Discussion
Previous studies focused on the biotopes of soil, phyllosphere and faeces in a grazed steppe ecosystem respectively while the connections of microbiota among these biotopes, which played a role to carry out the energy flow in the ecosystem, were neglected. This study investigated the diversity, structure and the interaction of the microbial communities in these biotopes at the molecular level. It provided direct evidence to reveal the microbial distribution, migration and ecosystem function of the microbiota among these biotopes.
Soil, phyllosphere and faeces all maintained distinct microbiota and the type of biotopes explaining the majority of this microbial variation. As for the α-diversity, the soil fungal richness was higher contrasted to the fungal microbiota from the other biotopes (Table S1). This community diversity variation was likely due to the complicate and distinctive soil environment. Soil encompasses numerous microcosms to harbour distinct microbial communities rather than one single environment. Although these soil environments were limited in the only micrometer or millimeter scale, they provided niches of unique abiotic habitat and biotic communities to filter the microbial activities, diversities and compositions (Fierer, 2017). In addition, compared to soil bacteria, soil fungi were considered more adaptive to the vast different environments. Fungi were regarded as the mediators of slower carbon cycling because of their low growth rate and tolerance of low temperature and aridity (Rinnan & Baath, 2009; Rousk & Bååth, 2007; Sun, Li, Avera, Strahm, & Badgley, 2017). Furthermore, the great priority effect influenced on fungi and made it easier to survive in the extensive environments (Peay, Belisle, & Fukami, 2012). Meanwhile, fecal richness was of the lowest level in the fungal and bacterial community among all the biotopes (Table S1). The fecal microbial communities were considered similar in those residing to the livestock intestinal tract (Romero-Perez, Ominski, McAllister, & Krause, 2011). Obviously, the extreme environment of intestinal tract limited the gut and fecal fungal and bacterial richness via filtering the microbiota with pH, moisture and temperature stress and anaerobic environment. With respect to the β-diversity, all the microbiome also showed distinct microbial clusters among the three biotopes. This was due to the environmental filter driving the microbial community assembling in a dynamic ecosystem. Rothschild et al revealed that the environment dominated in shaping microbiome although it was only considered as a minor factor in the previous studies (Rothschild et al., 2018). The biotope preferences of the microbiota were associated with the environmental and ecological characteristics such as physicochemical properties and physiological capabilities (Bates et al., 2011; Jiao, Xu, Zhang, & Lu, 2019; Stams & Plugge, 2009).
Although all the three biotopes harboured distinct microbial communities from each other, the overlaps among these microbiotas revealed the microbial migration in the grazed steppe ecosystem (Fig. 2). The largest proportion of OTUs overlap occurred between soil and phyllosphere. These overlaps mainly contributed by bacteria and fungi suggested that the majority of phyllosphere bacteria and fungi originated from the soil. In another words, soil showed its potential as a microbial source reservoir of phyllophere (Zarraonaindia et al., 2015). This was consistent with the common view that phyllosphere microbiota was influenced by the microbiota of surrounding environmental source initially and then they were selected by the leaf taxa (Copeland, Yuan, Layeghifard, Wang, & Guttman, 2015; Lajoie, Maglione, & Kembel, 2020; Maignien, DeForce, Chafee, Eren, & Simmons, 2014). However, the specific environmental source influencing phyllosphere microbiota most strongly depends on the growing condition of the leaf species. Each leaf shaped a unique environment to thrive for specific microorganisms (Vacher et al., 2016; Vorholt, 2012). For instance, Copeland found the phyllosphere microbiota of several annual crops mainly originated from soil microbiota in the field condition (Copeland et al., 2015). Nevertheless, on the A. thaliana leaves, the phyllosphere microbiota mirrored airborne microbiota initially in the sterile soil (Maignien et al., 2014). A small proportion of fecal OTUs also occurred in phyllosphere while most of them were at low level of abundance. Phyllosphere microbiota was colonized on the leaf surface exposed to rapidly fluctuating temperature, moisture and relatively alternative humidity and nutrient resource (S. E. Lindow & Brandl, 2003). Thus, although the fecal microbiota originated from a specific environment with extreme stresses quite different from most of other biotopes, there were still some OTUs shared by these two biotopes since phyllosphere could offer a temporary microenvironment similar to the fecal microbiota. However, the drastic fluctuation of environment condition made phyllosphere an extreme and hostile habitat for the microbial community (S. E. Lindow & Brandl, 2003). For instance, a large majority of bacterial colonist had to face threat of being washed or killed by water, peroxide and UV light (Beattie & Lindow, 1995; Wilson & Lindow, 1994b). The availability of nutrient source such as glucose, fructose, and sucrose were also limited for the phyllosphere epiphytic colonization (Wilson & Lindow, 1994a). Hence, phyllosphere microbiota maintained at a low level of abundance because it had to tolerate such environment. Nevertheless, OTUs rarely occurred in both soil and faeces biotopes. Soil lacked of the drastic fluctuation of environmental condition like phyllosphere to fit the extreme environment condition of faeces. To put it another way, the rapid fluctuation of environment condition limited the growth and activity of phyllosphere microbiota, but it made phyllosphere a most extensively adaptive biotope for microbiota. Therefore, phyllosphere played the role of bridge to link the microbial communities among biotopes of soil, phyllosphere and faeces in such a grazed steppe ecosystem.
Referring to the community structure, the microbial migration among biotopes was predominated by specific taxa. ADONIS and ANOSIM described the increasing community similarity among three biotopes with the successive order of bacteria, fungi and archaea (Table 1). The similarity for bacteria was mainly contributed by the shared taxa between soil and phyllosphere. With high relative abundance, Actinobacteria and Proteobacteria occurred in both biotopes of soil and phyllosphere as the dominant phyla (Fig. 3). Whipps and Laforest‐Lapointe also found α‐ and γ‐Proteobacteria the most common classes within phyllosphere. However, the proximity of the relative abundance of these two phyla in phyllosphere varied from the low abundance of Actinobacteria within tree phyllosphere while the patterns of some Conifer species were consistent with our study (Laforest-Lapointe, Messier, & Kembel, 2016; Whipps et al., 2008). Actinobacteria were the most dominant bacteria in the soil biotope to decompose the plant litters and debris (Ventura et al., 2007). In addition, the genus of Streptomyces , which belonged to Actinobacteria, played role in inhibiting the fungal pathogens within rhizosphere. It was supposed to perform the similar function in the biotope of phyllosphere (Kunova et al., 2016). Nevertheless, the bacteria in the faeces biotopes showed significant discrepancy from the other two biotopes. The dominant phyla Firmicutes and Bacteroidetes occurred rarely in the soil and phyllosphere. The observed dominance of these phyla was due to the extreme environment in the intestinal tract of livestocks. Firmicutes were resistant to harsh environmental conditions, especially the desiccation stress, through their Gram-positive cell walls and spore-forming ability (Schimel, Balser, & Wallenstein, 2007; Van Horn et al., 2014). Bacteroidetes were the primary gateway to take advantage of glycans and carbonhydrate for the metabolism of the microbiota in the anaerobic and extreme environment in intestinal tracts (Faith, McNulty, Rey, & Gordon, 2011; Fischbach & Sonnenburg, 2011; Sonnenburg et al., 2010; Wu et al., 2011). In contrast to the diversity in bacteria domain, the archaea showed homogeneous within the three biotopes. Thaumarchaeota predominated in both soil and phyllosphere. This distribution was similar to the previous study in the two biotopes respectively (Fierer et al., 2012; Rinta-Kanto et al., 2016; Taffner, Cernava, Erlacher, & Berg, 2019). Thaumarchaeota were known as a novel phylum serving for the aerobic ammonia oxidization and the nitrification in archaea independent to Crenarchaeota and Euryarchaeota (Brochier-Armanet, Gribaldo, & Forterre, 2012; Jiao et al., 2019; Konneke et al., 2005). In the fecal archaea, Euryarchaeota replaced Thaumarchaeota as the dominant phylum. Methanobacteria and Methanomicrobia (belonged to Euryarchaeota) composed the primary assemblages. Both these classes were methanogenic archaea colonized by gut to reduce carbon dioxide into methane through present hydrogen in the intestinal tracts (Roccarina et al., 2010; van de Pol et al., 2017). Therefore, with the environmental filtering and functional selection, bacteria and archaea migrated between the biotopes of soil and phyllosphere, rather than faeces.
The only communal phyla migrating among all the three biotopes were Ascomycota and a small amount of Basidiomycota belonging to fungi. It indicated that these fungal phyla possessed robust environmental resistance and extensive environmental adaptability. In most of the previous study focusing on the fungal communities of soil, phyllosphere and faeces, Ascomycota all occupied the significant place of the dominant phyla order (Jia, Wang, Fan, & Chai, 2018; Perazzolli et al., 2014; Yelle, Ralph, Lu, & Hammel, 2008). In all the three biotopes of soil, phyllosphere and faeces, the Ascomycota behaviours were saprophytic and played a role in decomposing the various substrates such as plant and animal residues to transform them into nutrition for the direct usage of relative biota. Meanwhile, Nectriaceae, a parasitic and saprophytic family belonging to Ascomycota, generally occurred in the livestocks and monocotyledons. Nectriaceae were pivotal for maintaining the balance of the grazed grassland ecosystem because of their function of regulating the host genotypes and environments (Jia et al., 2018). Only a few sequences classified into Basidomycota were shared by all the three biotopes. Most of them belonged to Cryptococcus .Cryptococcusoccurred commonly in the animal to cause the systemic infection and its primary risk factor was HIV infection (Cafarchia et al., 2006; Idnurm et al., 2005). TheCryptococcusfrom animal faeces could be carried into other biotopes such as soil and phyllosphere with the niches of sufficiently utilizable carbon and nitrogen source (Nielsen, De Obaldia, & Heitman, 2007). However, whether the overlap of Cryptococcus could be applied to other terrestrial area was not clear since this overlap occurred rarely in taxa classification and this result may depend on the timing of the sampling. Thus, Ascomycota was the definitely phylum migrating in the ecosystem to link all the biotopes. Yet more types of livestock should be brought into the survey to confirm this result since only sheep were investigated in this study.
Physical environmental factors explained more variation of the microbial abundance among the biotopes while nitrogen was the most remarkable nutrient source for the microbial distribution. SEM (Fig. 6) showed that temperature explained the most of the impacts on fungi and bacteria abundance while pH was the only factor influenced fungi, bacteria and archaea. Mantel test (Table S4) also demonstrated that temperature influenced microbial diversity significantly. This discrepancy might be owing to the environmental sensitivity of the dominant species such as Actinobacteria, Zygomycota and Thaumarchaeota, so they were replaced by more extensively adaptive and stress-tolerant phyla such as Firmicutes, Ascomycota and Euryarchaeota in the intestinal tracts with harsh environmental conditions (Egidi et al., 2019; Jiao et al., 2019; Roccarina et al., 2010; Schimel et al., 2007; van de Pol et al., 2017; Van Horn et al., 2014). Nitrogen was also found to be the vital nutrition factor revealed by SEM (Fig. 6). This was mainly caused by Ascomycota, the communal and dominant phylum migrating among all the biotopes since it was highly associated with the nitrogen immobilization (Egidi et al., 2019). In addition, Fusarium oxysporum, Fusarium solani, Cylindrocarpon tonkinense and Gibberella fujiuroii, belonging to Ascomycota, all performed function of denitrification (Levy-Booth, Prescott, & Grayston, 2014).
Although the soil showed its potential of microbial reservoir for phyllosphere, only fungi migrated among all the biotopes of soil, phyllophere and faeces. Thus, the biotopes in an ecosystem were linked by the energy flow through the fungal community. De Vries et alreported the fungal-based process and the carbon and nitrogen cycle governed led weak resilience rather than strong resistance of grassland ecosystem (de Vries et al., 2012). Our network analysis (Fig. 4) confirmed it by exhibiting the significant discrepancy between the microbial community of grazed and ungrazed grassland after long-term moderate grazing. It was also noteworthy that fungi showed the most marginal variation in the co-occurrence networks. This was consistent with the slower response of fungi due to their lower growth rate and extensively adaptability in contrast to bacteria (Bardgett & van der Putten, 2014; Rinnan & Baath, 2009; Rousk & Bååth, 2007; Sun et al., 2017). Therefore, the fungal migration among biotopes made the ecosystem robust for the short-term but instable for the long-term.