INTRODUCTION
Soft corals (Coelenterata: Anthozoa: Alcyonacea) belong to dominating shallow-water tropical benthic animals. Within coral reef communities, the soft corals frequently predominate over reef-building corals by their species diversity, bottom covering and biomass. Owing to their rapid growth, soft corals occupy hard substrates, frequently substituting many reef-building corals (Sorokin, Reference Sorokin1993). There are certain differences in feeding behaviour between soft corals and reef-building corals, since soft corals having a special anatomic structure are believed to possess specific mechanisms of catching fine suspended food particles (Lewis, Reference Lewis1982). Many soft coral species contain endosymbiotic microalgae (zooxanthellae) that are able, according to different estimates, to compensate for up to 50% of energy demand of the host animal. Nitrogen and phosphorus demand of the coral host are covered through heterotrophic feeding. The soft corals are able to assimilate dissolved organic matter, as reef-building corals, or to capture living and dead organic matter from suspension and receive the major part of nitrogen and phosphorus through ingestion of the particles suspended (Anthony, Reference Anthony1999).
The results of different investigations on feeding ecology of anthozoans demonstrate considerable contradictions. Sorokin (Reference Sorokin1993) studied ingestion of 14C-labelled microalgae in 24 anthozoan species and showed a high retention of algae only in the zoanthid Zoanthus sociatus and the gorgonian coral Mopsella aurantia, while assimilation of microalgae in the soft coral Dendronephthya sp. was low. Farrant et al. (Reference Farrant, Borowitzka, Hinde and King1987) reported very slight 14C incorporation, when the soft coral Capnella gaboensis was fed on 14C-labelled microalgae. According to Fabricius & Dommisse (Reference Fabricius and Dommisse2000), all these contradictions are associated with feeding plasticity of soft corals in highly productive environment of coral reefs. Thus, the application of additional independent approaches to the studies of soft coral nutrition is an important task.
Starting with the work of Lee et al. (Reference Lee, Nevenzel and Paffenhofer1971), studies of feeding ecology of marine organisms successfully used fatty acids (FAs) as biomarkers that opened an additional opportunity to determine food sources of invertebrates in marine environment. The concept of using FAs as trophic markers is based on the observation that marine primary producers synthesize definite FAs, which can be transferred unchanged to the following level within the food chain. These FAs can be considered as trophic markers for marine animals and applied in the elucidation of the food strategy of marine invertebrates (Dalsgaard et al., Reference Dalsgaard, St John, Kattner, Muller-Navarra and Hagen2003; Escribano & Perez, Reference Escribano and Perez2010).
It is widely accepted that zooxanthellate corals can meet their energetic demands both via heterotrophy (plankton and particulate organic matter) and via autotrophy (primary production of zooxanthellae). Lipids are regarded as the main components in energy balance of soft corals (Zaslow & Benayahu, Reference Zaslow and Benayahu1999). The FA composition of lipids provides useful information on either autotrophic or heterotrophic origin of food (Dalsgaard et al., Reference Dalsgaard, St John, Kattner, Muller-Navarra and Hagen2003). The FA composition of soft corals significantly differed from FA profiles of reef-building corals and hydrocorals of the genus Millepora (Imbs et al., Reference Imbs, Latyshev, Dautova and Latypov2010a). First of all, the soft corals were distinguished by the presence of C16 and C24 polyunsaturated FAs (PUFAs) (Imbs et al., Reference Imbs, Demidkova, Dautova and Latyshev2009). Total FAs and PUFAs were successfully applied for chemotaxonomy of the soft corals (Imbs & Dautova, Reference Imbs and Dautova2008). Possible pathways of PUFAs biosynthesis in soft corals were suggested (Imbs et al., Reference Imbs, Yakovleva, Latyshev and Pham2010c). PUFAs were used as the markers of zooxanthellae and the indicators of possible translocation of PUFAs from zooxanthellae to the host in the soft corals (Imbs et al., Reference Imbs, Yakovleva and Pham2010b). Unfortunately, the application of FAs as food source indicators for corals is very limited, and the capability of FAs as trophic markers for the soft corals remains unclear. More than 50 FAs were detected in total lipids of the soft corals (Pham et al., Reference Pham, Luu, Imbs and Dautova2008). It is obvious that food cannot be an only source of these FAs; some of them come from symbionts and their own coral biosynthesis.
This paper reports the FA composition of total lipids of the soft coral species from the genus Sinularia May, 1898 and phytoplankton collected at the same biotope. The aims of this study were as follows: (1) to differentiate coral FAs according to their possible origins; and (2) to find potential food sources of the trophic FA markers.
MATERIALS AND METHODS
Sampling site and sample processing
Soft corals Sinularia leptoclados (Ehrenberg, 1834), S. flexibilis (Quoy & Gaimard, 1833), S. aff. deformis Tixier-Durivault, 1969, S. lochmodes Kolonko, 1926, S. cf. muralis May, 1899, S. densa (Whitelegge, 1897), S. notanda Tixier-Durivault, 1966 and S. cruciata Tixier-Durivault, 1970 were collected at a depth of 3 m on a reef of the Den Island in Van Phong Bay, the South China Sea, Vietnam (12°35′N 109°18′E). Immediately after collecting, coral colonies were placed in a box with seawater and transported within an hour to the laboratory for lipid extraction. Before extraction, samples were washed in filtered seawater to free them from epibionts and mucus.
Plankton samples were taken with the use of a 5-l bathometer over the reef. The samples were concentrated to 40–50 ml by filtration through nucleopore filters (2 µm); and the plankton was precipitated by centrifugation. Biomass of algae was estimated by the volume method, using measurement data of cell volume for every species. To determine species composition of phytoplankton, the samples were studied under a light microscope in a Nojott chamber.
Analysis of fatty acids
For lipid analysis, three different colonies of each coral species and three plankton samples were taken. All samples were homogenized and total lipids were extracted by chloroform–methanol according to Bligh & Dyer (Reference Bligh and Dyer1959). Fatty acid methyl esters (FAMEs) were prepared by a sequential treatment of the total lipids with 1% of sodium methylate/methanol and 5% HCl/methanol (Carreau & Dubacq, Reference Carreau and Dubacq1979), and then purified by preparative silica gel thin-layer chromatography in benzene. N-acylpyrrolidide derivatives of FAs were prepared by direct treatment of the FAMEs with pyrrolidine/acetic acid (10:1, by volume) in a capped vial (1 hour, 100°C) followed by ethereal extraction from the acidified solution and purification by preparative TLC developed in ethyl acetate (Andersson, Reference Andersson1978).
A gas chromatography (GC) analysis of FAMEs was carried out on a Shimadzu GC-17A chromatograph (Shimadzu, Kyoto, Japan) with a flame ionization detector. A Supelco SUPELCOWAX 10 (Bellefonte, PA) capillary column (30 m × 0.25 mm i.d.) was used at 210°C. The injector and detector temperatures were 240°C. Helium was used as a carrier gas. Individual peaks of FAMEs were identified by a comparison with authentic standards (PUFA-3 mix from menhaden oil was purchased from Supelco, Bellefonte, PA) and using a table of equivalent chain-lengths (ECL) values (Stransky et al., Reference Stransky, Jursik and Vitek1997).
The structures of FAs were confirmed by gas chromatography–mass spectrometry (GC–MS) of their methyl esters and N-acylpyrrolidide derivatives. The GC–MS analysis of the FAMEs was performed on a Shimadzu GCMS-QP5050A instrument (Kyoto, Japan). A Supelco MDN-5S (Bellefonte, PA) capillary column (30 m × 0.25 mm i.d.) was used at 160°C with a 2°C/minute ramp to 240°C that was held for 20 minutes. Injector and detector temperatures were 240°C. Helium was used as a carrier gas. GC–MS of N-acylpyrrolidides was performed on the same instrument; the injector and detector temperatures were 300 and 240°C, respectively, and the column temperature was 210°C with a 3°C/minute ramp to 300°C that was held for 40 minutes.
Electron impact mass spectrometry (EIMS) (16:2n-7 pyrrolidide) 70 eV, m/z (rel. int.): 305 [M]+ (16.7), 113 (100.0), 126(58.7), 140 (7.2), 154 (3.2), 166 (3.4), 180 (10.0), 194 (5.0), 206 (2.0), 220 (2.7), 234 (3.7), 248 (1.6), 262 (1.1), 276 (1.1), 290 (0.3).
EIMS (18:2n-7 pyrrolidide) 70 eV, m/z (rel. int.): 333 [M]+ (27.1), 113 (100.0), 126 (73.1), 140 (6.5), 154 (4.9), 168 (12.9), 182 (3.9), 194 (2.1), 208 (12.9), 222 (5.5), 234 (2.7), 248 (1.9), 262 (2.1), 276 (1.8), 290 (1.5), 304 (1.2), 318 (0.4).
Analysis of data
The data were assessed statistically by a one-way analysis of variance and Tukey honestly significant difference-test. Differences with P < 0.05 were taken as significant. All tests were performed using Statistica 6.0 software.
RESULTS
Fatty acid composition of the total lipids obtained from the eight zooxanthellate coral species of the genus Sinularia are shown in Table 1. The main FAs were 14:0, 16:0, 7-Me-16:1n-10, 16:1n-7, 16:2n-7, 18:0, 18:1n-9, 18:2n-7, 18:4n-3, 20:4n-6, 20:5n-3, 22:6n-3, 24:5n-6 and 24:6n-3. The FA profiles of the corals studied were quite similar, except for Sinularia densa, which had a PUFA level significantly lower (P < 0.05) than other species. This decrease was mainly caused by the low content of 20:4n-6 and the high content of 16:0. PUFAs dominated in the majority of the species. On the average, saturated, monounsaturated, and polyunsaturated FAs contributed 35.6, 6.2 and 54.0% of total FAs, respectively.
Table 1. Composition of main fatty acids (% of total fatty acids, mean ± SD, N = 3) of total lipids of the soft corals (the genus Sinularia) and phytoplankton.
tr, trace (< 0.1%); SFAs, saturated fatty acids; MUFAs, monounsaturated fatty acids; PUFAs, polyunsaturated fatty acids; other, the sum of 12:0, i-15:0, ai-15:0, i-16:0, ai-16:0, i-17:0, 16:2n-6, 20:1n-9, 22:1n-7, 22:1n-9, 22:4n-3.
The content of saturated FAs varied in the wide range from 26.4% in Sinularia lochmodes tо 47.1% in S. densa (Table 1). Among saturated FAs, 16:0 and 18:0 dominated in all the species, with their content being 20.6–37.4% and 2.4–12.9%, respectively. Saturated odd-chain and methyl-branched FAs, such as 15:0, 17:0, i-15:0, ai-15:0, i-16:0, ai-16:0 and i-17:0 were found in minor amounts.
The content of monounsaturated acids varied from 4.7 to 8.2% of total FAs (Table 1). The main monounsaturated FAs were 16:1n-7 and 18:1n-7, but their content did not exceed 3.5 and 3.1% of total coral FAs, respectively. Sinularia contained only small amount of cis-vaccenic acid (18:1n-7). Peculiar branched 7-methyl-6-octadecenoic acid (7-Ме-16:1n-10) was identified in all coral species studied. This acid was characterized by GC–MS as previously described (Carballeira & Shalabi, Reference Carballeira and Shalabi1994); an abundant ion at m/z 138 in the mass spectra of methyl ester was very characteristic for 7-Ме-16:1n-10. The ratio of 16:1n-7/16:0 did not exceed the value of 0.1 for all Sinularia species.
Among PUFAs, the acids of n-6 series predominated in all coral species, with the content of 20:4n-6 being within the range of 10.2 to 23.8% (Table 1). The level of C18 n-6 PUFAs, such as 18:2n-6 and 18:3n-6 was lower than 1% of total FAs. The content of n-3 PUFAs did not exceed 15.9%. The main n-3 PUFA was 18:4n-3, which amounted to an average 5.4% of total FAs for seven corals species except Sinularia leptoclados. The content of 20:5n-3 and 22:6n-3, which dominated in marine organisms, was not higher than 2.4% and 3.9%, respectively. The ratio n-6/n-3 PUFAs > 1.6 was a characteristic for all the corals studied (Table 1). Sinularia leptoclados had the highest proportion of n-6 PUFAs (the ratio n-6/n-3 = 4) because of the lowest content of 18:4n-3.
In all coral species, two rare PUFAs of n-7 series, namely, 16:2n-7 and 18:2n-7, were identified by a GC–MS analysis. The mass spectra of methyl esters of 16:2n-7 and 18:2n-7 acids gave molecular ion peaks [M]+ at m/z 266 and 294, respectively. The mass spectra of N-acylpyrrolidide derivative of 16:2n-7 and 18:2n-7 gave molecular ion peaks [M]+ at m/z 305 and 333, respectively. In the mass spectrum of 16:2n-7 pyrrolidide, the gap of 12 amu between m/z 154 and 166, as well between m/z 194 and 206 confirmed the presence of double bonds in positions 6 and 9. The gap of 12 amu between m/z 182 and 194, as well as between m/z 222 and 234 was observed in the mass spectra of the 18:2n-7 pyrrolidide that indicated localization of double bounds at 8th and 11th carbon atoms of the original FA. The mass-spectra of N-acyl-derivatives of n-7 PUFAs obtained were similar to that of 16:2n-7 and 18:2n-7, which were identified as minor FAs in marine invertebrates (Kawashima & Ohnishi, Reference Kawashima and Ohnishi2004). The total content of n-7 PUFAs ranged from 7.8% in S. leptoclados to 20.2% in S. lochmodes (Table 1). As a rule, the amount of 16:2n-7 was higher than that of 18:2n-7.
Other C16 PUFAs detected in Sinularia were 16:3n-4 and 16:4n-1. On the average, these acids did not contribute more than 2.2% of total FAs.
Two very long-chain tetracosapolyenoic fatty acids (TPFAs) were observed in total lipids of the soft coral studied. The content of 24:5n-6 was significantly higher (P < 0.05) than that of 24:6n-3; these acids contributed on the average 6.7 and 1.2% of total acids, respectively.
Bacillariophyta and Dinophyta predominated in the phytoplankton species, which were collected in the habitat of Sinularia species. Diatoms and dinoflagellates contributed 67–97% and 3–28% of the total species number, respectively. Diatom algae gave 97–99% of the total phytoplankton biomass. Saturated FAs, first of all, 16:0 and 14:0, and monounsaturated acid 16:1n-7 predominated in total FAs of phytoplankton (Table 1). The content of saturated FAs amounted to about 50% of total FAs. Other principal FAs were 18:0, 18:1n-9, 20:5n-3, and 22:6n-3. The content of some n-6 PUFAs, such as 18:2n-6 and 20:4n-6 did not exceed an average value of 2.3%. Very long-chain FAs were not detected. In phytoplankton probes, the amount of saturated FAs was significantly higher and the amount of n-6 PUFAs was significantly lower than that in coral samples (P < 0.05). PUFAs of n-3 series were predominant PUFAs in the plankton probes. The n-6/n-3 ratio was 0.4, and the 16:1n-7/16:0 ratio was 0.62. Instead of 16:2n-7, which was detected in the coral lipids, plankton lipids contained 16:2n-4 (Table 1).
DISCUSSION
Sinularia belong to mass soft corals in the tropical waters of Vietnam. A lot of new metabolites were found in Sinularia corals (Lakshmi & Kumar, Reference Lakshmi and Kumar2009; Li & Pattenden, Reference Li and Pattenden2011). The FA profiles of the species investigated were similar to those of other representatives of the genus Sinularia from Vietnam (Imbs et al., Reference Imbs, Latyshev, Dautova and Latypov2010a). High content of 20:4n-6 and 24:5n-6 as well as a small amount of the monoenoic acids, were the chemotaxonomic features of the Sinularia studied and distinguished this soft coral genus from the majority of reef-building corals (Imbs et al., Reference Imbs, Latyshev, Dautova and Latypov2010a). All the Sinularia species studied were collected on the same coral reef and habitat. Similarity of FA compositions in all these corals (Table 1) confirms that these organisms have similar food sources. Relative amounts of definite FA markers show contribution of zooxanthellae, plankton, particulate organic matter and dissolved organic matter to nutrition of these corals.
Zooxanthellate corals compensate for their main energy demand using organic matter synthesized by zooxanthellae (Sorokin, Reference Sorokin1993). Zooxanthellae release various low molecular weight compounds, including glycerol, organic acids, glucose, and amino acids, which translocate to the coral host tissues (Grant et al., Reference Grant, Remond, People and Hinde1997). Zooxanthellae are endosymbiotic dinoflagellates of the Symbiodinium group. Dinoflagellates are distinguished by the high level of PUFA markers of n-3 series (18:4n-3, 20:5n-3 and 22:6n-3) (Dalsgaard et al., Reference Dalsgaard, St John, Kattner, Muller-Navarra and Hagen2003). These PUFAs and 18:5n-3 are the markers of zooxanthellae too (Bishop & Kenrick, Reference Bishop and Kenrick1980; Imbs et al., Reference Imbs, Yakovleva, Latyshev and Pham2010c). The general marker of zooxanthellae of reef-building corals, soft corals, and hydrocorals is 18:4n-3, which amounts up to 18% of total lipid FAs and up to 39% of galactolipid FAs of pure zooxanthellae from some reef-building corals (Bishop & Kenrick, Reference Bishop and Kenrick1980). The study of the distribution of PUFAs between zooxanthellae and host tissue showed 18:4n-3 to be mainly located in zooxanthellae, whereas the concentration of this marker acid in the host tissue was very low (Papina et al., Reference Papina, Meziane and van Woesik2003; Treignier et al., Reference Treignier, Grover, Ferrier-Pages and Tolosa2008; Imbs et al., Reference Imbs, Yakovleva and Pham2010b, Reference Imbs, Yakovleva, Latyshev and Phamc). Dinoflagellates, which can be a possible food source of 18:4n-3 in Sinularia studied, contribute less than 3% of the total phytoplankton biomass. Lipids of zooxanthellae can amount up to 30% of total coral lipids (Imbs et al., Reference Imbs, Yakovleva and Pham2010b). Therefore, the zooxanthellae should be regarded as the source of 18:4n-3 in Sinularia. Thus, 18:4n-3 in Sinularia does not originate from a food source (phytoplankton), but can be the useful markers for the estimation of the contribution of zooxanthellae lipids to total lipids from zooxanthellate soft corals.
In all the coral species studied, some C16 PUFAs, namely, 16:2n-7, 16:3n-4 and 16:4n-1, were found (Table 1). It is known that C16 PUFAs, such as 16:2n-4, 16:3n-4 and 16:4n-1, are the markers of diatoms (class Bacillariophyceae) (Dunstan et al., Reference Dunstan, Volkman, Barrett, Leroi and Jeffrey1994; Liang et al., Reference Liang, Mai and Sun2000), which are potential sources of nutrition for Sinularia. In contrast to the high level of 16:2n-7, only the trace amount of 16:2n-4 was detected in the coral lipids studied. The biosynthesis pathway 16:2n-7 → 16:3n-4 → 16:4n-1 was proposed in zooxanthellae of Sinularia (Imbs et al., Reference Imbs, Yakovleva and Pham2010b, Reference Imbs, Yakovleva, Latyshev and Phamc). The study of the distribution of PUFAs between zooxanthellae and host tissue showed 16:3n-4 and 16:4n-1, as well as 18:4n-3 to be mainly located in zooxanthellae. The concentration of C16 PUFAs in the host tissue was significantly higher than that of 18:4n-3. It is possible that zooxanthellae are the main source of 16:3n-4 and 16:4n-1 in Sinularia, but these corals also receive some part of these C16 PUFAs through feeding on diatoms.
Recently, species of the genus Sinularia have been divided into two groups chemotaxonomically according to the ratio 16:2n-7/18:3n-6 (Imbs et al., Reference Imbs, Latyshev, Dautova and Latypov2010a). All the species investigated belong to the group, where the ratio 16:2n-7/18:3n-6 is more than 1. The high level of 16:2n-7 together with the low level of 18:3n-6 was explained by the difference in substratum specificity of Δ6 desaturase in zooxanthellae, which converted 16:1n-7 to 16:2n-7 more effectively than 18:2n-6 to 18:3n-6. Subsequent elongation of 16:2n-7 up to 18:2n-7 probably takes place in the coral host tissue (Imbs et al., Reference Imbs, Yakovleva, Latyshev and Pham2010c). Thus, the presence of 16:2n-7 and 18:2n-7 indicates the features of PUFA biosynthesis in soft corals, but does not depend on their trophic sources.
The presence of TPFAs (24:5n-6 and 24:6n-3) is a chemotaxonomic feature of octocorals (Svetashev & Vysotskii, Reference Svetashev and Vysotskii1998; Imbs at al., Reference Imbs, Latyshev, Dautova and Latypov2010a). These acids were also found in echinoderms, bryozoans and brittle stars (Takagi at al., Reference Takagi, Kaneniwa and Itabashi1986; Carballeira et al., Reference Carballeira, Sostre and Rodrigues1997; Mansour et al., Reference Mansour, Holdsworth, Forbes, Macleod and Volkman2005). The origin of TPFAs in marine organisms is unknown. However, biosynthesis of TPFAs in mammals has been studied. The substrates for biosynthesis of 24:5n-6 and 24:6n-3 are 20:4n-6 and 20:5n-3, respectively (Sprecher, Reference Sprecher2000). The significant predominance of 24:5n-6 in TPFAs of Sinularia can be connected with the high content of 20:4n-6 (Table 1). Both zooxanthellate and azooxanthellae soft corals contain TPFAs (Imbs et al., Reference Imbs, Latyshev, Dautova and Latypov2010a), whereas plankton does not contain these. Thus, TPFAs of Sinularia cannot originate from trophic sources, but are produced by their own enzymes in the coral host tissue.
Numerous examinations of the interactions between corals and microbes have shown that there is dynamic microbiota living on the surface and, possibly, within the tissue of the corals. Soft corals can also obtain bacteria with particulate organic matter. Saturated odd-chain and methyl-branched FAs plus 18:ln-7 were regarded as biomarkers of bacteria (Kaneda, Reference Kaneda1991). In the Sinularia species studied, the low content of the ‘bacterial’ FAs (less than 1% of total FAs) indicated a negligible contribution of bacteria from trophic sources or coral microbiota. Probably, methyl-branched monoenoic acid 7-Me-16:ln-10 does not characterize the coral bacterial community as a whole, but can serve as a biomarker of some specific microbial groups associated with corals (Imbs et al., Reference Imbs, Latyshev, Zhukova and Dautova2007). The biosynthetic origin of this acid is still unclear.
The ability of soft corals to feed heterotrophically was documented (Sorokin, Reference Sorokin1993). For example, the gorgonian coral Paramuricea clavata graze daily from the environment 1–22% of diatom algae, 1–9% of nanoeukaryotes, 1–26% of dinoflagellates, 2–99% of ciliates and 2–10% of detritus material (Ribes et al., Reference Ribes, Coma and Gili1999). Soft corals control the quantity of phytoplankton obtained for nutrition. In the azooxanthellate soft coral Dendronephtya sp., the incorporated microalgae contributed a maximum of 26% to the daily organic carbon demand even at a maximum of phytoplankton assimilation. The zooxanthellate soft coral Alcyonium digitatum utilized about 18% of the total biomass of microalgae from the environmental waters (Widdig & Schlichter, Reference Widdig and Schlichter2001). These experiments indicated that soft corals fed on plankton of distinct type and size. Zooxanthellate soft corals received nutrition due to the photosynthesis of their symbiotic algae (Symbiodinium sp.) and heterotrophic sources. Zooplankton, bacterioplankton, small particulate matter, and dissolved organic matter are regarded as preferable sources of heterotrophic food for soft corals; some of the soft corals (e.g. Sinularia) limit a phytoplankton intake (Lewis, Reference Lewis1982; Sorokin, Reference Sorokin1991, Reference Sorokin1993; Anthony, Reference Anthony1999; Migne & Davoult, Reference Migne and Davoult2002).
Plankton is a good source of C20-22 PUFAs, such as 20:4n-6, 20:5n-3, and 22:6n-3 for marine invertebrates (Dalsgaard et al., Reference Dalsgaard, St John, Kattner, Muller-Navarra and Hagen2003). The species composition of plankton from Van Phong Bay, where the corals were collected, was characterized by the great diversity, but diatoms contributed the most part of total phytoplankton biomass. The diatom diet of marine organisms is characterized by the ratio of 16:1n-7/16:0 >1 and a high level of 20:5n-3 (Reuss & Poulsen, Reference Reuss and Poulsen2002). The 16:1n-7/16:0 ratio in all the soft corals studied was significantly lower than in lipids of phytoplankton collected on the reef (Table 1). The low values of this ratio were accompanied by the low level of 20:5n-3 for all corals. These data confirmed that diatom microalgae, in spite of their abundance in plankton, were not the dominant source of food for Sinularia. The dinoflagellate diet of marine organisms, which are another possible source of food for Sinularia, is mostly characterized by the ratio of 18:5n-3/18:3n-3 >1 and a great quantity of С22 PUFAs (Viso & Marty, Reference Viso and Marty1993). However, we did not find a noticeable amount of 18:5n-3 in the corals studied. A minor contribution of Dinophyceae to nutrition of soft corals was also confirmed by a low level of 22:6n-3 (on the average, 2.9% of total FAs).
An analysis of the composition of marker FAs showed that eukaryotic microalgae play a secondary role in feeding of soft corals. Probably, these animals compensate for a lack of food by other components of seston, and this conclusion is consistent with data of other authors (Farrant et al., Reference Farrant, Borowitzka, Hinde and King1987; Sorokin, Reference Sorokin1993). In reef ecosystems, the main source of nitrogen and phosphorus for benthic animals is heterotrophic protists that are an integral part of benthic food nets. They play an intermediate role in the transport and modification of organic matter from planktonic prokaryotes and eukaryotes to organisms of the higher trophic levels (Bak et al., Reference Bak, Joenje, de Jong, Lambrechts and Nieuwland1998). Octocorals are passive predators. The main mechanism ensuring capture of food by corals is mucus secretion (Brown & Bythell, Reference Brown and Bythell2005). The mucus secreted by soft corals absorbs ‘sea snow’, bacteria and microalgae; this medium is very rich in protists. Heterotrophic protists in these specific substrates are the potential source of food for alcyonarians (Wild et al., Reference Wild, Woyt and Huettel2005).
A lot of soft corals species contain large amounts of 20:4n-6, but the origin of this acid has not been reliably established. Elevated content of 20:4n-6 is associated with microorganisms, some diatom species, or macrophytes. However, the majority of researchers believe protozoans to be a most probable source of 20:4n-6 for marine organisms (Fullarton et al., Reference Fullarton, Dando, Sargent, Southward and Southward1995). High content of 20:4n-6 in the soft corals studied can be also explained by presumable nutrition on heterotrophic protists, which form a ‘microbial loop’ and are the main carriers of nutrients within benthic communities in marine ecosystems. Protists have no definite marker FAs, while their composition relies much on the food ingested, namely bacteria and microalgae (Ederington et al., Reference Ederington, MacManus and Harvey1995). However, heterotrophic protists are capable of modifying diet FA composition by PUFA synthesis.
Biosynthesis of PUFAs de novo in protists is poorly studied, but recent investigations have shown that they are a group of organisms capable of synthesizing PUFAs of both n-6 and n-3 series (Napier, Reference Napier2002). PUFA synthesis was described for some free-living fresh-water protists, as well as for marine parasitic and benthic ones, but free-living marine heterotrophic protists were characterized by PUFA composition only (Desvilettes & Bec, Reference Desvilettes, Bec, Arts, Brett and Kainz2009). The marine parasitic protist Perkinsus marinus synthesizes presumably n-6 PUFAs through the Δ8 desaturase pathway using the combination of Δ8, Δ5, and Δ4 desaturases associated with elongases (Chu et al., Reference Chu, Lund, Harvey and Adlof2004). The ratio between n-6 and n-3 PUFA in protist lipids depended on environmental temperature. In the culture of P. marinus, the rate of synthesis of 20:4n-6 increased five times in comparison to that of n-3 PUFAs, when the temperature increased from 10 to 28°C (Lund et al., Reference Lund, Chu and Harvey2004). Several n-6 PUFAs of the Δ8 desaturase pathway (18:2n-6, 20:2n-6, 20:3n-6 and 20:4n-6) were the main PUFAs in the culture of the marine protist Euplotes crassus and freshwater ciliate Tetrahymena pyriformis with minimal content of n-3 PUFAs (Zhukova & Kharlamenko, Reference Zhukova and Kharlamenko1999). Thus, protists are able to synthesize 20:4n-6 and can supply Sinularia with this n-6 PUFA as a trophic source. However, we cannot exclude the capability of coral polyps to synthesize 20:4n-6.
Sinularia flexibilis from the Den Island contained a significantly larger amount of 20:4n-6 (16.9%) than S. flexibilis from the Re Island (9.0%) studied before (Imbs at al., Reference Imbs, Latyshev, Dautova and Latypov2010a); whereas the average level of 20:4n-6 in Sinularia specimens from the Den Island and from the Re Island was the same. For S. flexibilis, the intraspecific distinction in 20:4n-6 level may be caused by a variation in the sources of food for corals.
CONCLUSION
The problem of heterotrophic feeding remains to be intensively studied in the ecology of Alcyonaria. Determination of the balance between autotrophic and heterotrophic feeding of zooxanthellate species, as well as between different sources of food for azooxanthellate species, is recognized as an important aspect of this problem. However, abundance and rapid change of sources of food for soft corals require applying new methods. We assume that the method of FA biomarkers can assist in the elucidation of the problem and estimation of the importance of heterotrophic feeding. The different sources of generally accepted FA markers in zooxanthellate soft corals were proposed by the example of common species of the genus Sinularia. It was important that some phytoplankton FA markers, presented in coral total lipids, mainly belonged to zooxanthellae lipids and did not originate from food sources. We suppose that protists are a possible key item of soft coral diet and transfer of lipids from primary producers to consumers can play an important role for carbon cycling in the marine food web.
ACKNOWLEDGEMENTS
The authors are grateful to T.N. Dautova and M.S. Selina for the identification of coral species and plankton groups. We also thank the staff of the Open Russian–Vietnamese Laboratory of Biochemistry for their facilities and hospitality. This work was supported by the Russian Foundation of Basic Research (grant numbers 09-04-01040 and 11-04-98505).
