Temporal and Spatial Dynamics of Dark Septate Endophytes in the Roots of Lycium ruthenicum in the Desert Region of Northwest China

: With the intensiﬁcation of desertiﬁcation in northwest China, drought has become a serious environmental problem restricting plant growth and ecological restoration. Recently, dark septate endophytes (DSEs) have attracted more attention because of their ability to improve plants’ resistance to drought. Here, we investigated DSE colonization and species diversity in roots of Lycium ruthenicum collected from Anxi and Minqin, in northwest China, during July, September, and December 2019. This study aimed to evaluate the inﬂuence of seasonality and sampling sites on DSEs. In different seasons, DSE colonization varied with the phenology of L. ruthenicum . At different sites, DSE colonization signiﬁcantly differed. Four isolates were reported in desert ecosystems for the ﬁrst time. The results showed microsclerotial colonization was directly affected by changing seasons, while hyphal colonization and species diversity were directly affected by sampling sites. The soil organic carbon, pH, alkaline phosphatase, and alkali-hydrolyzable nitrogen were the main predictors of DSE colonization and species diversity. We conclude that DSE colonization and diversity showed signiﬁcant spatial–temporal heterogeneity and were closely related to soil factors. This research provides a basis for the further understanding of the ecological functions of DSEs and their application potential for vegetative restoration and agricultural cultivation in drylands.


Introduction
Drought and desertification are rapidly intensifying in northwest China, where approximately 27% of the land is exposed to desertification [1]. Drought has become a serious environmental problem restricting plant growth and ecological restoration [2,3]. Therefore, it is particularly urgent to select appropriate plants with adaption to the arid environment. Fortunately, in order to survive in these conditions many plant species have evolved specialized ways to adapt to desert ecosystems [4]. Thus, special attention has been given to the utilization of these native plants for promoting ecological restoration in desert ecosystems [5]. Lycium ruthenicum, as an economically important traditional medicinal plant, belongs to Solanaceae and it is widely distributed in arid and semi-arid regions of northwest China as well as regions of Eastern Europe [6]. With the special characteristics of drought and salt resistance, L. ruthenicum is particularly well suited for increasing the nutrient and organic matter content of surface mineral soils in this arid ecosystem and may have the potential to improve soil conditions, facilitate ecosystem development, and delay further desertification. In addition to a direct drought resistance of plants, studies have shown that endophytic fungi may also affect the response of plants to drought stress, variation. The altitude, monthly average precipitation, and temperature of sampling sites are shown in Table S1 in Supplementary Materials. The studied soils were entisols and aridisols [32]. The selected plots are characterized by desert sands where L. ruthenicum is abundant. Psammophytic shrubs and subshrubs also occur at each plot.
Three sample plots were selected, and the distances between the plots were ≥10 km. Five replicate soil and L. ruthenicum root samples in each plot were randomly sampled from a depth of 0-30 cm. The distances between the plants that were sampled was ≥100 m. The samples were sealed in plastic bags and transported to the laboratory in an insulated container. Subsequently, all samples were sieved (<2 mm mesh) to remove litter and coarse roots. Fresh fine roots were extracted from each sample and were immediately processed for DSE morphological observation and DSE isolation. Soil samples for enzyme analyses were dried at 15-25 • C and stored at 4 • C until analyses. Other subsamples were air-dried and used for determination of soil physicochemical properties.

Soil Analysis
Soil pH was measured with a digital pH meter (PHS-3C, Shanghai Lida Instrument Factory, Shanghai, China) on a soil/water (1:2.5) suspension. Soil organic carbon (SOC) was measured by the combustion method [33]. The samples were heated in a muffle furnace (TMF-4-10 T, Shanghai Gemtop Scientific Instrument Corp., Shanghai, China) for 4 h at 550 • C. Alkali-hydrolyzable nitrogen (AN) was estimated using the alkaline hydrolysis diffusion method. The samples with sodium hydroxide and boric acid indicators were placed in a Petri dish and incubated at 40 • C for 24 h. Available phosphorus (AP) was determined by the chlorostannous reduced molybdophosphoric blue color method [34]. Soil alkaline phosphatase activity was determined by the method reported by Tarafdar and Marschner [35]. Soil urease activity was determined using the method of Hoffmann and Teicher [36], and urease activity was expressed as µg NH 4 + -N released from 1 g soil within 1 h [36].

Quantification of Fungal Colonization
Fresh roots were cut into 0.5 cm segments and cleaned using sterile water. The segments were placed in 10% (w/v) potassium hydroxide for 1 h at 100 • C and stained with 0.5% (w/v) acid fuchsin for 30 min at 90 • C [37]. For each sample, 30 segments were selected randomly and examined microscopically at 20× and 40× magnification [38]. The colonization rate of DSEs (%) (total, hyphal, and microsclerotial) and DSE colonization intensity were analyzed, according to Trouvelot et al. [39] as follows: Colonization rate (%) = (number of colonized root segments/total number of root segments) × 100%.

Isolation of Endophytic Fungi
For each plant individual, 15 randomly selected 0.5 cm fine root segments were disinfected with 75% ethanol for 5 min and 10% sodium hypochlorite for 2 min, and then washed with sterile water 3 times. The roots were placed in potato dextrose agar (PDA) culture medium supplemented with the antibiotics ampicillin and streptomycin sulphate and kept at 27 • C in the dark. In the meantime, water from the final rinse was plated in PDA as a negative control to ensure successful root surface sterilization. After incubation for 15 days, colonies with dark mycelium were isolated onto PDA to observe colony morphology and microscopic morphological characteristics [40].

DSE Species Diversity
The isolation rates (IRs) were calculated as the total number of plant segments infected by ≥1 type of fungus divided by the total number of incubated segments [42]. The isolation frequency (IF) was calculated as the number of strains of a fungus divided by the total number of strains isolated [43].
Diversity of DSEs was assessed using indices of the Shannon-Weaver index (H) [44]. Dominance was calculated with Simpson's dominance (D) measures [45]. An evenness index (J) was used for the determination of uniformity of the endophytic fungi [46]. The formulae were as follows: where the ratio "Pi" is the frequency of colonization of the taxon in the sample.
where "s" is the total number of fungi isolated.

Statistical Analysis
The DSE colonization of L. ruthenicum was assessed using one-way analysis of variance (ANOVA), and two-way analysis was performed to analyze the effects in different seasons and sampling sites on DSE colonization and soil variables; comparisons among means were performed using the least significant difference method (p < 0.05). Pearson correlation analysis was used to test the relationships among soil variables and DSE colonization and to evaluate the influence of sampling time, sample sites, and the interaction on DSE colonization. SPSS 21.0 software (SPSS Inc., Chicago, IL, USA) was used for all of the above analyses. Principal component analyses (PCA) of DSE colonization were assessed by Canoco 4.5 based on the principle of the cumulative contribution rate of variance being greater than 80% if correlation matrix eigenvalues exceeded 1. Structural equation model (SEM) was established by AMOS 21.0 to test the effects of environmental variables on DSE hyphal, microsclerotial colonization rate, and DSE diversity.

Spatial-Temporal Distribution of DSE Colonization
Typical DSE hyphae and microsclerotia structures were observed in the roots of L. ruthenicum, at Anxi and Minqin ( Figure 1). Brown to dark brown septate hyphae invaded the epidermal, cortical cells ( Figure 1A,G), or vascular tissue ( Figure 1C,I). Chainlike ( Figure 1E,K), tufted ( Figure 1D,F,L) and discrete ( Figure 1J) microsclerotia filled single cortical cells or colonized more than one cell.

Spatial-Temporal Distribution of DSE Colonization
Typical DSE hyphae and microsclerotia structures were observed in the roots of L. ruthenicum, at Anxi and Minqin ( Figure 1). Brown to dark brown septate hyphae invaded the epidermal, cortical cells ( Figure 1A,G), or vascular tissue ( Figure 1C,I). Chainlike ( Figure 1E,K), tufted ( Figure 1D,F,L) and discrete ( Figure 1J) microsclerotia filled single cortical cells or colonized more than one cell. With the change of seasons, the diameter of hyphae became thicker, the septum became shorter, and swollen hyphae formed at the end of the hyphae ( Figure 1A-C). The DSE colonization showed significant (p < 0.05) differences among seasons ( Figure 2). Total colonization and colonization intensity showed a V-shaped distribution, and the lowest value was in September. Hyphal colonization was July > December > September, and the highest values measured in July were 86.7% (at Anxi) and 54.4% (at Minqin). However, the microsclerotial colonization increased gradually with season changes, with the following order: July < September < December, and the highest values measured in December were 45.0% (at Anxi) and 34.4% (at Minqin). For different sampling sites, hyphal colonization, total colonization, and colonization intensity at Anxi were significantly (p < 0.05) higher than that at Minqin in July, while colonization intensity at Anxi were significantly (p < 0.05) lower than that at Minqin in September. However, microsclerotial colonization showed no significant differences among Anxi and Minqin in July, September, or December ( Figure 2). With the change of seasons, the diameter of hyphae became thicker, the septum became shorter, and swollen hyphae formed at the end of the hyphae ( Figure 1A-C). The DSE colonization showed significant (p < 0.05) differences among seasons ( Figure 2). Total colonization and colonization intensity showed a V-shaped distribution, and the lowest value was in September. Hyphal colonization was July > December > September, and the highest values measured in July were 86.7% (at Anxi) and 54.4% (at Minqin). However, the microsclerotial colonization increased gradually with season changes, with the following order: July < September < December, and the highest values measured in December were 45.0% (at Anxi) and 34.4% (at Minqin). For different sampling sites, hyphal colonization, total colonization, and colonization intensity at Anxi were significantly (p < 0.05) higher than that at Minqin in July, while colonization intensity at Anxi were significantly (p < 0.05) lower than that at Minqin in September. However, microsclerotial colonization showed no significant differences among Anxi and Minqin in July, September, or December ( Figure 2). Two-way ANOVA showed that DSE hyphal, microsclerotial, total colonization, and colonization intensity were significantly (p < 0.01) influenced by seasons. DSE hyphal, total colonization, and colonization intensity were significantly (p < 0.05) influenced by the sampling sites. DSE total colonization and colonization intensity were significantly (p < 0.05) influenced by the interaction of seasons and sampling sites (Table 1).  Two-way ANOVA showed that DSE hyphal, microsclerotial, total colonization, a colonization intensity were significantly (p < 0.01) influenced by seasons. DSE hyph total colonization, and colonization intensity were significantly (p < 0.05) influenced the sampling sites. DSE total colonization and colonization intensity were significantly < 0.05) influenced by the interaction of seasons and sampling sites (Table 1).

Morphological Characteristics and Identification of Endophytic Fungi
A total of 114 DSE strains were isolated from the roots of L. ruthenicum. The colon and microscopic morphologies of the DSEs isolated are illustrated in Figure S1. The D colonies were black, dark brown, or grey. Microscopic observation showed that all t DSE hyphae were dark and had septum. DSE HGQ-1 ( Figure S1A, HGQ-2 (Figure S1 Temporal and spatial changes of DSE colonization in L. ruthenicum roots. Significant (p < 0.05) differences among different months marked with different letters. Significant (p < 0.05) differences among different sites marked with * (* indicates a significant difference at p < 0.05 and ** indicates a significant difference at p < 0.01).

Spatial-Temporal Distribution of Soil Factors
Soil factors significantly (p < 0.05) varied among seasons and sampling sites. The lowest pH was recorded in July at both Anxi and Minqin and the lowest soil temperature was measured in December ( Figure 4). Soil urease and alkaline phosphatase were significantly (p < 0.05) higher in September than that in July and in December. Available Agronomy 2021, 11, 648 9 of 17 phosphorus was significantly (p < 0.05) lower in July than that in September and December. Soil moisture content in September was significantly (p < 0.05) higher than that in July and December at Anxi, while there was no significant difference at Minqin. For different sampling sites, SOC was significantly higher at Anxi than Minqin in July (p < 0.01) and December (p < 0.001). Alkali-hydrolyzable nitrogen was significantly higher at Anxi than Minqin in September (p < 0.05) and December (p < 0.01). Available phosphorus and soil temperature were significantly (p < 0.05) higher at Anxi than Minqin in July. Soil moisture contents were significantly (p < 0.05) higher at Anxi than at Minqin in each sampling season. Two-way ANOVA showed that all soil factors, except SOC, were significantly (p < 0.05) influenced by the seasons. SOC, available phosphorus, alkali-hydrolyzable nitrogen, alkaline phosphatase, and soil moisture contents were significantly (p < 0.05) correlated by the sampling sites. Soil moisture content, alkali-hydrolyzable nitrogen, and alkaline phosphatase were significantly (p < 0.05) influenced by interaction of seasons and sampling sites (Table S4). at Minqin in each sampling season while the evenness index was higher at Minqin than at Anxi ( Table 2 and Table S3).

Spatial-Temporal Distribution of Soil Factors
Soil factors significantly (p < 0.05) varied among seasons and sampling sites. The lowest pH was recorded in July at both Anxi and Minqin and the lowest soil temperature was measured in December ( Figure 4). Soil urease and alkaline phosphatase were significantly (p < 0.05) higher in September than that in July and in December. Available phosphorus was significantly (p < 0.05) lower in July than that in September and December. Soil moisture content in September was significantly (p < 0.05) higher than that in July and December at Anxi, while there was no significant difference at Minqin. For different sampling sites, SOC was significantly higher at Anxi than Minqin in July (p < 0.01) and December (p < 0.001). Alkali-hydrolyzable nitrogen was significantly higher at Anxi than Minqin in September (p < 0.05) and December (p < 0.01). Available phosphorus and soil temperature were significantly (p < 0.05) higher at Anxi than Minqin in July. Soil moisture contents were significantly (p < 0.05) higher at Anxi than at Minqin in each sampling season. Two-way ANOVA showed that all soil factors, except SOC, were significantly (p < 0.05) influenced by the seasons. SOC, available phosphorus, alkali-hydrolyzable nitrogen, alkaline phosphatase, and soil moisture contents were significantly (p < 0.05) correlated by the sampling sites. Soil moisture content, alkali-hydrolyzable nitrogen, and alkaline phosphatase were significantly (p < 0.05) influenced by interaction of seasons and sampling sites (Table S4). . Spatial and temporal distribution of soil factors of L. ruthenicum. p < 0.05, significant differences among different mouths marked with different letters; p < 0.05, significant differences among sampling sites marked with *; * indicates a significant difference at P < 0.05, ** indicates a significant difference at p < 0.01, and *** indicates a significant difference at P < 0.001. Figure 4. Spatial and temporal distribution of soil factors of L. ruthenicum. p < 0.05, significant differences among different mouths marked with different letters; p < 0.05, significant differences among sampling sites marked with *; * indicates a significant difference at P < 0.05, ** indicates a significant difference at p < 0.01, and *** indicates a significant difference at P < 0.001.

Principal Component Analysis of DSE Colonization
The PCA results for DSE colonization ( Figure 5) showed that the cumulative contribution of the two principal components (PCs) related to DSE colonization reached 91.0% (PC1 and PC2). Along the PC2 axis, the Anxi plots in July and December occurred on the right, and the plot in September occurred on the left. Along the PC1 axis, the Minqin plots in September and December were below the axis, and the plot in July was above the axis. The result indicated DSE colonization was significantly affected by seasons and sampling sites.

Principal Component Analysis of DSE Colonization
The PCA results for DSE colonization ( Figure 5) showed that the cumulative contribution of the two principal components (PCs) related to DSE colonization reached 91.0% (PC1 and PC2). Along the PC2 axis, the Anxi plots in July and December occurred on the right, and the plot in September occurred on the left. Along the PC1 axis, the Minqin plots in September and December were below the axis, and the plot in July was above the axis. The result indicated DSE colonization was significantly affected by seasons and sampling sites.

DSE Colonisation Status
DSEs, as root-associated fungi, have a wide ecological distribution and high colonization in a variety of desert plants [13,18,21,47]. In the present study, we found the

DSE Colonisation Status
DSEs, as root-associated fungi, have a wide ecological distribution and high colonization in a variety of desert plants [13,18,21,47]. In the present study, we found the roots of L. ruthenicum to be infected by typical dark septate hyphae and microsclerotia structures, which suggests that DSEs might be vital components of the roots of L. ruthenicum in desert habitats. The DSE morphology and colonization varied with seasons and sampling sites, which may be an effective adaptive adjustment of L. ruthenicum to adversity environments [20]. With a change of seasons, the morphology of DSE hyphae became correspondingly thicker and shorter, which was consistent with Zuo et al. [21]. Hyphae, being a channel for nutrient exchange between DSEs and plants, might protect the host against drought stress through a more effective absorption and translocation of limited nutrients that might otherwise not be accessible to the roots of the plant [48,49]. Since the sampling sites are desert regions with low rainfall, any additional feature that can improve nutrient status is a necessity. In this study, the PCA results revealed that DSE colonizations were correlated to the seasons. The hyphal colonization reached the highest in summer while the microsclerotinia colonization reached the maximum in winter, results that are consistent with the results of Rayment et al. [30]. Several studies have indicated that DSEs can promote plant biomass, improve root growth and development, and increase the number of root hairs, especially in arid environments [16,20,50]. L. ruthenicum grow actively in July and need abundant nutrients, while the DSE hyphae may promote the absorption and utilization of nutrients for host plants. DSE microsclerotia may play important roles in substrate storage, and act as a protective device for host plants to withstand severe conditions [51]. In December, with a reduction in temperature, the leaves and stems of L. ruthenicum gradually wither, and thus an increased number of microsclerotia may help host plants to resist the increasingly challenging environment [26]. In general, the changes of DSE colonization were consistent with the phenology of host plants. Moreover, except for the direct effect of season on microsclerotia colonization, season influences hyphal and microsclerotial colonization by adjusting soil factors such as pH, available phosphorus, soil temperature, and other abiotic factors, which may also lead to seasonal heterogeneity of DSE distribution [30].
For different sampling sites, the DSE hyphal colonization, total colonization, and colonization intensity showed obvious spatial heterogeneity, which was consistent with the results of Xie et al. [19], who reported that DSE infection was spatially predictable. In this work, SEM indicated that the hyphal colonization was directly affected by sampling sites. U'Ren et al. [52] declared climatic factors could drive fungi special heterogeneity. Nevertheless, in terms of climatic conditions, Anxi and Minqin were similar, both typically arid with low precipitation. Thus, we speculate that other factors can drive fungal spatial heterogeneity, such as soil characteristics. Li et al. [18] investigated the colonization of Ammopiptanthus mongolicus in three sampling sites in desert and concluded that the dynamics of DSEs had a highly spatial pattern, and were influenced by soil nutrient availability, which supported our results. Our results showed that sampling sites indirectly affect DSE microsclerotial colonization by influencing SOC, which is known as a vital factor in promoting the colonization of endophytic fungi [53,54]. In addition, soil moisture content is another trigger that causes increases in fungi colonization [55]. In this study, the soil moisture content, which was mainly the result of the unpredictable rainfall events, had positive effects on DSE colonization. The SOC and soil moisture content was higher at Anxi than at Minqin. Thus, relatively abundant soil moisture content and SOC can lead to the colonization differences among sampling sites.

DSE Species Diversity
In the present study, DSE strains were subdivided into 13 genera and 14 species based on their morphological and molecular identification. To our best knowledge, A. chartarum, C. olivacea, S. cylindrosporus, and T. verrucosus were reported here in desert ecosystems for the first time. The presence of these DSEs may be related to the arid environment and specific host plants. DSEs are generally believed to improve the ability of plants to cope with environmental stress. Meanwhile, previous studies have shown that endophytic fungal communities may have host specificity [56]. Nevertheless, the mechanism of interaction between DSEs and host plants is not clear [56,57]. In further studies we should pay more attention to the ecological functions of these DSEs and their relationship with host plants.
A. chlamydospore was the dominant isolate of L. ruthenicum, which was frequently detected among different seasons and sites. Alternaria sp. has been frequently identified in the roots of desert plants and seemingly has no host specificity [13,[58][59][60]. A recent study has shown that Alternaria sp. can promote the growth of Hedysarum scoparium [21]. H. striatispora, E. pedicellatum, Fusarium sp., and Microascus sp. were previously isolated in arid or semiarid environments [13,21,61]. Interestingly, M. alveolaris is not only the dominant species but also an endemic species at Minqin. The results of our study suggested that DSEs are generalists of host plants in desert regions, where they harbor abundant fungal resources.
The DSE species composition inhabiting the roots of L. ruthenicum were shown to differ geographically, and there was no overlapping or crossing strains for dominant isolates and endemic isolates at the Anxi and Minqin sites. The DSE isolation rate and species diversity were higher at Anxi than at Minqin, which suggests that environmental selection and habitat heterogeneity may mainly account for the difference [62,63]. In our research, SEM revealed that the species diversity was directly related to sampling sites. Previously, Knapp et al. [12] suggested that endophytic fungi composition may change with locations of host plants, and this postulate has been corroborated by other relevant studies [19,21,60,64]. Moreover, our results showed that sampling sites also affect the species diversity by influencing available phosphorus and alkaline hydrolyzable nitrogen. Another study found that soil phosphorus and nitrogen content was the key factor impacting the abundance and diversity of fungi [63]. When the carbon supply in the substrate is sufficient, phosphorus, nitrogen, and other nutrient elements become the limiting factors for microbial biomass [65]. Under low-phosphorus stress, plants can release insoluble soil phosphorus [66] for microbial to utilize, which ultimately changes the physical and chemical properties of the soil by improving nutrition. Changes in these properties further affect the fungi species composition and diversity. In addition, our results showed that sampling sites indirectly affect species diversity by influencing hyphae colonization. In terms of DSE colonization, the colonization of DSEs was significantly higher at Anxi than at Minqin; thus, this may be one of the explanations for the higher species diversity at Anxi than at Minqin.
Martins et al. [67] reported that the diversity and richness of endophytic fungi isolated from olive trees in spring were higher than in autumn, and the diversity of endophytic fungi was affected by seasonal variation. However, in our study, despite seasonal differences in species diversity, the species diversity was not directly correlated to the seasons, which suggests that other factors could drive fungal temporal heterogeneity. We speculate the variation of DSE species compositions over time might be due to changes in the chemistry of plant tissues in the course of the phenological growth stages of the tree and to the interspecific competition among the fungi [68]. Thus, more research is required to confirm the mechanism by which interactive associations form between the DSEs and host plants.

DSE and Soil Factors
Root-associated fungi play a pivotal role in nutrient exchange between plants and soil [69] and can alter the physical and chemical properties of the soil [70]. Numerous studies have confirmed that there is a significant relationship between soil nutrient and fungal infection [13,19,21]. Here, SOC was positively correlated with DSE microsclerotial colonization. Several studies have identified soil carbon sources as key ecological drivers of microbial community dynamics [71]. SOC provides the dominant carbon metabolic sources for endophytic fungi [72]. Abundant organic carbon can promote DSE colonization and can even improve the viability of host plants [53,54]. It has been well recorded that pH is an important factor that strongly affects colonization of soil fungi [73,74]. The optimal pH of most fungi are in the range of 5.5-7.5. With an increase in soil pH, absorption ability of host plants is limited [75,76]. Favorable changes in soil pH are likely to improve the nutritional status and health of the host by influencing fungal distribution [77]. Correlation analysis showed that DSE hyphal colonization was negatively correlated with soil pH, which was consistent with several studies about Solanaceae [78,79]. In addition, our data suggest that soil alkali-hydrolyzable nitrogen has a negative correlation with DSE microsclerotia, in part supporting the finding that the concentration and content of root N are related to infection by DSE structures [15]. Moreover, alkali-hydrolyzable nitrogen was also negatively correlated with species diversity. Li et al. [20] determined plants infected with DSE fungi increased N absorption. Although DSEs cannot improve uptake of N through direct transfer, they can allow increased access to a host plant [80]. For growth, host plants need N to provide nutrients [81], which is mainly acquired from soil substance. In the case of insufficient soil N source, host plants likely need more endophytic fungi to assist absorption of nutrients. In general, DSE colonization and composition may be affected by host plant identity, which is a substrate for the survival of fungi [56]. Meanwhile, environmental factors may also lead to spatial and temporal heterogeneity of DSEs [31,82,83]. Our research showed that changing seasons and sampling sites can directly impact DSE colonization and species composition and can indirectly influence DSEs by adjusting soil microenvironments.
Our results demonstrated that desert plants can be colonized by DSEs, which may have the potential to promote the growth of desert plants. This finding provides a new direction for the DSE fungi application to cultivation of L. ruthenicum. Meanwhile, agricultural ecological environments are facing drought stress. Some studies have shown that DSEs can promote plant growth under drought stress [8]. Therefore, it is crucial to understand the spatial and temporal distribution of DSEs in L. ruthenicum, which could provide the basis for the application of DSEs in agricultural drylands.

Conclusions
Overall, root-fungal associations were found between L. ruthenicum and DSEs in desert regions of northwest China. The DSE colonization and species diversity showed significant spatial and temporal heterogeneity and was closely related to soil factors. Among the identified DSE species, A. chartarum, C. olivacea, S. cylindrosporus, and T. verrucosus were reported here for the first time in a desert environment. The morphology and colonization of DSEs might be useful indicators for evaluating soil quality and function of desert ecosystems. Future research should investigate the function of DSE associations in L. ruthenicum to improve the understanding of the role of DSE fungi in desert ecosystems. Therefore, a better understanding of the complex interactions between L. ruthenicum and DSEs would help to enhance ecological restoration in desert ecosystems and for practical application in the agriculture cultivation of L. ruthenicum in northwest China.  Table S1: Environmental condition in the sites in Gansu, China, Table S2: DSE isolates from medicinal plant roots and results of sequence Blast from type material database, Table S3: DSE species diversity of L. ruthenicum, Table S4: Two-way analysis of variance of the effects in different months and sites on soil factors, Table S5: Correlation relationship among DSE colonization and soil factors.