Two new species of dictyostelid cellular slime molds in high-elevation habitats on the Qinghai-Tibet Plateau, China

Dictyostelid cellular slime molds (dictyostelids) are key components of soil microbes. The Qinghai-Tibet Plateau is characterized by unique and important forest types because of the considerable range in elevation which exists. During the period of 2012, 2013 and 2016, 12 species of dictyostelids were yielded from samples collected in this region, including two new species and three new records for China. Six other species were new records for this region. Ontogeny, morphology, ultrastructure and systematic molecular analyses (SSU & ITS) of D. minimum and D. multiforme confirm that they are Goup 4 new species. The ornamentation of the surface of dictyostelids’ spores is the first time to be observed until now. In the SSU phylogenetic tree generated in the present study, Synstelium, not assigned to order and family before, was assigned to the clade Acytosteliaceae in the Acytosteliales firstly. To our knowledge, the study reported herein is the first investigation of dictyostelid biodiversity carried out at elevations above 2000 m. Sorocarp size, sorus size, spore length, ratio of sorus and sorophore, and ratio of sorus and spore size were positively correlated with increasing elevation and no linear correlated with forest type, according to the results of linear regression analysis.

Dictyostelids, the second largest group of slime molds, have both animal-like (protozoan) and fungus-like characteristics. The vegetative phase consists of single-celled amoeboid forms that live in the soil, where they feed upon bacteria and other microbes, grow, and multiply until the available food supply is exhausted. When this happens, the amoeboid forms aggregate together in large numbers to form multi-celled pseudoplasmodia, which then give rise to fruiting bodies (sorocarps) that consist of supportive stalks and unwalled sori containing propagative spores 1,2 .
Although the first dictyostelid was described by Brefeld in 1869 3 , relatively little work was done on the Dictyosteliaceae until 1941, when Harper, Arndt and Raper began their studies of these organisms 2 . The first ecological survey was carried out in the forests of southern Wisconsin in the United States, when Cavender and Raper 4 sampled six sites located along a moisture gradient. Their samples were processed with the use of a quantitative method of isolation 5 . This experiment showed that dictyostelids are affected by environmental factors, especially moisture 6 . It is now known that there are a number of factors which can affect the distribution and abundance of dictyostelids, including the physiographic regime sampled 7,8 , soil pH, soil type, climatic conditions, forest type 6,9,10 , elevation 7,10-12 , and latitude 13 . Some species of dictyostelids have been grouped into abundance categories, including very common, common, rare, and very rare 9, [14][15][16] .
Environmental factors such as elevation and pH appear to have a predominant effect on patterns of biodiversity in dictyostelids, while the effects of forest management are secondary 7 . Species biodiversity is generally very low under dry conditions, although Romeralo et al. 17 found Polysphondylium violaceum Bref. to be prominent in drier vegetation types. In the Iberian Peninsula, dictyostelids abundance was reported to be highest in colder and wetter environments, which suggests that this group favors relatively cold places with high levels of water availability 6 .  Dictyostelids have evolved a number of different structures in response to the different environments in which they occur. Some species (e.g., Dictyostelium septentrionalis [Cavender 1978], now D. septentrionale) form a thick sorophore to help keep this structure erect as an adaptation for fruiting in cool temperatures; other species (e.g., D. rhizopodium Raper & Fennell, now Hagiwaraea rhizopodium [Raper & Fennell] S.Baldauf, S.Sheikh & Thulin) form root-like basal crampons in order to help the sorocarp remain erect longer in tropical habitats; and members of the genus Acytostelium produce small sorocarps, which is probably an adaptation associated with their narrow niche requirements 17 .
According to the traditional morphology-based classification, dictyostelids were placed in the class Dictyosteliomycetes under the phylum Protozoa. This class was considered to include one order, two families and four genera 18 that are distinguished morphologically by differences in sorophore composition and branching pattern. However, data from a phylogenetic analysis based on 18S ribosomal RNA (18S rRNA) and α-tubulin, indicated that none of these three genera are monophyletic, with the dictyostelids divided into four groups 19 and eight groups 20 , respectively. However, Sheikh et al. 21 proposed a new classification based on unique 18S rRNA sequence  The primary objective of the study reported herein was to increase our knowledge of dictyostelid biodiversity and abundance at high elevations (>2000 m) and in ecologically complex habitats. Samples for isolation of these organisms were collected from nine sites on the Qinghai-Tibet Plateau. The dictyostelids recovered from these samples were studied, both for their morphological features and also with molecular markers. The species biodiversity of dictyostelids on the Qinghai-Tibet Plateau is discussed in relation to previous studies and concepts relating to the overall ecology of these organisms.

Results
After being processed, the samples collected from the Qinghai-Tibet Plateau yielded 16 isolates representing 12 species of dictyostelids ( Fig. 1, Table S1). Three of these (Dictyostelium  Table S2), based on the concepts of Schaap et al. 19 , Romeralo et al. 20 , and Sheikh et al. 21 .
Taxonomy and molecular phylogeny. Dictyostelium minimum Li Y., P. Liu et Y. Zou, sp. nov. (Fig. 2) MycoBank: MB823443. When cultured at 17 C on non-nutrient agar with Escherichia coli, sorocarps white, erect, stout, very short, only 95-720 μm high (average 324 μm), solitary, non-phototropic. Sorophore stout, tapering from base to tip, usually consisting single tier of cells from the middle portion to the tip, consisting of two or several tiers of cells from the middle portion to the base, base clavate or acuminate with accessory structures and two or several tiers of cells. The interior of the sorophore is hollow according to the SEM observations of the sorophore. Sori white, globose or citriform, 33-148 μm (average 72 μm) diam. Spores globose, without polar granules, 3.7-5.4 μm diam. Cell aggregations mound-like, without radiate streams, usually 39-73 × 29-57 μm (average 52 × 42 μm). Pseudoplasmodia not migrating without sorophore formation.
Etymology. This name refers to the small size of the sorocarps. Holotype. HMJAU MR244. Isolated in 2012 (Strain 2794) from a soil sample collected from a mixed forest of Picea asperata, Pinus densata, Quercus semecarpifolia, Pterocarya stenoptera, and Betula delavayi, located at an elevation of approximately 3100 m in Lulang, Tibet.
Known distribution. Currently known only from China.
Commentary. There are primarily five species of dictyostelids that have globose spores 42  This species belongs to Dictyostelid Group 4 [19][20][21] in SSU rDNA phylogeny (Fig. 6). It forms a clade together with D. multiforme. In the ITS phylogenetic tree (Fig. 7), D. minimum forms a clade with D. crassicaule and D. pseudobrefeldianum H.Hagiw. However, the most noteworthy difference between them is the presence of globose spores in D. minimum. The SSU and ITS rDNA sequences for D. minimum are available at GenBank, with the accession numbers MG490369 and MG490372, respectively.
Dictyostelium multiforme Li Y., P. Liu et Y. Zou, sp. nov. (Fig. 3 [19][20][21] in the SSU rDNA phylogeny (Fig. 6). It forms a clade along with D. minimum. Morphologically, the spore shape and size, sorophore cell tiers and sorocarp appearance are different between D. multiforme and D. minimum. In the ITS phylogenetic tree (Fig. 7 Ontogeny. Life cycle of Dictyostelium minimum. Spore germination (Fig. 4A) begins with the appearance of a minute pore dissolved in the spore wall after 15 h on agar. At the same time, several myxamoebae (Fig. 4B) are released from the spores. Myxamoebae are colorless, transparent, and irregular. When a "critical mass" of myxamoebae occurs, myxamoebae aggregate to one center to form a single aggregation (Fig. 4C), with no cell streams 36 h after the spores have been inoculated on agar. The aggregations (Fig. 4D) grow larger and rise up in the center after 39 h. The aggregations culminate (Fig. 4E) and began to form pseudoplasmodia with indistinct sorophores 42 h after inoculation. The sorophore (Fig. 4F) begins to form and grow longer at about 45 h. Two hours later, the slug (Fig. 4G) begins to form and with no movement in this process. Three hours later, the sorogen is produced and is prostrate on the surface of the agar, and the young sorocarp begins to fruit (Fig. 4H). The top of the sorogen became tapered, then changes to globose, and the sorophore becames longer and curved on the Life cycle of Dictyostelium multiforme. Elliptical spore germination (Fig. 5A) begins with the dissolution of the spore wall to release myxamoebae (Fig. 5B,C) after 21.5 h on agar. Myxamoebae are colorless, transparent, and irregular. Their shape changed as they moved about, feeding upon bacteria, growing larger and aggregating as a result of being attracted by cAMP to form a single mould-like aggregation (Fig. 5D) without radiated cell streams 28 h after the spores had been inoculated on agar. After the aggregating of more and more myxamoebae, the cell aggregations (Fig. 5E) rise up in the center begin to form pseudoplasmodia with indistinct sorophores 30 h after inoculation. Two hours later, the sorophores begin to form (Fig. 5F). Sorophores grow longer, which gives them the appearance of a fruiting sorophore (Fig. 5G) at 33 h, pseudoplasmodia form and undergo a very short period of movement. The sorophore is slightly curved, which gives it an "S" shape in the central portion. After the formation of the sorophore, the sori begin to grow gradually to form globose sori at 37 h (Fig. 5H). Forty-one h after inoculating spores on the agar, the sorocarps (Fig. 5I) finally fruit. Sorocarps are solitary, erect or semi-erect and lack branches.
The whole life cycle of D. multiforme extends of a period of less than 2 d (41 h). The spores germinated and released myxomoebae 21.5 h after inoculation. The aggregations formed 28 h after the inoculation, whereas the pseudoplasmodia formation needed 33 h. After that, the sorophores and sori begin to grow orderly and finally form fruiting sorocarps after 41 h. The formation of myxomoebae of this species is later than that of D. minimum; however, the formation of sorocarps from pseudoplasmodia in this species is shorter than that of D. minimum.

Ecology. Dictyostelids-elevation relationships. Environmental factors such as elevation clearly have an effect
on the biodiversity and morphological features of species of dictyostelids. Considering the 11 species (15 isolates) of dictyostelids isolated in the present study, except Heterostelium tikalense, all have traditional simple sorophore, so the relationships between morphological characteristics of those 15 isolates and their elevations were analyzed. From the results of linear correlation analysis of elevation with the six predictors considered in the present study, increases in elevation led to increasing sorocarp size, sorus size, spore length, ratio of sorus and sorophore, and ratio of sorus and spore size ( Fig. 8A-C,E,F). In contrast, increasing elevation was correlated with decreasing spore width (Fig. 8D).

Dictyostelids-forest type relationships.
In the present study, 13 isolates of dictyostelids were obtained from different forest types including mixed forest, broadleaf forest, coniferous forest, grassland, and alpine grassland. Another three isolates of dictyostelids were isolated from animal dung. After make an analysis of the morphological charactertistics of those dictyostelids with those forest types, we found there was no linear correlation of forest type with sorocarp size, sorus size, spore length, spore width, ratio of sorus and sorophore, and ratio of sorus and spore size.

Discussion
Effects of elevation on the biodiversity of dictyostelids. Ecological characteristics such as elevation clearly have an effect upon the biodiversity of dictyostelids. However, studies of the distribution and abundance of dictyostelids associated with similar habitats at high elevations (>2000 m) are exceedingly limited. Although Cavender et al. 12 and Landolt et al. 10

Effects of elevation on the morphology of dictyostelids.
Elevation also affected the morphological features of the species of dictyostelids being considered. There was a positive correlation of increasing of elevation with sorocarp size, sorus size, spore length, ratio of sorus and sorophore, and ratio of sorus and spore size, although the R 2 values for linear correlations of spore length were weak (Fig. 8). Dictyostelids carry bacteria during spore dispersal and thus can "seed" a new food crop, which is a major advantage if edible bacteria are lacking at the new site 44 . Presumably, the larger the size of the sorocarp and sori would allow them carry more bacteria if edible bacteria are not abundant enough in habitats such as those found at higher elevations on the plateau.
Substrate types. Dictyostelids have been recovered from animal dung 45 , forest soil 17 , grassland soil 46 , canopy soil 47,48 , and soil in an agricultural field 49 . In the present study, dictyostelids were recovered from both animal dung and forest soil. From these samples, four species (Dictyostelium brevicaule, D. minimum, D. sphaerocephalum, Cavenderia aureostipes) were recorded from mixed forest soil, three species (D. vermiforme, D. sphaerocephalum, C. fasciculata) from coniferous forest soil, three species (D. sphaerocephalum, C. exigua, Heterostelium tikalense) from broadleaf forest soil, two species (D. brefeldianum, D. sphaerocephalum) from alpine grassland soil, one species (C. antarctica) from grassland soil, and two species (D. crassicaule, D. multiforme) from animal dung. However, there were no positive linear correlations of forest type with sorocarp size, sorus size, spore size, ratio of sorus and sorophore, and ratio of sorus and spore size. There are the highest genus diversity and species diversity of dictyostelids in broadleaf forest and mixed forest seperately in this study.
Environmental conditions affect the distribution of dictyostelids. Dictyostelium minimum has very small sorocarps, and the sorophore is stout and thick, which is a character also found in species which occur in habitats characterized by cool temperatures (20 C or lower) in order to keep the sorocarp erect during slow development 17 . Presumably, the small size of the sorocarps would allow them to compete more successfully under conditions that are marginal for larger species, such as those microhabitats with a limited bacterial food supply 10 .
In the present study, we used both 17 C and 23 C to culture the new species D. minimum; however, it grew well at 17 C and but was unstable and sometimes even unsuccessful at 23 C. Dictyostelium sphaerocephalum tends to be more abundant in habitats where conditions are less favorable [50][51][52] . Examples include extremely cool, dry, or disturbed habitats 15 such as in the tundra 50,53 . In our study, D. sphaerocephalum was also recovered at three collecting sites located at higher elevations (3000-4600 m). The occurrence of this species in alpine grassland soil, mixed forest soil and coniferous forest soil provide additional data to support the concept that D. sphaerocephalum is a truly widespread species 7 .
In contrast, D. crassicaule, C. exigua, and C. fasciculata were reported as restricted species in soils of the world's forests 43 , herein they were all found again from forest soil and animal dung, which supports the point of view 43 that animal vectors and plant associations have a major role in determining the distribution of dictyostelids.
Futhermore, the highest relative abundance of dictyostelids in this study are Localities (ii) Medog and (v) Lulang (Fig. 1). We found Medog has the highest annual temperature, and Lulang has the highest annual precipitation within localities of this study. Presumably, the species abundance of dictyostelids correlated with the environmental conditions such as temperature and precipitation a lot.
Ultrastructure of the two new species. From SEM and TEM images obtained for D. minimum and D. multiforme, we found that the spore wall was characterized by the presence of small bulges. The bulges of D. minimum are nearly globose and short. The ridged bulges of D. multiforme are deeper and longer than those of D. minimum. These data suggest that the surface of dictyostelid spore may differ with respect to ornamentation. Molecular phylogeny. Detailed analysis involving the alignment of SSU of these two new species D. minimum and D. multiforme indicated that AAG and ACTG occur in the nucleotide positions 664-666 and 671-674, respectively (Fig. 6); this indicates that they clearly belong to genus Dictyostelium 21 . However, the globose spores of D. minimum add a new spore feature to be considered in the new classification of Dictyostelium. In the SSU rDNA phylogenetic tree (Fig. 6), D. minimum and D. multiforme were in the same clade. However, when the phylogenetic relationships of Group 4 species were analyzed with the ITS rDNA (Fig. 7), D. minimum and D. multiforme occurred in different clades.
In the SSU phylogenetic tree published by Sheikh et al. 21 , Synstelium was not assigned to any family or order; herein we found it was most closely related to clade Acytosteliaceae in the Acytosteliales. Consequently, Synstelium was recognized in the Acytosteliaceae and Acytosetliales in the present study. All other taxonomic levels (genus, family and order) are consistent with those given by Sheikh et al. 21 .
Three species (D. rosarium, Tieghemostelium lacteum, and Raperostelium ibericum) with globose spores are members of Group 4, Group 3 A, and Group 3 C, respectively 21 . In the present study, another species (D. minimum) with globose spores was obtained which belongs to Group 4 according to the new classification of Sheikh et al. 21 . These four globose spore speices all belongs to Dictyostliales of two families (Group 3 and Group 4). As such, it is not possible to use only a single morphology-based taxonomic feature such as spore shape to differentiate species in the dictyostelids.

Materials and Methods
Sampling. Samples used for isolation of dictyostelids were collected from nine sites on the Qinghai-Tibet Isolation and cultivation. The isolation methods used in the present study followed those described by Cavender and Raper 5 . Each sample was weighed and diluted for an initial dilution of 1:10 by adding ddH 2 O. This dilution was shaken to disperse the material and to suspend the amoebae and spores of dictyostelids. Afterwards, a 0.5 mL aliquot of this dilution was added to each of five duplicate culture plates prepared with hay infusion agar 2 . Approximately 0.4 mL of a heavy suspension of the bacterium E. coli was added to each culture plate as a food source. The plates were incubated at temperatures of 17 and 23 C with a 12 h light and dark cycle. Each plate was examined at least once a day for two weeks after the appearance of initial aggregations. Each isolate was purified and cultivated for taxonomic studies and preservation on non-nutrient water agar plates with E. coli pregrown for 12-24 h. Spores from these plates were frozen in HL 5 media 54 and stored at −80 C in HMJAU, Changchun, China.
Morphological features and life cycle observations. Dictyostelid isolates were identified with the use of the descriptions provided by Raper 2 , whose nomenclature also was followed except for those species recently assigned to new genera in the system of classification proposed by Sheikh et al. 21 . In the primary isolation plates, the locations of each early aggregating clone and sorocarp that developed were marked. The characteristic stages in the life cycle, including cell aggregation and the formation of pseudoplasmodia, and sorocarps were observed under a Zeiss dissecting microscope (Axio Zoom V16) with a 1.5× objective and 10× ocular. Slides with sorocarps were prepared with water as the mounting medium. Features of spores, sorophores, and sorocarps were observed and measured on the slides by using a Zeiss light microscope (Axio Imager A2), with 10× ocular and 10, 40, and 100× (oil) objectives. Photographs were taken with Zeiss Axiocam 506 color microscope camera.
Observation of spore germination. Hanging drop cultures as described by Keller and Schoknecht 55 were prepared for the observation of spore germination. Spores obtained from a sorus were mixed with a droplet of sterile water on the undersurface of a 22-mm square cover glass. The cover glass was then inverted over a depression slide. Vaseline was used to ring the edges of the cover glass. Spores were freely suspended in the water droplet. Features of the myxamoebae were observed and photographed by a Zeiss laser confocal microscope (LSM 710).
Scanning electron microscopy. Spores of new species Dictyostelium minimum and D. multiforme were prepared for scanning electron microscopy according to the method of Boyde and Wood 56  Transmission electron microscopy. The sorocarps of new species D. minimum and D. multiforme were prepared for transmission electron microscopy according to standard techniques 57 . The sorocarps were collected with a prefixation in 4% glutaraldehyde for more than 4 h at 4 C, followed by postfixation in 2% osmium tetroxide for 2 h at 4 C, with both fixatives buffered in 0.05 N Na-cacodylate buffer at pH 7.4. Afterwards, the samples were dehydrated in water, ethanol and acetone, embedded in SPI-PON 812 for 12 h at 35 C, for 12 h at 45 C, and for 24 h at 60 C. After these processes had been completed, the sections to be observed were cut on a Leica EM UC7, stained in uranyl acetate and lead citrate, observed and photographed by a Hitachi transmission electron microscope (H-7650).
DNA isolation, PCR amplification and sequencing. After amoebae had cleared E. coli on the water agar media, the spores of dictyostelid isolates to be studied were collected with a sterile tip, then those spores were mixed with the lysed buffer of the NuClean Plant Genomic DNA Extraction Kit from CW Biotech (Beijing, China) and the following steps were carried out according to the instructions provided along with this kit. The genomic DNA solution was used directly for the PCR amplification. The SSU and ITS rDNA markers were amplified using the primers 18SF-A and 18SR-B, D542F and D1340R, and ITS1 and ITS4 (Table S3). PCR products were sent to Sangon Biotech Co., Ltd. (Shanghai) for sequencing.
Phylogenetic analysis. The newly-generated sequences were checked and then submitted to GenBank.
Sequences for all closely related species were downloaded from GenBank (Table S2) for phylogenetic analysis. The ITS and SSU sequences were aligned and compared separately using the program Muscle v.3.6 58,59 , then manually adjusted in MEGA 7.0 60 . Maximum likelihood (ML) analyses were performed using RAxML v7 61 . In the ML analyses, the best-fit substitution models were estimated using GTR submission model and a gamma correction for rate variation among sites (GTRGAMMA), using the CIPRES server. The statistical support of clades was assessed with 1000 rapid-bootstrap (BS) replications.
Ecological statistics analysis. Sorocarp size, sorus size, spore length, spore width, ratio of sorus and sorophore, and ratio of sorus and spore size data for the 11 species of Dictyostelium and Cavenderia being considered (15 isolates in total) in this study were inputted to the IBM SPSS 19.0 version software as predictors for linear regression analysis of those six predictors (sorocarp size, sorus size, spore length, spore width, ratio of sorus and sorophore, and ratio of sorus and spore size) with the two dependent variables of elevation and forest type, with each of the latter considered separately.
Nomenclature. According to the International Code of Nomenclature for algae, fungi, and plants, the electronic version of this article in Portable Document Format (PDF) will represent a published work. In addition, new names contained in this study have been submitted to MycoBank and will each be allocated a unique MycoBank number which will be accessible through MycoBank, Index Fungorum, GBIF and other international biodiversity initiatives where they will be made available to the Global Names Index.