SummaryCYP51 encodes the cytochrome P450 sterol 14a-demethylase, an enzyme essential for sterol biosynthesis and the target of azole fungicides. In Fusarium species, including pathogens of humans and plants, three CYP51 paralogues have been identified with one unique to the genus. Currently, the functions of these three genes and the rationale for their conservation within the genus Fusarium are unknown.Three Fusarium graminearum CYP51s (FgCYP51s) were heterologously expressed in Saccharomyces cerevisiae. Single and double FgCYP51 deletion mutants were generated and the functions of the FgCYP51s were characterized in vitro and in planta.FgCYP51A and FgCYP51B can complement yeast CYP51 function, whereas FgCYP51C cannot. FgCYP51A deletion increases the sensitivity of F. graminearum to the tested azoles. In DFgCYP51B and DFgCYP51BC mutants, ascospore formation is blocked, and eburicol and two additional 14-methylated sterols accumulate. FgCYP51C deletion reduces virulence on host wheat ears.FgCYP51B encodes the enzyme primarily responsible for sterol 14a-demethylation, and plays an essential role in ascospore formation. FgCYP51A encodes an additional sterol 14a-demethylase, induced on ergosterol depletion and responsible for the intrinsic variation in azole sensitivity. FgCYP51C does not encode a sterol 14a-demethylase, but is required for full virulence on host wheat ears. This is the first example of the functional diversification of a fungal CYP51.

Arbuscular mycorrhizal (AM) fungal spore communities and distribution patterns were surveyed in montane scrub grassland, alpine steppe, and alpine meadow sites at altitudes ranging from 3,500 to 5,200 m a.s.l. on the Tibetan Plateau. Thirty-two representative soil samples were collected from the root zone of the dominant and common plant species in late May 2004. Twenty-three AM fungal species representing six genera (Acaulospora, Entrophospora, Glomus, Pacispora, Paraglomus, and Scutellospora) were detected and species richness varied from 5.3 ± 0.8 to 10.5 ± 2.5 per site. Some AM fungal species were restricted to one vegetation type and Glomus mosseae, Glomus intraradices, and Scutellospora calospora were detected in all three vegetation types. Glomus species were found to be the most frequent and abundant in all three vegetation types. Acaulospora occurred mostly in the alpine steppe and alpine meadow. Scutellospora occurred mostly in montane scrub grassland. At the species level, Glomus mosseae was dominant in the montane scrub, Acaulospora laevis and Pacispora scintillans were dominant in the alpine steppe, and Acaulospora laevis, Pacispora scintillans, and Glomus claroideum dominated the alpine meadow. It was evident from the distribution pattern of AM fungi in the different vegetation types that the abundance and diversity of AM fungal species were lowest in the montane scrub grassland than the other two plant communities. Climatic conditions, especially temperatures, and intensity of land use may be the most important factors influencing the AM fungal community.

Defect detection on solid wood surface has two main problems: (1) the real-time performance of the available methods are poor despite good detection accuracy, and (2) the defect extraction process is complicated. Here, we propose a mixed, fully convolutional neural network (Mix-FCN) to detect the location of wood defects and classify the types of defects from the wood surface images automatically. The images were collected first by a data acquisition device developed in our laboratory. We then employed TensorFlow and Python language to construct a VGG16 model. We used two kinds of datasets (dataset1 and dataset2) to maximize the limited, collected data and enable the Mix-FCN to converge rapidly during training. The weights of the filters in front of the Mix-FCN during training were initialized from the trained VGG16 model. The weights of the VGG16 net were learned by dataset1. Our model was trained, validated, and tested by dataset 2. Overall classification accuracy (OCA), pixel accuracy (PA), mean intersection over union, detection rate, missing alarm, false alarm rate, and precision were used to evaluate the network, and the performance was good based on the seven evaluation indicators. We achieved 99.14% OCA and 91.31% PA, and a batch of 50 images required only 0.368 s of detection time. Our proposed method has better accuracy and less detection time compared to the previous methods of wood detection. INDEX TERMS Deep learning, full convolutional neural network, transfer learning, wood defects detection.

Powdery mildew is a highly destructive winter wheat pathogen in China. Since the causative agent is sensitive to changing weather conditions, we analyzed climatic records from regions with previous wheat powdery mildew epidemics (1970 to 2012) and investigated the long-term effects of climate change on the percent acreage (PA) of the disease. Then, using PA and the pathogen’s temperature requirements, we constructed a multiregression model to predict changes in epidemics during the 2020s, 2050s, and 2080s under representative concentration pathways RCP2.6, RCP4.5, and RCP8.5. Mean monthly air temperature increased from 1970 to 2012, whereas hours of sunshine and relative humidity decreased (P < 0.001). Year-to-year temperature changes were negatively associated with those of PA during oversummering and late spring periods of disease epidemics, whereas positive relationships were noted for other periods, and year-to-year changes in relative humidity were correlated with PA changes in the early spring period of disease epidemics (P < 0.001). Our models also predicted that PA would increase less under RCP2.6 (14.43%) than under RCP4.5 (14.51%) by the 2020s but would be higher by the 2050s and 2080s and would increase least under RCP8.5 (14.37% by the 2020s). Powdery mildew will, thus, pose an even greater threat to China’s winter wheat production in the future.