Introduction

Larval dispersal is of great importance to the ecology and evolution of sessile marine species because it determines the patterns of connectivity among populations (Jones et al. 2009). Larvae dispersal distance depends on many factors, both physical, such as currents and winds (Siegel et al. 2009), and biological, such as larval energy reserves (Burgess et al. 2009), survival and settlement competence dynamics (Connolly and Baird 2010), buoyancy and swimming ability (Whalan et al. 2008), and a full understanding of dispersal potential requires all these factors to be considered (Jones et al. 2009).

Many coral larvae have an extraordinary capacity to delay metamorphosis; in the absence of suitable settlement cues, a proportion of larvae can remain alive and competent for months (Richmond 1987; Graham et al. 2008; Connolly and Baird 2010). Coral propagules can be either symbiotic or aposymbiotic (symbiont-free), depending on whether or not the zooxanthellae are transmitted into the egg, prior to release. Coral larvae do not feed in the plankton; therefore, aposymbiotic larvae depend exclusively on energy reserves provided by the parent. In contrast, symbiotic larvae can obtain energy from zooxanthellae (Richmond 1987) to survive longer and thus possibly disperse further.

Lipids are the main biochemical component, making up between 34 and 85.5% of the biomass of newly released propagules of scleractinian corals (Table 1). The most abundant lipids in coral propagules are wax esters (WEs), phospholipids (PLs), and triacylglycerols (TAGs) (Table 1). Phospholipids are major constituents of cell membranes but are thought not to be involved in metabolism. Wax esters and triacylglycerols constitute a reserve of fatty acids that can be oxidized to provide energy or that can be incorporated in cell structures. Richmond (1987) hypothesized that a high abundance of wax esters may enable long-distance dispersal, because wax esters take longer to metabolize than triacyglycerols. Lipids also provide buoyancy, particularly wax esters that have a lower density than other lipid classes (Gurr et al. 2002; Lee et al. 2006), and may therefore promote fertilization by transporting gametes to the ocean surface (Harii et al. 2007). Photosynthetic zooxanthellae provide an extra energy source to symbiotic coral larvae, possibly contributing to the extension of larval duration; in the absence of zooxanthellae, coral larvae exhaust their energy reserves more rapidly (Harii et al. 2010).

Table 1 Average lipid content of eggs or newly released planulae (% in dry weight) and their triacylglycerol (TAG) and wax esters (WE) composition (% of total lipids) in coral species with different modes of larval feeding
Full size table

The fatty acids composition of coral eggs and larvae is still poorly studied (Arai et al. 1993); therefore, the specific role of each fatty acid during coral larval dispersal remains unknown. However, studies on other marine organisms have shown that palmitic (16:0) and vaccenic (18:1ω9) acids are the preferred source of metabolic energy in animals (Bergé and Barnathan 2005), whereas highly unsaturated fatty acids, particularly the essential docosahexaenoic (DHA), eicosapentaenoic (EPA), and arachidonic (ARA) acids, are critical to growth, survival, reproduction, and food conversion efficiency (Brett and Müller-Navarra 1997; Bergé and Barnathan 2005). Additionally, the fatty acids composition of eggs is known to reflect the feeding ecology of the adults (Dalsgaard et al. 2003).

Here, the potential role of lipids in coral larval dispersal was explored by measuring the change in total lipid, lipid classes, and fatty acid composition of the eggs and larvae of the coral Goniastrea retiformis species during the first 30 days of development.

Methods

Gametes of six colonies of Goniastrea retiformis were collected at Oku, Okinawa, in June 2002 following Babcock et al. (2003). Five hundred eggs were sampled right after spawn (day 0). Larvae were cultured in the Sesoko Tropical Biosphere Station, Okinawa, Japan, and maintained in three replicate 20-L plastic buckets in 0.2 μm filtered seawater at initial densities of approximately 1,000 larvae L−1 under a quantum irradiance of 160 μmol photons m−2 s−1 with a 12:12 h light/dark photo-period at temperatures between 25 and 27°C. Water was changed on each sampling occasion. Between 200 and 500 larvae were taken from each bucket after 1, 3, 7, 17, and 30 days. Larvae were counted into 25-mL plastic jars modified to form a sieve. The vials were drained, and the 60-micron plankton mesh sieve removed and placed in liquid nitrogen. Larvae were first freeze-dried before lipids were extracted with CHCl3–MeOH (1:2) (Bligh and Dyer 1959). The combined lipid extracts from each sample were partitioned with CHCl3 and H2O (Meziane and Tsuchiya 2002). Lipids were recovered in the lower CHCl3 phase, the solvents were removed under vacuum, and the lipid extracts were stored in CHCl3. The lipid extract was then weighed to estimate the biomass per larva. To analyze lipid classes, aliquots of lipid extract were applied to 10 cm/10 cm high-performance thin-layer chromatography (HPTLC) plates (Merck Ltd, Tokyo, Japan). The plates were first developed full-length with hexane–ether–acetic acid (80:20:0.1). After drying in a stream of air, plates were immersed in a mixture of phosphoric acid, 33% acetic acid, sulfuric acid, and 0.5% copper sulfate (5:5:0.5:90 v/v) for 40 s then heated at 130°C for 12 min. The chromatograms were scanned on a GT9000 (Seiko Epson Co., Nagano, Japan) in gray scale mode using Adobe Photoshop (Adobe Systems). Percent compositions of lipid classes were determined on the basis of band intensity by an image analysis program of NIH Image (Research Services Branch) that contains the lipid standards band intensity; the total amount of each lipid class was calculated by reporting the obtained surface to the total lipid concentration measured at the beginning of the extraction. The authentic standards used were cholesterol for sterols, oleic acid for free fatty acid (FFA), triolein for triacylglycerol (TAG), and stearyl oleate for wax ester (WE) following Oku et al. (2002).

In order to estimate the proportion of fatty acids methyl esters (FAMEs) in the total lipids, a second plate was prepared after saponification and esterification (Metcalfe and Schmitz 1961). All lipid extracts’ bands were scrapped and eluted with CHCl3–MeOH (2:1 v/v) at 40°C for 60 min. The FAMEs were separated and quantified by gas chromatography (GC 14.B, Shimadzu, Japan) equipped with flame ionization detector. Separation was performed with an FFAP-polar capillary column (30 × 0.32 mm internal diameter, 0.25 mm film thickness). Hydrogen was used as a carrier gas. After injection at 60°C, the oven temperature was raised to 150°C at a rate of 40°C min−1, then to 230°C at 3°C min−1, and finally held constant for 30 min. The flame ionization was held at 240°C. FAMEs’ peaks were identified by comparing their retention times with those of authentic standards (Supelco Inc.). An image analysis program (NIH 6 image) was used to estimate the relative contribution of the fatty acids in the lipids by integrating the chromatogram. One-way ANOVAs were used to test for significant differences in total lipid, lipid classes, and percent of each fatty acid among samples followed by Tukey’s post hoc multiple comparisons. One-way ANOVAs were also used to test for significant differences in triacylglicerol/wax ester ratio between species, in adults, and propagules. Statistical analyses were performed using Statistica 7.0 (StatSoft) with a significance level of 0.05.

Results

The eggs of G. retiformis have 78.8 ± 0.49% lipid by dry weight, which is similar to that of other broadcast spawning coral species (Harii et al. 2010, Table 1). In comparison, adult Goniastrea contain between 11 and 42% lipid (Harland et al. 1993), suggesting that lipids are important for larval biology. While the triacylglycerol/wax ester (TAG/WE) ratio in the tissue of adult corals is not significantly different between brooders and broadcast spawners (F 1,15 = 0.0316, p = 0.96, Table 2), however, the TAG/WE ratio of propagules is dependent upon whether or not they contain symbionts (F 1,7 = 8.177, p = 0.02, Table 1). While all propagules have a high proportion of WEs (23–86.3%), propagules that lack symbionts have much less TAGs (0.3–1.9%, Table 1) than propagules that possess symbionts (5.1–31%, Table 1). In G. retiformis, wax esters are the main class of lipid in the eggs (86.34 ± 0.23%, Table 3) followed by polar lipids (9.25 ± 0.00%), sterols (4.11 ± 0.15%), and TAGs (0.30 ± 0.08%). This pattern is similar to the eggs of other corals that lack zooxanthellae, which typically contain a very high content of WEs (58–81.8%), polar lipids (9.3–22.6%), and little TAGs (0.3–1.9%) (Table 1). In contrast, larvae with symbionts such as those of Pocillopora damicornis and Montipora digitata have a high proportion of TAG (5.1–31%, Table 1).

Table 2 Average lipid content (% in dry weight) and TAG and WE composition (% in total lipids) in the tissue of adult corals
Full size table
Table 3 Average (±SE) lipid content and composition (μg larva−1) during larval development (from egg to 30-day-old larva) of Goniastrea retiformis
Full size table

Total lipid decreased as larvae aged (F 5,7 = 10.72, p = 0.004), ranging from 8.4 ± 0.7 μg in the egg to 3 ± 0.5 μg in a 30-day-old larva (Table 3; Fig. 1). During early development (until day 3), lipids decreased rapidly; from day 4 to days 17–21, there was a decrease in lipid composition, but later in development, lipids decreased more rapidly again. WEs remained the larvae’s main lipid type throughout development, but declined significantly (F 5,7 = 14.27, p = 0.001, Fig. 1). Polar lipids and sterols also decreased with larval age (F 5,7 = 8.47, p = 0.007 and F 5,7 = 18.19, p = 0.001, respectively, Table 3) but not as rapidly as WEs. Free fatty acids changed significantly over larval development (F 5,7 = 8.86, p = 0.006), initially increasing and later decreasing (Table 3). TAGs were conserved through development (F 5,7 = 3.69, p = 0.059).

Fig. 1
figure 1

Lipid content and composition (average ± SE) during larval development of Goniastrea retiformis (PL polar lipids; FFA free fatty acids; TAG triacylglycerols; WE wax esters)

Full size image

Goniastrea retiformis eggs were rich in palmitic acid (16:0, 37.51 ± 1.07% of the total fatty acids). Other saturated fatty acids in the eggs were 14:0 and 18:0 (Table 4). The major monounsaturated fatty acids were oleic and vaccenic acids (18:1ω7/9 and 16:1ω7/9, respectively) (Table 4). Polyunsaturated fatty acids (PUFAs) were abundant (36.76 ± 1.22%). The most abundant PUFAs in the propagules were 18:3ω6 (8.78 ± 0.41%), 20:4ω6 (8.54 ± 0.06%), and 22:5ω3 (8.17 ± 1.18%), followed by 18:2ω6, 20:3ω6, 22:4ω3, 20:5ω3 (EPA), 22:6ω3 (DHA), and 18:4ω3 (Table 4). The proportion of most PUFAs, for example 18:2ω6, 18:3ω6, 20:3ω6, and 20:4ω6, declined significantly with larval age (p < 0.05, Table 4). The proportion of the fatty acids 14:0, 16:1ω7/9, 20:1ω9, and the PUFAs 18:4ω3, 22:4ω3, 22:5ω3, and 22:6ω3 remained constant.

Table 4 Average (±SE) percent of composition of fatty acids during larval development of Goniastrea retiformis (from egg to 30-day-old larvae) and p value (one-way ANOVA)
Full size table

Discussion

The eggs of corals typically contain a high content of wax esters (WEs), but larvae with symbionts have a much higher proportion of triacylglycerols (TAGs) than propagules that lack zooxanthellae (Table 1).This suggests that the TAGs are located in the cells of the symbiont and not in the host tissue. Secondly, wax esters are used to power the dispersal of lecithotrophic larvae, being a prime source of fatty acids either for oxidation to provide energy or for incorporation in phospholipids in cell membranes (Lee et al. 2006). In contrast, symbiotic larvae may not need as much WEs because they have zooxanthellae that can provide them with other sources of carbon to power dispersal. Thirdly, we speculate that WEs may also contribute to fertilization ecology (Harii et al. 2007). Their low density may allow the gametes of broadcast spawners to remain at the surface and thus maximize fertilization. The reduction in wax esters relative to other lipids though larval development will also result in reduced buoyancy, thus facilitating larval settlement. On the other hand, propagules that have symbionts may have low WEs because many of these species are able to settle shortly after release (Harii et al. 2002; Nishikawa et al. 2003), and therefore, buoyancy may not be advantageous.

The pattern of lipid consumption observed in G. retiformis was similar to that in Acropora tenuis (Harii et al. 2007): rapid decrease in lipids during early development (until days 2–3), followed by period where lipids are not considerably used (until days 17–21), and again a rapid consumption later in development. The rapid decrease during early development might reflect lipid oxidation fuelling intensive cell division during embryogenesis and motility (Okubo et al. 2008). However, the proportion of lipids consumed during the first days after spawning differs between aposymbiotic and symbiotic larvae. Aposymbiotic larvae of the coral G. retiformis consume 64.3% of its lipids over 30 days; however, 26% of these are consumed in the first 3 days (Fig. 1). Similarly, of all lipids consumed by A. tenuis larvae over a period of 30 days, 58% of the consumption happens in the first 5 days (Harii et al. 2007). The larvae of corals with zooxanthellae also consume lipids, but their rate of consumption is lower: in the first 6 days of development, Pocillipora damicornis only consumes 16% of the total lipids consumed in 30 days (Harii et al. 2010).

The fatty acid composition of the eggs of G. retiformis (Table 4) was similar to those of the adults of this species (Harland et al. 1993) and of the eggs of A. millepora, A. tenuis, and M. digitata (Arai et al. 1993). Palmitic acid, the most abundant fatty acid in G. retiformis eggs, is a trophic marker of zooxanthellae (Papina et al. 2003; Zhukova and Titlyanov 2003). Even though the eggs of G. retiformis do not contain zooxanthellae, zooxanthellae provide palmitic acid to the adult corals (Papina et al. 2003), and therefore, palmitic acid becomes highly represented in their tissue and eggs (Arai et al. 1993; Yamashiro et al. 1999). When adult corals are exposed to a combination of high temperatures and irradiance that disrupt symbiosis, they become deprived of a significant amount of palmitic acid (Oku et al. 2003). Because this is the main fatty acid in coral eggs, the larval quantity and quality following a bleaching event is likely to be negatively impacted, as has been observed on the broadcast spawning soft coral Lobophytum compactum (Michalek-Wagner and Willis 2001). The high abundance of polyunsaturated fatty acids (PUFAs) (36.76 ± 1.22%) is probably due to the fact that symbiotic zooxanthellae in cnidarians contain a lot of PUFAs (Zhukova and Titlyanov 2003; Mortillaro et al. 2009) and the symbionts transfer them to the host tissue (Harland et al. 1993; Papina et al. 2003, 2007; Pitt et al. 2009), after which they are incorporated into their eggs. The EPA/DHA ratio (2.64) suggests that G. retiformis has a medium trophic level, similar to other marine invertebrates such as shrimps and crabs (Rosa et al. 2007). The eggs of G. retiformis contain several fatty acid trophic markers such as 16:0 (zooxanthellae), 14:0 (copepods), 18:1ω9 (marine animals), and 16:1ω7 (diatoms), a high percentage of C18 and C20 PUFA (26.35%, which indicates herbivory), including 18:2ω6 (green algae), and long-chain fatty acids >C22 (microalgae) (Auel et al. 2002; Scott et al. 2002; Dalsgaard et al. 2003). This fatty acid profile indicates that adult corals have a diverse diet and do not rely exclusively on their symbionts. Specifically, several corals, including G. retiformis, have significant amounts of arachidonic acid (ARA) in their tissues and eggs (Arai et al. 1993; Yamashiro et al. 1999; this study). This fatty acid cannot be produced de novo and does not seem to be provided by the symbiont (Papina et al. 2003; Zhukova and Titlyanov 2003). Therefore, ARA is likely to be obtained by heterotrophic feeding on plankton. This indicates that, in order for corals to reproduce in captivity, the provision of live food rich in this essential fatty acid is key. Often corals in captivity are fed with Artemia nauplii that naturally have a low DHA and ARA content (Figueiredo et al. 2009) and therefore need to be previously enriched with these fatty acids to constitute a suitable diet for corals.

In an aposymbiotic larva such as those of G. retiformis, the total biomass of fatty acids can only decrease or remain constant through time because animals lack the biochemical pathways to produce them de novo (Moran and Manahan 2003). Consequently, when the proportion of a particular fatty acid decreases, it must either have been consumed or converted into another fatty acid, but when the proportion increases, it is not possible to infer whether the fatty acid has been produced through conversion or has simply been consumed at a lower rate than other fatty acids (or not consumed at all) (Rønnestad 1995). In G. retiformis, larval development seems to depend on PUFAs because the proportion of most PUFAs declined significantly with larval age. Polyunsaturated fatty acids confer fluidity, flexibility, and selective permeability to cellular membranes, affecting many cellular and physiological processes, including cold adaptation and survival, modulation of ion channels, endocytosis/exocytosis, defense against pathogens, and activities of membrane-associated enzymes that are sensitive to the biophysical properties of lipid membranes (Bergé and Barnathan 2005).

In conclusion, the high lipid content of eggs and larvae, particularly wax esters, potentially provides a source of energy to support long periods in the plankton. Additionally, the high abundance of wax esters in the eggs makes them buoyant, suggesting that selective pressure to increase fertilization success may also shape the biochemical composition of coral eggs.