Intraspecific and interspecific variation in thermotolerance and photoacclimation in Symbiodinium dinoflagellates

Abstract

Light and temperature are major drivers in the ecology and biogeography of symbiotic dinoflagellates living in corals and other cnidarians. We examined variations in physiology among 11 strains comprising five species of clade A Symbiodinium. We grew cultures at 26°C (control) and 32°C (high temperature) over a duration of 18 days while measuring growth and photochemical efficiency (F_v/F_m). Responses to thermal stress ranged from susceptible to tolerant across species and strains. Most strains exhibited a decrease in cell densities and F_v/F_m when grown at 32°C. Tolerance to high temperature (T₃₂) was calculated for all strains, ranging from 0 (unable to survive at high temperature) to 1 (able survive at high temperature). There was substantial variation in thermotolerance across species and among strains. One strain had a T₃₂ close to 1, indicating that growth was not reduced at 32°C for only this one strain. To evaluate the combined effect of temperature and light on physiological stress, we selected three strains with different levels of thermotolerance (tolerant, intermediate and susceptible) and grew them under five different light intensities (65, 80, 100, 240 and 443 µmol quanta m⁻² s⁻¹) at 26 and 32°C. High irradiance exacerbated the effect of high temperature, particularly in strains from thermally sensitive species. This work further supports the recognition that broad physiological differences exist not only among species within Symbiodinium clades, but also among strains within species demonstrating that thermotolerance varies widely between species and among strains within species.

1. Introduction

The ongoing burning of fossil fuels has substantially increased atmospheric concentrations of carbon dioxide [1], causing global temperatures to rise at unprecedented rates [2,3]. These changes are negatively impacting the planet's more sensitive biota. Episodes of high sea surface temperature (SST) have increased in frequency and intensity, affecting the health and viability of reef-building corals, thus degrading the ecosystems they build and sustain [4,5]. Temperatures that are 1.0–1.5°C above normal are physiologically stressful and can disrupt the mutualisms between corals and their dinoflagellate endosymbionts, Symbiodinium. This leads to the whitening, or ‘bleaching’, of corals. While severe events cause significant mortality, not all symbiotic corals are equally affected [6–8], even among conspecifics from the same population [9]. Certain populations, as well as particular species and colonies of the same species, exhibit differential responses to heat stress. Often, these differences in coral mortality are in large part due to physiological differences among the species of Symbiodinium they harbour [10–17] and possibly genotypic differences among individual strains.

Physiological and genetic diversity within Symbiodinium taxa is vast [14,18,19]. Numerous divergent clades (currently designated A through I) comprise the genus Symbiodinium [20] and it is apparent that each clade contains a diversity of species with distinct ecological attributes and physiologies [21–23]. For example, clade C probably contains hundreds of undescribed species that are mutualistic with a wide array of cnidarian hosts [24]. Similarly, there are tens of species within clade B, many of them endemic to the western tropical and subtropical Atlantic [23,25–29]. While the diversity within Symbiodinium—one of the most species-rich dinoflagellate genera—is well recognized, the physiological variability within and between species remains poorly characterized.

Stress tolerance among certain Symbiodinium may help coral populations cope with rising temperatures [9,30–33]. Tolerance of high temperatures was first observed and described in Symbiodinium glynnii, a species in clade D [34–36]. Subsequent studies indicate that Symbiodinium spp. from other clades also tolerate higher temperatures [15,33,37,38]. In the Persian Gulf where SSTs can reach a maximum of 34–36°C, several coral species harbour Symbiodinium thermophilum, a clade C strain. In the Red Sea, coral reefs experience the most extensive seasonal and latitudinal thermal gradients (21–33°C) worldwide. Along this gradient, the coral species Pocillopora verrucosa also exhibit thermotolerant symbionts belonging to clade A (ITS2 type A1 and A21) at higher temperatures (27–33°C) [38]. Beyond interspecies variation, the degree of physiological thermotolerance among strains within populations or among populations across a species' range of distribution remains largely unknown. Individual strains within species of Symbiodinium can exhibit different photophysiologies [18], and such variation will factor into the adaptation of Symbiodinium to rapid climate change. Thus, it is critical to also assess inter-individual physiological variation.

An important aspect of Symbiodinium spp. physiology is their ability to acclimate under physiological stress. For example, members of clade A acclimate differently from one another to various light levels [10–12]. The range of thermal acclimation also differs among Symbiodinium spp. [10–12,14,16,17]. Still, the tempo and mode of acclimatory responses within and between most species under the dual effects of light and temperature requires further study.

We conducted experiments to study the acclimation potential to temperature and light within and among species of Symbiodinium clade A—a phylogenetically well-characterized group [21,22]. We used 11 cultures from different populations corresponding to five species: Symbiodinium microadriaticum (=type A1), Symbiodinium pilosum (=type A1), Symbiodinium tridacnidorum (=type A3^Pacific), Symbiodinium linucheae (=type A4) and Symbiodinium necroappetens (=type A13). We specifically tested the occurrence of thermotolerance among species, as well as if intraspecific variation is comparable to that found among species. We also evaluated the effects of temperature and light on Symbiodinium growth and photochemistry. Our results suggest that thermotolerance is not restricted to individual species, but instead it is widespread throughout clade A, and that species often adjust to variations in light. We also observed that species surviving under stressful light and thermal conditions often incur physiological costs that slow down growth and accelerate photodamage.

2. Methods

(a). Symbiodinium cultures: growth conditions

To quantify physiological responses, we cultured Symbiodinium in ASP-8A medium [39] at 26°C with full-spectrum fluorescent lights at 100 µmol quanta m⁻² s⁻¹ with a 12 : 12 (light : dark) photoperiod. Light intensities were measured inside the culture flasks with a 4π sensor (Biospherical, USA). Cultures were maintained at logarithmic phase by replacing the culture media every 1.5–2 weeks. Symbiodinium strains were obtained from the Trench collection maintained by Todd LaJeunesse at The Pennsylvania State University and the BURR collection at SUNY Buffalo maintained by Mary Alice Coffroth. Herein, we follow a binomial nomenclature for Symbiodinium species [21,22,40] but table 1 for culture collection and strain designations.

Table 1.

Cultures used in this study, culture collection code, ITS2 type, animal species from which the culture was obtained, and geographical location where the culture was collected.

code	species name	strain name	type	collected from	origin	thermotolerance
Smic1	S. microadriaticum	CassKB8	A1a	Cassiopeia xamachana	North Pacific	tolerant
Smic2	S. microadriaticum	RT-061	A1b	Cassiopeia xamachana	Caribbean	tolerant
Smic3	S. microadriaticum	RT-362	A1b	Cassiopeia andromeda	Red Sea	sensitive
Snec1	S. necroappetens	RT-80	A13b	Condylactis gigantea	Caribbean	tolerant
Snec2	S. necroappetens	MAC4-225	A13a	Porites astreoides	Caribbean	sensitive
Slin1	S. linucheae	RT-379	A4b	Plexaura homamalla	Caribbean	tolerant
Spil1	S. pilosum	RT-024	A2b	Bartholomea annulata	Caribbean	sensitive
Spil2	S. pilosum	RT-104	A2b	Heliopora sp.	West Pacific	tolerant
Spil3	S. pilosum	RT-130	A2b	Meandrina sp.	Caribbean	tolerant
Stri1	S. tridacnidorum	RT-292	A3b	Tridacna maxima	West Pacific	sensitive
Stri2	S. tridacnidorum	CassEl-1	A3a	Cassiopeia sp.	North Pacific	sensitive

^aBURR collection at SUNNY Buffalo.

^bTrench collection at The Pennsylvania State University.

(b). High-temperature acclimation: experimental design

To test for acclimation to high temperature, we used 11 cultures from clade A that included five species: three cultures of S. microadriaticum (ITS2 type A1), two cultures of S. necroappetens (ITS2 type A13), three cultures of S. pilosum (ITS2 type A2), two cultures of S. tridacnidorum (ITS2 type A3^Pacific) and one culture of S. linucheae (ITS2 type A4) (table 1). We inoculated cultures at an initial concentration of 1 × 10⁶ cells ml⁻¹ in 15 ml tubes containing 8 ml of culture (day 0). To elevate the temperature to 32°C, we increased from 28.5°C and then gradually increased the temperature by 0.5°C every 6 h until 32°C was reached (day 2). We counted cell densities in triplicate every 3 days, and measured maximum quantum yield of photochemistry (F_v/F_m) daily (see below) for 17 days in four replicates per strain per temperature (control: 26°C and high temperature: 32°C).

(c). High-temperature acclimation and photoacclimation: experimental design

To test for the effects of interaction between temperature and light on Symbiodinium health, we selected three cultures with different levels of known physiological thermotolerance: a tolerant S. microadriaticum (Smic2) with no significant cell densities changes at 32°C, an intermediate tolerant S. microadriaticum (Smic1) with significant cell densities changes at 32°C and a sensitive strain from S. tridacnidorum (Stri1). We inoculated cultures at 1 × 10⁶ cells ml⁻¹ in 15 ml tubes, containing 8 ml of culture, initially grown at 100 µmol quanta m⁻² s⁻¹ at 26°C. We set the high-temperature treatment at 32°C (without ramping) and the control temperature treatment at 26°C. Light levels were adjusted to five different intensities ranging from under-saturating levels to over-saturating light levels: 65, 80, 100, 240 and 443 µmol quanta m⁻² s⁻¹, using mesh fabric to produce shade. We incubated 15 replicates per culture at each of the 10 treatments.

(d). Photochemical efficiency

To quantify photochemical activity, we measured the maximum quantum yield of charge separation of photosystem II (F_v/F_m) with a MiniPAM fluorometer (Walz, USA) after 30–50 min of dark acclimation at the end of the light cycle. We measured daily F_v/F_m in five independent tubes without disturbing the sample. For the species comparisons of 11 strains, these five measurements were performed in at least two independent runs, and in three independent runs for four of these cultures (Snec1, Smic1, Stri2 and Spil1).

(e). Population growth

To evaluate the performance of 11 strains during high-temperature acclimation, we measured cell density by fixing 1 ml aliquots of culture in 10% Lugol's iodine and determined cell densities using a haematocytometer. At least nine replicate aliquots were counted for each sample at each time point. Cell densities were measured for days 0, 6, 9, 13 and 17. We measured them at least once for all strains; Smic1, Snec1, Spil1 and Stri1 were estimated twice (these strains were chosen randomly).

We used these cell density estimates to estimate relative population growth rates (r, units: d⁻¹) at 26 and 32°C, as well as tolerance to 32°C for each strain. We estimated r as the slope of the relationship between log-transformed cell densities and time, a standard technique for growth experiments with photosynthetic microbes and algae that exhibit exponential growth from low initial densities [41]. To avoid taking the logarithm of zero, we added 111 cells ml⁻¹ to each observed count, which is 10% of the smallest observed cell density. We then estimated tolerance to 32°C for each strain, T₃₂, as the ratio of the relative growth rate at 32°C, r₃₂, to the relative growth rate at 26°C, r₂₆. In some cases, strains had negative growth rates at 32°C (i.e. they declined towards extinction). In these cases, we set the tolerance value to a defined minimum value of zero. A tolerance value equal to one indicates that growth is not reduced at 32°C, whereas a value of zero indicates that the isolate cannot grow at this high temperature.

(f). Statistical analysis

We examined how cell densities and photochemical efficiency (F_v/F_m) were affected by temperature, comparing Symbiodinium strains (11), temperatures (28.5, 29, 29.5, 30, 30.5, 31, 31.5 and 32°C) and time (days 0, 6, 9, 13 and 17). We also analysed the effect of light and temperature, comparing Symbiodinium strain (2), temperature (26 and 32°C) and light intensity (65, 80, 100, 240 and 443 µmol quanta m⁻² s⁻¹). One-way ANOVAs and ANCOVAs (fixed-effects, LS) were used to analyse the effect of temperature and/or light; all two-way interactions were included. We used JMP Pro 11 for all analyses (SAS Institute, Cary, NC, USA, 2014).

We tested for significantly negative effects of high temperature on strain growth using bootstrapping tests (105 iterations, alternative hypothesis: T₃₂ < 1) in which we randomly resampled data with replacement stratified by each strain, temperature and sampling date [42]. Strains with no thermotolerance are shown in red, while strains with some thermotolerance are shown in a gradient from purple (T₃₂ > 0) to blue (T₃₂ = 1).

3. Results

(a). Variation in temperature sensitivity within and among species

Significant differences in F_v/F_m were measured among species (figure 1) and within species (figure 3). Our linear models identified all factors (strain, day, temperature) and two-way interactions as having a significant effect on Symbiodinium cell densities and F_v/F_m (table 2 and figures 2 and 3). Temperature×day explained 42.5% of variance in cell densities responses, and 68.1% variance in F_v/F_m values (see F ratios, table 2). The relative population growth rate at 32°C was significantly different among strains (figure 4).

Figure 1.

Average photochemical efficiency (F_v/F_m) at 32°C of five Symbiodinium spp. over a 17-day period. Slin, S. linucheae (one strain); Smic, S. microadriaticum (three pooled strains); Snec, S. necroappetens (two pooled strains), Spil, S. pilosum (three pooled strains) and Stri, S. tridacnidorum (two pooled strains). Bars depict one standard deviation of the mean. (One-way ANOVA, F_4,1158 = 191.5, p < 0.0001, Tukey–Kramer HSD.) Grouping from post hoc comparisons are depicted with letter above each bar.

Figure 3.

Photochemical efficiency (F_v/F_m) of 11 strains of clade A under different growth temperatures, at 100 µmol quanta m⁻² s⁻¹. Blue line indicates linear regression of data under control temperature (26°C) and red line indicates linear regression of data under high temperature (32°C). Dotted lines indicate confidence intervals. We found that significant differences at the p < 0.0001 level between the control and the high-temperature treatments for all species except Spil2 and Spil3, which showed no statistical difference.

Table 2.

ANCOVA analysis examining cell density response (log₁₀ transformed) and photochemical efficiency (F_v/F_m) of 11 strains of dinoflagellates to high temperature. Cell densities and F_v/F_m showed similar results where temperature was the major source of variability. Cell densities were measured on days 0, 6, 9, 13 and 17 in three replicates per culture per treatment (n = 18). F_v/F_m was measured every day for 17 days with six technical replicates. F_v/F_m was measured at least two times for all strains and three times for four of them in independent repetitions of the experiment. Cultures were grown at 100 µmol quanta m² s⁻¹ under two temperatures: control (26°C), and high temperature (32°C). d.f., degrees of freedom.

response	source	d.f.	F ratio	Prob > F
cell densities (log transformed)	strain	10	264.87	less than 0.0001
	temperature	1	228.32	less than 0.0001
	day	1	2135.78	less than 0.0001
	strain×temperature	10	58.61	less than 0.0001
	strain×day	10	235.38	less than 0.0001
	temperature×day	1	946.77	less than 0.0001
whole model test: F33,2678 = 272.37, p < 0.0001, R2 = 0.770
photochemical efficiency	strain	10	413.379	less than 0.0001
	temperature	1	534.5921	less than 0.0001
	day	1	3938.295	less than 0.0001
	strain×temperature	10	29.871	less than 0.0001
	strain×day	10	203.1659	less than 0.0001
	temperature×day	1	658.8297	less than 0.0001
whole model test: F33,2278 = 378.1, p < 0.0001, R2 = 0.846

Figure 2.

Cell density (log₁₀ transformed) of 11 strains of clade A under different growth temperatures, at 100 µmol quanta m⁻² s⁻¹. Blue line indicates linear regression of data under control temperature (26°C) and red line indicates linear regression of data under high temperature (32°C). Dotted lines indicate confidence intervals. To avoid taking the logarithm of zero, we added 111, which is 10% of the lowest observed non-zero density, to each value before transformation. A value closer to 2 indicates zero cells. We found that significant differences at the p < 0.0001 level between the control and the high-temperature treatments for all species except Smic2, Spil2 and Spil3, which showed no statistical difference.

Figure 4.

Relative population growth rate at 32°C with simplified phylogeny of Symbiodinium clade A inferred from ITS1, ITS2 and 5.8S rDNA sequences [21,22,40]. (One-way ANOVA, F_10,310 = 3604.73, p < 0.0001, Tukey–Kramer HSD.) Error bars denote standard error.

Grouping strains by species (figure 1), we could conclude that there are three groups of species with different thermotolerance: S. microadriaticum and S. pilosum (group A), S. linucheae and S. necroappetens (group B) and S. tridacnidorum [group C—Tukey–Kramer honestly significant difference (HSD)]. However, fine scale analysis at the level of strains within species showed higher thermotolerance diversity (figures 2–5).

Figure 5.

Thermotolerance (T₃₂) versus photochemical efficiency (F_v/F_m) at high temperature (32°C). Error bars denote standard error.

Using our tolerance estimate, T₃₂, we categorize strains into sensitive (T₃₂ = 0) or tolerant (T₃₂ > 0) (table 1 and figure 5). All strains were significantly affected by high temperature (Spil2 p = 0.0257, Smic2 p = 0.0006, p = 0 for the rest of the strains). An exception was Spil3 (p = 0.3704), which was not affected by high temperature (T₃₂ = 0.933). Our thermotolerant strains, in order of tolerance, are Spil3, Spil2, Smic2, Snec1, Smic1 and Slin1. The most tolerant strains (Spil3, Spil2 and Smic2) showed no significant difference in cell densities (log₁₀ transformed) at high temperature when compared with controls (figures 2 and 5). Only the two most tolerant strains, Spil3 and Spil2, showed no significant difference on F_v/F_m at high temperature compared with control temperature (figure 3). The relative population growth at high temperature was positive for all tolerant strains (figure 4), indicating the ability to grow at high temperature. For the thermotolerant strains Snec1, Smic1 and Slin1, cell densities and F_v/F_m were significantly different among treatments indicating a cost in fitness when acclimated to high temperature (figures 2 and 3).

The sensitive strains, in order of thermosensitivity, are: Stri1, Stri2, Snec2, Smic3 and Spil1 (figures 2–5). Thermosensitive strains showed negative values of relative population growth (figure 4) indicating the inability to survive at high temperature (T₃₂ = 0, figure 5). The most thermosensitive strain, Stri1, decreased its population size faster than Stri2 and so on. Sensitive strains Stri1, Stri2 and Snec2 showed higher sensitivity because cultures died during the experiment. Although cultures of Smic3 and Spil1 contained available cells by the end of the experiment (figure 2), the negative trend indicates the culture would likely die if the experiment was conducted for a longer duration (figure 4). All these strains failed to thrive under high temperature, indicating an inability to cope with high temperature.

Species thermotolerance cannot be estimated by most species in this study. This is particularly true for S. microadriaticum, S. necroappetens and S. pilosum. For example, for S. microadriaticum, the two strains Smic1 and Smic3 decreased in cell density, with Smic3 showing greater decrease in density than Smic1 (average 3.9 × 10⁴ cells ml⁻¹, versus 22.6 × 10⁴ cells ml⁻¹ at day 17 of the experiment, figure 2). In both cases, F_v/F_m also decreased (figure 3). While cell densities were low for both, these cell cultures were still physiological active (mean F_v/F_m of 0.489 for Smic1 and 0.310 for Smic3). However, the relative population growth rate at high temperature indicates that Smic3 was negatively affected by high temperature, becoming unable to thrive. By contrast, cell densities of Smic2 did not change significantly with temperature, although there was a significant decrease in F_v/F_m indicating that it had acclimated better to the heat stress. For S. necroappetens, the tolerant Snec1 strain decreased significantly in cell density and showed a significant decrease in F_v/F_m under high temperature (figures 2 and 3). By contrast, the sensitive strain Snec2 died under high-temperature conditions. For S. pilosum, the cell densities for Spil1 remained constant until day 13, when growth rates began to decrease (figure 2). Photochemical efficiency for Spil1 was significantly lower than controls at high temperature (figure 3). Thus, the cells remained physiologically active (F_v/F_m = 0.534 versus 0.339), but this activity was insufficient to allow the culture to grow (figure 4). The F_v/F_m values for thermotolerant strains Spil2 and Spil3 were not significantly different under high-temperature treatment, and both maintained similar cell densities when grown under control and treatment conditions (electronic supplementary material, table S1; figure 2).

Strains of S. tridacnidorum (Stri1 and Stri2) were thermosensitive and both died when exposed to high temperature (figures 2, 4 and 5). Cell densities and F_v/F_m in these cultures decreased soon after the high-temperature treatment was initiated (figure 2). Symbiodinium linucheae (Slin1) exhibited lower growth rates under the control temperature, compared to other clade A species. The relative population growth rate was positive at high temperature (figure 4), demonstrating its ability to survive under the temperature treatment (figures 2 and 3). For these two species, we cannot categorize them as thermosensitive or thermotolerant, given the limited number of strains, we were able to include in this study.

(b). The synergistic effect of light and temperature in photochemistry and growth

Three strains with different thermal tolerances were chosen to evaluate the synergistic effects of light and high temperature (i.e. Smic2, Smic1 and Stri1). Under five different light levels (from low light to high light), all three strains had a linear photoacclimation response independent of temperature (table 3 and figure 6). F_v/F_m was highest at low light and decreased at high light. This linear response had similar slopes for both control and high temperature for all strains. The levels of photochemical efficiency were variable among strains. For the high tolerant strain Smic2 (T₃₂ = 0.44), F_v/F_m changed equally under both temperatures and different light intensities. The intermediate tolerant strain Smic1 (T₃₂ = 0.32) and sensitive strain Stri1 (T₃₂ = 0) decreased F_v/F_m under high temperature but maintained the same slope at both control and high-temperature conditions. The values measured for F_v/F_m decreased at a faster rate for Stri1 than for Smic1. After 7 days of experiment, F_v/F_m remained constant for both tolerant strains. The sensitive strain under high temperature vastly decreased F_v/F_m and no live cells remained at the end of the experiment (electronic supplementary material, figure S1).

Table 3.

ANCOVA analysis examining the combined effect of temperature and light on Symbiodinium in photochemical efficiency (F_v/F_m). F_v/F_m was measured after 12 h of exposure to light/temperature treatments with six technical replicates. Cultures were grown under control (26°C), and high temperature (32°C), and five light intensities 65, 80, 100, 240, 423 µmol quanta m⁻² s⁻¹. d.f., degrees of freedom. Whole model test: F_9,140 = 52.987, p < 0.0001, R² = 0.773.

response	source	d.f.	F ratio	Prob > F
photochemical efficiency	strain	2	67.85	less than 0.0001
	temperature	1	138.118	less than 0.0001
	relative light	1	49.565	less than 0.0001
	strain×temperature	2	66.3663	less than 0.0001
	strain×relative light	2	8.1871	0.0004
	temp×relative light	1	4.3963	0.0378

Figure 6.

Relative photochemical efficiency of three strains with different thermotolerances under different light conditions. Relative light was calculated by subtracting the total light dose of the experimental conditions (65, 80, 100, 240 and 443 µmol quanta m² s⁻¹) from the initial light conditions (100 µmol quanta m² s⁻¹). Relative photochemical efficiency was calculated with the initial F_v/F_m minus the value after 12 h at the light/temperature conditions. (a) Tolerant strain (Smic2), (b) intermediate tolerant strain (Smic1) and (c) sensitive strain (Stri1). T₃₂, thermotolerance value. Blue line indicates linear regression of data under control temperature (26°C) and red line indicates linear regression of data high temperature (32°C). Dotted lines indicate confidence intervals.

4. Discussion

This work substantiates the growing realization for the existence of broad physiological variation in thermal tolerance among and within species of Symbiodinium. Physiologies of strains that are sensitive, intermediate or tolerant appear to be independent of the species under study. There are species comprised strains that are generally thermotolerant (S. linucheae), while strains from other species (e.g. S. tridacnidorum) are sensitive. Other species within clade A (S. microadriaticum, S. necroappetens and S. pilosum) contain strains that diverge significantly in thermal tolerance. Thus, acclimation to high temperature is dependent on the species identity and the physiological attributes possessed by a particular genotype within a species, i.e. strain. This work also supports previous findings that high light and temperature interact to accelerate physiological breakdown [43], especially among strains with limited or no tolerance to high temperature.

(a). Widespread thermotolerance variation within species from clade A lineage

These findings indicate that at least some strains from each of the clade A species examined here can persist under 32°C, with the possible exception of S. tridacnidorum. It appears that thermotolerance is not phylogenetically restricted to certain species, but may indeed be extensive across members of clade A (figure 4). While earlier studies reported limited ‘within clade’ variation in thermal stress [13,14,33,38], our work with additional strains provides evidence for more widespread variation at intraspecific taxonomic levels (i.e. individual strains within species) than previously expected [44].

Extensive variation to thermal tolerance within species hints at genetically encoded variation among populations. Comparing molecular mechanisms at supraspecific taxonomic ranks (i.e. across species and across clades) may be less fruitful than infraspecific comparisons (i.e. within species), given the amount of genetic divergence within the genus. Major Symbiodinium clades diverged between 50 and 60 Ma [13,45]; a divergence time similar to bird [46] and primate families [47]. It is thus not surprising that contradictory results have been reported on the molecular mechanisms underlying thermotolerance [48,49]. Comparative transcriptomics of sensitive (Symbiodinium C3 K) versus thermotolerant (clade D) species (i.e. from separate clades) could not postulate a clear mechanism for thermal acclimation due to the long divergence time between clade C and D (40 Ma according to [45]). Similarly, Rosic et al. [50] were unable to see converging molecular patterns of thermotolerance between S. pilosum (=type A2), Symbiodinium psygmophilum (=type B2), Symbiodinium goreaui (=type C1) and Symbiodinium trenchii (=type D1a), representing species from highly divergent lineages. In this study, we demonstrate that comparative analysis at infraspecific taxonomic ranks (i.e. genotypes within and between closely related species) will be necessary if we want to understand underlying genetic attributes responsible for variation in thermotolerance.

In all clade A species examined, we find that 26°C is closer to the optimal growth temperature than 32°C. Growth rate appears to be much more affected than F_v/F_m by high temperature. The adverse effects of high temperature (above 32°C) on photosynthesis have been extensively characterized for Symbiodinium grown in culture [10,13–15,17]. For instance, in a strain of S. microadriaticum, photosynthesis impairment occurs at temperatures above 30°C, stopping completely at higher temperatures of 34–36°C [10]. Our results show that for 11 strains comprising five species of Symbiodinium clade A, a temperature of 32°C was high enough to distinguish between sensitive and thermotolerant strains. Notably, thermotolerance is not phylogenetically constrained and instead, there are sensitive, intermediate and tolerant populations within species. Although we were able to statistically categorize thermotolerance into these groups, a gradient in thermotolerance can be observed (figure 5). Some strains might fall into the tolerant side of the gradient, such as members of S. pilosum and S. microadriaticum, while others such as strains from S. tridacnidorum fall into the sensitive side of the tolerance gradient. Nevertheless, the thermotolerance of more strains from each species should be evaluated in order to make broader generalizations.

(b). Combined effect of light and temperature in physiological acclimation

Most investigators recognize the importance of considering the role of multiple stressors in order to improve predictions regarding physiological responses to climate change [51]. Acclimation to high temperature in Symbiodinium often comes with physiological trade-offs, including decreased photochemical activity and reduced growth [52]. It has been observed that during coral bleaching, high temperatures are usually interacting with high levels of light [4] and variation in available nutrients [53]. Combinations of various environmental factors will elicit different physiological responses by different genotypes and species. Having the ability to acclimate to multiple stressors may come with a cost to the organism's physiology. Studying these environmental factors separately and in combination aids the identification of possible trade-offs and any consequences for the functional performance of a particular host–symbiont combination. Our results showed that some tolerant strains had the ability to acclimate well to both high light and high temperature. The strain with intermediate tolerance tested here was affected by high temperature during photoacclimation. By contrast, thermal stress in the sensitive strain Stri1 was compounded by exposure to high light. Thus, the combined effect of light and temperature exacerbates stress, particularly in thermosensitive species and genotypes.

The ability to acclimate to multiple interacting factors including temperature and light should be further studied, especially among closely related Symbiodinium lineages. These physiological attributes will likely determine the future ecological success of certain Symbiodinium taxa to changing climatic conditions [54]. Understanding further the synergistic effects of temperature and light on Symbiodinium physiology is crucial for predicting the response of coral–algal symbioses to climate change.

5. Conclusion

A comparative approach—examining closely related species and individual genotypes within species—revealed broad physiological diversity among and within Symbiodinium spp. Thermotolerance is not species- or clade-specific, but rather widespread and varied among members of the genus. Moreover, the further delineation of species within the genus should improve research on mechanisms of thermotolerance, especially when comparing the performance of certain host–symbiont combinations under various environmental conditions.

Data accessibility

Data are available from the Dryad Digital Repository at: https://doi.org/10.5061/dryad.429vh [55].

Competing interests

We declare we have no competing interests.

Funding

This project was funded by CONACYT PhD award 305321, Institutional Research and Career Development Award NIH/NIGMS K12 GM000680, NSF grants nos OCE 1442206 and OCE 1642311 and Penn State startup funds.