Endogenous fungi isolated from three locoweed species from rangeland in western China

Leguminous locoweeds cause toxicosis to grazing animals in western China and western USA. Swainsonine, a toxic alkaloid, is produced by the endophytic fungus Alternaria section Undifilum sp. living within the locoweed plants. Nothing is known of the other endogenous fungi associated with locoweed and it is unknown if the presence of Alternaria sect. Undifilum sp., a potential mutualist, in a locoweed influences the fungal microbiome associated with the plant. To help address these questions, endogenous fungi associated with three locoweed species (Oxytropis glabra, Sphaerophysa salsula, and Astragalus variabilis) collected from grasslands from western China were evaluated. Fungi were isolated from the tissues and identified by morphological features and sequencing of the internal transcribed spacer (ITS) regions. A total of 1209 fungal isolates were obtained from 1819 tissues for an isolation rate of 66.5%. Alternaria sect. Undifilum oxytropis, Alternaria spp. and Fusarium spp. were most commonly isolated. Plant host species, plant part, and environment influenced the endogenous fungal communities isolated from the locoweed plants. There were significant differences in the diversity of fungal species isolated from O. glabra from two sites, and no differences between the diversity of fungi isolated from A. variabilis from two sites. Alternaria sect. Undifilum was found most frequently associated with toxic locoweeds. Plants or plant parts that did not yield this endophyte had more plant pathogenic fungi associated with them. This is the first report of the diversity of fungi associated with these locoweeds and the first to suggest a beneficial role for Undifilum.


INTRODUCTION
Locoweeds are perennial, herbaceous, poisonous legume plants containing the fungal-produced toxic alkaloid, swainsonine, which causes the neurological disorder locoism in grazing animals (Dorling et al., 1980; *Corresponding author.Email: creamer@nmsu.eduTel: 15756463068. Author(s) agree that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License James et al.,1981;Allred, 1991;Cook et al., 2009b).Locoweeds are primarily species of Astragalus and Oxytropis genera and have been reported from the Americas (western USA and South America) and Asia (Allred, 1991;Robles et al., 2000;Yu et al., 2010).Swainsonine is an indolizidine alkaloid that was first identified from Swainsona, a toxic legume that is found in Australia (Colegate et al., 1979).Swainsonine inhibits alpha-mannosidase and mannosidase II causing a lysosomal storage disease and neurologic impairment (Dorling et al., 1980).Locoism symptoms include a lack of muscular coordination and inability to eat or drink (James et al., 1970;Stegelmeier et al., 1999).Losses due to locoism depend on the severity of poisoning, but have been estimated at $300 million annually in the western United States due to poisoned cattle, sheep, and horses (Torell et al., 2000;Turner et al., 2012).
A large diversity of locoweeds is found in China.There are 46 species of locoweeds reported in China, including 23 species of Oxytropis, and 23 species of Astragalus (Lu et al., 2012b).Thirteen of the 46 species have been reported to cause severe damage to animal husbandry.Toxic locoweeds are primarily distributed over an area of 11 million ha in arid and semi-arid grasslands of Tibet, Inner Mongolia, Qinghai, Gansu, Xinjiang, Ningxia, Shanxi, and Sichuan provinces (Zhao et al., 2003;Lu et al., 2012b).
A fungal endophyte is defined as a fungus that lives within a plant.Fungal endophytes include saprophytic fungi, pathogenic fungi, and mutualistic fungi (Petrini, 1991).Fungal endophytes are widely found in roots, stems, and leaves of higher and lower plants.The diversity of fungal endophytes includes species diversity in colonization of host plants and diversity in the distribution of fungal endophytes within different tissues of host plants.This diversity could be due to differences in the host plants and natural factors, such as geographical location and climate, in addition to presence of other organisms (Tiina et al., 2013).
Different locoweed species have been associated with different Alternaria sect.Undifilum sp.Oxytropis lambertii and O. sericea harbor A. U. oxytropis, Astragalus mollissimus contains A. U. cinerea, and Astragalus lentiginosus contains A. U. fulva (Pryor et al., 2009;Baucom et al., 2012).Alternaria sect.Undifilum spp.can readily be cultured from dried plant tissue and most produce conidia in culture.The cultures are very slow growing, expanding less than an average of 0.2 mm/day (Pryor et al., 2009).They can be readily isolated from leaves, petioles, stems, flowers and seed (Braun et al., 2003).The fungi are vertically transmitted through seed, but found only in the seed coat and underlying layer (Oldrup et al., 2010).
Alternaria sect.Undifilum spp.can usually be cultured from locoweed samples with swainsonine concentrations greater than 0.01% swainsonine, and not from plants with lower swainsonine levels (Ralphs et al., 2008).While the endophyte has been found in all plant parts, it is highest in the crown, and has been difficult to isolate from roots (Cook et al., 2009a;Cook et al., 2011).Endophyte amounts change seasonally in above ground parts of plants until the plants reach maturity (Achata et al., 2012;Cook et al., 2012).Gao et al. (2012) cultured Alternaria sect.Undifilum oxytropis from Astragalus variabilis and Oxytropis glabra plants collected from Inner Mongolia that contained swainsonine concentrations greater than 0.01%.Swainsonine was detected in only 2 of 50 samples of Sphaerophysa salsula, and at concentrations less than 0.01% and A. U. oxytropis was not cultured from any of the samples.
While the role of Undifilum sp. in swainsonine production has been clearly established, the ecological role of A.U. oxytropis is not quite so clear.The fungus does not cause any obvious disease symptoms on the plant hosts and the plant host does not show a resistance response toward the fungus even at the cellular level (Reyna et al., 2012).The fungus does not deter invertebrate feeding, or aid in plant growth under heat stress or nitrogen deficiency (Thompson et al., 1995;Oldrup et al., 2010).Does Alternaria sect.Undifilum provide benefit to the locoweed plant, or does it function primarily as a commensal (Creamer and Baucom, 2013)?Perhaps it could help deter disease from pathogenic fungi.An ecological role for Alternaria sect.Undifilum could be suggested if it influenced the fungi associated with its host plant.
Highly stringent isolation procedures developed to culture the slow-growing Alternaria sect.Undifilum preclude the growth of other fungi (Braun et al., 2003).As a result, there is little known about the fungi, other than Alternaria sect.Undifilum sp., that is associated with locoweeds.Therefore, using lower stringency culture techniques, fungi were isolated from two toxic locoweeds, O. glabra, A. variabilis, and a rarely-toxic locoweed, S. salsula, that were collected from semi-arid meadows in Ningxia Hui Autonomous Region and Inner Mongolia Autonomous Region in north central China.Fungi were identified based on morphological characteristics and phylogenetic analyses of rDNA ITS sequences and the diversity of the fungi were compared between plant species, locations, and plant parts.This study is the first to identify the diverse fungi associated with locoweeds and the first to suggest a beneficial role for Alternaria sect.Undifilum.

Sample collection
Samples were collected from four areas, Yinchuan city in Ningxia Hui Autonomous Region and BayanHot, AoLun Prague, and Jilantai towns, in the Inner Mongolia Autonomous Region, which were selected to be representative of different environments (Table 1).O. glabra was collected from Yichuan city and Bayan Hot, A. variabilis was collected from AoLun Prague and Jilantai, and S. salsula was collected from BayanHot.One whole healthy-looking locoweed plant was collected from each of three sites within each of the four areas.At sampling, the type of vegetation, landscape, habitat, altitude, latitude, and climate were recorded and the location determined by GPS (Global Position System) (Table 1).The samples were dried with allochroic silica gel and maintained at-20°C.

Isolation of fungal endophytes
The dried plants were washed thoroughly with running tap water, then twice more with double distilled water to remove debris.The plants were cut into separate parts, i.e. leaves and petioles, roots, and seed pods were removed.Stems, leaves, and seeds of O. glabra, A. variabilis and S. salsula were assessed.Roots from A. variabilis, and petioles from S. salsula were also tested.Similar tissues from plants of the same species were combined, that is, O. glabra leaves harvested from the three plants collected from BayanHot were pooled for testing to maximize diversity from each plant species and location.The tissues were treated with a minimal cursory treatment of 30 s in 75% ethanol, followed by 1-3 min in 2%NaClO, and then 3-5 times in sterile water.Tissues were dried on sterile paper towels, and stems, petioles, and roots were cut into 5 mm sections, leaves were cut into 5 mm x 5 mm sections, and seeds were quartered.The tissues (6 to 7 sections) were placed in the same potato dextrose agar (PDA, Qingdao High-tech Industrial Park Haibo Biotechnology Co., Ltd, Qingdao, Shandong, China) plate at 28°C for 7 to 30 days.To determine if loosely associated microbes had been removed from the plant tissue surface, tissue prints were done with sterile water on control media.No bacterial contamination was observed, suggesting that cursory surface sterilization was effective.When hyphae grew from the cut tissues, the hyphae were removed to new media, and transferred 2 to 3 times.The hyphae were harvested and stored at 4°C.

Morphological identification of fungal endophytes
Hyphal tips of all fungi were plated onto PDA, and cultured in an Lu et al. 157 inverted position at C. Colonies were observed daily and size, color, texture, growth rate, edge, and shape were recorded.Subsequently, a third of a coverslip was inserted into the edge of the colonies in the medium at a 45° angle, and fungal mycelium, conidiophore structure, spore morphology, color, and ontology, were observed using light microscopy.Primary identification was carried out based on the taxonomic guides of Barnett and Hunter (1977), Shao et al. (1996), and Wei (1979).Ecological roles for fungi were determined using the USDA Fungal Databases from the Systematic Mycology and Microbiology Laboratory (http://nt.arsgrin.gov/fungaldatabases).Fungi were classified as pathogens, mutualists, or saprophytes based on the majority of reports for the fungus.Fungi not identified to species were not classified.

Identification of rDNA ITS sequences
Hyphae grown on PDA were collected and weighed.Genomic DNA was extracted from the fungi, with cetyltrimethylammonium bromide (CTAB) lysis buffer (Sun et al. 2006).Universal primers that amplify fungal rDNA, ITS1 (5'-TCCGTAGGTGAACCTGCGC-3') and ITS4 (5'-TCCTCCGCTTATTGATATGC-3'), were used as upstream and downstream primers (White et al., 1990;Deng et al., 2006) to amplified the conserved ribosomal rDNA-ITS sequences using the genomic DNA as template.PCR reactions were carried out in a volume of 25 μL, containing 3 μL of DNA, 10 μmol•L -1 of ITS1 and ITS4 (1 µL each), 12 µL of 2X EasyTaq PCR SuperMix (TransBionovo, Beijing) and 8 µL of ddH2O.Amplification was carried out in a Bio-RAD Gene Cycler TM PCR with an initial denaturation at 95°C for 30 min, 35 cycles of 95°C for 30 s, 55°C for 30 s, 72°C for 45 s, and a final extension of 72°C for 10 min.Amplification products (5 μL) were separated by electrophoresis on a 1.2% low-melting agarose gel, and the bands were visualized using a gel imaging system.Selected DNA fragments were purified from the agarose gel using a DNA extraction kit and sequenced (Sangon Biotech Co., Ltd., Shanghai).Species identification was based on sequence analysis of approximately 500 bp from the nuclear ribosomal internal transcribed spacer region (rDNA-ITS), (Arnold and Lutzoni 2007).

Phylogenetic relationship between different fungi
After sequencing, homology comparison was done using Blast and 5.8S rDNA-ITS sequence in Gen Bank, then the selected high similarity sequences were used to produce a phylogenetic tree by applying ClustalX 1.83 version and Neighbor-Joining method (MEGA5.0)with bootstrap support based on 1,000 replicates, and evolutionary distances computed using the Maximum Composite Likelihood method.

Data analysis
The isolated fungal communities were characterized for diversity by plant and by site.The -diversity of the fungal community associated with locoweeds was analyzed using Shannon-Wiener diversity index (H' = -×ln Pi) and Simpson index (D =- 2 ) (k= the total species of fungi in certain plants, Pi= the proportion of a specific fungus out of all isolated fungi) (Gazis and Chaverri, 2010;Kharwar et al., 2011).The Evenness index (J=H/lnS) was used to estimate the evenness of species distribution  (Kharwar et al., 2011).
The -diversity of the fungi associated with locoweed species and sites was determined using two similarity indices, Sorenson Index (CS = 2/(a+b)) and Jaccard Index (CJ =j/(a+b-j)), which were used to compare the similarity of fungal species composition between the two sites (j= the common species of fungi between the two sites, a and b= the species of fungi in the two sites, respectively) (Gazis

Isolation and identification of fungi associated with locoweeds
Fungi were isolated from all tissues using all combinations of bleach and ethanol treatments tested.Generally, fewer fungi were isolated from toxic locoweeds with increasing times of bleach and ethanol.Although only tested for A. variabilis, roots yielded the highest fungal isolation rates, 87%.Isolation rates from the nontoxic S. salsula were generally lower and more variable, ranging from a high of 41.6% from leaves to a low of 3.9% from seed.One hundred sixty-eight fungal isolates were obtained from 211 tissues (leaves, stems, seeds) from O. glabra from BayanHot with an isolation rate of 79.6%, and 417 fungal isolates were obtained from 642 tissues from O. glabra from Yinchuan with an isolation rate of 65.0%.Fungal isolation rates from BayanHot and Yinchuan were 87.8 and 69.0% from seeds, 69.4 and 67.5% from leaves, and 67.3 and 55.2% from stems, respectively (Table 2).The fungi isolated from O. glabra differed between the two areas.Only five fungal species were isolated from plants from BayanHot, while 21 species were isolated from plants from Yinchuan.
Two hundred eighty five fungal isolates were obtained from 397 tissues (leaves, stems, seeds, roots) from the A. variabilis from Jilantai.The isolation rate was 71.8%.Two hundred forty five fungal isolates were obtained from 354 tissues of A. variabilis from AoLun Prague.The isolation rate was 69.2%, with higher isolation rates from stems and seeds than from leaves or roots from both locations (Table 2).The number of fungi isolated from A. variabilis was somewhat similar between the two areas.Fifteen fungi were isolated from plants from Jilantai and 19 fungi were isolated from plants from AoLun Prague.
Ninety-four fungal isolates were obtained from 215 tissues of S. salsula (leaves, petioles, stems, and seeds) from BayanHot.The isolation rate was 43.7%, from leaves, 53.1% from seeds and 50.9% from stems (Table 2).The 94 cultures of fungi that were isolated included 19 species, distributed in 4 classes, 5 orders, 8 families, (2 families undetermined) and 10 genera on the basis of morphological characteristics and rDNA-ITS sequence analyses (Figure 1).
Overall there were 10 orders of fungi isolated, with the largest number of fungi within the Pleosporales and Hypocreales (Table 3).There were 9 species of Alternaria, 7 species of Fusarium, and 4 species of Aspergillus isolated.Alternaria porri was the only fungus isolated from all locations, while Alternaria sect.U. oxytropis, Alternaria alternata, and Fusarium tricinctum were each isolated from 4 locations.

Community composition and diversity of isolated fungi
The distribution of fungi isolated from O. glabra differed between BayanHot town and Yinchuan city.Only 5 fungal species were isolated from O. glabra from BayanHot, while 21 were isolated from the plants collected from Yinchuan.The endophyte, A. U. oxytropis, was distributed in all tissues of O. glabra from both locations except for the leaves in Yinchuan city.Relative isolation rates of A. U. oxytropis were all higher than 65.0% in O. glabra from BayanHot, with a relative isolation rates in seeds, leaves, and stems of 80.6, 80.0 and 68.6%, respectively.In Yinchuan, however, A. U. oxytropis was isolated only from the seeds and stems of O. glabra, and the relative isolation rates were 58.9 and 44.8%, respectively.Additionally, Alternaria spp., Aspergillus spp., and Fusarium spp.were isolated from most tissues of O. glabra, from both locations, and relative isolation rates were also higher (Table 4).The -diversity (number of unrelated species of fungi isolated) for O. glabra (Figure 3) and A. variabilis (Figure 2) was similar and lower than for S. salsula (Figure 1).
Alternaria sect.Undifilum spp.were distributed in all the tissues of A. variabilis collected from both locations and was the dominant species in most tissues.The total relative isolation rates of Alternaria sect.Undifilum sp. from A. variabilis in Jilantai were 46.2, 60.0, 58.7 and 33.9% from leaves, stems, seeds, and roots, respectively, and in AoLun Prague was 62.7, 38.7, 68.6 and 20.0%, from leaves, stems, seeds, and roots, respectively.Hence, A. U. oxytropis was the dominant fungal species isolated from A. variabilis and O. glabra.Alternaria sect.Undifilum gansuense, instead of A. U. oxytropis, was isolated from roots of A. variabilis collected from AoLun Prague, and the dominant species was Alternaria alternata, which was similar to that isolated from leaves of O. glabra in Yinchuan (Table 4, Figure 2).
For the four tissues of S. salsula, Alternaria spp.were dominant in the stems, leaves, and seeds, while Fusarium chlamydosporum was dominant in petioles.The relative isolation rates of Alternaria spp.from stems, leaves, and seeds were 65.2, 46.2, and 44.8%, respectively, while the rate for F. chlamydosporum from petioles was 25.0% (Table 4).

Fungal diversity indices
According to the Shannon-Wiener index for the fungi isolated from the 3 locoweeds in the 4 ecological communities, the diversity of fungi varied among plants and locations (Table 5).The diversity of fungi from O. glabra in BayanHot was quite low (1.0) compared to that of O. glabra in Yinchuan (2.2).Enriched diversity of fungi from A variabilis was the same (2.0-2.1) in both sampling sites.The Shannon-Wiener index of fungi from the three tissues sampled from all plants (leaves, stems, seeds) for S. salsula was the highest at 2.5, followed by A. variabilis at 2.0-2.1, and O. glabra at 1.0-2.2.
The Simpson index (range 0 to 1) reflects the dominance in a community, so increases as diversity decreases.The Simpson index values for fungi isolated from locoweed tissues gave similar results as those calculated for the Shannon-Wiener index.The Simpson index of S. salsula was the lowest, and that for O. glabra was the highest, reflecting the highest and lowest diversity, respectively.
Evenness reflects the uniformity of distribution of different species in the community.The evenness of distribution of fungi was not homogenous in the sampling sites.The evenness index of fungi was highest for the community isolated from A. variabilis and lowest for the community from S. salsula.However, the evenness of distribution of fungal endophytes in O. glabra differed between the two sampling sites (Table 5).
The Jaccard and Sorenson indices provide an estimate of similarity or shared species between samples.The Jaccard indices were 0.4 or less comparing fungi isolated from S. salsula to those from O. glabra or A. variabilis, suggesting low similarity.However, the Sorenson indices were 0.6 for fungi from stems and leaves of O. glabra from Bayan Hot and leaves, stems, and leaves of A. variabilis from Jilantai compared to the fungi from S. salsula leaves, suggesting more similarity among those populations.
For fungi isolated from O. glabra collected from BayanHot, the Sorenson and Jaccard indices among stems and leaves, seeds and leaves, and seeds and       stems were not less than 0.6, suggesting that the similarity of fungi among leaves, stems, and seeds at this location was very high.For fungi isolated from O. glabra collected from Yinchuan, the Sorenson indices for stems and leaves and seed and leaves was less than 0.5, suggesting that there was lower similarity in fungal species.
Comparing the fungi isolated from tissues of O. glabra at two sampling sites and A. variabilis at two sampling sites gave generally low similarity, with a few exceptions.The Sorenson index was 0.6from stems of A. variabilis collected from AoLun Prague compared with O. glabra, and 0.6 for stems of O. glabra collected from Yinchuan compared with A. variabilis (Table 4).
Among the fungi isolated from leaves, stems, and seeds of A. variabilis collected in Jilantai, Sorenson and Jaccard indices were higher than 0.6.In contrast, the Sorenson and Jaccard indices were lower for AoLun Prague.The Jaccard index was less than 0.5 among seeds, leaves and stems of A. variabilis collected in AoLun Prague.The Sorenson index was more than 0.5, indicating that similarity of fungi was higher in A. variabilis collected from AoLun Prague (Table 6).

Ecological roles of isolated fungi
The relative rate of isolation of mutualist endophytes significantly impacted recovery of pathogenic fungi (Table 4).There was an inverse relationship between isolation of A. U. oxytropis and plant pathogenic fungi.O. glabra from BayanHot yielded very high levels of A. U. oxytropis (68.6-80.6% of total fungal isolations) and very low levels of plant pathogens (12-15% of total fungal isolations).In the absence of A. U. oxytropis, the pathogen levels were very high; O. glabra leaves from Yinchuan yielded 89% pathogens.Overall the relative pathogen and saprophyte levels from O. glabra from Yinchuan were higher and A. U. oxytropis levels lower than from O. glabra from BayanHot.The leaves from Yinchuan had 89% pathogens and 11% saprophytes, the stems 34% pathogens and 19% saprophytes, and the seeds 32% pathogens and 6% saprophytes.A. variabilis yielded moderate pathogen levels at both locations.Pathogen levels from leaves, stems, seeds, and roots from Jilantai were 50, 33, 37 and 50%, respectively, and at AoLun Prague were 23, 45, 23 and 73%, respectively.Saprophyte levels were higher in roots than other tissues from both locations, 16% from Jilantai and 17% from AoLun Prague.S. salsula yielded high levels of pathogens from all tissues, 88, 63, 87 and 72% from leaves, petioles, stems, and seeds, respectively.Higher levels of saprophytes (25%) were only isolated from petioles.

DISCUSSION
Plant host species, plant part, and environment influenced the endogenous fungal communities isolated from the locoweed plants.Diverse fungi were found associated with the locoweed plants.While one fungal species (Alternaria porri) was found associated with all three plants tested, generally different species of plants were colonized by different fungi.There were a few dominant species found, particularly Alternaria sect.U. oxytropis from O. glabra and A. variabilis.There were also abundant rare fungal species found in small numbers from a single plant species, for example 12 for O. glabra from Yinchuan, 8 from S. salsula from BayanHot and 7 from A. variabilis from AoLun Prague.Since there are very few reports of fungi associated with locoweeds, for many of the fungi isolated, this constitutes the first report of most of the association of those fungal species with these plant species.Only A. U. oxytropis, A. U. gansuense, and Fusarium proliferatum have been previously reported to be associated with locoweeds in China (Gao et al., 2012;Zhou et al., 2012;Liu et al., 2016).Other than these three fungal species, only 8 species found in this study have even been reported to be a pathogen of any legume (Tu, 1985;Hernandez-Bello et al., 2006;Asan, 2011;Sharma et al., 2012).Interestingly, most of the pathogens found have not been reported to infect legumes, including A. porri, a pathogen of onion, which was found associated with all three plant species.Most of the saprophytic fungi isolated are considered ubiquitous, found in soil or on decaying plant material of many different plant types and environments (Petrini, 1991).Our experimental objective was to screen for culturable strains of fungi associated with locoweeds.
It is likely that other nonculturable fungi are associated with the locoweeds, but not identified in this study.
Environment had an influence on the endogenous fungal communities, particularly for the two toxic locoweeds, for which location and environment appeared to strongly influence the fungal community.The isolation rates and species of fungal endophytes differed within O. glabra collected from different locations, in which, five species of fungi were isolated from O. glabra in BayanHot town, while 21 species were isolated from Yinchuan city and 12 of those 21 species were unique to that site.BayanHot is in Inner Mongolia and Yinchuan is in Ningxia province, thus, providing different locations and environmental factors.The environmental factors that differed between the two sites include surrounding flora, habitat, soil type, rainfall, temperature, and altitude.Because of the limited number of sampling sites, the quantitative relationship between a particular environmental factors and fungal endophyte species could not be definitively determined.In contrast, the numbers of fungi isolated from A. variabilis from Jilantai (14, with 4 unique species) were similar to that isolated from AoLun Prague (17, with 7 unique species).Both collection sites for A. variabilis are located in Inner Mongolia.The distribution of fungi within a plant species can be influenced by location, season, age, climate, and geographical conditions.Arnold and Lutzoni (2007) analyzed the distribution of communities of fungal endophytes in forests between the Canadian arctic to the central part of Panama, and found that the colonization rate of fungal endophytes is significantly influenced by latitude.However, within a narrow scope of latitude, fewer fungal endophyte species were found in arid areas than in semi-deciduous forests (Hoffman and Arnold, 2008).Siles and Margesin (2016) found that the relative size fungal communities associated with alpine forests increased with altitude and the composition changed as well.The authors noted that changes in fungal richness and diversity and community structure were primarily influenced by pH and carbon/nitrogen in the soil, which was the result of environmental factors at the sites tested.
For α-diversity index, Shannon-Wiener and Simpson indices were used in this study.Two factors influence the Shannon-Wiener index: diversity (variety and amount) and uniformity of fungal localization among species.In this study, the Shannon-Wiener index of S. salsula collected from BayanHot of the Inner Mongolia Autonomous Region was the highest, while that of O. glabra was the lowest.These results could be because the presence of the fungal endophyte, Alternaria sect.U. oxytropis has an impact on other fungi infecting the same plant tissues.The Simpson index is also an indicator of diversity, with a larger value indicating greater diversity, but is a dominance index giving more weight to common species, such that a few rare species do not affect the diversity value.The Simpson diversity index of fungal endophytes in O. glabra in the two locations was significantly different, but that of A. variabilis was not significantly different between locations.In contrast, the Shannon-Wiener index of fungal endophytes isolated from A. variabilis collected in AoLun Prague was higher than in Jilantai.Compared to forage plants in similar ecological environments in other countries, the diversity index we found was slightly low (Porras-Alfaro et al., 2008;Khidir et al., 2009).This may be due to the small number of samples.
β-diversity, as measured through the Jaccard and Sorenson indices, is the difference in the composition of species between different communities in different habitats, or the replacement rate of species along environmental gradients, which is also called betweenhabitat diversity.The main ecological factors controlling β-diversity are thought to be soil, topography, and interference of external factors.Fewer common species are found in different communities or different sites in the same environmental gradient and hence, β-diversity is larger.The β-diversity index can indicate the degree to which species are isolated by habitat, which can be used to compare habitat diversity in different areas, and constitutes the overall diversity or the biological heterogeneity in certain locations together with the αdiversity index.For β-diversity index, Jaccard and Sorenson diversity indices were used to compare the degree of similarity of composition of species of fungal endophytes between two locations.In this study, the species and quantity of distribution of fungal endophytes differed between hosts and tissues.Compared to the results calculated with Jaccard and Sorenson indices, there were uniform patterns of distribution of fungal endophytes in different tissues from the same plant, and the same plant at different sample sites.Stems and seeds showed the highest index (0.8) in similarity of flora in S. salsula collected in BayanHot, followed by petioles and seeds, indicating a similarity in flora of fungal endophytes among stems, petioles, and seeds in S. salsula.The Sorenson and Jaccard indices between the tissues colonized by fungi, in O. glabra and S. salsula and between all the tissues colonized by fungi in A. variabilis and S. salsula, were low, suggesting that there are no similarities in the community of fungi between O. glabra or A. variabilis and S. salsula.
The fungal communities differed somewhat by plant part.Although roots were only sampled from A. variabilis, they yielded higher proportions of saprophytes than the aerial portions of the plants.Similar results were demonstrated for roots and aerial parts of wheat (Comby et al., 2016).David et al. (2016) demonstrated that the influence of location and host species on fungal endophyte community composition also depended on plant part, in that environment and host species were very important in endophyte community composition from leaves of beach grasses, but only the sand dune environment was important for determining the endophyte community from roots.
Alternaria sect.U. oxytropis was consistently isolated from the seeds of the toxic locoweeds and less consistently with the leaves and stems.This colonization is consistent with transmission of A.U. oxytropis, which is maternally transmitted through seed (Oldrup et al. 2010).Braun et al. (2003) found that slow growing Alternaria sect.Undifilum spp.were isolated from A. mollissimus, O. lambertii, and Oxytropis sericea locoweed populations collected from New Mexico.They found that the infection rates of fungal endophytes in stems, leaves, flowers, and seeds were 97.2, 72.9, 100 and 93.1%, respectively.Fast-growing fungi were also isolated from roots when sterilization was minimal, but they were not characterized.Cook et al. (2009a) quantified fungal endophytes in 10 plants each of O. sericea, A. mollissimus, and A. lentiginosus, using quantitative PCR, and showed that the amount of Alternaria sect.Undifilum sp.differed significantly among the plant species and between individual plants within a species.
The isolation rate of Alternaria sect.U. oxytropis from plant sections coincided with fewer plant pathogens isolated from the same sections.Fewer plant pathogens and a higher isolation rate of A. U. oxytropis were obtained from O. glabra from BayanHot than from Yinchuan.Comparing plant species collected from BayanHot, the non-toxic S. salsula, which did not yield A. U. oxytropis, had a much higher relative proportion of pathogens, than did O. glabra.The diversity of the fungi was also much lower from O. glabra than S. salsula from BayanHot.Together this suggests that A. U. oxytropis may influence the decrease in pathogen isolations.This could be due to competition for space or nutrients or it might be related to its swainsonine toxin production.Since swainsonine data was not obtained from these plants, this can't be verified.Infection with fungal endophytes has been shown to change microbial species composition and reduce pathogens (Clay, 1990;Arnold et al., 2003;Pan and May, 2009;Busby et al., 2016).Infection with E. coenophiala, a shoot-specific fungal endophyte of tall fescue influenced soil fungal communities, decreasing the relative abundance of Ascomycota and increasing the abundance of Glomeromycota (Rojas et al., 2016).Fungal endophyte infection in mature leaves of the tropical tree Theobroma cacao prevented infection by Phytophthora sp.pathogens (Arnold et al., 2003).The nature of the beneficial interaction needs to be assessed with other locoweeds and species of Alternaria sect.Undifilum to determine if this is a universal characteristic or specific only to this situation.

Conclusion
The diversity of fungi associated with locoweeds, the sampling sites, plant part and isolation rates were considered in this study.Location/environment, plant species, and presence/absence of the fungal endophyte Alternaria sect Undifilum all had strong impacts on the diversity of fungi associated with the locoweed plants.This initial catalog of fungi associated with locoweeds lays the foundation for future research on the microbial community associated with these toxic plants.The endogenous fungal community associated with locoweeds in other locations such as the western USA and associated with other plant species containing different Alternaria sect.Undifilum spp.should be examined to determine if the community profile is ubiquitous.This is the first report of a beneficial role for this fungal endophyte.The beneficial role for Alternaria sect.Undifilum sp.found here has implications for management strategies, since this endophyte has been found associated with a variety of plants in the western USA and Australia, as well as China.Current efforts to manage the toxicosis associated with grazing plants infected with the fungus have ranged from using herbicide to kill the plant, to avoiding grazing on infected (toxic) plants, and using fungicide to cure locoweeds of the toxic endophyte.If the beneficial role for the endophyte is linked to its production of swainsonine, then efforts to develop endophyte-free plants should be pursued.If the beneficial role for the endophyte is not tied to swainsonine production, then development of a fungal mutant that did not produce swainsonine might possibly help its plant host survive pathogen infection.

Figure 2 .
Figure 2. Phylogenetic tree constructed using Neighbor-joining (NJ) based on 5.8S rDNA-ITS sequences of fungi isolated from Astragalus variabilis.

Figure 3 .
Figure 3. Phylogenetic tree constructed using Neighbor-joining (NJ) based on 5.8S rDNA-ITS sequences of fungi isolated from Oxytropis glabra

Table 1 .
Location and environment of locoweed collection sites.

Table 2 .
Isolation of fungi from locoweeds through surface sterilization.

Table 3 .
Species of fungi isolated from Locoweeds.

Table 5 .
Diversity index of fungi from locoweeds and sampling sites.

Table 6 .
Similarity index of fungi isolated from different locoweed tissues and species.