating onion seedlings. Accordingly we (1) evaluated the effect of the mite on onion seedling germination and survival without fungi, (2) compared the attraction of the mite to species and isolates of various fungi, (3) assessed the effect of a relatively non-pathogenic isolate of Fusarium oxysporum on mite fecundity, and (4) determined the effects of the mite and of F. oxysporum separately and together, on onion seedling germination and sprout development. A significant reduction of seedling survival was recorded only in the 1,000 mites/pot treatment, after 4 weeks. Mites were attracted to 6 out of 7 collected fungi isolates. Mite fecundity on onion sprouts infested with F. oxysporum was higher than on non-infested sprouts. Survival of seedlings was affected by mites, fungi, and their combination. Sprouts on Petri dishes after 5 days were significantly longer in the control and mite treatments than both fungi treatments. During the 5-day experiment more mites were always found on the fungi-infected sprouts than on the non-infected sprouts. Future research using suppressive soils to suppress soil pathogens and subsequent mite damage is proposed.

Whole-cell bacterial biosensor for volatile detection from Pectobacterium-infected potatoes enables early identification of potato tuber soft rot disease.

Half of the harvested food is lost due to rots caused by microorganisms. Plants emit various volatile organic compounds (VOCs) into their surrounding environment, and the VOC profiles of healthy crops are altered upon infection. In this study, a whole-cell bacterial biosensor was used for the early identification of potato tuber soft rot disease caused by the pectinolytic bacteria Pectobacterium in potato tubers. The detection is based on monitoring the luminescent responses of the bacteria panel to changes in the VOC profile following inoculation. First, gas chromatography-mass spectrometry (GC-MS) was used to specify the differences between the VOC patterns of the inoculated and non-inoculated potato tubers during early infection. Five VOCs were identified, 1-octanol, phenylethyl alcohol, 2-ethyl hexanol, nonanal, and 1-octen-3-ol. Then, the infection was detected by the bioreporter bacterial panel, firstly measured in a 96-well plate in solution, and then also tested in potato plugs and validated in whole tubers. Examination of the bacterial panel responses showed an extensive cytotoxic effect over the testing period, as seen by the elevated induction factor (IF) values in the bacterial strain TV1061 after exposure to both potato plugs and whole tubers. Moreover, quorum sensing influences were also observed by the elevated IF values in the bacterial strain K802NR. The developed whole-cell biosensor system based on bacterial detection will allow more efficient crop management during postharvest, storage, and transport of crops, to reduce food losses.

Monitoring of infection volatile markers using CMOS-based luminescent bioreporters.

Over the past two decades, whole-cell biosensors (WCBs) have been widely used in the environmental field, with only few applications proposed for use in agricultural. This study describes the development and optimization of a WCB for the detection of volatile organic compounds (VOCs) that is produced specifically by infected potato tubers. First, the effect of calcium-alginate matrix formation (beads vs. tablets) on the membrane uniformity and sensing efficiency was evaluated. Then, important parameters in the immobilization process were examined for their effect on the sensitivity to the presence of VOCs. The highest sensitivity to the target VOC was obtained by 20 min polymerization of bacterial suspension with optical density of 0.2 at 600 nm, dissolved in low-viscosity sodium alginate (1.5% w/v) and exposure to VOC at 4 °C. After optimization, the lowest limit of detection for three infection-sourced VOCs (nonanal, 3-methyl-1-butanol, and 1-octen-3-ol) was 0.17-, 2.03-, and 2.09-mg/L, respectively, and the sensor sensitivity was improved by 8.9-, 3.1- and 2-fold, respectively. Then, the new optimized immobilization protocol was implemented for the CMOS-based application, which increased the sensor sensitivity to VOC by 3-fold during real-time measurement. This is the first step in creating a sensor for real-time monitoring of crop quality by identifying changes in VOC patterns.

Differentiation of Pectobacterium and Dickeya spp. phytopathogens using infrared spectroscopy and machine learning analysis.

Pectobacterium and Dickeya spp. are soft rot Pectobacteriaceae that cause aggressive diseases on agricultural crops leading to substantial economic losses. The accurate, rapid and low-cost detection of these pathogenic bacteria are very important for controlling their spread, reducing the consequent financial loss and for producing uninfected potato seed tubers for future generations. Currently used methods for the identification of these bacterial pathogens at the strain level are based mainly on molecular techniques, which are expensive. We used an alternative method, infrared spectroscopy, to measure 24 strains of five species of Pectobacterium and Dickeya. Measurements were then analyzed using machine learning methods to differentiate among them at the genus, species and strain levels. Our results show that it is possible to differentiate among different bacterial pathogens with a success rate of ~99% at the genus and species levels and with a success rate of over 94% at the strain level.

Identification of fungal phytopathogens using Fourier transform infrared-attenuated total reflection spectroscopy and advanced statistical methods.

The early diagnosis of phytopathogens is of a great importance; it could save large economical losses due to crops damaged by fungal diseases, and prevent unnecessary soil fumigation or the use of fungicides and bactericides and thus prevent considerable environmental pollution. In this study, 18 isolates of three different fungi genera were investigated; six isolates of Colletotrichum coccodes, six isolates of Verticillium dahliae and six isolates of Fusarium oxysporum. Our main goal was to differentiate these fungi samples on the level of isolates, based on their infrared absorption spectra obtained using the Fourier transform infrared-attenuated total reflection (FTIR-ATR) sampling technique. Advanced statistical and mathematical methods: principal component analysis (PCA), linear discriminant analysis (LDA), and k-means were applied to the spectra after manipulation. Our results showed significant spectral differences between the various fungi genera examined. The use of k-means enabled classification between the genera with a 94.5% accuracy, whereas the use of PCA [3 principal components (PCs)] and LDA has achieved a 99.7% success rate. However, on the level of isolates, the best differentiation results were obtained using PCA (9 PCs) and LDA for the lower wavenumber region (800-1775 cm(-1)), with identification success rates of 87%, 85.5%, and 94.5% for Colletotrichum, Fusarium, and Verticillium strains, respectively.