Community structure of fungal pathogens causing spikelet rot disease of naked oat from different ecological regions of China

Spikelet rot disease (SRD) is an emerging disease of the grain surface of naked oat in China that affects both grain yield and quality. The typical symptom is discoloration from the black structures of the causal fungi. Here, we investigated the fungal communities on the grain surfaces of cultivar Bayou 13 grown in ten ecological oat-producing regions of China, to identify the main pathogens of naked oat SRD. Our results showed that the growth of Alternaria spp. and Davidiella spp. exhibited a competitive relationship and was mainly affected by the elevations of all 10 ecological regions. The dominant pathogens were Davidiella spp. in Shannan Prefecture in Tibet and Haidong Prefecture in Qinghai Province and Alternaria spp. in the other eight regions. The ratios of black pathogens of interest to all pathogens in Shannan Prefecture and Haidong Prefecture were significantly lower than those of the other eight regions, thus indicating that SRD mainly occurred in regions below 2000 m (elevation). We isolated black fungal pathogens from grain surfaces and deduced that they were Alternaria spp. by sequence comparison. The blackened appearance of the grain surfaces was more evident under spray inoculation with a spore suspension of Alternaria than under the control in greenhouse experiments. The recovered pathogen was the same as the pathogen used for inoculation. We thus concluded that Alternaria alone causes naked oat SRD and mainly infects naked oat in regions below 2000 m, which provides a basis for the recognition and management of SRD of naked oat.

. We isolated black fungal pathogens from grain surfaces and deduced that they were Alternaria spp. by sequence comparison. The blackened appearance of the grain surfaces was more evident under spray inoculation with a spore suspension of Alternaria than under the control in greenhouse experiments. The recovered pathogen was the same as the pathogen used for inoculation. We thus concluded that Alternaria alone causes naked oat SRD and mainly infects naked oat in regions below 2000 m, which provides a basis for the recognition and management of SRD of naked oat.
Oat (Avena sativa) is one of the most important cereal crops in the world and is cultured in all agricultural regions of the world with a moderate climate 1 . The hexaploid cultivated species A. sativa is represented by both hulled (A. sativa subsp. sativa L.) and naked forms (A. sativa subsp. nudisativa (Husnot.) Rod. et Sold.) 2 , which are believed to have originated in China 3 . The yield and composition of nutritional and functional active ingredients were higher in naked oat than in hulled oat 4,5 . Naked oat is the main cultivated oat species in China, accounting for 92% of the total area planted in oat (106.7 hm 2 of 120.0 hm 2 ) 6 . Naked oat is mainly distributed in arid and cold regions, including Inner Mongolia, Yunnan Province, Gansu Province, Sichuan Province, Hebei Province, Shanxi Province, Qinghai Province, Jilin Province, and Tibet 7 . In these regions, rainfall is scarce and mostly concentrated from July to September. The occurrence of continuous rainfall during the harvest season leads to high moisture and warm temperatures being important factors in the spread of plant diseases 8 . In this period, many fungi on glumes and panicles of weakened plants or moribund tissues of cereal crops result in grain discoloration 9 . Extensive occurrence of spikelet rot disease (SRD, Supplementary Figure S1) on oat grains in fields is a result of rainfall just before harvest and is a regularly observed phenomenon 10 .
SRD of naked oat is characterized by black fungal discoloration, especially on the surface of the kernels as well as beneath the pericarp of the grain ears 11 . This discoloration is mainly considered a blemish of the grain but not a factor affecting quality 10 . This disease has generally received comparatively little attention and is considered a minor disease that does not warrant specific control measures 10 when compared with oat diseases such as rusts, blights, smuts, spots, and culm and root rots 1,[12][13][14] . Recently, the most general usage of naked oat has been for livestock feed, but oat consumption as a human food has recently increased, perhaps due to the reported health benefits arising from the nutritional value of the grains 14 . However, fungal infection of grains in the field may result in yield nutritive value reduction and the production of toxins that are harmful to humans and animals 10,15,16 . Thus, a high grain quality is of great significance to the safe production of naked oat products.

Results
Metagenomic sequence characteristics of fungal pathogens on grains from different ecological regions. Metagenomic sequencing of the fungal communities on the grain surfaces of Bayou 13 obtained from 10 ecological regions during the harvest season yielded 1,458,785 valid sequences, with 99.29% (1,448,477) being high-quality sequences. The proportion of high-quality sequences was more than 97% for all regions. The samples from Shannan Prefecture produced the fewest valid and high-quality sequences (114,171 and 113,867, respectively). The samples from Dingxi City in Gansu yielded the most valid and high-quality sequences (200,080 and 199,194, respectively) ( Species diversity and richness of fungal pathogens on grains from different ecological regions. In total, 1880 OTUs were produced from the sequencing of fungal pathogens on grains from 10 regions. When the calculated diversity indices were compared, the Chao1 and ACE indices were similar to the OTUs. The highest Chao1 and ACE index values were observed in samples from Dingxi City, and the lowest Chao1 and ACE index values were observed in samples from Datong City. The highest values for the Shannon indices were observed in samples from Baicheng City, and the lowest values were observed in samples from Shannan Prefecture, Tibet (Table 2).
In total, 69 genera were detected on grain surface samples, with the dominant pathogens belonging to the genera Alternaria Hierarchical cluster analysis at the genus level revealed that 13 genera, including Alternaria, Pyrenophora, and Gibberella, belonged to one cluster, and 13 others, including Davidiella, Botrytis, and Bipolaris, belonged to www.nature.com/scientificreports/ another cluster. When region-based clustering was performed, it was found that Haidong Prefecture formed a cluster of its own, and all other regions formed another cluster (Fig. 2). Based on the richness of each species in the various samples, the correlations among species were calculated using the richness information at the genus level for the top 50 species. A correlation network consisting of significantly correlated species was constructed for the prediction of interactions among species. From Fig. 3, it can be seen that, with the exception of one negative correlation (i.e., competitive relationship), all other correlations between Davidiella and Alternaria were positive (i.e., collaborative relationships).

Culture of fungal pathogens from grains from different ecological regions.
Based on the information obtained on the fungal pathogen community (Supplementary Table S2, Supplementary Figure S2), the infestation ratio of all pathogens on the grains from Haidong Prefecture in Qinghai Province was significantly higher than that from Shannan Prefecture in Tibet and significantly lower than that in the other regions (F = 64.96, df = 29.9, P < 0.05) (Fig. 4). Moreover, the infestation ratios of black pathogens from Shannan Prefecture in Tibet and Shannan and Haidong Prefectures were significantly lower than those from the other regions (F = 49.557, df = 29.9, P < 0.05) (Fig. 4). www.nature.com/scientificreports/ Verification of SRD-causing fungal pathogens on grain surfaces. The black pathogen strains purified on the grain surfaces from the 10 ecological regions were identified as Alternaria spp. through sequencing analysis and morphological identification (Fig. 5). After the granules and glumes of the three groups had been cultured on PDA medium for 5 days, greenhouse experiments were performed, and the infestation ratios of granules (F = 142.855, df = 8.6, P < 0.05) and glumes (F = 49.258, df = 8.6, P < 0.05) in the control group were significantly higher than those in the normal growth group (untreated) and lower than those in the experimental group (Fig. 6). The black pathogen strains were again isolated from the granules and glumes of the experimental group. The isolates were observed to be consistent with the original isolates, thus fulfilling Koch's postulates.

Discussion
Through the investigation of the community structure of fungal pathogens on the surface of Bayou 13 oat grains from 10 ecological regions, our results showed that the fungal pathogens mainly belonged to 69 genera of Ascomycota and Basidiomycota, with the dominant pathogens belonging to Alternaria spp. and Davidiella spp. The same results were found in Lolium perenne and Trifolium repens 32 . These two fungal genera are widely distributed in different proportions in plants, aquatic ecosystems and soils [33][34][35] . Ascomycetes are more easily amplified than Basidiomycota with the ITS2 subsite 36 , while Davidiella was the dominant genus amplified with the ITS2 region 37 . Furthermore, we observed that these two genera exhibited a competitive relationship. A negative interaction www.nature.com/scientificreports/   www.nature.com/scientificreports/ between Fusarium spp. and Alternaria spp. was found on the surface of wheat and barley grains of reduced quality 10,22 . Fungal distribution is mainly affected by fungal interactions such as spatial competition 38 . This may explain why, except for Shannan Prefecture and Haidong Prefecture, the proportion of Ascomycota in the other eight regions was significantly higher than that of Basidiomycota, and Alternaria spp. were the dominant fungi in the eight regions. In addition to being proven with plate culture, fungal cooccurrence has been investigated by metagenomic approaches and network analysis 39,40 . We also found that Alternaria and Davidiella spp. were negatively and positively correlated with elevation, respectively. Elevational gradients exert a strong influence on the relationships among crops and their microbiota in alpine regions 41,42 . This result thus indicated that elevation mainly affected the relative occurrence of Alternaria and Davidiella spp., presenting an inverse relationship with blackened surfaces of Chinese naked oat grains. On the other hand, the infestation ratios on grains from Shannan Prefecture and Haidong Prefecture were significantly lower than those from the other eight regions. According to the Chinese Three Gradient Terrains 43 , Shannan Prefecture and Haidong Prefecture belong to the first terrain, Kelan, Dingxi, Datong, Jining, and Zhangjiakou belong to the second group, and Chifeng, Hinggan League, and Baicheng belong to the third group. Clearly, visible differences in black pigmentation were also found among the surface fungi of naked oat grains between the first terrain and the other terrains. This result thus indicated that SRD of naked oat characterized by black fungal discoloration mainly occurred in regions below 2000 m (elevation). Our previous reports showed that for grains of naked oat planted in Kelan, the infection level of Alternaria spp. on the black surface (38.7%) was five times as much as that on the normal surface (7.6%), while the infection level of Davidiella spp. on the black surface (3.9%) was similar to that on the normal surface (2.39%) 27 . Alternaria are dematiaceous fungi characterized by dark colonies ranging from gray to olive/brown 44 . Taking these results together, the black pathogen(s) causing SRD in eight of the regions examined may be Alternaria spp.
Further experimental results showed that the black pathogens cultured on PDA were Alternaria spp., which was consistent with the sequencing results; additionally, the greenhouse experiment results confirmed that Alternaria spp. was the cause of SRD of naked oat. Naked oat SRD differs from rice SRD 17,18 in that it is associated with one pathogen. Oat leaf spot disease is only caused by Alternaria alternata [45][46][47] . Alternaria is one of the main mycotoxigenic fungal genera found in cereals worldwide 48 , causing diseases in over 400 host plants and postharvest spoilage of several crops 49,50 . The prevalence of this fungus in cereals indicates a high disease incidence, with more than 90% of the grains affected in the field 51 and more than 50% of the grains affected in the greenhouse experiment in our study. Reports of cereal diseases characterized by leaf blackening and blight associated with Alternaria are continually published 16 , and cereal grains are constantly affected by Alternaria spp. and their toxins 51 . Although Alternaria toxins are often neglected in grains, possibly due to the lack of severe economic losses directly caused by the genus, recent studies have shown that their frequency and ability to produce a wide range of toxins is significant 10,16 . With governments attaching great importance to the sustainable development of the environment and agriculture and people's concern for health, research and application of biological control are increasing. Microbial agents, botanical agents and resistance inducers are receiving increasing attention regarding their roles in the biocontrol of plant diseases caused by Alternaria 52 . Oat grains can be processed into many kinds of foods 53 . Thus, the finding that SRD of naked oat is caused by only Alternaria spp. presents an obvious advantage for the biological control of this disease.

Conclusions
SRD has become an emerging naked oat disease in China, as losses caused by the disease have been regularly increasing over recent years. The pathogens of SRD have been investigated in many regions of Tibet, Shanxi Province, Gansu Province, Inner Mongolia, Qinghai Province and Hebei Province, China. Using metagenomic analysis, molecular techniques and field validation, one causal agent was identified as Alternaria spp., and this disease mainly occurred at altitudes below 2000 m. Our results can help improve the recognition, diagnosis, and management of this important disease. It is emphasized in this study that this disease is caused by a single pathogenic taxon, and further research into the application of antagonistic microbes against Alternaria spp. to prevent the occurrence of pre-and postharvest SRD of naked oat are required to ensure the production of certified oat materials.  Supplementary Table S1. In Datong City, Dingxi City, Baicheng City, Chifeng City, and Hinggan League, Bayou 13 plants were sown in early April and harvested in early August. In Kelan County, Zhangjiakou City, Shannan Prefecture, Haidong Prefecture, and Jining City, Bayou 13 plants were sown in mid-late May and harvested in mid-September. In 2017, the seeds were sown in drills 3-5 cm deep, spaced ca. 25 cm apart and at an average rate of 450 seeds m -2 in these regions. Irrigation and fertilization were scheduled according to the oat requirements, soil storage capacity, and climate. The cropping system followed the guidelines of the National Oat and Buckwheat Industrial Technology System 6 . For each of the ten sites, we randomly selected three sampling plots; the plot dimensions were 10 m × 5.0 m, and each plot contained 20 rows. At the ripening stage of the Bayou 13 plants, 500 g of grains per plot for each region was collected by the Center for Agricultural Genetic Resources Research unit and stored in a polyethylene bag at − 80 °C. www.nature.com/scientificreports/ Collection of fungal pathogens on the grain surfaces and DNA extraction. The surfaces of oat grain samples were washed three times with phosphate-buffered saline (PBS), and the resultant pathogen-containing solutions were centrifuged at 15,000×g for 30 s. After the supernatant was discarded, total pathogen genomic DNA samples were extracted using Fast DNA SPIN extraction kits (MP Biomedicals, Santa Ana, CA, USA) following the manufacturer's instructions and stored at − 20 °C prior to further analysis. The quantity and quality of extracted DNA were measured using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and agarose gel electrophoresis, respectively. Sequence analysis. The Quantitative Insights Into Microbial Ecology (QIIME, v1.8.0) bioinformatic pipeline was employed to process the sequencing data, as previously described 54 . Briefly, raw sequencing reads with exact matches to the barcodes were assigned to respective samples and identified as valid sequences. The lowquality sequences were filtered with previously reported criteria 55,56 . Paired-end reads were assembled using FLASH 57 . After chimera detection, the remaining high-quality sequences were clustered into operational taxonomic units (OTUs) at 97% sequence identity by UCLUST 58 . A representative sequence was selected from each OTU using default parameters. OTUs were taxonomically classified by BLAST, and the representative sequence set was searched against the Unite Database 59 using the best hits 60 . An OTU table was also generated to record the abundance and taxonomy of each OTU in each sample. OTUs containing less than 0.001% of total sequences across all samples were discarded. To minimize the difference in sequencing depth across samples, an averaged, rounded rarefied OTU table was generated by averaging 100 evenly resampled OTU subsets under 90% of the minimum sequencing depth for further analysis.

Materials and methods
Bioinformatics and statistical analysis. Sequence data analyses were mainly performed using QIIME Culture of fungal pathogens from surfaces of oat grains from different ecological regions. From the samples of each region, 30 oat grains from the initial 500 g of grains were randomly selected. A conventional tissue separation method was used for isolating and culturing samples 64 . After the germ had been removed from each grain using a sterilized blade, the grains were disinfected in 75% ethanol for 30 s, washed three times with sterile water, placed on sterile filter paper to absorb the excess water on the grain surfaces, and evenly arranged using sterile forceps on potato dextrose agar (PDA) medium (potato [Kunming, China], 200 g l −1 ; glucose [Solarbio, Beijing, China], 20 g l −1 ; agar, 15 g l −1 ). For each ecological region, three PDA plates with 10 grains each were prepared, i.e., three parallel experiments were conducted. The loaded PDA plates were incubated in an upright position at 25 °C without light for 5 days 65 . The number of black grains was counted, and the morphology of the pathogens on the surfaces of naked oat grains was preliminarily observed using an Olympus BX53 microscope (Olympus, Japan).

Identification of SRD-causing fungal pathogens on the surfaces of naked oat grains. Accord-
ing to the single spore isolation method 66 , single spores derived from all colonies around each grain from each region were selected from the culture medium. www.nature.com/scientificreports/ Each group was at least five meters apart. The preserved Alternaria was cultured on PDA at 25 °C under a 12:12 LD photoperiod for 7 days. Subsequently, suspensions of conidia were prepared using sterile water containing 0.1% Tween-20, and the conidial suspensions were passed through two layers of a sterile cheesecloth to remove hyphal fragments 68 . Spore concentration was calculated using a hemocytometer and adjusted to 1 × 10 6 spores/ ml. During the maturation period, the experimental group was uniformly sprayed with the spore suspension, the control group was sprayed with sterile water containing 0.1% Tween-20, and the normal growth group was left untreated. Using a hand sprayer, 50 ml of spore suspension or 0.1% Tween-20 solution for each batch of six pots was directly sprayed on the rachises; each rachis was then put in a transparent plastic bag to prevent drying 69,70 . Spraying was performed once every morning, noon, and night for 2 weeks during the maturity stage. From each plot, 50 granules and 50 glumes were randomly selected, surface disinfected using 75% ethanol for 30 s, washed three times with sterile water, evenly arranged on CMA medium, and incubated at 25 °C. After 5 days, the number of black granules and glumes was recorded. The pathogens were isolated from the black areas of grains and glumes. The colony and conidial characteristics of the recovered pathogen were observed using an Olympus BX53 microscope (Olympus, Japan) and compared with those of the pathogen used for inoculation.
Statistical analysis. The infestation ratio was calculated using the following equation: Data collected from the infestation ratio of grains, granules, or glumes were used for one-way analysis of variance (Fisher's protected least significant difference), and normalized data were transformed using an arcsine square root function before analysis. Among ecological indices such as elevation, sunshine hours, temperature, rainfall, latitude and longitude from Supplementary Table S1, two models of key indices and two genera, Alternaria and Davidiella spp. were generated using stepwise backward selection 71 . All analyses were conducted in SPSS for Windows Version 16.0 (SPSS Inc., Chicago, Illinois, United States of America). www.nature.com/scientificreports/