Continuous Cropping Regulates Fungal Community Structures and Functional Groups in Rhizosphere Soil of Tibet Barley

Continuous cropping regulates the community structure of rhizosphere soil microbes; however, the effects of continuous cropping on the community structure and function of rhizosphere fungal communities of plants cultivated in high-altitude regions are not well understood. In this study, 18S rRNA gene high-throughput sequencing was applied to examine the rhizosphere fungal community structure during continuous cropping of Tibet barley (a principal cereal cultivated on the Qinghai-Tibetan Plateau). The results showed that the Chao1 and phylogenetic diversity (PD) indices declined as cropping years increased. Additionally, relative abundance of the genera Cystolobasidium, Mucor, and Ustilago increased, whereas the abundance of Fusarium decreased during continuous cropping. Furthermore, identication of ecological groups using FUNGuild revealed that saprotrophs, pathogens, and symbiotrophs were the dominant groups in rhizosphere soil, and these three trophic modes all increased signicantly with continuous cropping. During continuous cropping, the fungal plant-pathogens Parastagonospora and Ustilago also increased remarkably, as did the endophyte fungi Verticillium. Collectively, continuous cropping of Tibet barely increased the potential for plant-pathogenic and endophyte fungi in rhizosphere soil. Thus, the development of sustainable farming practices to reduce the abundance of harmful fungi is vital to Tibet barley growth and production during continuous cropping of plants cultivated in high-altitude regions.


Introduction
Continuous cropping, which is the agricultural practice of long-term cultivation of the same crop in identical soil, is a vital agricultural issue because it leads to reduced the soil quality and augmentation of harmful pests and soilborne microbial communities, resulting in yield decline of plants 1 . Rhizosphere soil microorganisms affect the growth, development, and health status of plants 2 , and the in uences of continuous cropping on microbial community structures in rhizosphere soil have gained considerable attention [3][4][5][6][7] . When compared with bacteria, rhizosphere fungi are important decomposers that play vital roles in balancing the rhizosphere ecosystem 8 . Moreover, many fungi are closely associated with plant diseases, and research has shown that accumulation of fungal pathogens is often responsible for continuous cropping problems 9 .
High-altitude ecosystems are generally characterized by low temperature, variable rainfall, reduced atmospheric pressure, and soil nutritional stress 10 . In high-altitude regions, low temperature is the major factor in uencing microbial diversity and function 11 . Plants and microbes have co-evolved and interact with each other in the environment; thus, high-altitude regions are important to understanding the interactions between certain microbes and plants cultivated in cold environments 12 . Moreover, lower species richness and diversity have been observed in soil from higher elevations 13,14 . In a study of three vertical climate zones on the Qinghai-Tibetan Plateau, China (av. 4,000 m above sea level), soil bacteria exhibited more apparent elevational zonation features and decreased diversity patterns with increasing elevation 14 . Studies have also been conducted to explore the dynamics of rhizosphere fungal communities under continuous cropping of plants cultivated at normal altitudes, such as soybean 15 , potato 5 , strawberry 16 , peanut 17 , tea 18 , and cucumber 19 . However, the fungal community structures of rhizosphere soil in high altitude regions are less understood. Additionally, changes in potentially bene cial and pathogenic fungi have not been studied during continuous cropping of plants cultivated in high-altitude regions.
Tibet barley (Hordeum vulgare L., qingke), which serves as a staple food and an important livestock feed in the Tibetan Plateau 20 , is drought and cold tolerant, exhibiting strong adaptability to the extreme climate of the plateau 21 . The cropping area of Tibet barely accounts for approximately 43% of the grain crop area on the Tibetan Plateau 22 . We previously found that continuous cropping reduced the yield of Tibet barley (accepted data). Thus, understanding the in uence of continuous cropping on the fungal community in the soil rhizosphere is essential to designing effective farm management practices to relieve soil depletion associated with Tibetan barley cultivation. Therefore, we assume that the continuous cropping of Tibet barley may directly in uence the diversity and function of fungal communities. In this study, we considered the fungal communities in Tibet-barley elds with 2 to 6 years of monoculture history. The objectives of our study were to explore the dynamics of diversity and composition of the rhizosphere fungal community composition during continuous cropping. Furthermore, we evaluated changes in fungal groups with potential ecological functions during continuous cropping. The results presented herein will provide a better understanding of the response of rhizosphere soil fungal communities to continuous cropping of plants cultivated in high-altitude regions.

Dynamics of fungal community diversity and composition
A total of 783,935 high-quality sequences of 18S rRNA were obtained and each sample was evenly rare ed to 38,031 reads before diversity calculation. Continuous cropping was found to have a negative effect on the species richness and phylogenetic diversity of rhizosphere fungal communities of Tibet barley. The Chao1 and PD indices of the rhizosphere fungal community signi cantly decreased from 2 to 5 years, but slightly increased after 5 years of continuous cropping ( Table 1). The Chao1 index decreased from 301.5 ± 10.23 at 2 years to 224.8 ± 6.6 at 5 years of continuous cropping, then climbed slightly to 233.0±21.3 after 6 years. The PD indices decreased from 19.8 ± 0.6 to 13.4 ± 0.5, then increased slightly to 14.5 ± 1.1 after 6 years. However, no distinct pattern of species diversity (represented by Shannon and Simpson indices) was observed during continuous cropping (Table 1). PCoA was used to depict the degree of fungal community differentiation after continuous cropping for various lengths of time. The rst two principal coordinates explained 60.31% of the total variance of the rhizosphere soil fungal community structures (Figure 1). PCoA revealed that fungal communities from the second and third years (CC2Y and CC3Y) were separated from those from later years (CC4Y, CC5Y, and CC6Y) along the rst PCoA axis. Communities from the last three years were separated along the second PCoA axis (Figure 1). Nearly 79% of the total sequences of the 18S rRNA gene could be taxonomically classi ed at the phylum level. Ascomycota (56.74%±4.91%), Basidiomycota (17.45%±7.60%), and Mucoromycota (3.21%±1.91%) were the dominant phyla ( Figure S1), accounting for more than 75% of all sequences. The predominant families in the rhizosphere soil were Nectriaceae, Mucoraceae, and Pleosporaceae ( Figure 2A). Nectriaceae was the only abundant family that showed a decreasing trend (from 10.06% ± 0.71% to 3.28% ± 0.14%) throughout the experiment. The abundances of all other dominant families (Cysto lobasidiaceae, Mucoraceae, Phaeosphaeriaceae, Pleosporaceae, Sarcosomataceae, Sarocladiaceae, and Ustilaginaceae) increased during continuous cropping. Among these, Ustilaginaceae increased 46.8-fold, from 0.12% ± 0.02% to 5.62% ± 1.02% (Figure 2A). At the genus level, Fusarium, Mucor, Parastagonospora, Sarocladium, and Ustilago mostly appeared across all samples ( Figure 2B). The relative abundance of Fusarium uctuated, but showed an overall decrease from 10.06% ± 0.71% to 3.28% ± 0.14% throughout the experiment. Cysto lobasidium and Ustilago were most abundant after 6 years of continuous cropping ( Figure 2B).
Taxa shared in rhizosphere soil during continuous cropping Shared taxa were classi ed at the phylum, class, order, and genus level for all samples. Taxa that showed more or less than 1% mean relative abundance were de ned as dominant or rare, respectively ( Figure 3). Among the rhizosphere fungal communities, 42 shared genera were identi ed during continuous cropping, comprising about 31.82% (dominant shared genus, ~16.67%, rarely shared genus, ~15.15%) of the total sequence abundance ( Figure 3). Fusarium was the only dominant shared genera identi ed in all samples with different years of continuous cropping, while Parastagonospora, Sarocladium, Tetracladium, and Cysto lobasidium were the dominant shared genera after 4 years of cropping ( Figure  4).

Functional group succession during continuous cropping
The functions of the soil fungal community were classi ed into ecological guilds using FUNGuild 23 . There were 132 OTUs, which accounted for 19.7% of the total sequences that could be allocated into fungal functional groups using the FUNGuild annotation tool. Among assigned OTUs with function, saprotrophs were the dominant trophic type, followed by pathogens, and then symbiotrophs ( Figure S2). The relative abundances of saprotrophs, pathogens, and symbiotrophs all increased signi cantly with continuous cropping. Speci cally, saprotrophs increased from 7.73% ± 0.24% to 14.90% ± 1.10%, pathogens from 4.88% ± 0.33% to 15.47% ± 0.79%, and symbiotrophs from 1.83% ± 0.01% to 5.97% ± 0.44% ( Figure S2). In the pathogen guild, plant pathogens were most abundant and continuously increased from 2.76% ± 0.18% to 13.20% ± 1.94%. Furthermore, animal pathogens, fungal parasites, and lichen parasites all increased after continuous cropping ( Figure 5A). Endophytes and unde ned saprotrophs were the predominate trophic modes among symbiotrophs and saprotrophs, respectively. The detected guilds of endophytes and unde ned saprotrophs also showed increasing trends during continuous cropping ( Figure 5B and Figure 5C). We next focused on the taxonomy of plant pathogens and endophytes because of their high abundance and remarkable increases during continuous cropping ( Figure 6A and Figure 6B). Among plant pathogens, unclassi ed, Parastagonospora, Ustilago, and Verticillium were the dominant genera, and these increased signi cantly, especially Ustilago, which showed an increase in abundance of approximately 44-fold, from 1.26‰ ± 0.24‰ to 55.29‰ ± 10.27‰ ( Figure 6A). Moreover, the occurrence of Parastagonospora and Verticillium increased by approximately two-fold and three-fold, respectively ( Figure 6A). Among endophytes, Verticillium was the only genera that showed an average abundance higher than 1‰ of the total sequences, increasing from 2.91‰ ± 0.20‰ to 8.28‰ ± 0.96‰ ( Figure 6B).

Discussion
Continuous cropping in uenced fungal diversity and taxonomy in rhizosphere soil Continuous cropping in the same soil may lead to cumulative effects that could affect fungal diversity. In the present study, we found that continuous cropping of Tibet barley resulted in a remarkable change in soil fungal community diversity and richness. The Chao1 and PD indices decreased as the number of crop years increased (Table 1). Similar results have also been observed in rhizosphere soils of tea 24 , strawberry 16 , American ginseng 25 , and sweet potato 5 . Decreases in fungal diversity and richness have been recognized as an essential threat to the ecosystem, resulting in loss of soil function, which may have a negative effect on plant production during continuous cropping 26, 27 .
Ascomycota and Basidiomycota, which have been identi ed as the most abundant phyla in rhizosphere soils of Tibet barley ( Figure S1), are reportedly the main decomposers in agricultural waste composting 28,29 . At the genus level, the abundance of Fusarium decreased with increased cropping years, while the abundance of Ustilago increased tremendously after 6 years of continuous cropping ( Figure  2B). In soil, the genus Fusarium comprises many species of environmental, agricultural, and human health importance, but its notoriety mainly results from its pathogenicity toward a wide range of plants 30 .
Continuous cropping of soybean increased the relative abundance of Fusarium oxysporum in soil 6 , and Fusarium was linked to soil sickness during cucumber continuous cropping 19 . Moreover, Coptis chinensis monoculture was found to increase Fusarium, which, in turn, promoted root rot disease and limited crop yield 7 . Ustilago, which are potential plant pathogens found in soil, were found to increase after continuous soybean cropping 15 . The divergence between Fusarium and Ustilago may re ect their adaption to changed rhizosphere soil during continuous cropping of Tibet barley.

Effects of continuous cropping on functional groups of the rhizosphere fungal community
We speculated that continuous cropping of Tibet barley might select speci c functional groups as they adapt to root exudates and residues in the rhizosphere. Numerous studies have reported that continuous cropping led to the accumulation of organic acids and phenolic acids secreted by roots, promoting the growth of pathogenic microorganisms and further in uencing the structure of the rhizospheric soil microbial community 31 . Our results showed that the continuous cropping of Tibet Barley signi cantly in uenced fungal functions in rhizosphere soil. FUNGuild fungal functional prediction revealed that saprotrophs were the dominant trophic type in rhizospheric fungal communities of the Tibet barley plants in continuous cropping elds ( Figure S2), which was similar to the results of a study of Coptis chinensis 3 . These ndings could be attributed to the central roles of saprotrophs in organic decomposition during continuous cropping 23 . Generally, symbiotrophic fungi are extremely bene cial to the health, nutrition, and quality of most crops 32,33 . However, pathotrophic fungi usually obtain nutrients by attacking host cells; thus, they are considered to cause disease or have negative effects on plant performance 34 . In this study, the abundances of pathogens increased signi cantly with continuous cropping ( Figure S2). Among pathotrophs, plant pathogens increased with increased continuous cropping time ( Figure 5A). These results were consistent with the ndings reported by 17 , who stated that continuous cropping increased the abundance of pathogenic fungi in peanut soil. Similarly, the abundances of soil-borne pathogens were increased in continuous cropping elds of cucumber 35 , tomato 36 , and potato 37 . These ndings suggested that the environmental conditions under continuous cropping were likely to facilitate pathogen proliferation, and that Tibet barley plants might suffer a higher risk of fungal diseases with extended continuous cropping years. Speci cally, we found that the relative abundances of the plant pathogens Parastagonospora and Ustilago were signi cantly increased during continuous cropping of Tibet barley ( Figure 6A). Parastagonospora nodorum causes the wheat disease Septoria nodorum blotch and is responsible for substantial yield losses in many regions around the world 38 . As previously reported, corn smut diseases can be caused by Ustilago maydis infection 39 . Moreover, root-endophytic microbial communities have been shown to be linked to apple replant disease 40 . In this study, the occurrence of Verticillium, which was classi ed as an endophyte, also signi cantly increased with increased duration of continuous cropping of Tibet barley ( Figure 6B). Verticillium dahliae not only delayed, but also reduced wilt symptoms in cotton 41 . Thus, Verticillium should function as bene cial fungi for the roots of Tibet barley. However, the effects of rhizosphere pathogens and endophytes on plants have not yet been examined during continuous cropping of Tibet barley; accordingly, future studies should be conducted to elucidate the mechanisms through which their effects occur.

Conclusion
In summary, the dynamics of fungal community structure and function during 6 years of continuous monoculture of Tibet barley, an important highland crop, were examined. 18S rRNA gene high-throughput sequencing was applied to obtain the fungal sequences and provide detailed insight into the distribution patterns of fungal communities in Tibet barley rhizosphere soil during continuous cropping. We found that continuous cropping of Tibet barley markedly reduced the fungal richness and phylogenetic diversity. Moreover, the plant-pathogen fungal genera Parastagonospora and Ustilago were enriched during continuous cropping. However, Verticillium, which should bene t plant roots and reduce wilt symptoms, increased. In summary, our research helps ll gaps in knowledge regarding the response of fungal communities to continuous cropping of plants cultivated in areas with high altitude, but further research is needed to elucidate the mechanisms of changes in abundance of fungal pathogens and endophytes in rhizosphere soil of Tibet barley.

Study site description and soil sampling
The Tibet-barley eld was located at the Quankou Experimental Station, Menyuan Hui Autonomous County, Qinghai province, China (37°21′N, 101°44′E). The experimental eld has a mean annual precipitation of 530-560 mm, an average annual sunshine duration of 2,575 h, an average annual temperature of 1. Rhizosphere soils were collected for analysis as previously described 42 . Brie y, plants were carefully pulled from the ground, after which the conglutinating soil from the roots was mildly crushed and shaken to collect the soil located within the plant root surfaces. Quadruple soil and plant samples were collected from the four experimental elds. The rhizosphere soil was immediately transported on ice to the laboratory, where it was stored at -20°C until DNA extraction and chemical analysis.
Soil DNA extraction, PCR ampli cation, and Illumina MiSeq sequencing Community DNA was extracted using a MoBio Soil DNA isolation kit (MO BIO Laboratories Inc., USA) according to the manufacturer's protocols. The 18S rRNA gene was ampli ed using the fungal-speci c primer pair NS1 (5′-GTA GTC ATA TGC TTG TCT C-3′) and Fung (5′-ATT CCC CGT TAC CCG TTG-3′) 43 with a 12-base unique barcode attached to the forward primer. PCR reactions and amplicon puri cation were performed as described by Cheung et al. 44 . Finally, the puri ed products were quanti ed using a Qubit 3.0 (Thermo Fisher Scienti c, USA) and then sequenced on an Illumina MiSeq system at Sangon Biotech Co., Ltd. (Shanghai, China).

Processing of sequencing data
Quantitative Insights into Microbial Ecology (QIIME) 45 was applied to process the 18S rRNA gene sequences after paired-end fragments were merged by FLASH 46 . Brie y, the demux plugin (https://github.com/qiime2/q2-demux) was used to demultiplex and classify sequences into the different samples by unique barcode. The q2-feature-classi er plugin (https://github.com/qiime2/q2-dada2) was then used for quality control and chimeric sequences removal with the default settings. Next, UCLUST 47 was applied to cluster sequences into operational taxonomic units (OTUs) at 97% similarity. The classi er-sklearn plugins (https://github.com/qiime2/q2-feature-classi er) were used for taxonomic assignment of each sequence with the UNITE database 48 . The OTU table was rare ed to the minimum sample count (38,031 sequences for each sample) to reduce the effects of sequence numbers on diversity calculation.
The Shannon and Simpson's indices were calculated to represent alpha species diversity in the fungal community. Chao1 was used to describe the fungal species richness, and the PD index was used to determine the phylogenetic diversity 49,50 . To compare the fungal community beta diversity, Principal Coordinates Analysis (PCoA) was performed using the phyloseq package 51 and plotted with the ggplot2 package 52 .
Identi cation of fungal taxa shared in rhizosphere soil during continuous cropping Taxa shared across samples were obtained from the taxa-abundance matrix (OTU table) generated by the QIIME software at the genus level. Dominant and rare OTUs in sequence libraries and shared taxa were de ned at a threshold of 1% relative abundance 53 . Shared taxa in samples were visualized using the Pheatmap package in R 54 . For phylogenetic analysis of shared taxa at the genus level, representative sequences belonging to the OTUs of each shared taxon were extracted using QIIME and taxonomically classi ed after re-alignment with the UNITE database.
Predicting function of fungal community in rhizosphere soil FUNGuild (https://omictools.com/fungi-functional-guild-tool) was applied to taxonomically parse fungal OTUs based on their ecological functional characteristics 55 , and only guild con dence rankings assigned to"highly probable" and "probable" were kept. We classi ed three fungal functional groups according to trophic mode: symbiotrophy, saprotrophy, and pathotrophy. The OTUs that could not be identi ed into functional groups were categorized as unclassi ed and are not present in our results description. The relative abundance was estimated with the data of reads for each functional group by dividing the total reads assigned by FUNGuild. Fungal OTUs belonging to multiple functional groups were also treated as members of other groups. The taxonomic information describing OTUs assigned as functional guilds was classi ed after re-alignment with the UNITE database using the QIIME software.

Statistical analysis
Statistical analyses were conducted using SPSS 18.0 software (SPSS Inc., Chicago, IL, USA). Statistical signi cance was estimated by two-way analysis of variance followed by the least signi cant difference test to determine the signi cance of differences between groups. Differences were considered to be signi cant when the P-value was less than 0.05.

Nucleic acid sequences
The 18S rRNA gene sequences were deposited into the NCBI Sequence Read Archive with the accession number PRJNA669607.    Shared genera and their absolute abundance. Sequence abundances were log-transformed and colored from red-to-blue to indicate higher-to-lower abundance, respectively.

Figure 5
Dynamics of trophic models in pathotrophs (A), symbiotrophs (B), and saprotrophs (C) during continuous cropping of Tibetan barley. Values represent means ± standard deviation (n = 4). Different lowercase letters within the same column indicate signi cant differences among different continuous cropping durations at P < 0.05 according to two-way analysis of variance.

Figure 6
Taxonomic composition of fungi classi ed as plant pathogens (A) and endophytes (B) during continuous cropping. Values represent means ± standard deviation (n = 4). Different lowercase letters within the same column indicate signi cant differences among different continuous cropping durations at P < 0.05 according to two-way analysis of variance.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.