Diversity and distribution of Symbiodiniaceae detected on coral reefs of Lombok, Indonesia using environmental DNA metabarcoding

Background Dinoflagellates of family Symbiodiniaceae are important to coral reef ecosystems because of their contribution to coral health and growth; however, only a few studies have investigated the function and distribution of Symbiodiniaceae in Indonesia. Understanding the distribution of different kinds of Symbiodiniaceae can improve forecasting of future responses of various coral reef systems to climate change. This study aimed to determine the diversity of Symbiodiniaceae around Lombok using environmental DNA (eDNA). Methods Seawater and sediment samples were collected from 18 locations and filtered to obtain fractions of 0.4–12 and >12 µm. After extraction, molecular barcoding polymerase chain reaction was conducted to amplify the primary V9-SSU 18S rRNA gene, followed by sequencing (Illumina MiSeq). BLAST, Naïve-fit-Bayes, and maximum likelihood routines were used for classification and phylogenetic reconstruction. We compared results across sampling sites, sample types (seawater/sediment), and filter pore sizes (fraction). Results Phylogenetic analyses resolved the amplicon sequence variants into 16 subclades comprising six Symbiodiniaceae genera (or genera-equivalent clades) as follows: Symbiodinium, Breviolum, Cladocopium, Durusdinium, Foraminifera Clade G, and Halluxium. Comparative analyses showed that the three distinct lineages within Cladocopium, Durusdinium, and Foraminifera Clade G were the most common. Most of the recovered sequences appeared to be distinctive of different sampling locations, supporting the possibility that eDNA may resolve regional and local differences among Symbiodiniaceae genera and species. Conclusions eDNA surveys offer a rapid proxy for evaluating Symbiodiniaceae species on coral reefs and are a potentially useful approach to revealing diversity and relative ecological dominance of certain Symbiodiniaceae organisms. Moreover, Symbiodiniaceae eDNA analysis shows potential in monitoring the local and regional stability of coral–algal mutualisms.

Symbiodiniaceae cannot be directly identified using conventional microscopy. The need for collection and isolation from multiple locations increases the difficulty of assessing this taxonomic group. Advances in the use of environmental DNA (eDNA) and multitaxon sequencing techniques (metabarcoding) have allowed the study of Symbiodiniaceae communities through the collection of environmental samples, such as water and sediment (Arif et al., 2014;Shinzato et al., 2018;Fujise et al., 2021). The advantages of the eDNAbased approach include ease of use, noninvasive nature, broad spatial scale, and cost effectiveness (Deiner et al., 2017).
This study aimed to develop a rapid proxy for estimating the diversity of Symbiodiniaceae in water and sediment samples from the coral reef ecosystems around Lombok Island in Indonesia using eDNA. We make comparisons across sampling sites, eDNA source (seawater/sediment), and filter pore sizes (fraction). A better understanding of the diversity and composition of Symbiodiniaceae in Indonesian coral reefs is important for conservation and management of marine ecosystems.

Study sites
This study was conducted in coral reef habitats around Lombok Island, West Nusa Tenggara Province, Indonesia. This island is the constituent of the marine ecoregion of Nusa Tenggara (Lesser Sunda), which has a coral reef area of about 272,123 ha (Giyanto Abrar et al., 2017). The western part of Lombok Island is directly adjacent to Lombok Strait and its southern part is the Indian Ocean. The study areas were located 5-100 m from shore, with depths ranging 1-10 m and a mean tidal range of about 1.8 m. Samples were collected from 5th to 12th July 2018 (Table 1).

eDNA sample collection
During the survey, eDNA seawater and sediment samples were collected by scuba diving from six reef stations within each coastal area (West Lombok, East Lombok, and North Lombok). Two samples (one seawater and one sediment) per station were collected per day from three stations, in total, 72 samples in Lombok ( Fig. 1 and Tables 1 and 2). The distance between the sampling stations was at least 1500 m to avoid overlap. At each station, 4 L of seawater was collected from the water column (∼2 m above the reef substrate) as well as a sediment sample (water + sediment in 1:1 ratio) in sterilized bottles. Before sampling, the bottle was rinsed with a 30% commercial bleach solution, followed by distilled water. The collected eDNA samples were stored in a cool box and brought to basecamp at Lombok Island as soon as possible (less than 12 h). Each sample was filtered twice using a peristaltic pump (Thermo Fisher Scientific) through 47 mm diameter polycarbonate membrane filters (Sterlitech) with two different pore sizes: 12 µm first and then 0.4 µm. According to Turner et al. (2014), a combination of ≥ 0.2 µm filtration pore size and water volume enables optimal eDNA capture and maximize detection probability. In addition, a large pore size is required to avoid clogging the filters. The sediment samples, were shaken first and then

DNA extraction, amplification, and sequencing
The filtered eDNA samples were extracted and amplified at the Marine Biodiversity and Systematic Laboratory of IPB University and sequenced at the University of Rhode Island (URI) Genomics and Sequencing Center, United States of America. DNA was extracted from the filters using ZymoBiomics Miniprep Kit (Zymo Research, Irvine, CA, USA) following the manufacturer's instructions. V9 hypervariable regions of the eukaryotic small sub unit (SSU) 18S ribosomal RNA (rRNA) genes were amplified using a polymerase  et al., 2009;Stoeck et al., 2010), Illumina adapters, linker sequences, index, and pad (Kozich et al., 2013). The PCR profile used was as follows: 3 min at 94 • C, followed by 35 cycles of 94 • C for 45 s, 48 • C for 30 s, and 72 • C for 30 s, and a final extension at 72 • C for 5 min. Each 49 µL of PCR reaction comprised 25 µL of MyTM HS red mix (Bioline Ltd., London, UK), 1 µL of (10 µM) forward primer, 1 µL of (10 µM) reverse primer, and 1 µL of DNA template. The final volume was adjusted to 49 µL using ddH2O. 1x reaction was 0.2 µM. The PCR product was checked via the electrophoresis final master mix concentration in 1 × reaction was 0.8 ×, and the final primer concentration in of 5 µL of aliquots on 1% agarose gel in 0.5X TBE buffer. Library preparation and sequencing were performed at URI. A second PCR was performed to add the dual indices and Illumina sequencing adapters from the TruSeq PCR-Free LT kit to the target amplicons, using Kapa HotStart HiFi 2x ReadyMix DNA polymerase (Kapa Biosystems Ltd., London UK). The PCR profile used was as follows: initial denaturation at 95 • C for 3 min, followed by 9 cycles of 95 • C for 30 s and 55 • C for 30 s, and final extension at 72 • C for 5 min. The presence and length (bp) of the PCR product or amplicon were tested by electrophoresis. Successful amplicons were then purified using paramagnetic Kapa pure beads (bead-to-sample volumetric ratio in 1.6:1). A Qubit fluorometer with Qubit dsDNA HS Assay reagent (Invitrogen, California, US) was used to quantify all libraries. The prepared samples were combined in equal concentrations and then pooled with a 20% denatured and diluted PhiX Illumina control library. The final pooled library was sequenced on an Illumina MiSeq with the MiSeq v2 500-cycle kit (Illumina, San Diego, CA, United States). After quality checking, only 41 out of 72 samples were found to be of sufficiently high quality for sequencing ( Table 2). The low quality of some libraries may be due to eDNA degradation during sample transport and extraction.

Data processing and bioinformatic analyses
The obtained forward and reverse raw sequence data were converted to demultiplexed fastq files (see additional information on data availability). The sequence read quality was checked using FastQC v.0.11.8 (https://www.bioinformatics.babraham.ac.uk) at each analysis step. Cutadapt v.1.18 (Martin, 2011) was used to trim the reverse and forward primer sequences and remove short reads with lengths <100 bp and low quality reads with a Phred Q score of <20. Qiime2.2019.10 pipeline (Caporaso et al., 2010;Bolyen et al., 2019) was employed for further data processing. DADA2 v.2018.11.0 (Callahan et al., 2016) (via q2-dada2) was applied for denoising, joining denoised paired-end reads, filtering out chimeric sequences and singletons, and dereplicating sequences to produce amplicon sequence variants (ASVs). Owing to the high quality of the sequences obtained after Cutadapt procedure, trimming and truncating were not performed during DADA2 processing.

ASV identification
Symbiodiniaceae species were identified from the eDNA sequences by classifying all ASVs (File S1) using the q2-feature-classifier (Bokulich et al., 2018)

Statistical analyses
The relative abundance data for the putative Symbiodiniaceae taxa (File S5) were from DADA2 results and used as input for Venn diagram and statistical analyses. Venn diagram analyses were performed using the online application in http://bioinformatics.psb.ugent. be/webtools/Venn/ to compare the Symbiodiniaceae individuals across different locations (coastal area), eDNA source (seawater and sediment samples), and fractions (filter pore size). We determined the most commonly distributed subclades and distinctive subclades to each location/station. Statistical analyses were used to compare Symbiodiniaceae abundance, diversity, and features observed across different sites, sample types, and fractions. All these statistical analyses were carried out on Qiime2.2019.10 pipeline (Caporaso et al., 2010;Bolyen et al., 2019). Alpha diversity (observed features and Shannon's   Fig. 3). *Reference database. ** Confidence level. *** Percentage identity. entropy) and beta diversity (Bray-Curtis dissimilarity) were estimated using q2-diversity after the samples were rarefied (subsampled without replacement) to 28 sequences per sample. The comparison of all samples were grouped by location, eDNA source, and fraction to examine differences in abundance and alpha diversity employing the Kruskal-Wallis test (Kruskal & Wallis, 1952) and beta diversity applying the Permanova test (Anderson, 2001) using 9999 permutations.

Obtained sequences, ASVs, and eukaryote classification
From the 72 samples across 18 stations, DNA was successfully extracted from 41 samples at 16 stations, yielding a total of 3,168,655 raw sequences and about 30,205-240,604 sequences per sample (Table 2 and Fig. S1). DADA2 yielded a total of 20,486 ASVs (File S1). The mean length of the obtained sequences was 127.81 ± 22.03 bp. The ASV classification demonstrated the potential diversity of eukaryotes in the reef waters of Lombok Island. According to the total ASVs classified to taxon level 4 from the SILVA database, the dominant taxon was unclassified eukaryotes (43.35%), followed by Metazoa (9.47%), Ochrophyta (7.83%), Dinoflagellates (4.5%), and Discicristata (4.4%) (Fig. 2). Table 3 summarizes the results of Symbiodiniaceae classification performed using a Eukaryote classifier (File S2) and BLAST (File S3) and phylogenetic analyses. The probabilistic classifier detected and classified the Symbiodiniaceae taxa at the family level. Twenty-two ASVs (named OTU.sym1 to OTU.sym22) were found to be putative Symbiodiniaceae with confidence levels ranging 0.743-0.999 (Table 3). BLAST results indicated that some ASVs were neither Symbiodiniaceae nor classified at the genus level. A partial phylogenetic reconstruction of the families in order Suessiales was conducted using the reference sequences obtained from the searched databases (File S4) and the putative Symbiodiniaceae ASV sequences from the study (Fig. 3A). Only 16 out of the 22 ASVs were identified as members of the monophyletic group of the Symbiodiniaceae family clade on the basis of the score of 100/100/1/99 for SH-alrt/lbt/abayes/ufboot. Three of the six remaining ASVs were categorized in the clades representing genera in Family Suessiaceae, two ASVs were in the Yihiella clade (OTU.sym11 and OTU.sym13), and one was in the Ansanella clade (OTU.sym20). The remaining three ASVs were designated to the Suessiaceae family but were not classified at the genus level (OTU.sym3, OTU.sym9, and OTU.sym14).

Symbiodiniaceae distribution and diversity
Venn diagrams show the overlap of the 16 ASVs belonging to Symbiodiniaceae according to location (Fig. 4A) and sample type (media and fractions) (Fig. 4B) (see also File S6). The presence/absence table shows the Symbiodiniaceae proportion per subclade by site-sample type-filter pore size combination (Table 4 and File S7). This table illustrates the common and unique subclades of Symbiodiniaceae. The unique subclades were the sequences distinctive of sampling location, medium, and fraction. Three subclades were most common (C.sym1, D1.sym2, and G2.sym4), and the remaining subclades were unique ( Table 4). The unique subclades (<11.11% of subclade presence in all samples) showed site-or sample type-specificity. C.sym1 was the most common (77,78%) and was detected at more sites-media-fractions than D1.sym2 (44.44%) and G2.sym4 (33.33%). In term of medium, the sediment samples yielded more Symbiodiniaceae subclades than seawater (12 vs. 7 subclades), with nine unique ASVs found in the sediment medium. On the basis of Symbiodiniaceae relative abundance, genus Cladocopium was the most dominant (Fig. 5). In general, the Symbiodiniaceae communities of Lombok were characterized with low alpha diversity and high beta diversity (Fig. 6). However, comparison of Symbiodiniaceae abundances, observed features, and diversity does not show significant difference between locations, media, and fractions (see File S8).

DISCUSSION
The results illustrate the potential of eDNA to detect Symbiodiniaceae. The eDNA of Symbiodiniaceae can be obtained from different sources including freeliving Symbiodiniaceae (Hirose et al., 2008;Littman, Van Oppen & Willis, 2008) and Symbiodiniaceae living in symbioses with various host organisms (Freudenthal, 1962;Loh, Cowlishaw & Wilson, 2006;Barneah et al., 2007;LaJeunesse et al., 2010a;LaJeunesse et al., 2010b;LaJeunesse et al., 2018;Pochon & Gates, 2010;DeBoer et al., 2012;Pochon, Putnam & Gates, 2014;Ramsby et al., 2017). Additionally, these eDNA sources could come from within and outside the sample site (Goldberg et al., 2016). Symbiodiniaceae DNA   could be obtained from prey organism feces and through the shedding of host cells in the water and sediment (Rees et al., 2014;Grupstra et al., 2021). The SSU 18S rRNA gene primer set has long been used in the biomolecular studies of Symbiodiniaceae (Rowan & Powers, 1991;Loh, Cowlishaw & Wilson, 2006). Hypervariable regions V4 and V9 isolated and then amplified by the SSU 18S rRNA gene universal primer were successful in detecting and identifying Symbiodiniaceae from water samples (Stoeck et al., 2010). This study used the same V9-SSU 18S rRNA gene primer set for oceanic planktonic Symbiodiniaceae by the Ocean TARA Expedition. The substitutions in the hypervariable terminal loop region amplified by this primer allowed us to distinguish Symbiodiniaceae genera and subclades (Decelle et al., 2018). Other primers such as ITS, LSU 28S, and chloroplast primers can be used to provide high taxonomic resolution for Symbiodiniaceae (Venera-Ponton et al., 2010;Takabayashi et al., 2012;Arif et al., 2014). Nevertheless, this study succeeded in detecting and identifying Symbiodiniaceae at the genus level.
The use of universal eukaryote primers with eDNA samples can reveal information on the rich diversity of marine life and compensate for the high cost of next-generation sequencing (Smart et al., 2016;Bálint et al., 2018). Universal primers allow us to broadly look at the system and complete more than a single study using the same data (Madduppa et al., 2021). The lack of field blanks (non-reef sampling areas) and filter blanks (distilled water or sterile seawater samples), might influence our study results. Lack of control/blanks can lead to contamination of the eDNA source, or false-positive data. However, the comparitive analyses across the given samples allowed the evaluation of the possibility of exogenous and local eDNA sources. Moreover, the presence of contaminant DNAs was likely suppressed by rinsing the instruments (e.g., bottle samples and filtering tools) with bleach to make them as sterile as possible.
To the authors' knowledge, this work is the first study of Symbiodiniaceae using eDNA in Indonesia and Southeast Asia. Symbiodiniaceae in the Southeast Asia region have been identified from scleractinian stony corals, sea slugs, giant clams, and other bivalves, sea anemones, sponges, zoantharians, antipatharian black corals, and Heliopora blue corals. At least seven Symbiodiniaceae genera have been discovered in Southeast Asia (Table 5). Various primers, such as nuclear primers, mitochondrial organelle primers, and chloroplast primers, and a range of molecular techniques such as single stranded conformational polymorphism, restriction fragment length polymorphism, and denaturing gradient gel electrophoresis, have been used in the identification and characterization of the genetic diversity of Symbiodiniaceae in the region but did not detect as many genera as the present study did (see Table 5). No report was found about the genus Effrenium and Clade I in Southeast Asia. However, clade E (AF238261.1) in our phylogeny (Fig. 3) was assigned to clade D1 by Kimes et al. (2013). E. voratum is the only species from Effrenium that was previously described and is only found in temperate waters (Jeong et al., 2014). LaJeunesse, Parkinson & Trench (2012) predicted that the Southeast Asia region might have a higher diversity of Symbiodiniaceae species than other regions in the world. Previous and current findings supports this prediction (Loh, Cowlishaw & Wilson, 2006;Bo et al., 2011;DeBoer et al., 2012;Purnomo, 2014). Therefore, other under-sampled coral reef areas in Indonesia should be further explored.
The detected Symbiodiniaceae in the study sites are probably coral endosymbionts. Some species of Symbiodinium, Breviolum, Cladocopium, and Durusdinium are the main coral endosymbiont genera, and species of Fugacium and Gerakladium are rare endosymbionts in corals (LaJeunesse et al., 2010a;LaJeunesse et al., 2010b;Rouzé et al., 2017). The main coral endosymbionts, especially in Indo-Pacific, are species of Cladocopium and Durusdinium; meanwhile members Symbiodinium and Brevolium are common in corals in the Caribbean (Baker, 2003;LaJeunesse et al., 2004;LaJeunesse et al., 2010a;LaJeunesse et al., 2010b;LaJeunesse, 2005;Stat & Gates, 2011). Many members of Cladocopium (e.g., ITS2 subclade C1) generally have high rates of carbon fixation, provide a high fitness benefit, translocate high amounts of carbon to host corals, and positively impact host coral growth rates. By contrast, some species of Durusdinium tend to be opportunistic, even though they can help corals to survive or quickly recover from bleaching when sea surface temperatures rise (Stat, Morris & Gates, 2008;Stat & Gates, 2011;Lesser, Stat & Gates, 2013;Bay et al., 2016).
This study detected the three most common subclades, namely C.sym1, D1.sym2, and G2.sym4. These subclades may represent the most common species or types of Symbiodiniaceae. BLAST results showed that C.Sym1 was similar to C. goreaui (99.24%), formerly clade C type C1, which is a generalist Symbiodiniaceae found in many coral hosts in the Great Barrier Reef (LaJeunesse, 2005;Bongaerts et al., 2015). The sequence of D1.sym2 detected by BLAST has 100% sequence similarity with the molecular marker of D. trenchii, a Symbiodiniaceae species that increases the tolerance of corals to bleaching stress (Stat & Gates, 2011). Previous studies have suggested the importance of a minimum density of D. trenchii as a minority component alongside a dominant endosymbiont from the genus Cladocopium in the Symbiodiniaceae community within a coral colony (Bay et al., 2016). However, Swain et al. (2017 found that each genus of Symbiodiniaceae has the potential for heat-resistant species or variants. For example, C. thermophilum is a thermotolerant variant of Cladocopium type C3 (Hume et al., 2015).
This study fully resolved the ASV of subclade G2.sym4 within the Foraminifera Clade G (formerly clade G type G2). This genus can be isolated from the foraminifera, particularly in Subfamily Soritinae (Pochon et al., 2007). Bo et al. (2011) also isolated a subclade close to type G2 from Indonesian octocorals. Foraminifera Clade G is a common endosymbiotic Symbiodiniaceae in sponges, such as bio-eroding sponge (Cliona orientalis) in Australia   (Schönberg & Loh, 2005;Ramsby et al., 2017). However, G2.sym4 appears to be a common type and is also found in the sediment samples. Therefore, this subclade may be an endosymbiont of benthic foraminifera. Foraminifera communities around Lombok are diverse, widely distributed, and present in the seabed in shallow coastal waters around the island (Auliaherliaty, Dewi & Priohandono, 2004;Natsir, 2009;Natsir, 2010;Dewi et al., 2012). However, no studies of foraminifera endosymbiotic Symbiodiniaceae in Indonesia have been published.
The detected Halluxium in this study is the first record in the Southeast Asia region. To date, Halluxium has only been found in Guam, Heron Island (Great Barrier Reef, Australia), and the Caribbean (Pochon, LaJeunesse & Pawlowski, 2004;Pochon et al., 2007;Nitschke et al., 2020). This genus and Clade I are generally foraminifera-specific endosymbionts. Meanwhile, Breviolum or Effrenium species living as foraminifera endosymbionts have never been reported (Pochon & Pawlowski, 2006;Pochon & Gates, 2010).
The richer Symbiodiniaceae subclades in sediment than in seawater indicate the potential occurrence of benthic Symbiodiniaceae. These Symbiodiniaceae can have important implications for the coral reef ecosystems of Lombok. The benthic sediment can be a source of free-living Symbiodiniaceae that live outside the host (Hirose et al., 2008;Littman, Van Oppen & Willis, 2008;Fujise et al., 2021). Some of these can (re-) establish stable host-algal mutualisms (transient free-living), and others are true free-living, such as E. voratum (Yamashita & Koike, 2013;Jeong et al., 2014). Some transient free-living Symbiodiniaceae can come from expelled coral endosymbionts. Corals regularly expel some of their endosymbionts into the seawater column (Fujise et al., 2014), most of which are deposited in sediments. The other source of transient free-living Symbiodiniaceae is reef fishes. Corallivorous, detritivorous, and herbivorous fishes can contribute to the release and distribution of transient free-living Symbiodiniaceae in their habitat through their feces (Castro-Sanguino & Sánchez, 2012;Grupstra et al., 2021). The availability of such Symbiodiniaceae in the environment is essential. During larval stage and/or recruitment time, most corals horizontally obtain transient free-living Symbiodiniaceae from the nearby environment (Coffroth et al., 2006;Fujise et al., 2021). The presence of such Symbiodiniaceae can also influence juvenile coral survival (Suzuki et al., 2013).
This study found that 13 of the 16 subclades were distinctive of different sampling locations. These subclades may represent the species or types of Symbiodiniaceae originating from local sources. Environmental genetic materials are prone to degradation (Barnes & Turner, 2016), so they tend to accumulate around the source. Therefore, eDNA is representative of the local biotic genetic material. Shinzato et al. (2018) showed the feasibility of studying nearby coral species and their symbiotic algae detection using eDNA; therefore, it might also be used to monitor coral ecosystem health. However, such data must be carefully interpreted because of some issues regarding the possible sources of eDNA from outside the sample site due to biological factors and human activities (Goldberg et al., 2016).
The eDNA method also has some limitations, such as the dependence on the presence and concentration of eDNA in the water sample, capture efficacy, extraction efficacy, sample interference (e.g., inhibition), and assay sensitivity (see Goldberg et al., 2016). Seawater eDNA samples can degrade beyond the detection threshold within 1 day to weeks (Dejean et al., 2011;Thomsen et al., 2012). Water quality conditions, such as high temperatures, neutral pH, and moderately high UV-B, tend to increase the eDNA degradation rate (Strickler, Fremier & Goldberg, 2014). However, the degradation rate of eDNA in aquatic environments is different from that in sediments. The nature and proportion of minerals, organic substances, and charged particles adsorbing eDNA fragments influence the rate of eDNA degradation in sediments and protect them from further destruction. A previous study showed that the degradation rate of eDNA in sediment is about 57 times slower than that in seawater (Torti, Lever & Jørgensen, 2015;Turner, Uy & Everhart, 2015;Sakata et al., 2020). Limited information is available regarding the factors that influence the rate of symbiont DNA shed by coral reef taxa and maintained in the water column over spatial scales.

CONCLUSIONS
This study demonstrates that eDNA surveys can describe the potential diversity of Symbiodiniaceae in the reefs around Lombok. Six genera (or genera-equivalent clades) of Symbiodiniaceae were identified. eDNA survey has higher sensitivity than traditional methods and thus offer a rapid proxy for evaluating Symbiodiniaceae communities across different coral reefs. This approach can also be used to enhance the understanding of the diversity and relative ecological dominance of certain Symbiodiniaceae members. Moreover, the presence of distinctive Symbiodiniaceae individuals in different locations support the potential application of eDNA for monitoring the local and regional stability of coral-algal mutualisms. Further confirmation through isolation from a variety of sources (including possible hosts) and microscopic observations is warranted to strengthen the evidence for local eDNA sources.
• Beginer Subhan performed the experiments, authored or reviewed drafts of the article, and approved the final draft.
• Hawis Madduppa conceived and designed the experiments, performed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the article, and approved the final draft.

Field Study Permissions
The following information was supplied relating to field study approvals (i.e., approving body and any reference numbers): eDNA seawater sampling in this study was permitted within the framework of the

Data Availability
The following information was supplied regarding data availability: