Abstract
Background
The ecological phenomenon of fungal fairy rings is usually found in grasslands and caused by the growth of specific fairy ring fungi in soil. The fairy rings are classified into three zones (DARK, DEAD, and OUT), and they have the potential to increase crop yield. Among these fairy rings, distinct characteristics of type I fairy rings can be seen in the rings formed by Leucocalocybe mongolica (LM). Our studies addressed changes in the soil microbial structure due to LM fairy rings to enhance understand of this ecological phenomenon.
Methods
In the present study, we report the soil microbial analysis results (fungi and bacteria), including those of metabarcoding (16s rRNA, ITS), microbial quantity, and metagenomics surveys of soils collected from various fairy ring zones, of 6 LM fairy rings. All sampling sites cover the grasslands of Mongolian Plateau in China.
Results
First, we found through metabarcoding surveys that the difference in microbial diversity is relatively less in bacteria and that the abundance of fairy ring fungi (LM) is relatively high in DEAD zones. We also identified eight bacterial and fungal families, including Sphingobacteriaceae and Sphingomonadaceae that were enriched within the soils of fairy ring zones. Second, we found that the abundance of soil bacteria in the DEAD zones is sharply increased along with the growth of fairy ring fungi (LM). Third, we found through shotgun sequencing that fairy ring-infected zones, DARK and DEAD, exhibit greater genetic diversity than OUT zones. Finally, we showed that the fairy ring ecosystem is the center for a rich grassland microbial community.
Conclusions
The reported data can improve our understanding of type I fairy rings and will be further insightful to the research on crop production.
Supplementary Information
The online version contains supplementary material available at 10.1007/s42770-021-00478-3.
Keywords: Metabarcoding survey, Fairy ring, Soil microbial diversity, Habitat survey, Grasslands
Introduction
Soil biodiversity is mostly determined by the richness of soil microorganisms, but most of these microbial taxa remain unidentified, which necessitates cultivation and edaphology studies [1]. Fungal fairy ring (Fig. 1) is a unique type of fungal growth pattern usually found in grasslands and formed by soil-dwelling fungi. Fairy rings have ecological value because of their ability to promote the growth of crops and plants [2, 3]. This ability can enhance the richness of plant–soil species in grassland ecosystems [4, 5]. Recent studies have shown that both beneficial and toxic metabolites generated by the fairy ring fungi might participate in the enrichment process of plant–soil species [6–9].
Fig. 1.
Geographical location and sampling spot of LM fairy rings. a Illustrates the geographical location of the fairy rings. Brown dots in the figure on the right represent the sampling points, and the map we used was obtained from the website of the Ministry of Natural Resources of China (Audit number: GS(2019)1694 (http://bzdt.ch.mnr.gov.cn)). b Illustrates the division of the three fairy ring zones. c Represents the location of the sampling spot for each fairy ring. We used blue, yellow, and red spots to mark the location of the corresponding flags that also represent the sampling spots for the DARK, DEAD, and OUT zones
Fairy rings can be divided into three types (I–III) [10, 11]. Among these types, the type I ring interests most researchers because of the effect of this ring type on plant growth [4, 11–15]. Type I fairy rings have an obvious bare area comprising a dead vegetation zone that we refer to as the DEAD zone, and a green, dark plant-rich zone that we usually refer to as the DARK zone (Fig. 1). Examples of type I rings include those formed by Agaricus arvensis [4]. Type II rings do not have a DEAD zone (e.g., fairy rings formed by Calvatia cyathiformis) [11]. Type III rings have neither DEAD zones nor DARK zones and can only be recognized from a circle outline of mushrooms (e.g., rings formed by Macrolepiota procera) [10].
Leucocalocybe mongolica (LM) is a high-value, type I fairy ring-forming fungi (Fig. 1) that usually produces distinctive fairy rings, suggesting that LM can be used to determine the microbial mechanism underlying fairy ring ecology. Therefore, the soil microbial study of LM fairy rings can help in understanding this ecological phenomenon.
Sequencing technology is considered the most effective approach for studying soil microbial diversity [16]. Because the type I fairy rings are characterized by a representative integrated fairy ring structure (both DARK zone and DEAD zone), the metabarcoding surveys of type I fairy rings have been successfully applied to other type I fairy ring fungi [4, 17]. However, the aforementioned studies have important limitations. First, none of the studies have assessed differences in soil microbial communities that form the fairy ring of LM. Second, none of the studies have assessed genetic differences between different fairy ring zones. Third, all the studies have assessed fungal and bacterial diversity in the fairy ring soil only through the relative quantification and not through the absolute quantification method.
Therefore, the main objectives of our study were as follows:
To assess differences in soil microbial communities from three fairy ring zones of LM
To assess genetic differences among the three fairy ring zones of LM
To absolutely quantify the microbiota from three fairy ring zones of LM
Materials and methods
Materials
Soil samples were collected from six LM fairy rings identified within the two prefecture-level cities of the Inner Mongolia autonomous region, China, of which fairy rings #5, #11, #12, and #13 were located in Hulunbuir City and fairy rings #7 and #9 were located in Xilin Gol League (Fig. 1a). The sampling area was divided into three zones: DARK, DEAD, and OUT (Fig. 1b). For each area, we selected 3–8 sampling spots along the arc-shape of the fairy ring and used blue, yellow, and red flags to represent the sampling place of each of these zones (Fig. 1c). Finally, according to each fairy ring, we combined 3–8 soil samples obtained from one area to represent the soil sample of the corresponding area of each fairy ring. For each sampling spot, the soil was collected from the top layer by using a soil sample collector (depth, 0–10 cm; diameter, 5 cm). Then, the collected soil sample was screened using a sieve (0.3-mm filter). The soil samples were packed into sterile cryo-storage tubes and frozen in liquid nitrogen, as described previously [18]. In total, 18 soil samples were collected, wherein 3 soil samples were obtained from each fairy ring (one sample from each zone) to represent each of the three fairy ring zones.
DNA extraction, PCR amplification, and sequencing for a metabarcoding survey
Microbial DNA was extracted using the HiPure Soil DNA Kit (Magen, Guangzhou, China) according to the manufacturer’s protocols. The 16S rDNA V3-V4 region of the ribosomal RNA gene was amplified through PCR by using primers 341F (CCTACGGGNGGCWGCAG) and 806R (GGACTACHVGGGTATCTAAT) for bacteria [19] and primers ITS3_KYO2 (GATGAAGAACGYAGYRAA) and ITS4 (TCCTCCGCTTATTGATATGC) for fungi [20]. The purified amplicons were pooled in equimolar ratios and paired-end sequenced (PE250) on an Illumina platform (NovaSeq 6000 sequencing) according to the standard protocols, as described previously [18].
Bioinformatics analysis
(a) Read filtering: Raw data containing adapters or low-quality reads were further filtered using FASTP, as described previously [21]. (b) Read assembly: The cleaned paired-end reads were merged as raw tags by using FLASH (version 1.2.11) [22]. (c) Raw tag filtering: Noisy sequences of raw tags were filtered using the QIIME (version 1.9.1) pipeline [23, 24]. (d) Operational taxonomic units (OTUs) analysis: The effective tags were clustered into OTUs with ≥ 97% similarity by using the UPARSE (version 9.2.64) pipeline [25]. (e) Taxonomy: Representative sequences were classified into organisms through a naive Bayesian model by using the RDP classifier (version 2.2) [26] based on the SILVA database (version 132) [27] for bacterial taxonomy and UNITE database (version 8.0) [28] for fungal taxonomy. The stacked bar plot of the community composition was visualized in the R project ggplot2 package (version 2.2.1) [29] and the Welch’s t-test in the R project Vegan package (version 2.5.3) [30]; the LEfSe cluster and LDA analyses were performed using LEfSe software (http://huttenhower.org/galaxy/). (f) Alpha diversity analysis: Chao1, Simpson, and all other alpha diversity indices were calculated using QIIME (version 1.9.1) [23]. (g) Beta diversity analysis: Sequence alignment was performed using Muscle (version 3.8.31) [31], and the principal component analysis (PCA) in OTUs was plotted using the R package “ade4.”
Microbial quantity
Microbial quantities were assessed through quantitative real-time PCR (qPCR) by using the DNA materials and primers described in “Bioinformatics analysis.” The qPCR cycling parameters were as follows: 95°C for 5 min, 40 cycles of 95°C for 10 s, and 60°C for 40 s. PCR reactions were performed in triplicate by using 15-μL aliquots containing 2-μL AceQ qPCR SYBR Green Master Mix (JZ121-02, Jizhenbio), 0.7 μL of each primer (10 μM), 100 ng (1 μL) of template DNA, and ddH2O to a final volume of 15 μL. To calculate the gene copy number in the soil, we constructed recombinant plasmids. The absolute quantitative standard curve was constructed from the plasmid copy number. Next, the gene copy number in the template DNA was calculated according to the linear equation and the Ct value conversion. In our study, we did three biological replicates and the gene copy number value we used represents the geometric mean values per gram of soil for each sample.
Metagenomics sequencing
For metagenomics sequencing, we used the DNA of three sequencing samples (12 DARK, 12 DEAD, and 12 OUT; Table 2). Then, the DNA was quantified using a Qubit Fluorometer and Qubit dsDNA BR Assay kit (Invitrogen, USA), and the quality of an aliquot was determined using 1% agarose gel electrophoresis. After DNA extraction, 1 μg of genomic DNA was randomly fragmented by Covaris, followed by purification using the AxyPrep Mag PCR clean-up kit. Thereafter, the libraries were constructed and sequenced on the Illumina Hiseq platform. For the bioinformatics analysis, all the raw data were trimmed using SOAPnuke v.1.5.2 [32]. High-quality reads were de novo assembled using Megahit software [33]. Assembled contigs with lengths of <300 bp were discarded in the subsequent analysis. Genes were predicted over contigs by using MetaGeneMarker (2.10) [34]. Redundant genes were removed using CD-HIT [35], with an identity cut-off of 95%. To obtain functional information, the protein sequences were aligned against the eggNOG database [36], CAZy database (2017-09) [37], and KEGG (Kyoto Encyclopedia of Genes and Genomes) database (89.1) [38] by using DIAMOND [39], with an E value cut-off of 1E−5. To generate functional abundance profiles, the reads were aligned to the genes by using Botwie2 [40] with the default setting.
Table 2.
Grouping type of different soil samples obtained from fairy rings
| Basis for grouping | Group | Soil samples ID | ||||
|---|---|---|---|---|---|---|
| Fairy ring zone | DARK | 5DARK | 11DARK | 12DARK | 13DARK | 7DARK |
| 9DARK | ||||||
| DEAD | 5DEAD | 11DEAD | 12DEAD | 13DEAD | 7DEAD | |
| 9DEAD | ||||||
| OUT | 5OUT | 11OUT | 12OUT | 13OUT | 7OUT | |
| 9OUT | ||||||
| Geography | Hulunbuir | 5DARK | 5DEAD | 5OUT | 11DARK | 11DEAD |
| 11OUT | 12DARK | 12DEAD | 12OUT | 13DARK | ||
| 13DEAD | 13OUT | |||||
| Xilin Gol | 7DARK | 7DEAD | 7OUT | 9DARK | 9DEAD | |
| 9OUT | ||||||
| Different fairy rings | F5 | 5DARK | 5DEAD | 5OUT | ||
| F11 | 11DARK | 11DEAD | 11OUT | |||
| F12 | 12DARK | 12DEAD | 12OUT | |||
| F13 | 13DARK | 13DEAD | 13OUT | |||
| F7 | 7DARK | 7DEAD | 7OUT | |||
| F9 | 9DARK | 9DEAD | 9OUT | |||
Results
Analysis of microbial diversity of fairy ring soil
During the field survey, we collected 18 soil samples from each of the three zones of the fairy rings (Fig. 1a). To analyze the diversity of bacteria and fungi in these soil samples, we performed 16S rRNA and internal transcribed spacer (ITS) metabarcoding sequencing. A total of 4,133,834 clean tags were obtained. The results indicated the maximum number of OTUs in the DARK zone of fairy ring #11 (11 DARK, 3466 OTUs) and fairy ring #7 (7 DARK, 1016 OTUs) (Table 1).
Table 1.
16S rRNA and ITS metabarcoding sequencing data
| Sample ID | 16S rRNA | ITS | ||
|---|---|---|---|---|
| Clean tags | OTUs | Clean tags | OTUs | |
| 11DARK | 105855 | 3466 | 124162 | 736 |
| 11DEAD | 105159 | 3194 | 124197 | 598 |
| 11OUT | 104748 | 3481 | 124176 | 815 |
| 12DARK | 105104 | 3396 | 124207 | 869 |
| 12DEAD | 104746 | 3093 | 124211 | 707 |
| 12OUT | 106226 | 3432 | 124171 | 908 |
| 13DARK | 104651 | 3304 | 124158 | 847 |
| 13DEAD | 105650 | 3398 | 124199 | 766 |
| 13OUT | 105845 | 3399 | 124200 | 972 |
| 5DARK | 104856 | 3183 | 124201 | 825 |
| 5DEAD | 105995 | 2921 | 124177 | 585 |
| 5OUT | 104985 | 3078 | 124169 | 861 |
| 7DARK | 105512 | 3591 | 124165 | 1016 |
| 7DEAD | 105198 | 3166 | 124171 | 698 |
| 7OUT | 107548 | 3265 | 124183 | 840 |
| 9DARK | 105335 | 3411 | 124194 | 901 |
| 9DEAD | 106028 | 3119 | 124187 | 803 |
| 9OUT | 105091 | 3089 | 124174 | 858 |
In sample ID, the number represents fairy ring number, whereas the term DARK/DEAD/OUT represents the sample area
Our main objective was to better understand differences in the sequencing results in different fairy ring areas. Therefore, the soil samples were grouped and analyzed according to different fairy ring zones (DARK, DEAD, and OUT), and used sample from different fairy rings to be biological repetition. Further, we regrouped and analyzed all the 18 soil samples according to different geographies and fairy rings to correct differences in the sequencing results of different fairy ring locations (Table 2).
We first evaluated microbial diversity in different fairy ring zones of LM. As shown in Figs. 2a and 3a, Acidobacteria (20.05%, 23.43%, and 31.93% in the DARK, DEAD, and OUT zones, respectively), Proteobacteria (20.19%, 16.61%, and 14.81%, respectively), and Bacteroidetes (18.82%, 18.66%, and 12.25%, respectively) were the three most abundant phyla. Similarly, Basidiomycota (25.21 %, 76.5 %, and 23.99 % in the DARK, DEAD, and OUT zones, respectively), Ascomycota (55.36 %, 17.42 %, and 54.77%, respectively), and Zygomycota (12.71%, 3.98%, and 14.83%, respectively) were the three most abundant fungal phyla. Chitinophagaceae (12.43%, 8.29%, and 9.11% in the DARK, DEAD, and OUT zones, respectively), Chthoniobacteraceae (5.25%, 7.79%, and 11.28%, respectively), and Sphingobacteriaceae (2.33%, 6.29%, and 0.52%, respectively) were the three most abundant families. Tricholomataceae (0.37%, 68.39%, and 0.72% in the DARK, DEAD, and OUT zones, respectively), Nectriaceae (14.65%, 5.32%, and 12.44%, respectively), and Mortierellaceae (12.63%, 3.84%, and 14.28%, respectively) were the three most abundant fungal families. According to other two grouping styles (grouping according to geographical location and fairy rings, as shown in Fig. 2b and c and Fig. 3b and c), the relative abundances of the bacteria and fungi were highly stable, indicating that microbial diversity is not affected by geographical location and fairy rings. In addition, at the genus level, we found that 60% of the OTUs belong to unclassified bacteria and the OTUs belonging to the fairy ring fungi (LM) were incorrectly annotated to the genus Lepista (Supplementary Table S1). These observations indicate that our further analysis would be more accurate at the family level.
Fig. 2.
Relative abundance statistics of bacterial taxonomy in different groups and individuals (at the phylum and family levels). (a) Grouping according to the fairy ring area. (b) Grouping according to different fairy rings. (c) Grouping according to different geographical locations
Fig. 3.
Relative abundance statistics of fungal taxonomy in different groups and individuals (at the phylum and family levels). (a) Grouping according to the fairy ring area. (b) Grouping according to different fairy rings. (c) Grouping according to different geographies
We also analyzed the bacterial and fungal biomarker families in three grouping types, which were identified from the obtained results, by using the linear discriminant analysis (LDA) value. Through these analyses, we attempted to identify dominant families related to the soil ecology of the fairy rings of LM. Because the LEfSe cladograms formed base LDA values at the family level, as shown in Fig. 4, we found that compared with the grouping style according to the geographical location (Fig. 4b and c) and fairy ring type (Fig. 4b and e), the grouping style according to the fairy ring zone (Fig. 4a and d) enables the identification of more abundant biomarker families, which indicates significant differences in the microbial distribution in different fairy ring zones. Analysis of the biomarker families based on the grouping style according to fairy ring zones indicated that 42 and 43 bacterial and fungal family biomarkers, respectively, have LDA values of ≥3 and relative abundance of at least >1% (Supplementary Table 2 and Fig. 4a and d). Among these biomarkers, 8 biomarker families unique to one fairy ring zone were selected after eliminating the interfering factors from the geographical location and fairy ring type categories; the excluded families are shown as biomarker families from different groups simultaneously (for the grouping type, see Supplementary Table 1). As shown in Table 3, we identified the following 4 bacterial families: Paenibacillaceae (1.16% in the DEAD zone, P = 1.69E-03); Sphingobacteriaceae (6.3% in the DEAD zone, P = 2.46E-03); Sphingomonadaceae (2.66% in the DARK zone, P = 8.41E-03); Streptosporangiaceae (1.16% in the DEAD zone, P = 2.21E-02). We also identified 4 fungal families: Chaetomiaceae (1.17% in the OUT zone, P = 4.46E-02), Leptosphaeriaceae (3.21% in the DARK zone, P = 2.21E-02), Phaeosphaeriaceae (1.60% in the DARK zone, P = 4.72E-03), and Tricholomataceae (68.39% in the DEAD zone, P = 3.32E-03).
Fig. 4.
LEfSe cladograms of bacteria (a, b, c) and fungi (d, e, f) form different grouping patterns. a and d Represent the grouping pattern according to the fairy ring zone; b and e represent the grouping pattern according to the fairy ring type; and c and f represent the grouping pattern according to the geography. Different colors represent different groups, and nodes with different colors represent species that play an important role in the groups represented by that color. The one-color circle represents a biomarker, and the figure in the upper right corner is the biomarker name. Yellow nodes represent microbial groups that do not play an important role in different groups. Circles from the inside to the outside indicate phylum, class, order, family, and genus levels of species
Table 3.
Species biomarker (family taxonomic level) data for different tested fairy ring areas
| Family | Group | LDA | P-value | Relative abundance (%) | |
|---|---|---|---|---|---|
| Bacteria | Paenibacillaceae | DEAD | 3.06 | 1.69E-03 | 1.16 |
| Sphingobacteriaceae | DEAD | 4.62 | 2.46E-03 | 6.30 | |
| Sphingomonadaceae | DARK | 4.29 | 8.41E-03 | 2.66 | |
| Streptosporangiaceae | DEAD | 3.79 | 2.21E-02 | 1.16 | |
| Fungi | Chaetomiaceae | OUT | 3.81 | 4.46E-02 | 1.17 |
| Leptosphaeriaceae | DARK | 4.28 | 2.21E-02 | 3.21 | |
| Phaeosphaeriaceae | DARK | 3.91 | 4.72E-03 | 1.60 | |
| Tricholomataceae | DEAD | 5.58 | 3.32E-03 | 68.39 |
The LDA value represents the result of the linear discriminant analysis. Relative abundance values are shown only for those species that accounted for at least 1%
We evaluated the alpha diversity of bacteria and fungi by using four diversity indices: Shannon, Simpson, Chao1, and Good’s coverage. As shown in Table 4, the value range of the Shannon index was from 6.062 (9 OUT zones) to 6.82 (12 DARK zones) for the bacterial species and from 1.037 (5 DEAD zones) to 5.052 (13 OUT zones) for the fungal species; the value range of the Chao1 index was from 3694 (5 DEAD zones) to 4255 (7 DARK zones) for the bacterial species and from 751.7 (11 DARK zones) to 1076 (7 DARK zones) for the fungal species. Moreover, all Good’s coverage index values were more than 0.98, indicating that the sequencing coverage tends to move toward saturation. These results showed that fungal diversity in different fairy ring zones is more variable than bacterial diversity.
Table 4.
Alpha diversity index statistics for the bacteria and fungi identified from the 18 fairy ring soil samples
| Sample ID | Bacteria | Fungus | ||||||
|---|---|---|---|---|---|---|---|---|
| Shannon | Simpson | Chao1 | Good’s coverage | Shannon | Simpson | Chao1 | Good’s coverage | |
| 5DARK | 6.574 | 0.005 | 3781 | 0.985 | 3.794 | 0.095 | 855.1 | 0.999 |
| 11DARK | 6.610 | 0.005 | 4107 | 0.985 | 4.187 | 0.054 | 751.7 | 0.999 |
| 12DARK | 6.820 | 0.003 | 4113 | 0.984 | 4.935 | 0.024 | 976.6 | 0.999 |
| 13DARK | 6.552 | 0.004 | 3994 | 0.984 | 4.068 | 0.055 | 919.8 | 0.999 |
| 7DARK | 6.522 | 0.006 | 4255 | 0.982 | 4.705 | 0.028 | 1076 | 0.999 |
| 9DARK | 6.437 | 0.010 | 4045 | 0.984 | 4.292 | 0.057 | 962.7 | 0.999 |
| 5DEAD | 6.277 | 0.006 | 3694 | 0.985 | 1.037 | 0.681 | 759.8 | 0.998 |
| 11DEAD | 6.292 | 0.006 | 4004 | 0.984 | 2.085 | 0.446 | 652.5 | 0.999 |
| 12DEAD | 6.342 | 0.005 | 3822 | 0.984 | 2.299 | 0.296 | 813.4 | 0.999 |
| 13DEAD | 6.347 | 0.007 | 4035 | 0.984 | 2.311 | 0.449 | 865 | 0.999 |
| 7DEAD | 6.208 | 0.010 | 3870 | 0.984 | 2.496 | 0.392 | 752.2 | 0.999 |
| 9DEAD | 6.144 | 0.010 | 3837 | 0.984 | 1.398 | 0.662 | 970 | 0.998 |
| 5OUT | 6.193 | 0.009 | 3825 | 0.984 | 4.460 | 0.038 | 960.6 | 0.999 |
| 11OUT | 6.407 | 0.007 | 4226 | 0.983 | 4.794 | 0.031 | 861.6 | 0.999 |
| 12OUT | 6.212 | 0.013 | 4211 | 0.983 | 4.651 | 0.036 | 1040 | 0.999 |
| 13OUT | 6.152 | 0.013 | 4076 | 0.984 | 5.052 | 0.022 | 1013 | 0.999 |
| 7OUT | 6.154 | 0.011 | 4065 | 0.984 | 4.482 | 0.033 | 898.5 | 0.999 |
| 9OUT | 6.062 | 0.012 | 3941 | 0.984 | 4.877 | 0.019 | 898 | 0.999 |
For the beta diversity analysis, results of the VENN analysis of OTUs indicated that 4519 of 6715 (67.3%) and 1221 of 3728 (32.75%) of the common OTUs are distributed in all fairy ring zones for bacteria and fungi, respectively (Figs. 5 and 6). The PCA analysis results showed that for bacteria, the samples forming DARK and OUT zones are clustered together, whereas the samples from the DEAD zone did not display a clear clustering tendency. For fungi, the samples from the DEAD zone clustered alone and samples from other zones clustered together.
Fig. 5.
Bacterial beta diversity analysis of 18 fairy ring soil samples. In the VENN analysis, the value represents the number of OTUs, and ID represents the fairy ring area. In the PCA analysis, the PC1 value on the x-axis represents the factor of OTU abundance, and the PC2 value on the y-axis represents the factor of OTU genetic distance
Fig. 6.
Fungal beta diversity analysis of 18 fairy ring soil samples. In the VENN analysis, the value represents the number of OTUs, and ID represents the fairy ring area. In the PCA analysis, the PC1 value on the x-axis represents the factor of OTU abundance, and the PC2 value on the y-axis represents the factor of OTU genetic distance
To summarize this section, we performed a metabarcoding analysis of 18 soil samples obtained from three fairy ring zones. Although we found differences in the bacterial distribution in the three fairy ring zones, the difference was not statistically significant. In addition, in the alpha diversity analysis, the difference in bacterial species across three fairy zones could not be effectively distinguished. The result suggests that our microbial diversity assay requires further development.
Analysis of microbial quantify of fairy ring soil
We further performed a microbial quantity analysis using a quantitative real-time PCR approach of the three fairy rings (#11, #12, #13) identified earlier (Fig. 1a). We used the gene numbers amplified by the metabarcoding markers that were calibrated using the standard curve to represent the absolute quantities of microorganisms (Fig. 7). We found that for all the fairy rings, the bacterial and fungal quantity in the DEAD zone is significantly increased relative to the DARK and OUT zones. Among these, the fungal quantity sharply increased because the fairy ring fungi (LM) breed in the DEAD zones. However, interestingly, we also identified a significant increase in the bacterial quantity along with the fungal quantity in the DEAD zone.
Fig. 7.
Microbial quantity of fairy ring soil. The x-axis shows the fungal and bacterial classification, and the y-axis represents the gene copy number per gram of soil. The figure legends are given on the right side
Metagenomics survey of a fairy ring
Although the aforementioned results indicate that the microbial composition and quantity of the three zones of fairy rings differ, it remains unclear whether soils in the three zones perform different biological functions. Hence, we performed metagenomic sequencing of one of the fairy rings (#12, Fig. 1) that was clearly identified from different zones. The DARK and DEAD zones had higher genetic diversity than the OUT zone (Table 5 and Fig. 8). The metagenome assembly lengths of the DARK and DEAD zones (169,206,079 bp and 153,408,521 bp, respectively) were similar, whereas those of the OUT zone were shorter (91,430,371 bp, 54.03% and 59.6% of those of the DARK and DEAD zones, respectively). The predicted gene number was also fewer for the OUT zone (86,944 genes) compared with the DARK and DEAD zones (278,973 and 235,931 genes, respectively). Among the 387,006 genes found across the three soil samples, 50,064 were common to all three zones. In the OUT zone, only 21.14% (18380/86944) of genes were unique to that zone (Fig. 8b).
Table 5.
Metagenomic sequencing result of three soil samples from the fairy ring #12
| Sample | 12DARK | 12DEAD | 12OUT |
|---|---|---|---|
| Clean data (Gb) | 12.58 | 12.58 | 12.58 |
| Contig number | 198,713 | 180,015 | 118,615 |
| Assembly length (bp) | 169,206,079 | 153,408,521 | 91,430,371 |
| N50 (bp) | 790 | 791 | 721 |
| N90 (bp) | 534 | 534 | 528 |
| Average size (bp) | 851 | 852 | 770 |
| Gene number | 278,973 | 235,931 | 86,944 |
Fig. 8.
Metagenomic annotation of the DARK, DEAD, and OUTER zones of the No. 12 fairy ring soil. a Represents the annotation results from the eggNOG and KEGG databases. The x-axis represents the number of genes, and the y-axis represents the functional classification. b Represents the VENN map of all predicted genes from the soil samples of the three zones. c represents the distribution of the CAZymes genes across the three zones. The x-axis represents the number of genes, and the y-axis represents the CAZymes modules of glycoside hydrolases (GHs), glycosyltransferases (GTs), polysaccharide lyases (PLs), carbohydrate esterases (CEs), and auxiliary activities (AAs)
To compare the biological function of the three fairy ring zones, we annotated the predicted genes by using the KEGG [38] and eggNog [36] databases (Fig. 8a). By using this approach, we found that the three soil samples have a similar pattern of gene classes, suggesting that the soils perform similar biological functions across all the fairy ring zones. We further investigated the carbohydrate transport and metabolism and the carbohydrate metabolism genes class by using the dbCAN [37] database. We again found that carbohydrate genes are the major class of genes in the above annotation, and still, the OUT, DEAD, and DARK zones were found to have similar distribution patterns (Fig. 8c).
Discussion
Dominant families regulate community diversity
Although the relative species compositions (bacteria and fungi) are similar across the three zones, eight unique enrichment families were found across different zones (Table 3). In bacteria, Sphingobacteriaceae (6.3% in DEAD zone) was the most abundant family in the DEAD zones, suggesting that this family participates in the formation of the DEAD zone. In a study on Agaricus arvensis fairy rings, reported the enrichment of Sphingobacteriaceae in the DEAD zones [4]. The finding that Sphingomonadaceae are enriched in the DARK zone (2.66%) is consistent with the findings reported in literature: Sphingomonadaceae were found to be enriched in the fairy ring of Agaricus arvensis [4], Agaricus gennadii [17], and in our pre-experiment [18]. The families Sphingobacteriaceae and Sphingomonadaceae are mostly constituted by chemoorganotrophic bacteria usually found in soil, and the ability of both the families to grow is attributed to the production, degradation, and conversion of biomolecules [41, 42]. This suggests that fairy ring fungi (LM) might possess the ability to stimulate the abundance of Sphingobacteriaceae and Sphingomonadaceae and that the two families may participate in the formation of DEAD and DARK zones of the LM fairy ring.
Tricholomataceae (68.39% in the DEAD zone, mycelium of LM) were enriched in the DEAD zone, and this is a natural phenomenon because fairy ring fungi (LM) are the keystone species exhibiting the ecological phenomenon. However, we noticed an interesting result: the Tricholomataceae component accounted for 65–78% of the DEAD zone, which is similar to the proportion detected in our pre-experiment. This suggests that metabarcoding sequencing cannot distinguish the mixed zones soil from the DEAD zone soils. This could lead to erroneous conclusions if relying entirely on relative quantitation analyses and thus stresses the importance of including absolute quantitative analysis. Five other enriched families were first found in the fairy ring ecosystem and might have important functions in each zone in which they settle. These species formed eight, high-abundance families, and this pattern was consistent across the six studied habitats. Thus, these families, particularly the Sphingobacteriaceae family (with the highest abundance in eight families), are likely to be the dominant species that regulate fairy ring community diversity.
Fairy rings are associated with significantly increased microbial quantity
We found that the microbial quantity assay is better suited, compared with community diversity, for revealing differences among the soil zones. Particularly, we detected that an increase in the fungal quantity also causes an increase in bacterial quantity in the DEAD zone. Previous studies have reported that the mycelium of dead fairy ring fungi releases nutrients into the soil, which increases the richness of plants in DARK zones [5, 17, 43, 44]. In our study, we first found that in fairy ring ecosystems, the abundance of soil bacteria in DEAD zones increases sharply prior to the development of plant richness. Although the cause of the increase in bacterial abundance in the DEAD zone is uncertain, this increase occurred after the growth of fairy ring fungi in the DEAD zone; hence, we assume that the two factors are related. In summary, these results might represent a crucial link to the phenomenon of plant richness in the DARK zone.
The first shotgun sequencing study of a fairy ring ecology
To the best of our knowledge, this is the first study to determine the soil metagenome of LM fairy ring ecology by using the shotgun method. The genetic diversity of the fairy ring soil ecosystem was revealed, and surprisingly, we found that the two zones DARK and DEAD affected by fairy ring fungi (LM) exhibit significant genetic diversity advantage compared with the OUT zone. In addition, we found that the CAZymes genes are abundant in all the predicted genes; these genes were also abundant in both eggNog and KEGG annotations. This may be because the soil microorganisms function as decomposers in nature, whereas the main function of CAZymes genes is to metabolize energy that probably accounts for their high abundance. A previous study on the forest soil metagenome reported a similar result of genes from level 1 functional categories of carbohydrate enrichment from all the soil samples [45]. Our results suggest that the CAZymes genes may be an important gene class contributing to the fairy ring ecosystem.
Fairy ring fungi create a center for microbial communities in grasslands
The metabarcoding analysis, based on a relative quantitative analysis, indicated similar levels of diversity in all three zones; however, the fairy ring fungus (LM) was found to be significantly enriched in the DEAD zone. In addition, the microbial quantity and metagenomics analyses revealed that the soil organisms form three distinct zones. On the other hand, we also found flourishing biological activity in the infection zones, DEAD and DARK zones, of the fairy ring, and the finding is supported by the following three pieces of evidence: (i) of the 8 biomarker families that formed three fairy ring zones located in different habitats, 7 families belong to the DEAD and DARK zones (Table 3), which indicates that the growth of these dominant family species is driven by the fairy ring fungi; (ii) the quantity analysis further showed that dominant family species increases not only in terms of percentage but also in absolute abundance, especially in the DEAD zones; and (iii) the metagenomics survey results indicated that the genetic diversity is abundant in the DARK and DEAD zones, with the OUT zone having fewer genes and no additional biological functions (the fairy ring fungi significantly improved the soil biological activity). These data indicate that fairy rings act as centers for bacterial and fungal communities and because of their importance in genetic diversity, they are termed as the ecological phenomena worthy of attention and preservation.
In this study, we systematically surveyed the soil microbial ecology of type I fairy rings formed by LM. First, we performed a metagenomics survey and assessed the microbial quantity in the fairy ring ecosystem. We showed that the growth of fairy ring fungi mycelium leads to the proliferation of soil bacteria and that fairy rings act as centers for bacterial and fungal communities. The reported data advance our understanding of type I fairy rings and will be insightful for the research on crop production.
Supplementary information
(XLSX 12 kb)
(XLSX 16 kb)
Acknowledgements
This study was financially supported by The Biodiversity Survey and Assessment Project of the Ministry of Ecology and Environment, China (No.2019HJ2096001006). We acknowledge TopEdit (www.topeditsci.com) for linguistic editing and proofreading during the preparation of this manuscript.
Data availability statement
The raw amplicon sequencing datasets (from the metabarcoding and metagenomics surveys) are available from the NCBI Sequence Read Archive under accession PRJNA641981.
Author contribution
M.D. defined the questions and the methodological approach then analyzed the data and wrote the manuscript; T.B. conceived the idea.
Declaration
Competing interest
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Fierer N. Embracing the unknown: disentangling the complexities of the soil microbiome. Nat Rev Microbiol. 2017;15(10):579–590. doi: 10.1038/nrmicro.2017.87. [DOI] [PubMed] [Google Scholar]
- 2.Yang C, Li J, Liu N, Zhang Y. Effects of fairy ring fungi on plants and soil in the alpine and temperate grasslands of China. Plant Soil. 2019;441(1-2):499–510. doi: 10.1007/s11104-019-04141-7. [DOI] [Google Scholar]
- 3.Tobina H, Choi J-H, Asai T, Kiriiwa Y, Asakawa T, Kan T, Morita A, Kawagishi H. 2-Azahypoxanthine and imidazole-4-carboxamide produced by the fairy-ring-forming fungus increase wheat yield. Field Crop Res. 2014;162:6–11. doi: 10.1016/j.fcr.2014.03.008. [DOI] [Google Scholar]
- 4.Zotti M, De Filippis F, Cesarano G, Ercolini D, Tesei G, Allegrezza M, Giannino F, Mazzoleni S, Bonanomi G. One ring to rule them all: an ecosystem engineer fungus fosters plant and microbial diversity in a Mediterranean grassland. New Phytol. 2020;227(3):884–898. doi: 10.1111/nph.16583. [DOI] [PubMed] [Google Scholar]
- 5.Bonanomi G, Mingo A, Incerti G, Mazzoleni S, Allegrezza M, Zobel M. Fairy rings caused by a killer fungus foster plant diversity in species-rich grassland. J Veg Sci. 2012;23(2):236–248. doi: 10.1111/j.1654-1103.2011.01353.x. [DOI] [Google Scholar]
- 6.Choi JH, Abe N, Tanaka H, Fushimi K, Nishina Y, Morita A, Kiriiwa Y, Motohashi R, Hashizume D, Koshino H, Kawagishi H. Plant-growth regulator, imidazole-4-carboxamide, produced by the fairy ring forming fungus Lepista sordida. J Agric Food Chem. 2010;58(18):9956–9959. doi: 10.1021/jf101619a. [DOI] [PubMed] [Google Scholar]
- 7.Choi JH, Ohnishi T, Yamakawa Y, Takeda S, Sekiguchi S, Maruyama W, Yamashita K, Suzuki T, Morita A, Ikka T, Motohashi R, Kiriiwa Y, Tobina H, Asai T, Tokuyama S, Hirai H, Yasuda N, Noguchi K, Asakawa T, Sugiyama S, Kan T, Kawagishi H. The source of “fairy rings”: 2-azahypoxanthine and its metabolite found in a novel purine metabolic pathway in plants. Angew Chem Int Ed Eng. 2014;53(6):1552–1555. doi: 10.1002/anie.201308109. [DOI] [PubMed] [Google Scholar]
- 8.Caspar J, Spiteller P. A free cyanohydrin as arms and armour of Marasmius oreades. Chembiochem. 2015;16(4):570–573. doi: 10.1002/cbic.201402453. [DOI] [PubMed] [Google Scholar]
- 9.Suzuki T, Yamamoto N, Choi JH, Takano T, Sasaki Y, Terashima Y, Ito A, Dohra H, Hirai H, Nakamura Y, Yano K, Kawagishi H. The biosynthetic pathway of 2-azahypoxanthine in fairy-ring forming fungus. Sci Rep. 2016;6(1):39087. doi: 10.1038/srep39087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Halisky PM, Peterson JL. Basidiomycetes associated with fairy rings in turf. Bull Torrey Bot Club. 1970;97(4):225–227. doi: 10.2307/2483462. [DOI] [Google Scholar]
- 11.Shantz HL, Piemeisel RL (1917) Fungus fairy rings in eastern Colorado and their effect on vegetation 11 (5):191-246
- 12.Edwards PJ. Effects of the fairy ring fungus Agaricus arvensis on nutrient availability in grassland. New Phytol. 1988;110(3):377–381. doi: 10.1111/j.1469-8137.1988.tb00275.x. [DOI] [Google Scholar]
- 13.Klironomos JN. Feedback with soil biota contributes to plant rarity and invasiveness in communities. Nature. 2002;417(6884):67–70. doi: 10.1038/417067a. [DOI] [PubMed] [Google Scholar]
- 14.Gramss G, Voigt KD, Bergmann H. Factors influencing water solubility and plant availability of mineral compounds in the tripartite fairy rings of Marasmius oreades (BOLT.: FR.) FR. J Basic Microbiol. 2005;45(1):41–54. doi: 10.1002/jobm.200410459. [DOI] [PubMed] [Google Scholar]
- 15.Marí T, Castaño C, Rodríguez A, Ibáñez M, Lobo A, Sebastià M-T. Fairy rings harbor distinct soil fungal communities and high fungal diversity in a montane grassland. Fungal Ecol. 2020;47:100962. doi: 10.1016/j.funeco.2020.100962. [DOI] [Google Scholar]
- 16.Nesme J, Achouak W, Agathos SN, Bailey M, Baldrian P, Brunel D, Frostegard A, Heulin T, Jansson JK, Jurkevitch E, Kruus KL, Kowalchuk GA, Lagares A, Lappin-Scott HM, Lemanceau P, Le Paslier D, Mandic-Mulec I, Murrell JC, Myrold DD, Nalin R, Nannipieri P, Neufeld JD, O'Gara F, Parnell JJ, Puhler A, Pylro V, Ramos JL, Roesch LF, Schloter M, Schleper C, Sczyrba A, Sessitsch A, Sjoling S, Sorensen J, Sorensen SJ, Tebbe CC, Topp E, Tsiamis G, van Elsas JD, van Keulen G, Widmer F, Wagner M, Zhang T, Zhang X, Zhao L, Zhu YG, Vogel TM, Simonet P. Back to the Future of Soil Metagenomics. Front Microbiol. 2016;7:73. doi: 10.3389/fmicb.2016.00073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Yang C, Zhang F, Liu N, Hu J, Zhang Y. Changes in soil bacterial communities in response to the fairy ring fungus Agaricus gennadii in the temperate steppes of China. Pedobiologia. 2018;69:34–40. doi: 10.1016/j.pedobi.2018.05.002. [DOI] [Google Scholar]
- 18.Duan M, Bau T (2020) Initial sample processing can influence the soil microbial metabarcoding surveys, revealed by Leucocalocybe mongolica fairy ring ecosystem. bioRxiv:2020.2006.2009.142018. 10.1101/2020.06.09.142018
- 19.Guo M, Wu F, Hao G, Qi Q, Li R, Li N, Wei L, Chai T. Bacillus subtilis improves immunity and disease resistance in rabbits. Front Immunol. 2017;8:354. doi: 10.3389/fimmu.2017.00354. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Toju H, Tanabe AS, Yamamoto S, Sato H. High-coverage ITS primers for the DNA-based identification of ascomycetes and basidiomycetes in environmental samples. PLoS One. 2012;7(7):e40863. doi: 10.1371/journal.pone.0040863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Chen S, Zhou Y, Chen Y, Gu J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34(17):i884–i890. doi: 10.1093/bioinformatics/bty560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Magoč T, Salzberg SL. FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics. 2011;27(21):2957–2963. doi: 10.1093/bioinformatics/btr507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, Fierer N, Pena AG, Goodrich JK, Gordon JI. QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 2010;7(5):335–336. doi: 10.1038/nmeth.f.303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bokulich NA, Subramanian S, Faith JJ, Gevers D, Gordon JI, Knight R, Mills DA, Caporaso JG. Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing. Nat Methods. 2013;10(1):57–59. doi: 10.1038/nmeth.2276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Edgar RC. UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nat Methods. 2013;10(10):996–998. doi: 10.1038/NMETH.2604. [DOI] [PubMed] [Google Scholar]
- 26.Wang Q, Garrity GM, Tiedje JM, Cole JR. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol. 2007;73(16):5261–5267. doi: 10.1128/AEM.00062-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Pruesse E, Quast C, Knittel K, Fuchs BM, Ludwig W, Peplies J, Glöckner FO. SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res. 2007;35(21):7188–7196. doi: 10.1093/nar/gkm864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Nilsson RH, Larsson K-H, Taylor AFS, Bengtsson-Palme J, Jeppesen TS, Schigel D, Kennedy P, Picard K, Glöckner FO, Tedersoo L. The UNITE database for molecular identification of fungi: handling dark taxa and parallel taxonomic classifications. Nucleic Acids Res. 2019;47(D1):D259–D264. doi: 10.1093/nar/gky1022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wickham H, Chang W (2008) ggplot2: An implementation of the Grammar of Graphics. R package version 07, URL: https://CRAN.R-project.org/package=ggplot2. Accessed 13 Dec 2019
- 30.Oksanen J, Blanchet FG, Kindt R, Legendre P, O’hara R, Simpson GL, Solymos P, Stevens MHH, Wagner H (2010) Vegan: community ecology package. R package version 1.17-4. URL: https://CRAN.R-project.org/package=vegan. Accessed 16 Dec 2019
- 31.Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 2004;32(5):1792–1797. doi: 10.1093/nar/gkh340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Chen Y, Chen Y, Shi C, Huang Z, Zhang Y, Li S, Li Y, Ye J, Yu C, Li Z, Zhang X, Wang J, Yang H, Fang L, Chen Q. SOAPnuke: a MapReduce acceleration-supported software for integrated quality control and preprocessing of high-throughput sequencing data. Gigascience. 2018;7(1):1–6. doi: 10.1093/gigascience/gix120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Li D, Liu CM, Luo R, Sadakane K, Lam TW. MEGAHIT: an ultra-fast single-node solution for large and complex metagenomics assembly via succinct de Bruijn graph. Bioinformatics. 2015;31(10):1674–1676. doi: 10.1093/bioinformatics/btv033. [DOI] [PubMed] [Google Scholar]
- 34.Zhu W, Lomsadze A, Borodovsky M. Ab initio gene identification in metagenomic sequences. Nucleic Acids Res. 2010;38(12):e132. doi: 10.1093/nar/gkq275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Fu L, Niu B, Zhu Z, Wu S, Li W. CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics. 2012;28(23):3150–3152. doi: 10.1093/bioinformatics/bts565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Huerta-Cepas J, Szklarczyk D, Heller D, Hernandez-Plaza A, Forslund SK, Cook H, Mende DR, Letunic I, Rattei T, Jensen LJ, von Mering C, Bork P. eggNOG 5.0: a hierarchical, functionally and phylogenetically annotated orthology resource based on 5090 organisms and 2502 viruses. Nucleic Acids Res. 2019;47(D1):D309–D314. doi: 10.1093/nar/gky1085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Yin Y, Mao X, Yang J, Chen X, Mao F, Xu Y (2012) dbCAN: a web resource for automated carbohydrate-active enzyme annotation. Nucleic Acids Res 40 (Web Server issue):W445-451. 10.1093/nar/gks479 [DOI] [PMC free article] [PubMed]
- 38.Kanehisa M, Goto S, Hattori M, Aoki-Kinoshita KF, Itoh M, Kawashima S, Katayama T, Araki M, Hirakawa M. From genomics to chemical genomics: new developments in KEGG. Nucleic Acids Res. 2006;34(Database issue):D354–D357. doi: 10.1093/nar/gkj102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Buchfink B, Xie C, Huson DH. Fast and sensitive protein alignment using DIAMOND. Nat Methods. 2015;12(1):59–60. doi: 10.1038/nmeth.3176. [DOI] [PubMed] [Google Scholar]
- 40.Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–359. doi: 10.1038/nmeth.1923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Lambiase A. The family sphingobacteriaceae. Prokaryotes. 2014;4:907–9014. doi: 10.1007/978-3-642-38954-2_136. [DOI] [Google Scholar]
- 42.Glaeser SP, Kämpfer P (2014) The family sphingomonadaceae. The Prokaryotes: Alphaproteobacteria and Betaproteobacteria:641-707 10.1007/978-3-642-30197-1_302
- 43.Edwards PJ. The growth of fairy rings of Agaricus arvensis and their effect upon grassland vegetation and soil. J Ecol. 1984;72(2):505–513. doi: 10.2307/2260062. [DOI] [Google Scholar]
- 44.Caesar-Tonthat TC, Espeland E, Caesar AJ, Sainju UM, Lartey RT, Gaskin JF. Effects of Agaricus lilaceps fairy rings on soil aggregation and microbial community structure in relation to growth stimulation of western wheatgrass (Pascopyrum smithii) in Eastern Montana rangeland. Microb Ecol. 2013;66(1):120–131. doi: 10.1007/s00248-013-0194-3. [DOI] [PubMed] [Google Scholar]
- 45.Sun S, Badgley BD. Changes in microbial functional genes within the soil metagenome during forest ecosystem restoration. Soil Biol Biochem. 2019;135:163–172. doi: 10.1016/j.soilbio.2019.05.004. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(XLSX 12 kb)
(XLSX 16 kb)
Data Availability Statement
The raw amplicon sequencing datasets (from the metabarcoding and metagenomics surveys) are available from the NCBI Sequence Read Archive under accession PRJNA641981.