Spatial Variation of Cladophora Epiphytes in the Nan River, Thailand

Cladophora is an algal genus known to be ecologically important. It provides habitats for microorganisms known to provide ecological services such as biosynthesis of cobalamin (vitamin B12) and nutrient cycling. Most knowledge of microbiomes was obtained from studies of lacustrine Cladophora species. However, whether lotic freshwater Cladophora microbiomes are as complex as the lentic ones or provide similar ecological services is not known. To illuminate these issues, we used amplicons of 16S rDNA, 18S rDNA, and ITS to investigate the taxonomy and diversity of the microorganisms associated with replicate Cladophora samples from three sites along the Nan River, Thailand. Results showed that the diversity of prokaryotic and eukaryotic members of Cladophora microbiomes collected from different sampling sites was statistically different. Fifty percent of the identifiable taxa were shared across sampling sites: these included organisms belonging to different trophic levels, decomposers, and heterotrophic bacteria. These heterogeneous assemblages of bacteria, by functional inference, have the potential to perform various ecological functions, i.e., cellulose degradation, cobalamin biosynthesis, fermentative hydrogen production, ammonium oxidation, amino acid fermentation, dissimilatory reduction of nitrate to ammonium, nitrite reduction, nitrate reduction, sulfur reduction, polyphosphate accumulation, denitrifying phosphorus-accumulation, and degradation of aromatic compounds. Results suggested that river populations of Cladophora provide ecologically important habitat for microorganisms that are key to nutrient cycling in lotic ecosystems.


Introduction
Cladophora (Kützing) Kützing is a cosmopolitan filamentous green algal genus widely distributed in freshwater, marine, and brackish shoreline environments across arctic and tropical regions [1][2][3][4]. Freshwater species occur as filaments attached to various kinds of substrates, from rock surface to shells of arthropods, and as free-floating algal masses. This alga survives in water bodies of a wide range of nutrients levels, from the ultra-oligotrophic to the highly eutrophic, and a wide variety of conditions of attachment substrate, light intensity, water temperature, and water velocity [5].

Materials and Methods
Algal samples were collected from Nan River, Nan, Thailand, during the late growing season in March 2020. Five collecting sites with four replicates per site were located in four different districts-Chiang Klang (19 • 13 34.0 N 100 • Figure 1. The collected algal samples were submerged in DNA/RNA Shield TM (Zymo Research, Irvine, CA, USA) and stored in sterile microcentrifuge tubes and Whirl-Pak ® (Nasco, Madison, WI, USA) before being brought back to the laboratory for the following analysis. Additionally, we obtained the secondary information of the water physical properties from the annual report available at Regional Environmental Office 2 Lampang (http://www.reo02.mnre.go.th/th/index; accessed on 27 August 2021). The information included values of the water quality index, dissolved oxygen, biological oxygen demand, and ammonia nitrogen content. Figure 1. Five sample collecting sites in Nan River, Nan Province, located in four different districts. CKD located in Chiang Klang district, PUA located in Pua district, TD1 and TD2 located in Tha Wang Pha district, and MND located in Mueang Nan district. Only three locations (CKD, PUA, and TD1) were Cladophora dominant. Yellow dot represents each collecting site. Bold blue line represents the Nan River. Orange arrows represent the flow direction. Other blue lines represent other smaller rivers and streams.

Algal Identification
At the laboratory, the algal samples were put into sterile petri dishes submerged in DNA/RNA Shield TM . Algal filaments were selected using sterile forceps and washed three times to remove loosely associated materials. Then, the alga samples were observed using an ECLIPSE E100 microscope (Nikon Corporation, Tokyo, Japan). Only samples from sites dominated by Cladophora were employed in this study. To select the Cladophora dominant samples, we used morphological characteristics of algal samples to identify the alga to the genus level [33]. To identify the dominant Cladophora species of each replicate, we used the assembled 18S rDNA amplicons to perform BLASTN search against the NCBI non-redundant nucleotide database (accessed on 13 September 2021).

DNA Extraction and Sequencing
The Cladophora samples were washed three times using DNA/RNA Shield TM to remove the dirt and debris. Then, genomic DNA was extracted using Quick-DNA TM Fecal/Soil Microbe Kits (Zymo Research, Irvine, CA, USA). The 16S rRNA region V3-V4 was amplified using primers 341F and 805R [38], the 18S rRNA region V4 was amplified using primers Reuk454FWD1 and V4r [39], and the ITS was amplified using primers ITS-1F [40] and ITS-2R [41], as provided in Table S1. The amplicons were then purified, indexed, and sequenced using an Illumina MiSeq at Omics Sciences and Bioinformatics Center (Chulalongkorn University, Bangkok, Thailand).

Data Analysis
The quality of the 250 bp paired-end raw reads was assessed by FastQC [42]. Lowquality nucleotide bases were removed using Trimmomatic v. 0.39 using SLIDINGWIN-DOW:4:30 [43]. The taxa and the taxonomic diversity of the epiphytic microbiota of Cladophora sp. were identified and assessed by comparing the annotated rDNA against the SILVA ribosomal RNA gene database release 138.1 [44] using SILVAngs pipeline. In the workflow, trimmed reads were aligned using SILVA INcremental Aligner (SINA). Reads shorter than 50 aligned nucleotides, with ambiguities larger than 2%, or with homopolymers larger than 2% were eliminated before entering the downstream processes. Then, putative contaminant reads, including PCR artifacts and low-quality aligned reads (with alignment identity less than 50% and alignment score less than 40), were filtered out.
Reads that passed the filtered steps were then de-replicated. Identical reads were identified and clustered into different Operational Taxonomic Units (OTUs) using cd-hitest [45] running in accurate mode, overlooking overhangs, and applying identity criteria of 1.00 and 0.98, respectively. Then for each cluster, the representative sequence was then classified by a local nucleotide BLAST search against the non-redundancy of the SILVA SSU Ref database release 138.1 using BLASTN with standard settings. The classifiable reads were mapped onto all reads that were assigned to the respective OTU. Lastly, to filter out chimeric reads, we searched the classified results against SILVA release 138 The raw data for the Cladophora microbiomes have been deposited in NCBI SRA BioProject PRJNA761577 and BioSample SAMN21356006 (CKD), SAMN21356007 (PUA), and SAMN21356008 (TD1).

Alpha Diversity
The taxonomic abundance of the microbial taxa was assessed using the results returned from the SILVAngs pipeline. To assess the alpha-diversity, we firstly calculated the values of Shannon index, Simpson index, Chao1, and richness using functions namely diversity and estimateR implemented in the vegan package v. 2.5-7 [50] in RStudio v. 1.4.1106 [51]. To evaluate whether the values obtained from calculations were statistically different, values of One-way ANOVA (p < 0.05) and Tukey's HSD were calculated using the agricolae package v. 1.3-5 [52]. Then, results were visualized using box plots implemented in the ggplot2 package v. 3.3.5 [53] in RStudio.

Beta Diversity
Beta-diversity index, Bray-Curtis dissimilarity, was calculated using the vegdist function in the vegan package [50] in RStudio. Then, to examine whether the values of Bray-Curtis dissimilarity were statistically different, we performed Analysis of Similarity (ANOSIM; function anosim, vegan package) and permutational Multivariate Analysis of Variance (perMANOVA; function adonis, vegan package) using 999 permutations. Results were then visualized using the Principal Coordinate Analysis (PCoA) with the cmdscale function.

Common Organisms in Lotic and Lentic Freshwater Cladophora Microbiome
To investigate the presence of taxa that are commonly present in freshwater Cladophora microbiomes, we re-analyzed the taxonomic classification using the SILVAngs pipeline described above for sequence read archive of metagenomic data from Cladophora collected from Lake Mendota (Madison, WI, USA) [6,9] and Lake Michigan (Portage, IN, USA) [7]. Then, we searched for the taxa commonly present in the lotic and lentic Cladophora microbiomes.

Results
Identification of Cladophora sp.
Only the algal samples collected from three sites (CKD, PUA, and TD1) were morphologically identified to genus Cladophora, having distinct characteristics of monosiphonous and branched filaments with a unique branching style, where the lateral branch arisen under the septum between two cells of the main axis. Each cell in the filament contained reticulate chloroplasts and thick cell walls ( Figure 2A). The algal basal region was attached to the substrates using holdfast. The filaments appeared as tufts, forming the macroscopic green algal bloom floating in water ( Figure 2B).
In addition to having similar morphology, the Cladophora samples collected from the three sites also have identical amplicon sequences ( Figure S1). BLASTN results suggested that the sequence was equally similar to published and unpublished freshwater Cladophora spp. reported from authors from several countries (Table S2).

SSU-16S rDNA Amplicon Analysis
Amplicon analysis showed that bacterial phyla and their relative abundance varied ( Figure 3). The five most abundant phyla in site CKD were Proteobacteria (40.04%), Bacteroidetes (21.07%), Firmicutes (10.59%), Verrucomicrobia (5.30%), and Planctomycetes (4.80%). The five most abundant phyla in site PUA were Proteobacteria (29.29%), Bacteroidetes (22.50%), Firmicutes (10.86%), Planctomycetes (6.88%), and Cyanobacteria (6.20%). The five most abundant phyla in site TD1 were Proteobacteria (31.77%), Bacteroidetes (24.37%), Firmicutes (6.67%), Verrucomicrobia (6.22%), and Cyanobacteria (4.98%). Bacterial taxa were classified into 689 distinct genera; 227 bacterial genera were present in all four replicates of the three collecting sites (Table S3). The three collected sites in this study were located in the upper region of the Nan River, where the water flowed from CKD to PUA and TD1, respectively. We initially sampled the water from each site for physical and chemical measurements. However, all the assessment facilities were shut down for a few months, which made it not possible to obtain such parameters from the collecting sites. However, we compared the abundance of the common bacterial genera and found that some bacterial genera occurred as a gradient from the upper stream (CKD) to the lower stream of the river (TD1). Genera that had higher relative abundance in the upper stream included Proteobacteria Acinetobacter, Aeromonas, and Vogesella, Epsilonproteobacteria Pseudarcobacter, Firmicutes Clostridium Sensu stricto 12, and Bacteroidetes Bacteroides and Flavobacterium. In contrast, genera that had higher relative abundance in the lower stream included Bacteroidetes Paludibacter, Proteobacteria Hydrogenophaga and Leptothrix, and Firmicutes Fusibacter (Table S3).

Alpha Diversity
Alpha-diversity indices revealed the diversity, evenness, and richness within each study site. Shannon's index indicates the diversity within site, in which richness, evenness, and rare OTUs played an essential role in the measure. Simpson's index indicates diversity as Shannon's index does, however, the rare OTUs play a minor role in the estimation. Chao1 and the number of identifiable OTUs reflect the richness within site.
At the genus level of the bacterial taxa ( Figure 5-B), Shannon's indices of CKD, PUA, and TD1 were 4.91, 5.11, and 5.12, which were not statistically different. Simpson's index of the three sites was 0.98. Choa1 of CKD, PUA, and TD1 were 488, 491, and 496, respectively, which were not statistically different. The richness values of CKD, PUA, and TD1 were 461, 474, and 470, respectively, which also were not significantly different.

Beta Diversity
Comparative analysis showed that the Bray-Curtis dissimilarity values of the bacterial phyla were not different among sites while the bacterial genera from the three study sites were significantly different from each other. The R value, which signifies dissimilarity between groups based on the Bray-Curtis dissimilarity values, resulted from both ANOSIM and permANOVA was 0.852 (p < 0.001). PCoA based on Bray-Curtis dissimilarity showed that the variance explained of PCo1 and PCo2 were 45.31% and 22.42% (67.73% in total), where the PCoA plot showed the prominent cluster of each site, representing the similarity of bacterial genera within each site ( Figure 6A).
For eukaryotes, at the genus level, the R value obtained from both ANOSIM and permANOVA was 0.630 (p < 0.001). Results from PCoA based on Bray-Curtis dissimilarity showed that the variance explained of PCo1 and PCo2 were 47.86% and 22.17% (70.03% in total), where the PCoA plot showed a cluster of PUA and TD1 that distinguished these sites from CKD ( Figure 6B).

Common Members of Freshwater Cladophora Microbiomes
The wide distribution range of the freshwater Cladophora sp. made it interesting to investigate whether the alga present in different geographical areas and different ecosystems (lotic and lentic) host a common group of microbes. Therefore, we compared the microbiomes of Cladophora found in this study to those of previously reported studies. Results showed that among the reports of Cladophora microbiomes, the Cladophora microbiome shared more common taxa with results reported from Lake Mendota, USA [6,9] than to Lake Michigan [7]. Interestingly, though different in abundance, 17 bacterial genera were present in microbiomes of all these freshwater Cladophora studies. These included Bacteroidetes Chryseobacterium, Flavobacterium, Terrimonas, Deinococcus-Thermus Deinococcus, Proteobacteria Acidovorax, Acinetobacter, Altererythrobacter, Aquabacterium, Bdellovibrio, Hydrogenophaga, Hyphomicrobium, Methylotenera, Novosphingobium, Pseudomonas, Pseudorhodobacter, Rhodobacter, and Sphingopyxis (Table S6).

Discussion
Along the river basin, Cladophora appeared as tufts/thalli attached on the rock surface across the river streams where the water level was about knee height. Among the collecting sites, Cladophora dominated the upper regions of the Nan River (CKD, PUA, and TD1), where the water temperatures were lower and the main substrates were rocks and pebbles.
Presence of similar algal morphology and identical sequences of selected gene markers suggested that the algal samples collected in this study belong to the same algal population. The best 100 BLASTN results of Cladophora amplicons in this study were sequences from published and unpublished Cladophora species collected from different countries, showing the same total score, percent identity, and E-value (Table S2). This suggested that the amplified region had been conserved across different Cladophora populations. We, therefore, could not annotate the algal sample further than the genus level.
It was interesting to us that although we found that our amplicons were identical to sequences reported from other locales, these amplicons were not identical to any of the reported freshwater Cladophora found in Thailand or the neighboring counties, e.g., Gen-Bank accession LT607372.1 [54] and accessions JQ071987.1-JQ072004.1 [55]. This suggests that there might be more than one population of the freshwater Cladophora endemic to the region, which has not yet been explored.

Microbiomes of Lentic Cladophora sp.
The nature of Cladophora, producing high-surface-area filaments attached to substrates in the river basin, makes it crucial in the ecosystem by providing microhabitats for other microbial organisms. Results from amplicon analysis revealed that the dominant epiphytic eukaryotes in the three sampling sites were Amorphea and Stramenopiles, which agreed with results from light microscopy, where we observed zooplankton and photosynthetic stramenopiles Cocconeis, Gomphonema, Synura, Synedra, and Navicula on the Cladophora cell surface.
Among the diatoms, Cocconeis and Gomphonema were present in all the sampling replicates, judged by both microscopic and molecular evidence, where the relative abundance of Cocconeis ranged from 4.9% at CKD to 27.4% at PUA, and the relative abundance of Gomphonema ranged from 4.6% at TD1 to 0.8% at PUA. The presence of dominant Cocconeis and Gomphonema on the filament surface was similar to results from previous studies [9,10], which hypothesized that this diatom's cell shape allowed it to avoid grazing and maintain firm attachment to the algal filament.
In addition to the dominant diatom Cocconeis, a few other diatoms were also present in all the sampling sites, e.g., Achnanthidium, Amphora, Craticula, Cymbella, Epithemia, Fistulifera, Gomphonema, Melosira, Navicula, Nitzschia, Pinnularia, Planothidium, Rhopalodia, and Thalassiosira. These diverse diatoms on the algal filaments might be explained by ability to tolerate shading that occurs in the Cladophora tuft. Within the algal tuft, the high density of algal filaments results in variation in light intensity. By having such ability, the diatoms can maintain their abundance by upregulating the expression of genes involved in transcription and photosynthesis to compensate for the lower light intensity [56].
Presence of these diatoms on the Cladophora surface promotes high diversity of the algal microbiome as these diatoms release their photosynthetic exudates or exopolysaccharides, which provide the building blocks of biofilms that encase the algal filaments, e.g., [57,58]. According to results from our amplicon analysis, there were 698 unique bacterial genera and 575 unique eukaryotic genera present in the Cladophora microbiome (Tables S3 and S4).
Alpha indices suggested that the diversity of bacterial phyla at PUA and TDI was higher than that of CKD, whereas the eukaryotic genera were more diverse at CKD than at PUA and TD1. In addition, clusters obtained from the PCoA plots suggested that the diversity of the eukaryotic genera and the bacterial genera was more similar within site than among sites. It is unclear at this stage why the diversity of these microbiomes was statistically different. However, we believe that it is not mainly affected by the trophic states of the water at the collecting site as the trophic states do not correlate with the diversity level of the organisms [59][60][61]. In this study, although we failed to collect the physical properties of the water from the sampling sites, secondary data obtained from the annual report of the local environmental organization showed that the trophic state of the water in the Nan River fluctuated. The values of water quality index, dissolved oxygen, biological oxygen demand, and ammonia nitrogen content appeared to fluctuate along the course of the river ( Figure S2). To understand this better, an investigation of the biomes presents at the riverbank or the adjacent floodplains might be needed. Variation in effluent inputs to the river might also be important.

Methane oxidation
Methylocystis [92] Methylovulum [189] Methylotrophy Gemmobacter [190] OM43 clade [191] Extracellular polymeric substance (EPS) secretion Terrimonas [192] Biosorption of heavy metals Sphaerotilus [193] The majority of the fungi present within the samples collected from TD1 were fungi that function in decomposition, e.g., Avachytrium, Entophlyctis, Glutinoglossum. Other fungi were those known to be involved in (1) parasitism in algae, amoebae, and other fungi, (2) plant-fungal interaction, (3) predation, and even (4) mutualism as in the lichen-forming fungi (Table 2). We believe that, while some fungi were present temporarily (e.g., lichenforming fungus, phylloplane fungi, and plant pathogens), a portion of these fungi live within the host algal microbiome. However, to know which fungi live in the associations and which fungi are there temporarily due to the influx from the nearby flood plain, more replications of Cladophora microbiomes from more study sites are needed. Table 2. Fungal genera commonly present in the Cladophora microbiomes and their putative ecological functions.

Common Members of Freshwater Cladophora Microbiomes
We re-analyzed the published data in archives [6,7,9] and investigated if there are shared taxon between the lotic and lentic freshwater Cladophora. As expected, only a few taxa were commonly present in all algal microbiomes as the collecting sites were located in different hemispheres and hydrological systems. All the common taxa were bacteria, which were known to be involved in important ecological activities, including cobalamin (vitamin B 12 ) biosynthesis (Flavobacterium and Pseudomonas), extracellular polymeric substance secretion (Terrimonas), aerobic chemoorganotrophy (Chryseobacterium and Novosphingobium), anoxygenic phototrophy (Rhodobacter), polyphosphate accumulation (Acinetobacter), denitrification (Acidovorax and Methylotenera), degradation of alkane (Aquabacterium and Hyphomicrobium), and aromatic compounds (Hydrogenophaga). The presence of these common bacteria might be due to their crucial roles for Cladophora dispersal and survival or the wide distribution of the bacteria themselves. To answer the question, a further investigation of their interactions is needed.
It is challenging to infer interactions between organisms living in the lotic environment as there is an influx and efflux of nutrients occurring continuously. This turbulence affects the richness and abundance of the organisms living in the algal microbiomes, e.g., [236][237][238]. Even so, data from this study revealed that Cladophora microbiomes were dominated by heterotrophic bacteria (32.57-37.17%), followed by photoautotrophs (8.20-14.68%), micrograzers (2.48-5.13%), parasites (1.52-2.38%), decomposing fungi (0.51-1.97%), and mesograzers (0.18-0.46%), as shown in Figure 7. The presence of similar relative abundances of organisms belonging to different trophic levels and decomposers suggested that there might be some unique interactions or activities present in the lotic freshwater Cladophora microbiomes that have not yet been investigated. In summary, we have shown that the Cladophora sp. that we sampled at three sites along the Nan River belongs to the same algal species. Our amplicon analyses revealed that this alga harbored diverse groups of microorganisms whose taxonomic richness and abundance varied among the sampled sites. About 50 percent of the identifiable taxa were shared among the Cladophora microbiomes. These organisms span different trophic levels in the food chain and putatively serve various ecological services.
Supplementary Materials: The followings are available online at https://www.mdpi.com/article/ 10.3390/plants10112266/s1, Figure S1: Alignment of 18S rDNA amplicons collected from CKD, PUA, and TD1 showing identical sequences of Cladophora sp. use in this study. Figure S2: The water quality along the Nan River collected from February 2019 to December 2020 from locations 1-4, span the collecting time in this study in March 2020. (a) values of the water quality index, (b) dissolved oxygen, (c) biological oxygen demand, (d) ammonia nitrogen content, and (e) the locations of the water sampling stations. Table S1: Primer sequences and PCR conditions for amplification of 16S rDNA, 18S rDNA, and ITS amplicons. Table S2: Top 100 BLASTN hit of the Cladophora 18S rDNA amplicon. Table S3: Bacterial genera and their abundance found in each replicate of collecting cites CKD, PUA, and TD1. Table S4: Eukaryotic genera and their abundance found in each replicate of collecting cites CKD, PUA, and TD1. Table S5: Fungi phyla and their abundance found in TD1. Table  S6: Common members of freshwater Cladophora microbiomes. Green boxes represent the presence of the corresponding taxa.