Introduction

Concerns for the future of coral reefs have been heightened by observations of widespread coral mortality following severe mass bleaching events (Berkelmans and Oliver 1999). Recent estimates of temperature increases [1.8–4.0°C by 2100; (IPCC 2007)] are consistent with predictions that conditions will exceed the temperature tolerances of corals in the Great Barrier Reef (GBR) annually by 2050 (Hoegh-Guldberg 1999). The capacity of damaged coral communities to recover through larval recruitment from surviving genotypes is critical to their future. Although the sensitivity of mature coral colonies to increased temperatures has been extensively studied, there have been relatively few investigations of the sensitivity of other life stages to unusually high yet foreseeable sea surface temperatures.

The majority of scleractinian coral species rely on external fertilization and larval development in the water column following annual spawning (Harrison and Wallace 1990). The first cleavage typically takes place 1–2 h. after fertilization and embryos develop symmetrically for the first 3–4 cleavages. Abnormal embryonic development is most easily detected by direct observation up until the 16-cell stage. The embryos then develop asynchronously until gastrulation and progressive differentiation and ciliation into motile, pear-shaped planula larvae. Larval cohorts are usually competent to undergo settlement and metamorphosis after 4–5 days of development. Rising seawater temperatures (SWT) can directly affect coral reproduction by causing embryonic abnormalities (Bassim et al. 2002) or reducing larval fitness and survival (Bassim and Sammarco 2003). Reproductive success can also be affected indirectly by increases in SWT and gametes from corals that have suffered previous bleaching may perform sub-optimally (Szmant and Gassman 1990; Michalek-Wagner and Willis 2001; Omori et al. 2001; Ward et al. 2002).

Little is known of direct temperature effects on the fertilization of Indo-Pacific coral species and how this might relate to the thermal tolerance of adult colonies. This laboratory study examined the influence of elevated SWT on fertilization of gametes from three coral families, each with different morphological characteristics. Mature colonies from these families differ in their relative susceptibility to bleaching, with field surveys indicating a hierarchy of sensitivity to elevated SWT: Acropora the most susceptible, followed by Mycedium and Favites being the most resistant to bleaching and bleaching-induced mortality (Marshall and Baird 2000). Fertilization was tested at a range of temperatures that represent average conditions during spawning seasons, 2°C below average and 5–6°C above average, the range of temperature increases that might be expected under climate change scenarios during this century (IPCC 2007).

Materials and methods

Two experiments were conducted on gametes of the reef-building corals (1) Favites chinensis and (2) Acropora millepora in order to determine broad thermal tolerance ranges fertilization in these species. The final experiment (3) was conducted on A. millepora, Favites abdita and Mycedium elephantotus. Abnormal embryo development and cell division rates were also measured in this final experiment. A. millepora is a short-branched plate coral, F. abdita and F. chinensis are massive corals and M. elephantotus forms laminar plates. Each of these species are common across the Indo Pacific and oocytes of each species are aposymbiotic.

Favites chinensis (Okinawa)

Colonies were collected from the reef flat off Sesoko Island, Okinawa (26°38′N, 127°52′E) and transported to Sesoko Station, Tropical Biosphere Research Center. The egg-sperm bundles were collected from individual colonies after spawning and gently agitated to separate the eggs and sperm into two layers. The buoyant eggs were then isolated from the sperm by suction and washed three times in sperm-free seawater, while a stock of 106 sperm ml−1 was prepared from a second colony.

Six-well polystyrene tissue culture plates (NuncTM, Denmark), each well containing 8 ml seawater, were equilibrated for 60 min along a thermal gradient (plates were rested on a 5 mm thick aluminum plate which was heated at one end with a thermostatically controlled heating unit). All fertilization experiments were performed under low light (less than 10 μmol quanta m−2 s−1). Eggs (∼120) and sperm were combined in the wells at a final sperm concentration of 105 sperm ml−1. The temperature of each well was measured directly with a calibrated digital thermometer and the treatments were grouped into seven ranges: 25.5, 26.4, 27.0, 27.8, 28.9, 30.2, 31.6°C (range ± 0.5°C). The ambient temperature in Okinawan waters at spawning is 27°C. The number of replicate wells for each temperature was 3. Early fertilization (first and second cleavage) was assessed after 3 h using a dissecting microscope.

Acropora millepora (GBR)

A second experiment was performed on A. millepora gametes. Parental colonies were collected from Davies Reef on the GBR (18°50′S, 147°37′E) at ∼6 m depth. Colonies were transferred to the Australian Institute of Marine Science outdoor, flow-through aquaria. The fertilization experiments were conducted as for F. chinensis but embryos were sampled at 2 and 4 h and fixed with 4% Bouins solution for later assessment of fertilization and gamete quality. Temperature treatments included 28.3, 30.4, 31.8, 32.8, 34.1°C (range ± 0.5°C) in three replicate wells. Ambient temperature at the time of spawning on the GBR (Davies Reef) was 28°C.

Three species with pre-acclimation of gametes (GBR)

Acropora millepora, F. abdita and M. elephantotus, were collected from Davies Reef as described above. Sperm stocks were prepared at 2 × 105 sperm ml−1 and the sperm and eggs were then transferred separately into six replicate 20 ml glass scintillation vials per temperature treatment (∼120 eggs or 2 × 105 sperm per 5 ml seawater). The vials were then floated in water baths set at 26, 28, 30 and 32°C (range ± 0.3°C) for 30 min in order to pre-expose the eggs and sperm separately prior to fertilization. The short pre-exposure better mimics the field situation where gametes would be exposed to elevated SWT prior to contact with partner gametes. The eggs and sperm were then combined at a final sperm concentration of 105 sperm ml−1. The embryos were sampled and fixed at 2 and 4 h as above for fertilization assessment.

Data analysis

Percentage data were arcsine transformed and differences in response to temperature were assessed using one-way analysis of variance (ANOVA). The Tukey Honest Significant Difference was performed on each transformed data set and p < 0.05 considered significant (Table 1). Statistical analyses were performed using Statistica 6.0, StatSoft Inc., Tulsa, OK, USA.

Table 1 Summary of one-way analysis of variation results
Full size table

Results and discussion

Fertilization rates were consistently high (79–91%) in Okinawan F. chinensis across all temperatures up to 31.6°C, which is almost 5°C above ambient SWT (Fig. 1, Table 1). Fertilization was greater than >90% for GBR A. millepora eggs up to 30.4°C after 2 h, but this dropped significantly to only 13% at 31.8°C, which is almost 4°C above ambient SWT (Fig. 1, Table 1). Fertilization success of A. millepora eggs remained almost identical after an additional 2 h of development. While normal development was characterized by radial holoblastic cleavage resulting in equally-sized blastomeres, aberrant development was clearly observed at higher temperatures in the form of asymmetrical and irregular cleavages (Fig. 2). At 34.1°C no fertilization was recorded in A. millepora eggs and cellular integrity was sometimes compromised. At this temperature, pairs of eggs occasionally fused, forming larger chimeras.

Fig. 1
figure 1

Proportions of normally fertilized Okinawan Favites chinensis eggs after 3 h and GBR Acropora millepora eggs after 2 and 4 h (% ± SE, most bars are smaller that symbols)

Full size image
Fig. 2
figure 2

Representative images of A. millepora eggs and embryos at different temperatures and times. Egg diameter approx 550 μm

Full size image

In the final experiment, high levels of fertilization were observed for all three GBR species after 4 h at 26 and 28°C [A. millepora (98 ± 1% SE), F. abdita (84 ± 2%) and M. elephantotus (94 ± 2%)] (Fig. 3). At 32°C, significant reductions in fertilization and increases in abnormal development were observed for A. millepora at the first cell cleavage stage (2 h) (Fig. 3, Table 1). The frequency of abnormalities in embryos exposed to 32°C increased over time for A. millepora, reaching 42 ± 9% (mean ± SE) by the conclusion of the experiment at 4 h (Fig. 2).

Fig. 3
figure 3

Proportion of normally fertilized eggs (% ± SE, most bars are smaller than symbols), proportion of embryos exhibiting abnormal development and the average number of cell divisions (not counting unfertilized eggs) of each coral species after 2 and 4 h at four different temperatures

Full size image

The gametes of both the Okinawan and GBR favids were not affected by SWT nearly 5°C higher than would be normally experienced during spawning (Figs. 1, 3). The fertilization of A. millepora, on the other hand, was negatively affected at 4°C above ambient SWT, irrespective of whether gametes were pre-conditioned at this temperature. The sperm were motile at all temperatures and it is not clear whether eggs or sperm were more compromised at 32°C. Thermal and irradiance exposure history of parental colonies can affect coral fertilization (Szmant and Gassman 1990; Omori et al. 2001; Ward et al. 2002); however, each of the parental colonies used in Experiment 3 were from the same broad reef habitat and no bleaching was recorded in the 12 months leading up to the spawning. Thermal tolerance of the gametes may instead be governed directly by genetic differences and/or be dependent on the host/symbiont relationship and physiology of parental colonies. Impairment of functional enzymes and proteins at high temperature is likely to be the cause of reduced fertilization and increased abnormality in embryo development. Adult corals employ an array of protective cellular mechanisms, including heat shock proteins, glutathione, ubiquitin and superoxide dismutases, in response to oxidative stress caused by elevated temperature and/or irradiance (Brown et al. 2002). It is possible that coral gametes employ similar cellular tools to maintain the structural integrity of critical proteins and enzymes, and that this protection varies among species.

Although in this experiment, only early embryogenesis of A. millepora was demonstrated to be sensitive to the elevated temperatures used, the effect of exposing F. abdita and M. elephantotus embryos beyond the 5th cell division may also reveal that these species are sensitive to increased SWT. For example, abnormalities and mortality in later stage embryos and larvae were demonstrated in the Atlantic massive coral Diploria strigosa exposed to elevated temperatures (Bassim et al. 2002; Bassim and Sammarco 2003). Diploria strigosa embryos developed more slowly than A. millepora and in this earlier work, abnormal embryonic development was not observed until the 4th cell cleavage (6 h). The proportion of aberrant D. strigosa embryos at 5 h was negligible but after 24 h had reached >50% at both 30 and 32°C (Bassim et al. 2002).

The effects of thermal stress on coral fertilization and early embryo development may be compared to the effects on other life stages. For instance, the exposure of D. strigosa larvae to SWTs of 30 and 32°C for up to 9 days resulted in higher mortality rates and impaired larval searching behavior, settlement and metamorphosis (Bassim and Sammarco 2003). In controlled laboratory exposures, adult Acropora formosa from the central GBR exhibited 50% bleaching after 5 days at 31–32°C while Pocillopora damicornis and Acropora elseyi were slightly more tolerant, bleaching at 32–33°C (Berkelmans and Willis 1999). The results of the present study and others (Edmunds et al. 2001; Bassim and Sammarco 2003) indicate that, despite the absence of symbiotic dinoflagellates, fertilization, embryogenesis and larval development of A. millepora is also susceptible to direct temperature stress.

The average number of cell cleavages was 1–1.5 at 2 h and this increased to 4–5 cleavages for each species by 4 h (Fig. 3). Cell cleavage rates for all species increased slightly but significantly (10–20%) with temperature by 4 h (Fig. 3, Table 1). Despite A. millepora exhibiting the highest frequency of abnormalities at 32°C, those cells which divided normally did so more rapidly (4.65 ± 0.07 at 32°C compared with 4.10 ± 0.07 at 26°C). Faster rates of cell division at 30 and 32°C was also reported for D. strigosa at 6 h but again embryo integrity was compromised at these higher temperatures (Bassim et al. 2002). Increases in SWT are likely to cause an increased metabolic rate, which may explain the higher cleavage rates as temperatures increase. If embryonic development is accelerated and the larval pre-competency period is reduced at higher temperatures, more limited larval dispersal would be expected. This phenomenon has been described for numerous fish and invertebrates (O’Connor et al. 2007) and can affect local community structure and biodiversity as well as reducing connectivity between populations. The combined effects of fertilization failure and accelerated embryonic development in some coral species are likely to exacerbate the ecological impacts of climate change by reducing biodiversity and hindering recovery of reefs damaged by mass coral bleaching.