Enhancing Green Fraction Estimation in Rice and Wheat Crops: A Self-Supervised Deep Learning Semantic Segmentation Approach.

The green fraction (GF), which is the fraction of green vegetation in a given viewing direction, is closely related to the light interception ability of the crop canopy. Monitoring the dynamics of GF is therefore of great interest for breeders to identify genotypes with high radiation use efficiency. The accuracy of GF estimation depends heavily on the quality of the segmentation dataset and the accuracy of the image segmentation method. To enhance segmentation accuracy while reducing annotation costs, we developed a self-supervised strategy for deep learning semantic segmentation of rice and wheat field images with very contrasting field backgrounds. First, the Digital Plant Phenotyping Platform was used to generate large, perfectly labeled simulated field images for wheat and rice crops, considering diverse canopy structures and a wide range of environmental conditions (sim dataset). We then used the domain adaptation model cycle-consistent generative adversarial network (CycleGAN) to bridge the reality gap between the simulated and real images (real dataset), producing simulation-to-reality images (sim2real dataset). Finally, 3 different semantic segmentation models (U-Net, DeepLabV3+, and SegFormer) were trained using 3 datasets (real, sim, and sim2real datasets). The performance of the 9 training strategies was assessed using real images captured from various sites. The results showed that SegFormer trained using the sim2real dataset achieved the best segmentation performance for both rice and wheat crops (rice: Accuracy = 0.940, F1-score = 0.937; wheat: Accuracy = 0.952, F1-score = 0.935). Likewise, favorable GF estimation results were obtained using the above strategy (rice: R2 = 0.967, RMSE = 0.048; wheat: R2 = 0.984, RMSE = 0.028). Compared with SegFormer trained using a real dataset, the optimal strategy demonstrated greater superiority for wheat images than for rice images. This discrepancy can be partially attributed to the differences in the backgrounds of the rice and wheat fields. The uncertainty analysis indicated that our strategy could be disrupted by the inhomogeneity of pixel brightness and the presence of senescent elements in the images. In summary, our self-supervised strategy addresses the issues of high cost and uncertain annotation accuracy during dataset creation, ultimately enhancing GF estimation accuracy for rice and wheat field images. The best weights we trained in wheat and rice are available: https://github.com/PheniX-Lab/sim2real-seg.

Self-Supervised Plant Phenotyping by Combining Domain Adaptation with 3D Plant Model Simulations: Application to Wheat Leaf Counting at Seedling Stage.

The number of leaves at a given time is important to characterize plant growth and development. In this work, we developed a high-throughput method to count the number of leaves by detecting leaf tips in RGB images. The digital plant phenotyping platform was used to simulate a large and diverse dataset of RGB images and corresponding leaf tip labels of wheat plants at seedling stages (150,000 images with over 2 million labels). The realism of the images was then improved using domain adaptation methods before training deep learning models. The results demonstrate the efficiency of the proposed method evaluated on a diverse test dataset, collecting measurements from 5 countries obtained under different environments, growth stages, and lighting conditions with different cameras (450 images with over 2,162 labels). Among the 6 combinations of deep learning models and domain adaptation techniques, the Faster-RCNN model with cycle-consistent generative adversarial network adaptation technique provided the best performance (R2 = 0.94, root mean square error = 8.7). Complementary studies show that it is essential to simulate images with sufficient realism (background, leaf texture, and lighting conditions) before applying domain adaptation techniques. Furthermore, the spatial resolution should be better than 0.6 mm per pixel to identify leaf tips. The method is claimed to be self-supervised since no manual labeling is required for model training. The self-supervised phenotyping approach developed here offers great potential for addressing a wide range of plant phenotyping problems. The trained networks are available at https://github.com/YinglunLi/Wheat-leaf-tip-detection.

Kurthia huakuii sp. nov., isolated from biogas slurry, and emended description of the genus Kurthia.

A novel facultatively anaerobic bacterium, designated strain LAM0618(T), was isolated from biogas slurry samples collected from the large-scale anaerobic digester of Modern Farming Corporation in Hebei Province, China. Cells of strain LAM0618(T) were Gram-stain-positive, motile, non-spore-forming and short-rod-shaped. The optimal temperature and pH for growth were 30 °C and 7.0, respectively. The strain did not require NaCl for growth but tolerated up to 70 g NaCl l(-1). The major fatty acids of strain LAM0618(T) were iso-C(15 : 0), anteiso-C(15 : 0), iso-C(14 : 0), C(16 : 0) and C(18 : 0). The predominant menaquinones of strain LAM0618(T) were menaquinone 7 (MK-7) and menaquinone 6 (MK-6). The main polar lipids of strain LAM0618(T) were phosphatidylglycerol, diphosphatidylglycerol, phosphatidylethanolamine and six unknown glycolipids. The genomic DNA G+C content was 41 mol% as determined by the Tm method. Analysis of the 16S rRNA gene sequence revealed that strain LAM0618(T) was a member of the genus Kurthia, and was most closely related to 'Kurthia massiliensis' DSM 24639, Kurthia zopfii DSM 20580(T), Kurthia gibsonii DSM 20636(T) and Kurthia sibirica DSM 4747(T), with 96.9, 95.7, 95.6 and 94.9 % sequence similarity, respectively. Based on its phenotypic and genotypic properties, strain LAM0618(T) is suggested to represent a novel species of the genus Kurthia, for which the name Kurthia huakuii sp. nov. is proposed. The type strain is LAM0618(T) ( = ACCC 06121(T) = JCM 19187(T)).

Rhizobium petrolearium sp. nov., isolated from oil-contaminated soil.

Two Gram-negative, aerobic, rod-shaped bacteria, designated strains SL-1(T) and F11, which had the ability to decompose polycyclic aromatic hydrocarbons (PAHs), were isolated from soil samples contaminated by oil. The cells were motile by polar or lateral flagella. According to comparison of 16S rRNA gene sequences, strains SL-1(T) and F11 were identical and showed the greatest degree of similarity (96.8%) to both Rhizobium oryzae Alt505(T) and Rhizobium mesosinicum CCBAU 25010(T); however, only Rhizobium oryzae with SL-1(T) and F11 formed a separate clade. There were low similarities (<90%) between the atpD and recA sequences of the two strains and those of the genus of Rhizobium. The bacteria grew at temperatures of 10-40 °C with an optimum of 30 °C. The pH range for growth was 6.0-10.0 and optimum pH was 7.0-8.0. Growth occurred at NaCl concentrations up to 3.0% (w/v). They were catalase- and oxidase-positive. The main cellular fatty acids were summed feature 8 (18:1ω7c and/or 18:1ω6c) and 16:0. The DNA G+C content was 62.2 mol%. Strain SL-1(T) showed 29 and 0% DNA-DNA relatedness, respectively, with the most related strains R. oryzae Alt505(T) and R. mesosinicum CCBAU 25010(T) according to phylogenic analysis of the 16S rRNA gene. According to physiological and biochemical characteristics and genotypic data obtained in this work, the bacteria represent a novel species of the genus Rhizobium, and the name Rhizobium petrolearium is proposed. The type strain is SL-1(T) ( = ACCC 11238(T) = KCTC 23288(T)) and it could nodulate Medicago sativa in nodulation tests.

Rhizobium pseudoryzae sp. nov., isolated from the rhizosphere of rice.

A Gram-stain-negative, aerobic, rod-shaped bacterium, designated strain J3-A127(T), was isolated from the roots of fresh rice plants (Oryza sativa). Cells were non-motile and no flagellum was detected. Comparison of 16S rRNA gene sequences indicated that the strain was phylogenetically related to species of the genus Rhizobium, with closest similarity to Rhizobium oryzae Alt 505(T) (96.4 %). The low levels of 16S rRNA gene sequence similarity (<90 %) found between the gyrB, atpD, recA and glnII gene sequences of strain J3-A127(T) and the type strains of recognized species of the genus Rhizobium also indicated that it represented a separate species. The temperature range for growth was 10-40 °C (optimum around 28 °C) and the pH range was 6.0-11.0 (optimum pH 7.0-8.0). Strain J3-A127(T) tolerated NaCl concentrations up to 5.0 % (w/v). The strain was catalase- and oxidase-positive. The main cellular fatty acids were summed feature 8 (C(18 : 1)ω7c and/or C(18 : 1)ω6; 46.7 %). The DNA G+C content of strain J3-A127(T) was 59.5 mol%. Strain J3-A127(T) did not form any nodules on four different legumes and the nodD and nifH genes were not detected by PCR. According to physiological and biochemical characteristics and genotypic data, strain J3-A127(T) is considered to represent a novel species of the genus Rhizobium, for which the name Rhizobium pseudoryzae sp. nov. is proposed. The type strain is J3-A127(T) ( = ACCC 10380(T) = KCTC 23294(T)).

Development and characterization of simple sequence repeat (SSR) markers for the mushroom Flammulina velutipes.

Ten polymorphic SSR markers of Flammulina velutipes were developed and characterized with FIASCO methods. The polymorphism information content (PIC) ranged from 0.13 to 0.69. This is the first report on development of SSR markers in F. velutipe.

Flavobacterium haoranii sp. nov., a cypermethrin-degrading bacterium isolated from a wastewater treatment system.

A Gram-negative, non-spore-forming, yellow-pigmented bacterium, strain LQY-7(T), was isolated from activated sludge treating synthetic pyrethroid-manufacturing wastewater. The taxonomic status of the strain was determined using a polyphasic taxonomic approach. Phylogenetic analysis based on 16S rRNA gene sequences revealed that strain LQY-7(T) was a member of the genus Flavobacterium but had low similarities with other species of this genus (95.0 % similarity with Flavobacterium indicum GPTSA100-9(T) and <94 % similarities with other Flavobacterium species). On the basis of phenotypic, genetic and phylogenetic data, strain LQY-7(T) should be classified as a representative of a novel species of the genus Flavobacterium, for which the name Flavobacterium haoranii sp. nov. is proposed; the type strain is LQY-7(T) (=ACCC 05409(T) =KCTC 23008(T)).

Rheinheimera tangshanensis sp. nov., a rice root-associated bacterium.

A Gram-negative, aerobic, rod-shaped bacterium, designated strain JA3-B52(T), was isolated from the roots of fresh rice plants (Oryza sativa). The cells were motile by means of polar single or lateral flagella. The colonies were non-pigmented. On the basis of 16S rRNA gene sequence comparisons, the strain was phylogenetically related to species of the genus Rheinheimera, having the greatest level of sequence similarity with respect to Rheinheimera texasensis A62-14B(T) (97.16 %). The bacterium grew at temperatures from 10 to 37 degrees C, with an optimum at 30 degrees C. The strain exhibited growth with 0-3.0 % (w/v) NaCl and at pH 6.0-8.5. The main cellular fatty acids were C(16 : 1)omega7c, C(17 : 1)omega8c, C (16 : 0), C(18 : 1)omega7c and C(12 : 0) 3-OH. The DNA G+C content was 47.0 mol%. The levels of similarity between the 16S rRNA gene sequence of strain JA3-B52(T) and those of the type strains of Rheinheimera species ranged from 95.38 to 97.16 %. The mean level of DNA-DNA relatedness between strain JA3-B52(T) and R. texasensis A62-14B(T), the strain most closely related to the isolate, was 20.4 %. On the basis of physiological and biochemical characteristics and genotypic data obtained in this work, strain JA3-B52(T) represents a novel species of the genus Rheinheimera, for which the name Rheinheimera tangshanensis sp. nov. is proposed. The type strain is JA3-B52(T) (=CGMCC 1.6362(T) =DSM 19460(T)).

Strain-typing of Lentinula edodes in China with inter simple sequence repeat markers.

To validate strain typing by inter simple sequence repeat (ISSR) analysis in Lentinula edodes cultivars, 17 Chinese L. edodes strains including 15 cultivated strains cultivated on a large scale and two wild strains were analyzed with the ISSR technique. With the use of two ISSR primers, a total of 32 DNA products were detected, of which, 31 DNA products (96.9% of the detected products) were polymorphic between two or more strains. The profiles of those two primers could be employed to differentiate all of the tested strains. A cluster analysis based on ISSR data revealed that the 17 strains could be classified into two distinct groups. One group consisted of eight strains in which the cultivated strains were H (high-temperature)-type or B (broad-temperature)-type, and the other group comprised cultivated strains that were of the L (low-temperature)-type or M (medium-temperature)-type. In contrast to the two wild strains, the genetic diversity of 15 cultivated strains was very rich based on a similarity coefficient analysis.