Characteristics of Macroalgal Consumption by Eight Herbivorous Coral Reef Fishes From the Xisha Islands, China, as Determined by Microscopy, 18S rRNA High-throughput Sequencing and Stable Isotope Analyses


 Herbivorous fishes play an important role in controlling the overabundance of macroalgae on coral reefs. Understanding the feeding selectivity and consumption of macroalgae by herbivorous fishes can be challenging in studies of their ecological role in the preservation and recovery of coral reefs. Coral reef decline, macroalgal overgrowth and overfishing are clearly visible in the Xisha Islands, China. However, there have been no studies of the feeding behaviors of herbivorous fishes in this area. We used microscopy, 18S rRNA high-throughput sequencing and stable isotope analyses to comprehensively examine the dietary spectrum of eight herbivorous reef fish species common in the Xisha Islands, including one parrotfish, two chub, two unicorn fish and three rabbitfish. Multi-technique analyses of intestinal contents revealed that Kyphosus vaigiensis, Naso unicornis and Siganus argenteus showed a high consumption potential of macroalgae, suggesting that they are the key browsers which should receive priority protection in in the Xisha Islands. Kyphosus cinerascens, K. vaigiensis, N. unicornis and S. punctatissimus fed on the entire macroalgal thallus, indicating their greater ecological importance compared with species which only consume the algal fronds. However, Calotomus carolinus can consume the red alga Pneophyllum conicum, which is widely distributed on Indo-Pacific coral reefs and can overgrow and kill live corals. Clearly, a diverse herbivorous fish fauna is very important in the Xisha coral reefs. These results not only demonstrated the various functions of different herbivorous fish species in macroalgal removal, but also provided insights into the management of herbivorous fishes on the coral reefs of the South China Sea.


Introduction
Coral reefs are often called the tropical rainforests of the ocean and are the most productive and biologically diverse ecosystem in the world, supporting more than 25% of all known marine species (Moberg and Folke 1999;Walker and Wood, 2005). However, the world's coral reefs are in serious decline due to numerous anthropogenic activities and global climate disturbances (Wilkinson, 2008;Liu et al. 2021). Nineteen percent of the world's coral reefs were lost in the 4 years of [2004][2005][2006][2007][2008], and only 46% of the world's reefs were considered to be in a relatively healthy condition during that period (Wilkinson, 2008). The decline of coral reefs can involve a shift from coral towards macroalgae as the dominant feature (Cheal et al. 2010;Kopp et al. 2010). The majority of China's coral reefs occur around the Xisha Islands in the Indo-Paci c region and have suffered recent dramatic declines in coral cover . Many parts of the Xisha Islands' reefs are now dominated by macroalgal growths (Chen et al. 2019). A predominance of macroalgae hinders the settlement, survival and growth of corals, so that a healthy coral reef ecosystem is very di cult to reestablish (Puk et al. 2016;Dell et al. 2020). Phase shifts in the balance of coral/macroalgal species have been anecdotally attributed insu cient herbivore activity, eutrophication, environmental disturbances (e.g. typhoons, ocean warming and coral bleaching, etc.), bottom-up factors (nutrients pollution) or a combination of these factors (Cheal et al. 2010;Russ et al. 2015;Neilson et al. 2018;Bruno et al. 2019;Adam et al. 2021).
Herbivorous reef shes are the major consumers of macroalgae on coral reefs (Cheal et al. 2010).
Parrot sh (Scarus and Sparisoma spp.), chub (Kyphosus spp.), unicorn sh (Naso spp.) and rabbit sh (Siganus spp.) are recognized as the important macroalgal grazers on coral reefs (Lefevre and Bellwood, 2011;Puk et al. 2016;Dell et al. 2020). The feeding preferences of herbivorous reef shes differ at the species level (Hoey and Bellwood, 2009;Duran et al. 2019;Dell et al. 2020). Parrot sh appear to prefer green and red algae, but some taxa (e.g., Naso spp. and Kyphosus spp.) generally target brown macroalgae (Puk et al. 2016). Of the over 50 herbivorous sh species identi ed off Lizard Island, northern Great Barrier Reef, only one, Naso unicornis, fed on the erect brown macroalga Sargassum across all habitats (Hoey and Bellwood, 2009). A study of the feeding behaviors of four rabbit sh species, indicated that Siganus argenteus and S. sutor were generalist herbivores, foraging on turf algae, macroalgae, seagrass and epiphytic algae, while S. corallinus and S. stellatus were specialist herbivores foraging primarily on turf algae growing on the reef substrate (Ebrahim et al. 2020). Herbivore species richness appears to be critical in generally lowering macroalgal abundance because of the complementary feeding habits of a diverse assemblage of herbivores (Burkepile and Hay, 2008). Therefore, it is essential to understand the dietary spectra of the various herbivorous reef shes to understand their cumulative effect on coral reef health. However, very little is currently known about the macroalgal feeding preferences of the herbivorous reef shes on the coral reefs off the Xisha Islands, and this limits our understanding of their different functional roles in controlling the spread of macroalgae.
The feeding ecology of herbivorous shes on coral reefs is usually determined by behavioral observations, counting the number of bites taken by sh feeding in the eld (Mantyka and Bellwood, 2007;Dell et al. 2020;Ebrahim et al. 2020), or by microscopic examination of intestinal contents (Dromard et al. 2015). These methods are convenient and provide detailed information on the dietary spectrum of different species (de Carvalho et al. 2019). However, there are some practical problems; small, fragile, or morphologically indistinct organisms are di cult to identify in intestinal samples, especially after digestion (Dromard et al. 2015;Kume et al. 2021). Stable isotope analysis is a powerful tool which can reveal feeding behavior over an extended period, although with some taxonomic limitations (Rodriguez-Barreras et al. 2020;Kume et al. 2021). High-throughput sequencing is emerging as a molecular method to estimate sh dietary composition by identifying the taxa eaten from genomic DNA recovered from sh intestines (Corse et al. 2010;Leray et al. 2015). It is more effective when used in conjunction with reference sequence databases such as GenBank (Devloo-Delva et al. 2019). Using a combination of these three methods allows comprehensive descriptions of the dietary composition of various sh species.
Coral cover off the Xisha Islands has declined dramatically over the past 15 years ) and macroalgae have become dominant in many of their coral reef areas (Chen et al. 2019). Furthermore, over shing has become a serious threat for these reefs (Zhao et al. 2016). We explored the dietary differences of eight different herbivorous reef shes common around the Xisha Islands, based on microscopic examination, 18S rRNA high-throughput sequencing and stable isotope analyses. We focused on comparisons of the macroalgal feeding selectivity of herbivores, including one parrot sh (Calotomus carolinus -CC), two chub (Kyphosus cinerascens -KC, and K. vaigiensis -KV), two unicorn sh (Naso brevirostris -NB, and N. unicornis -NU) and three rabbit sh (Siganus argenteus -SA, S. puellus -Spe, and S. punctatissimus -SPn). The results provided key information on the ecological function of these herbivorous shes in removing macroalgae from coral reefs, and gave insights into the most effective over shing prevention measures to keep the coral reef ecosystem at its most healthy.

Materials And Methods
Study site and sh sample collection The Xisha Islands, in the central South China Sea, are derived from coral reefs and comprise over 40 islands, reefs and cays, including the Dongdao Atolls, the Huaguang Atolls, the Xuande Atolls, the Yongle Atolls, and some smaller islands (Zhao et al. 2017;Ding et al. 2019;Zhao et al. 2019). The Qilianyu Islets and Cays (16°59′ N, 112°18′ E) of the Xisha Islands comprise an arced reef at, which extends in a NW-SE direction and curves to the NNE (Shen et al. 2017). Over 100 sh species have been recorded off the Qilianyu Islets and Cays, with the dominant species belonging to the families Pomacentridae, Labridae and Scaridae (Li et al. 2017). Eight herbivorous reef sh species common in the Qilianyu Islets and Cays were collected by SCUBA in June 2020. The details of the shes sampled in this study, along with their phylogenetic classi cation and feeding strategies, are shown in Table 1. The feeding strategy of each species was determined based on the previous studies. All of the shes collected were kept at -4°C and transported to the laboratory, where they were dissected using sterile scissors. The anterior intestinal contents were separated and divided into two subsamples. One subsample was used for microscopic examination, and the another was used for 18S rRNA high-throughput sequencing analysis. A small piece of muscle tissue was sampled from each sh and used for stable isotope analysis.

Intestinal contents for microscopical analysis examination
Intestinal contents of each sh species were collected for dietary assessment. Photographs of the food items were taken using a microscope (Zeiss SteREO Discovery.V20, Germany). Macrophytes (Macroalgae and seagrasses) were identi ed according to their morphological traits (Titlyanov et al. 2017;Huang, 2018).

Composition of intestinal contents measured by 18S rRNA high-throughput sequencing analysis
The total DNA of intestinal contents (0.2 g sample) of individual sh was extracted using a QIAamp ® Fast DNA Stool Mini Kit (Qiagen, Germantown, MD, USA) according to the manufacturer's protocols. The total DNA recovered from each sh species was pooled for PCR ampli cation which targeted the V4 region of the eukaryotic 18S rRNA gene using primers 528F (5'-GCGGTAATTCCAGCTCCAA-3') and 706R (5'-AATCCRAGAATTTCACCTCT-3'), where the barcodes were an eight-base sequence unique to each sample (Cheung et al. 2010). Amplicons were extracted from 2% agarose gels and puri ed using the AMPure XP Beads (Beckman, Agencourt, USA). These puri ed amplicons were pooled in equimolar paired-end sequences (2 × 250) on an Illumina platform (Gene Denovo Co., Guangzhou, China). Raw reads for each sh species were deposited into the NCBI Sequence Read Archive (SRA) database with the accession number PRJNA742779. The representative sequences a liated to the phyla Chlorophyta, Ochrophyta and Rhodophyta obtained in this study were deposited in the GenBank database under accession numbers MZ481947-MZ481964, MZ481932-MZ481946 and MZ481965-MZ482014, respectively.
Raw reads were ltered using FASTP (V0.18.0) to obtain high quality clean reads according to the following rules: (1) remove reads containing more than 10% of unknown nucleotides and (2) remove reads containing less than 50% of bases with a quality (Q-value) 20 . Paired-end clean reads were merged as raw tags using FLSAH (V1.2.11) with a minimum overlap of 10 bp and mismatch error rates of 2% or less (Magoc and Salzberg, 2011). Noisy sequences of raw tags were then ltered using QIIME (V1.9.1) (Caporaso et al. 2010) under speci c ltering conditions (Bokulich et al., 2013) to obtain high quality clean tags. Clean tags were searched against the reference database (http://drive5.com/uchime/uchime_download.html) to perform reference-based chimera checks, using the UCHIME algorithm (Edgar et al. 2011). All chimeric tags were removed to nally obtain effective tags for further analysis.
The effective tags were clustered into operational taxonomic units (OTUs) of ≥ 97% similarity using UPARSE (V9.2.64) (Edgar, 2013). A dominant sequence was selected within each cluster as a representative sequence. The representative sequences were then picked-out to annotate taxonomic assignments using the RDP classi er (V2.2) (Wang et al. 2007) based on the SILVA database (V132) (Pruesse et al. 2007). The OTU sequences belonging to class Actinopteri were discarded in further analysis. Moreover, if an OTU sequence could not be assigned to any phylum in the SILVA database, it was subjected to a further BLAST search (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The taxonomic classi cation was then nally con rmed by its similarity to a BLAST result.

Stable isotope analysis
Fish muscle tissues were dried at 60°C to a constant weight. Samples were then ground to a ne, homogeneous powder using an automatic sample grinder (Jxfstprp-24, Jingxin Co., Shanghai, China). Samples were taken for carbon and nitrogen stable isotope analysis using a continuous-ow isotope ratio mass spectrometer (Finnigan MAT 253, Thermo Scienti c, USA) coupled to an elemental analyzer (Flash EA 1112, Thermo Scienti c, USA). The C and N isotope ratios were determined as δ 13 C and δ 15 N, respectively, according to the following formula: where R is the corresponding ratio 13 C/ 12 C or 15 N/ 14 N, R sample is measured for sh, and R standard is an international standard (Pee Dee Belemnite for C isotopes and atmospheric N 2 for N isotopes).

Data analysis
Bar charts of the community composition and stable isotope signatures (δ 13 C and δ 15 N) of sh muscles were constructed using Origin 2018 software (OriginLab Co., Northampton, MA, USA). A phylogenetic tree of 18S rRNA sequences related to the phyla Chlorophyta, Ochrophyta and Rhodophyta was constructed using MEGA X and the neighbor-joining algorithm, and the maximum composite likelihood method with bootstrap analyses for 1,000 replicates (Kumar et al. 2018). A heat map of the relative abundance of the 18S rRNA sequences belonging to the phyla Chlorophyta, Ochrophyta and Rhodophyta was individually generated using the OmicShare tools (http://www.omicshare.com/tools). Dominant OTUs from the heat maps were selected to analyze taxonomic assignments using BLAST research.

Microscopic observations of sh intestinal contents
Intestinal contents of the eight sh species collected were analyzed according to microscopic observations. After thorough washing, anterior intestinal contents of shes were examined and a high proportion of occulent detritus was found. Coral sand was only detected in the intestinal contents of CC and SA (data not shown). Importantly, macrophytes (macroalgae or seagrasses) were easily observed in sh intestines ( Table 2). The number of macrophyte species consumed by the eight herbivorous reef shes varied. SA consumed the highest number of macrophyte species (six species), followed by NU (four species). However, only a few macrophyte species could be identi ed according to their morphological traits, while other fragments belonging to Chlorophyta, Ochrophyta or Rhodophyta were too small and were di cult to identify. The green algae Valonia ventricosa and Halimeda sp. were detected in the sh intestines of NU and SA, respectively. The brown alga Turbinaria ornata and the red alga Acanthophora spp. were observed in the intestinal contents of NU and SPn, respectively. In addition, the seagrass Halophila ovalis was identi ed as a food source of SA.

Molecular detection of macrophyte species composition in sh intestines
The dietary spectra of the eight shes were analyzed with 18S rRNA gene high-throughput sequencing at the phylum and class levels (Fig. 1). The 15 most abundant phyla accounted for 88.2%-99.3% of the total sequences found in the eight sh species' intestines (Fig. 1a). There were marked differences in the relative abundance of the dominant phyla in each species sample. Porifera was the most abundant taxon in CC, which was further identi ed as Demospongiae (sponges) (Fig. 1). The dominant phyla in KC were Bacillariophyta and Porifera. Ochrophyta, Streptophyta and Bigyra were signi cantly more abundant in KV, SA and SPn, respectively. However, Cnidaria, mainly Anthozoa (corals), were much higher in SPe, NB, NU and CC (Fig. 1). Sequences belonging to Streptophyta were identi ed as Halophila ovalis, a seagrass species which occurred in the intestines of CC, NB, SA and SPn.
Sequences referring to the phyla Chlorophyta, Ochrophyta and Rhodophyta found in the eight shes were further analyzed (Fig. 2). A total of 18, 15 and 47 OTUs were observed belonging to Chlorophyta, Ochrophyta and Rhodophyta, respectively (Fig. 2a). A phylogenetic tree of the observed OTUs was constructed and is shown in Fig. 3. Chlorophyta OTUs were most diverse in KV and SA (≥10 OTUs). The highest numbers of Rhodophyta OTUs (23 OTUs) were observed in KC and KV. Ochrophyta OTUs were most abundant in KV (11 OTUs). The relative abundance of consumed macroalgae found in sh intestines are shown in Fig. 2b. KV consumed the greatest abundance of macroalgae (76.9%), followed by KC (14.5%). NU preferred to take brown algae, while SPn consumed more green algae. The diet of SA contained 2.9% green algae, 0.8% brown algae and 1.9% red algae.
The differences in the macroalgal composition between the intestinal contents of the eight sh species were analyzed (Fig. 4 and Table 3). The species composition of green algae or red algae in the eight herbivorous shes diets was quite different (Fig. 4a and c). Regarding the green algae, Otu000030, identi ed as Dictyosphaeria cavernosa, was abundant in KV. Otu000055, which was very similar to Ulvella leptochaete, was a dominant component in SA and KC. Both Otu000067 and Otu000102 were a liated to the genus Cladophora. The occurrence of Otu000067 was higher in KC, while Otu000102 dominated in NB and SA. Regarding the red algae (Fig. 4c), Otu000056 dominated in SA and showed a high similarity to Spyridia lamentosa. Otu000077 was related to Centroceras hyalacanthum, and was a major component in SA and SPn. Otu000094 matched with Ceramium sinicola and was relatively abundant in KC. Otu000200 and Otu000260 were relatively abundant in KV and were similar to Peyssonnelia rumoiana and Peyssonnelia armorica, respectively. Otu000535 was identi ed as the red alga Peyssonnelia rosenvingei and was very abundant in CC. The intestinal contents of CC also contained a dominant OTU (Otu000554), which was matched with Pneophyllum conicum. The dominant OTUs in NU included Otu000309 and Otu000874, which were similar to Chondrophycus cf. undulates and Ceramium sinicola, respectively. Regarding the brown algae (Fig. 4b), Otu000009, related to Lobophora variegate, was dominant in the intestinal contents of all the shes, except SA. Otu000014 was also abundant in most shes, especially SA in which it was most abundant.

Discussion
Microscopic examination and high-throughput sequencing analyses of sh intestinal contents Microscopic examination and 18S rRNA high-throughput sequencing analyses were used to investigate the dietary composition of eight herbivorous reef sh species off the Xisha Islands, China. There were some similarities in the results of the two methods. The seagrass Halophila ovalis was observed in the intestinal contents of SA based on microscopic examination (Table 2), and high-throughput sequencing analysis found that this seagrass (Liliopsida) also dominated the intestinal contents of SA (Fig. 1b).
Microscopic examination revealed coral sand in the intestinal contents of CC and SA (data not shown). High-throughput sequencing analysis also showed a high abundance of the class Anthozoa (corals) (Fig. 1b).
However, 18S rRNA high-throughput sequencing analysis provided higher taxonomic resolution of dietary composition, compared with microscopic examination. For example, some taxa which were found to dominate the intestinal contents using molecular analysis, such as Porifera, Cnidaria, Bacillariophyta, etc. (Fig. 1a) were not detected by microscopic examination. The possible reasons were as follows. First, small organisms or digested organisms are too di cult to identify using a microscope (Dromard et al. 2015;Kume et al. 2021). In this study, microscopic examination found a large amount of detritus in the sh intestinal contents (data not shown). Dromard et al. (2015) also observed a very large proportion of unidenti ed detritus in the intestines of Scaridae using a microscope. Second, high-throughput sequencing is highly sensitive and can detect traces of DNA in mixed samples. This molecular technique can discriminate OTUs even with only a single nucleotide variation within the targeted ampli ed region (Albaina et al. 2016). Third, the number of sampled shes in this study was small and some taxa were not observed using the microscope. For example, previous studies have shown that SPe can consume sponges (Hoey et al. 2013). While microscopic examination did not nd any sponge remains in sh intestinal contents, Demospongiae (sponges) were identi ed as the main food for SPe using highthroughput sequencing (Fig. 1b). Leray et al. (2015) also showed that pyrosequencing analysis could broaden the recognition of the food webs of coral dwelling predatory sh and achieved unprecedented taxonomic resolution of their diets. The complimentary use of molecular analysis enabled our study to reveal the highly complex dietary spectrum of the herbivorous reef shes, which we studied, an impossible result if we had relied on microscopic examination alone.

Differences in macroalgal consumption among herbivorous coral reef shes
Both microscopic examination and high-throughput sequencing analyses of intestinal contents showed that the eight herbivorous reef shes studied fed on macroalgae. However, the estimates of their consumption potential of macroalgae differed according to the results of the two methods. Based on microscopic examination, we found that SA fed on the most species of macroalgae ( ve species), followed by NU (four species) (Table 2). However, according to the high-throughput sequencing results, KV removed the highest diversity of macroalgae species as well as the greatest quantity of macroalgae, followed by KC, SA and NU (Fig. 2). In addition, NU was found to occupy the lowest trophic position (Fig. 5). Accordingly, we clearly demonstrated that KV, NU and SA showed a high consumption potential of macroalgae from the coral reef. Similarly, previous studies have also shown that KV and NU were highly effective in consuming macroalgae, especially the brown macroalgae (Lefevre and Bellwood, 2011;Streit et al. 2015;Puk et al. 2016). In the Great Barrier Reef, Australia, SA has been shown to remove a range of substrate algae, including turf algae, macroalgae, seagrass and epiphytic algae (Ebrahim et al. 2020). Furthermore, SA have been shown to especially focus more on red and green macroalgae (Hoey et al. 2013). Similarly, we found that KV and NU were highly effective consumers of brown macroalgae, while SA favored red and green macroalgae. Of course, macroalgal grazing selectivity of herbivorous shes can vary with location, season, food availability and according to the other sh species present (Puk et al. 2016).
Ecological function of herbivorous shes in removing macroalgae from coral reefs In addition to identifying the diets of the eight herbivorous reef shes studied, our results demonstrated the ecological importance of these shes in removing macroalgae from coral reefs. Herbivorous shes are broadly classi ed into four functional groups: excavators, scrapers, grazers, and browsers (Hoey and Bellwood, 2009). Excavators, scrapers and grazers generally consume small macroalgae and algal turfs, while browsers target large, erect macroalgal species and play a critical part in controlling macroalgal spreading (Hoey and Bellwood, 2009;Michael et al. 2013;Ebrahim et al. 2020). Previous studies have shown that CC, KC, KV, NB, NU and SA belong to the browser group (Puk et al. 2016;Sura et al. 2021) and this was con rmed in our study. Based on microscopic examination, NU can consume V. ventricosa and T. ornate, while SA can consume Halimeda sp. (Table 2). According to the 18S rRNA high-throughput sequencing analysis, we found that CC, KC, KV, NB and NU consumed the brown alga Lobophora variegate, and that KV and NU foraged especially heavily on this macroalgae. However, SA fed more on Dictyota linearis ( Fig. 2 and Table 3). KV played an important role in removing the green alga Dictyosphaeria cavernosa, while SA preferred to feed on the red algae Spyridia lamentosa and Centroceras hyalacanthum. KV also preferred to consume Peyssonnelia spp. (Fig. 2 and Table 3). The macroalgae mentioned above all belonged to either the upright calcareous or eshy macroalgae. These two macroalgae groups are problematic as they probably inhibit coral settlement, while the crustose calcareous algae and algal turfs have only minor effects on coral settlement (Diaz-Pulido et al. 2010). Overgrowth of Lobophora and Dictyota is widely supposed to reduce coral settlement (Foster et al. 2008;Diaz-Pulido et al. 2010;Evensen et al. 2019;Vieira, 2020) and the spread of Dictyota can even cause coral disease outbreaks (Brandt et al. 2012). Unfortunately, Lobophora and Dictyota, are the dominant species on Indo-Paci c coral reefs (Cardoso et al. 2009;Titlyanov et al. 2017). We found that the nitrogen signature (δ 15 N) of NU was lower than that of the other herbivorous shes (Fig. 5), indicating a high consumption potential of macroalgae. In general, we suggested that KV, NU and SA were the key browsers removing macroalgae from reefs off the Xisha Islands. This result is similar to that of a previous view, which recommended that KV, NU. and Siganus canaliculatus were the predominant removers of macroalgae, the "true macroalgae browsers" of coral reefs (Puk et al. 2016).
Herbivorous shes which eat the entire macroalgal thallus have a more important functional impact on macroalgae removal from coral reefs than species which only consume algal fronds (Streit et al. 2015). In this study, KC, KV, NU and SPn were found to feed on the entire macroalgal thallus (Table 2). Streit et al. (2015) also showed that KV and NU consumed the entire macroalgal thallus in approximately 90% of bites taken. Body size, tooth shape, and feeding behavior of different herbivores may result in these functional differences. SP was found to feed on Acanthophora sp. (Table 2) which can inhibit the successful settlement of coral planulae (Vermeij et al. 2009). The red alga, Pneophyllum conicum is widely distributed on Indo-Paci c coral reefs and can overgrow and kill live corals (Antonius, 2001).
Coincidentally, we found that CC consumed P. conicum (Table 3). In addition, herbivorous shes can selectively remove some macroalgae, which often occur in the Xisha Islands, including Ulvella, Cladophora, Sphacelaria, Spyridia, Centroceras, Ceramium and Neosiphonia (Table 3). The spread of macroalgae can negatively affect coral reefs through the inhibition of coral fecundity and growth, the reduction of coral larval settlement and recruitment and the increasing prevalence of coral diseases (Dell et al. 2020). Burkepile and Hay (2008) identi ed herbivore species richness and feeding complementarity as key factors in the effective suppression of the spread of macroalgae in a coral ecosystem. In summary we recommend an end to over shing, in particular of key browser species, such as KV, NU and SA, in order to control the spread of macroalgae on the reefs off the Xisha Islands. The preservation of as great a diversity of herbivorous shes on coral reefs will serve to better protect and reestablish beleaguered coral reefs.    Figure 1 Relative abundances of eukaryote sequences found in sh intestines, determined with 18S rRNA highthroughput sequencing. a phylum level. Polylines show the sum of the relative abundance of the phyla Streptophyta, Chlorophyta, Ochrophyta and Rhodophyta. b class level.

Figure 2
Eukaryote sequences belonging to the phyla Chlorophyta, Ochrophyta and Rhodophyta obtained from intestinal contents of eight coral reef sh species. a OTU numbers of taxonomic groups. b Relative abundance of taxonomic groups.

Figure 3
Neighbor-joining tree of 18S rRNA sequences a liated to the phyla Chlorophyta, Ochrophyta and Rhodophyta obtained from the intestinal contents of eight coral reef sh species. The scale bar represents 5% estimated sequence divergence.