ABSTRACT

Theoretical models predict that ocean acidification, caused by increased dissolved CO2, will reduce the maximum thermal limits of fishes, thereby increasing their vulnerability to rising ocean temperatures and transient heatwaves. Here, we tested this prediction in three species of damselfishes on the Great Barrier Reef, Australia. Maximum thermal limits were quantified using critical thermal maxima (CTmax) tests following acclimation to either present-day or end-of-century levels of CO2 for coral reef environments (∼500 or ∼1000 µatm, respectively). While species differed significantly in their thermal limits, whereby Dischistodus perspicillatus exhibited greater CTmax (37.88±0.03°C; N=47) than Dascyllus aruanus (37.68±0.02°C; N=85) and Acanthochromis polyacanthus (36.58±0.02°C; N=63), end-of-century CO2 had no effect (D. aruanus) or a slightly positive effect (increase in CTmax of 0.16°C in D. perspicillatus and 0.21°C in A. polyacanthus) on CTmax. Contrary to expectations, early-stage juveniles were equally as resilient to CO2 as larger conspecifics, and CTmax was higher at smaller body sizes in two species. These findings suggest that ocean acidification will not impair the maximum thermal limits of reef fishes, and they highlight the critical role of experimental biology in testing predictions of theoretical models forecasting the consequences of environmental change.

INTRODUCTION

Climate change is causing concomitant increases in average temperature and carbon dioxide partial pressure (PCO2) in aquatic environments, as well as increasing the frequency of transient heatwaves (Hughes et al., 2003; IPCC, 2013; Pearce and Feng, 2013; Wernberg et al., 2013; Popova et al., 2016). Tropical organisms tend to have narrower thermal performance curves than temperate species and they are forecasted to be among the most sensitive to climate warming, either because they have evolved under relatively stable thermal conditions (Tewksbury et al., 2008; Lough, 2012) or because rate processes (e.g. metabolism) change proportionally faster in tropical systems as a result of warmer ambient temperatures (Payne and Smith, 2017). Regardless of what may cause heightened thermal sensitivity of organisms in tropical systems, little is known of how increased PCO2 (i.e. acidification) may interact with thermal stress in these or other environments. Theoretical models, appearing in the latest report from the Intergovernmental Panel on Climate Change (Pörtner et al., 2014), predict that elevated PCO2 reduces the aerobic metabolic scope of aquatic ectotherms at a given temperature and consequently narrows the species-specific thermal window over which organisms can function (Fig. 1A; Pörtner, 2008; Pörtner and Farrell, 2008). Thus, these models predict that elevated PCO2 will decrease thermal tolerance limits. While some aspects of these models are hotly debated, namely the assertion that thermal tolerance, niche patterns and fitness are governed by the temperature dependence of aerobic metabolic scope (Clark et al., 2013b; Jutfelt et al., 2014; Ern et al., 2016; Lefevre, 2016), insufficient attention has been focused on testing the prediction that thermal limits will be negatively affected by end-of-century levels of PCO2. Indeed, even in the absence of a change in aerobic metabolic scope, there could be other mechanisms such as cellular acid–base disequilibria that cause a negative impact of PCO2 on thermal limits. Prior to extensive investigations into putative mechanisms, it is prudent to first quantify whether elevated PCO2 does indeed impact thermal limits.

Fig. 1.

Theoretical models of the effects of temperature and elevated PCO2 on ectotherms. (A) Theoretical changes in the performance (i.e. fitness) of ectothermic animals as a function of temperature, where animals under present-day levels of PCO2 (blue curve; e.g. 400 µatm) have higher performance and a broader thermal performance curve than animals under future projected levels of PCO2 (red curve; e.g. 1000 µatm). Putative mechanisms to explain this hypothetical response include cellular acid–base disequilibria and/or a reduction in the scope for oxygen transport (i.e. aerobic metabolic scope) of the animal. Such a response to PCO2 would lower the critical thermal maximum (CTmax) of animals in an elevated PCO2 environment (leftward shift in dashed vertical line). (B) Thermal limits (e.g. CTmax) have been proposed to increase throughout ontogeny and reach peak values in late-stage juveniles, subsequently declining as animals approach reproductive maturation and spawning. Increased environmental PCO2 is predicted to decrease the thermal limits of all life stages (shift from blue to red curve) (adapted from: Pörtner, 2008; Pörtner and Farrell, 2008).

Fig. 1.

Theoretical models of the effects of temperature and elevated PCO2 on ectotherms. (A) Theoretical changes in the performance (i.e. fitness) of ectothermic animals as a function of temperature, where animals under present-day levels of PCO2 (blue curve; e.g. 400 µatm) have higher performance and a broader thermal performance curve than animals under future projected levels of PCO2 (red curve; e.g. 1000 µatm). Putative mechanisms to explain this hypothetical response include cellular acid–base disequilibria and/or a reduction in the scope for oxygen transport (i.e. aerobic metabolic scope) of the animal. Such a response to PCO2 would lower the critical thermal maximum (CTmax) of animals in an elevated PCO2 environment (leftward shift in dashed vertical line). (B) Thermal limits (e.g. CTmax) have been proposed to increase throughout ontogeny and reach peak values in late-stage juveniles, subsequently declining as animals approach reproductive maturation and spawning. Increased environmental PCO2 is predicted to decrease the thermal limits of all life stages (shift from blue to red curve) (adapted from: Pörtner, 2008; Pörtner and Farrell, 2008).

Thermal limits of aquatic organisms such as fishes can vary greatly across latitudinal scales, yet interspecific differences at regional scales are much less understood. This is particularly true for tropical systems, where the research effort has been small compared with temperate systems (but see Mora and Ospína, 2001; Ospína and Mora, 2004; Eme and Bennett, 2009). Moreover, relatively little is known about factors (beyond thermal acclimation) that drive thermal limits at an intraspecific level, although available evidence suggests that disease state, ontogenetic stage (Fig. 1B) and body size might be influential determinants (Vaughan and Coble, 1975; Brewer, 1976; Hutchison, 1976; Daufresne et al., 2009; Rijnsdorp et al., 2009; Forster et al., 2012; Cheung et al., 2013). In general, early life stages are thought to be the most sensitive to environmental stressors because their physiological/regulatory systems may not be fully developed (Fig. 1B; Brewer, 1976; Pörtner et al., 2005; Rijnsdorp et al., 2009; Freitas et al., 2010), yet much of the support for this idea stems from comparisons across studies/species rather than dedicated investigations across an ontogenetic range within a study/species (but see Komoroske et al., 2014). These knowledge gaps impede our capacity to identify sensitive species or life stages that may be of particular conservation concern.

With a rich history in comparative physiology (Becker and Genoway, 1979; Paladino et al., 1980; Lutterschmidt and Hutchison, 1997), the critical thermal maximum (CTmax; the upper temperature at which an animal loses motor function) has emerged as a potentially useful metric for understanding species’ distribution ranges and resilience to heatwaves and climate warming (Bennett and Beitinger, 1997; Sunday et al., 2012, 2014). While measurements of CTmax are typically conducted using thermal ramping rates that are relevant to heatwave scenarios or habitats such as tide pools, emerging evidence suggests that qualitative trends from CTmax measurements may also have relevance to the slower warming rates associated with progressive climate warming (Mora and Maya, 2006; Sunday et al., 2012). The controlled nature of CTmax measurements offers an avenue for investigating the influence of PCO2 on the maximum thermal limits of animals, yet we are aware of only two studies on only two species of fish (one published during the peer review of the present paper) that have taken a similar approach. Pimentel et al. (2014) exposed larvae of the subtropical Senegalese sole (Solea senegalensis) for 30 days post-hatching to an extreme ocean acidification scenario of pH 7.51 (estimated PCO2 ∼1600 µatm; N=30 per group). They reported that CTmax decreased (at a starting temperature of 18°C) or remained the same (at a starting temperature of 22°C) compared with a control group maintained at pH 8.02 (PCO2 ∼400 µatm). Ern et al. (2017) measured CTmax in adults of the tropical black-axil chromis (Chromis atripectoralis) following exposure to pH 8.23 (estimated PCO2 ∼340 µatm) or pH 7.81 (PCO2 ∼1086 µatm; N=5 per group) and did not detect any differences between treatment groups. The impacts of near-future levels of ocean acidification (end-of-century PCO2 ∼1000 µatm in a business-as-usual scenario; IPCC, 2013) on the thermal limits of fishes, including those in highly sensitive tropical systems, remain poorly understood.

Here, we studied three species of site-attached coral reef damselfishes (Acanthochromis polyacanthus, Dascyllus aruanus and Dischistodus perspicillatus) in a near-equatorial location to test the prediction that end-of-century levels of PCO2 reduce the maximum thermal limits (CTmax) of aquatic ectotherms. These species were chosen because they are co-occurring, abundant on the Great Barrier Reef, spread out across the damselfish phylogeny (Frédérich et al., 2013), and they were available in a wide size range (early-stage juveniles to adults) at the time of the study. Indeed, we were able to additionally examine whether CTmax and the impacts of PCO2 differ within species as a function of body size (i.e. ontogeny). While levels of PCO2 on coral reefs can vary across diurnal and seasonal cycles (Shaw et al., 2012), daily means are often around 450–550 µatm and an increase to ∼1000 µatm is predicted for the end of the century in a business-as-usual emissions scenario (IPCC, 2013). Using direct comparisons of species that co-exist in abundance across the tropical Western Pacific, we aimed to highlight interspecific and intraspecific patterns that may underlie susceptibility to the single and combined effects of acute warming and elevated PCO2.

MATERIALS AND METHODS

Animal collection and holding conditions

Animal collection and experiments were conducted under permits from the Great Barrier Reef Marine Park Authority (G13/35909.1) and with approval from the Animal Ethics Committee of James Cook University in association with the Australian Institute of Marine Science (A2314). Spiny chromis [A. polyacanthus (Bleeker 1855), N=63], humbug dascyllus [D. aruanus (Linnaeus 1758), N=85] and white damsel [D. perspicillatus (Cuvier 1830), N=47] of mixed sex were collected in the austral summer from around Lizard Island in the northern Great Barrier Reef, Australia (14°40′S, 145°28′E). Fish were collected on SCUBA either using a barrier net (10 mm stretch monofilament) or using hand nets and diluted clove oil spray. Body mass and standard length (SL) ranges were 0.10–24.08 g and 14.1–80.8 mm for A. polyacanthus, 0.02–9.45 g and 7.4–53.0 mm for D. aruanus, and 0.14–5.65 g and 14.9–53.6 mm for D. perspicillatus. Fish were moved from underwater holding cages into aerated buckets and transported within 90 min of capture to the aquarium facilities at the Lizard Island Research Station, where they were placed in tanks with flow-through seawater at ambient temperature (∼29°C; range 28.0–30.7°C). Species were kept separate and fish were divided into approximately even numbers between 22 holding tanks (10–30 l each, depending on fish size; 1–3 l min−1 flow-through). At 24–48 h post-capture, the PCO2 of half of the holding tanks (N=11) was gradually increased to 1089±326 µatm (mean±s.d., encompassing daily fluctuations) over 24 h using a CO2 dosing system (pH stat Computers, Aqua Medic, Bissendorf, Germany) connected to solenoid valves regulating administration of 100% CO2 gas into four aerated flow-through (∼10 l min−1) header tanks (30 l each). The remaining holding tanks (N=11) received water from four additional header tanks kept at ambient PCO2 levels (520±94 µatm, mean±s.d.). Water samples (N=9) collected from different reefs around Lizard Island confirmed that ambient levels of PCO2 ranged from 540 to 660 µatm in the morning of commencement of the CTmax experiments, values which are typical for coral reef environments (Shamberger et al., 2011). The PCO2 levels of the holding tanks were checked at least daily using a handheld CO2 meter (GM70, Vaisala, Helsinki, Finland) connected to a submerged gas-permeable PFTE probe (Qubit systems, Kingston, ON, Canada), as described previously (Jutfelt and Hedgärde, 2013). The Vaisala CO2 meter was factory calibrated in Vantaa, Finland, prior to experiments. The experimental design and CO2-dosing system followed best practices for ocean acidification research (Riebesell et al., 2010; Moran, 2014; Cornwall and Hurd, 2015). Fish were given 9–18 days to acclimate to the treatment conditions before trials commenced, which is sufficient time for individuals to establish new acid–base equilibria (Toews et al., 1983; Cameron and Iwama, 1987) and more than double the modal acclimation time used in previous PCO2 studies on coral reef fishes (Heuer and Grosell, 2014). Fish were fed to satiation with commercial pellet food once per day, but food was withheld for 15–16 h prior to experiments. Tanks were cleaned 2–3 times per week. Negligible mortality was observed during the acclimation period.

CTmax

The same CTmax protocol was repeated on two consecutive days to achieve desired sample sizes (D. aruanus: control N=41, high PCO2 N=44; A. polyacanthus: control N=31, high PCO2 N=32; D. perspicillatus: control N=24, high PCO2 N=23). The CTmax of each individual was measured in two rectangular tanks (L×W×D, 60×38×39 cm; water depth 27 cm) supplied with seawater from individual sumps (60×38×39 cm; water depth 28 cm). Water was continuously pumped from the sumps into the respective CTmax tanks (3.3 l min−1; Eheim submersible pumps), draining through a standpipe back into the sumps below. One sump contained control water (ambient PCO2) while the other contained water with elevated PCO2 (same as holding tanks) regulated by a dosing computer and solenoid valve (Aqua Medic, as above). Each sump received additional flow-through seawater throughout the trial to maintain water quality. PCO2 was measured >4 times throughout each trial to ensure the differential between the treatment groups was maintained. Each sump contained an identical 1 kW heater with a digital controller that maintained the set-point temperature within ±0.1°C, and these were supplemented with smaller aquarium heaters (300 W, Aqua One) at the higher temperatures during the trials (i.e. >36°C). Each sump was aerated to ensure oxygen levels remained at >90% air saturation at all times.

Fish of a broad size range that had been acclimated to either control or elevated PCO2 treatments were placed into the corresponding CTmax tanks in the evening and left to settle overnight (6–31 fish per species per treatment per day). Two pieces of opaque PVC piping (diameter 10 cm) were placed vertically in each CTmax tank during the whole trial to provide structure for the fish. Water temperature was kept at ambient conditions for the Lizard Island region (29–30°C) during the overnight settling period (∼12 h), and PCO2 was 530–620 µatm (control) or 970–1010 µatm (elevated PCO2) at the commencement of trials the following morning. At approximately 07:00 h, the 1 kW heater in each sump was set to 30.5°C and the temperature was subsequently increased 0.5°C every 30 min until all fish had reached their CTmax, which was defined as the temperature at which loss of equilibrium occurred uninterrupted for 10 s (Lutterschmidt and Hutchison, 1997). The temperature of each CTmax tank was logged continuously [0.125°C resolution using calibrated (with a mercury thermometer) iButton data loggers; Maxim Integrated, San Jose, CA, USA] and monitored manually with a handheld meter (0.1°C resolution; HQ40D, Hach Lange, Loveland, CO, USA). When an individual reached CTmax, the time, temperature (from the handheld meter), treatment (control or elevated PCO2), mass (to 0.01 g) and standard length (with callipers to the nearest 0.1 mm) were noted before the fish was placed in a recovery tank containing seawater at ambient temperature (∼29°C). Sixth-order polynomial regressions were plotted through the iButton temperature data as a function of time at the conclusion of each trial (R2 range 0.989–0.994) and used to calculate precise CTmax temperatures (i.e. interpolating between logged values).

Statistical analyses

The effect of PCO2 exposure on CTmax was tested in the three species using a general linear model (LM). We controlled for mass and trial number by specifying these terms first in the model and computing Type I sequential sum of squares. Mass was centred by subtracting the grand mean from all values to facilitate the interpretation of parameter estimates. We allowed a four-way interaction and all three-way interactions between predictors and performed model simplifications using two likelihood ratio (LR) tests. The four- and three-way interactions were subsequently removed (LR test, P>0.05 in both cases), and the final model included trial, mass, species and treatment as fixed factors, and all two-way interactions (in this order). We verified model assumptions using diagnostic plots (i.e. plot {base}, residualPlots {car} and qqPlot {car} functions in R v3.2.0; http://www.R-project.org/). Excluding five outliers considerably improved these diagnostics (Figs S1 and S2), hence results herein are from models where outliers have been excluded. Analysing the entire dataset (i.e. including the five outliers) produced qualitatively similar results (i.e. significance and direction of effects unchanged), as did a mixed-effects model with trial specified as a random factor. We conducted post hoc comparisons using the functions ‘lsmeans’ and ‘contrast’ in the R package ‘lsmeans’. The R script and data used for the analyses are available on figshare (https://doi.org/10.6084/m9.figshare.3408310.v1), allowing readers to assess and visualize the outputs of models with and without outliers.

RESULTS

Overall, our statistical model explained 96.5% of the variance in CTmax (adjusted R2). Species differed in their maximum thermal limits (F2,175=2412.10, P<0.001; Fig. S3A): D. perspicillatus had the highest CTmax (mean±s.e.=37.88±0.03°C), followed by D. aruanus (37.68±0.02°C) and A. polyacanthus (36.58±0.02°C). CTmax values were slightly higher for all species (0.28±0.02°C) in the first versus the second trial but the magnitude of this difference varied among species (species×trial: F1,175=24.92, P<0.001; Fig. S3B).

The effect of PCO2 exposure on CTmax differed among species (species×treatment: F2,175=7.87, P<0.001; Fig. 2). PCO2 exposure did not influence CTmax in D. aruanus (estimate±95% CI=0.043±0.028, P=0.123), whereas it increased CTmax by 0.21°C in A. polyacanthus (estimate±95% CI=0.208±0.032, P<0.001) and by 0.16°C in D. perspicillatus (estimate±95% CI=0.161±0.037, P<0.001).

Fig. 2.

CTmax of three damselfish species maintained in present-day (control; ∼500 µatm) or end-of-century (CO2; ∼1000 µatm) PCO2 conditions. (A) Box-and-whisker plots (median and interquartile range) showing the distribution of the raw data, where five outliers (excluded from the model) are indicated with dotted green circles. (B) Model predictions for Acanthochromispolyacanthus (dotted line; N=63), Dascyllusaruanus (dashed line; N=85) and Dischistodusperspicillatus (solid line; N=47). Error bars are 95% confidence intervals.

Fig. 2.

CTmax of three damselfish species maintained in present-day (control; ∼500 µatm) or end-of-century (CO2; ∼1000 µatm) PCO2 conditions. (A) Box-and-whisker plots (median and interquartile range) showing the distribution of the raw data, where five outliers (excluded from the model) are indicated with dotted green circles. (B) Model predictions for Acanthochromispolyacanthus (dotted line; N=63), Dascyllusaruanus (dashed line; N=85) and Dischistodusperspicillatus (solid line; N=47). Error bars are 95% confidence intervals.

The relationship between body mass and CTmax also differed among species (species×mass: F2,175=29.47, P<0.001; Fig. 3), irrespective of PCO2 treatment (3-way interaction not significant, LR test P>0.05): smaller A. polyacanthus and D. perspicillatus tolerated higher temperatures (CTmax∼mass slope±95% CI=−0.074±0.026 and −0.048±0.031, respectively), whereas the opposite trend existed for D. aruanus (CTmax∼mass slope±95% CI=0.021±0.012). Nevertheless, a single, very small D. aruanus (0.02 g) was excluded from the analyses to satisfy model assumptions but had an exceptionally high CTmax (38.38°C; small outlier in Fig. 3A).

Fig. 3.

Effects of body mass on the CTmax of three damselfish species maintained in present-day (control; ∼500 µatm) or end-of-century (CO2; ∼1000 µatm) PCO2 conditions. Sample sizes are given in Fig. 2. (A) Plots of the raw data with five outliers (excluded from the model) indicated with dotted green circles. (B) Model predictions with 95% confidence intervals indicated in grey; mass was centred (mean subtracted from each value) so that zero represents the mean mass across species.

Fig. 3.

Effects of body mass on the CTmax of three damselfish species maintained in present-day (control; ∼500 µatm) or end-of-century (CO2; ∼1000 µatm) PCO2 conditions. Sample sizes are given in Fig. 2. (A) Plots of the raw data with five outliers (excluded from the model) indicated with dotted green circles. (B) Model predictions with 95% confidence intervals indicated in grey; mass was centred (mean subtracted from each value) so that zero represents the mean mass across species.

DISCUSSION

Our study shows that end-of-century levels of PCO2 do not compromise the maximum acute thermal limits of coral reef damselfishes, providing empirical evidence against theoretical models that predict detrimental effects of ocean acidification on the maximum thermal limits of aquatic ectotherms (Pörtner, 2008; Pörtner and Farrell, 2008). We also show clear interspecific differences in CTmax among common damselfish species that co-exist in identical reef environments across the Western Pacific, and reveal differences in maximum thermal limits within species as a function of body mass (i.e. ontogeny). It is notable that our CTmax protocol utilized fish in communal tanks to avoid any anomalies that may arise when keeping social animals in isolation, and therefore our measures of CTmax are likely to be more representative of what may occur on natural reefs if the heating rate was to match the rate used here. While the mechanism underlying thermal tolerance is often assumed to be a mismatch between oxygen supply and demand (Pörtner and Knust, 2007; Pörtner and Farrell, 2008), existing evidence seems to refute this idea and highlights a need for exploration of alternative mechanisms, including the thermal dependence of protein denaturation, neural function and mitochondrial membrane integrity (Clark et al., 2013a,b; Iftikar and Hickey, 2013; Wang et al., 2014; Brijs et al., 2015; Ern et al., 2016). Even so, the new knowledge derived from our study has broad implications for understanding interspecific and intraspecific variability in responses to ocean acidification and acute thermal challenges.

Interspecific differences in environmental resilience

The three damselfish species we examined differed in their maximum thermal limits, despite residing in the same habitat. Dischistodus perspicillatus had the highest CTmax, followed by D. aruanus and A. polyacanthus (Fig. 2). The effect of PCO2 exposure on CTmax also differed among species, with high CO2-exposed D. perspicillatus and A. polyacanthus having a higher CTmax than control fish, while there was no effect of elevated PCO2 on CTmax of D. aruanus (Fig. 2). Although statistically significant, the increases in CTmax with PCO2 were small (0.16°C in D. perspicillatus and 0.21°C in A. polyacanthus). The biological significance of these differences may thus be negligible, particularly in light of a difference in CTmax of similar magnitude within species across trials (0.28°C).

To our knowledge, only two previous single-species studies on fish have examined CTmax in the context of ocean acidification. Pimentel et al. (2014) reported that a shift in pH from 8.02 (estimated PCO2 ∼400 µatm) to pH 7.51 (∼1600 µatm) caused a decline in the CTmax of S. senegalensis when individuals were initially held at 18°C, but did not affect CTmax when individuals were initially held at 22°C (heating rate of 0.033°C min−1 in both cases). While it is possible that the ∼1.5°C decrease in CTmax reported for S. senegalensis held at 18°C could be due to the young age of the individuals (up to 30 days post-hatching), our study does not lend support to the idea that elevated PCO2 (i.e. lower pH) disproportionately affects early life stages (Figs 2 and 3). Our conclusions are corroborated by a very recent study by Ern et al. (2017), in which one species of coral reef fish (C. atripectoralis) exhibited a CTmax of ∼38.1°C regardless of whether it was exposed to pH 8.23 (estimated PCO2 ∼340 µatm) or pH 7.81 (PCO2 ∼1086 µatm). Our results and those of Ern et al. (2017), combined with a lack of influence of pH on CTmax at the higher test temperature used by Pimentel et al. (2014), suggest that the concomitant increases in temperature and PCO2 occurring in the world's oceans may not compromise CTmax in most fishes.

One of our study species (D. aruanus) was previously reported to have a CTmax of 40.0°C when individuals were acclimated to similar temperatures to those used here (∼29°C) (Eme and Bennett, 2009). The higher CTmax in Eme and Bennett (2009) compared with that of D. aruanus in our study (mean CTmax 37.7°C) is probably a consequence of faster heating rates (0.31°C min−1 versus 0.017°C min−1 in our study), a smaller average size of the fish (16 mm SL versus 34 mm SL in our study), and differences in the duration constituting loss of equilibrium (‘at least 1 min’ versus 10 s in our study). Comprehensive analyses of the positive relationship between heating rate and CTmax have been presented previously (Mora and Maya, 2006; Peck et al., 2009), and this is likely to be an underlying factor in the difference in CTmax of A. polyacanthus reported here (∼36.7°C when heated at 0.017°C min−1) versus that in Ern et al. (2017) (∼38.4°C when heated at 0.033°C min−1). Indeed, CTmax for seven reef fish species investigated by Ospína and Mora (2004) ranged from 35.0 to 37.5°C when heating rates were 0.017°C min−1 (starting temperature 26.5°C), while CTmax of 24 reef fish species investigated by Menasveta (1981) ranged from 36.7 to 42.0°C when heating rates were 0.25°C min−1 (starting temperature 28.0°C). Notably, some reef-associated species such as the white mullet (Mugil curema) display exceptional thermal limits (CTmax 40.8°C; starting temperature 26.5°C) even when heating rates are relatively low (0.017°C min−1) (Mora and Ospína, 2001).

It should be emphasized that CTmax represents a thermal limit for acute, not chronic exposures. For example, Eme and Bennett (2009) found that D. aruanus did not survive an acclimation period of 14 days when temperatures exceeded 34–35°C, despite having a CTmax of 40°C. Even so, meta-analyses on CTmax data have shown that the differences identified among species have relevance for understanding thermal distribution patterns and the potential impacts of climate warming on fish populations (Sunday et al., 2012, 2014). Thus, the lower CTmax of A. polyacanthus compared with that of the other two species studied here is likely to have functional significance in the natural environment. Specifically, an increase in the frequency of coastal heatwaves (e.g. Wernberg et al., 2013) is likely to impact A. polyacanthus to a greater extent than D. aruanus or D. perspicillatus. Such interspecific differences should be considered not only in the context of mortality but also in relation to sub-lethal impacts that are known to arise at temperatures much lower than CTmax. For example, anomalously high temperatures can cause cessation of feeding (e.g. Ortubay et al., 1997) and delayed reproductive maturation (e.g. Jeffries et al., 2012) in fishes.

A notable difference in the life history of A. polyacanthus compared with that of most other damselfishes, including D. aruanus and D. perspicillatus, is the lack of a pelagic larval stage (Miller-Sims et al., 2008). Whether the lack of open ocean exposure of A. polyacanthus larvae underpins the lower CTmax in this species is unknown, although this idea is hard to reconcile given that coral reefs are generally warmer than the open ocean because of their shallow depth and proximity to land. Perhaps a more likely explanation for the lower thermal limits observed in A. polyacanthus is local adaptation to the temperature profiles experienced around Lizard Island. Gene flow among populations of A. polyacanthus is low compared with that of other damselfish species as a result of their limited dispersal (Doherty et al., 1994; Bay et al., 2006); juveniles only disperse away from their parents after approximately 3 months, and some related individuals remain together as adults (Miller-Sims et al., 2008). As a result, most A. polyacanthus are likely to experience similar thermal conditions to their parents, possibly selecting for narrower thermal limits compared with species such as D. perspicillatus and D. aruanus, which spend 2–3 weeks as pelagic larvae (Wellington and Victor, 1989). As pelagic larvae can disperse hundreds of kilometres across a large latitudinal range (e.g. Victor, 1987), the ability to withstand a broad temperature range should be favoured in species like D. perspicillatus and D. aruanus. Further study of the life history characteristics and physiological processes that underlie differences in CTmax across damselfish species may prove fruitful for deciphering the underlying mechanisms of the findings presented here.

Intraspecific/ontogenetic differences in environmental resilience

Our finding that small, early-stage juvenile damselfishes generally have a higher CTmax compared with that of larger conspecifics conflicts with the established belief that larval and early-stage juvenile fishes are especially sensitive to environmental stressors as a consequence of underdeveloped physiological/regulatory systems (Hutchison, 1976; Pörtner and Farrell, 2008). While the enhanced thermal limits of small individuals were statistically apparent in A. polyacanthus and D. perspicillatus, where several early-stage juveniles were represented in the dataset, the observation of exceptionally high CTmax in the single, early-stage juvenile D. aruanus is suggestive of a similar phenomenon in this species (small outlier in Fig. 3A).

Previous attempts to examine ontogenetic or size differences in thermal limits of reef fishes have failed to identify obvious trends. Ospína and Mora (2004) reported that CTmax was poorly related to body size in seven reef fish species, but they noted significant negative slopes in regressions of CTmax versus body size for rainbow wrasse (Thalassoma lucasanum; SL range 30–125 mm) and chameleon wrasse (Halichoeres dispilus; SL range 20–65 mm). Based on our findings, it is possible that the absence of early-stage juveniles from previous datasets has masked significant ontogenetic differences in thermal limits. That said, temperate delta smelt (Hypomesus transpacificus) generally exhibit a decline in CTmax with successive ontogenetic stages (Komoroske et al., 2014). Interestingly, we found some evidence that mature A. polyacanthus may have higher thermal limits than late-stage juveniles (but not as high as early-stage juveniles; Fig. 3A). Future experiments encompassing the entire mass range from larvae to full-sized adults are required to determine whether damselfishes may exhibit a U-shaped CTmax curve as a function of body mass, which would be the opposite to theoretical predictions (Fig. 1).

The underlying mechanisms associated with superior thermal limits in early-stage juvenile reef fishes remain unknown, but existing evidence suggests that CTmax is unlikely to be governed by a mismatch between oxygen supply and demand (Wang et al., 2014; Brijs et al., 2015; Ern et al., 2016). Our a priori assumption based on available data and theoretical models of climate change resilience was that late-stage juveniles and sub-adults would possess the highest thermal limits because they have fully developed physiological systems (unlike early-stage juveniles) but are not yet challenged by investment in reproduction (unlike mature adults) (Fig. 1B; Stewart and Allen, 2014). We also anticipated that if PCO2 influenced CTmax, the effects would be most pronounced in early-stage juveniles as a result of less-developed acid–base regulation (Brown and Sadler, 1989; Pörtner and Farrell, 2008). While intuitive, the assumption that larvae and early-stage juveniles are more sensitive to elevated PCO2 is primarily based on comparisons across unrelated studies/species (e.g. Pörtner et al., 2005, and references within) rather than single studies that compare different life stages using consistent approaches [but see Ishimatsu et al. (2004) for ontogenetic comparisons at very high PCO2 levels]. In this context, the present study is, to our knowledge, the first to quantify the interactive effects of elevated PCO2 and high temperature on fishes across a broad range of sizes and life stages.

Implications and conclusions

Our study demonstrates the existence of interspecific and ontogenetic differences in maximum thermal limits of coral reef damselfishes. Importantly, we show that end-of-century levels of PCO2 do not compromise acute maximum thermal limits. The heightened thermal limit and equivalent PCO2 tolerance of smaller individuals provide evidence of an unexpected capacity to maintain physiological homeostasis from a young age, at least over temporal scales of hours (high temperature) to weeks (high PCO2).

It is apparent that existing theoretical frameworks regarding the temperature and PCO2 sensitivities of fishes must be assessed with empirical evidence from robust experimental studies. While ongoing ocean acidification may influence coral reef fishes indirectly through detrimental effects on the calcification capacity of habitat-forming corals, we found no direct effects of end-of-century levels of PCO2 and acidification on upper thermal tolerance limits. Given the rapid rate of warming compared with the rate of increase in PCO2, we propose that thermal stress via heatwaves and chronic warming will be the primary concern to the conservation of most coral reef organisms. Nevertheless, major reductions in CO2 emissions are necessary to curtail the rate of warming (Hughes et al., 2017).

Acknowledgements

We thank F. Jutfelt for contributing experimental equipment, R. Slobodeanu for statistical advice, E. Walsh for fish illustrations, and the staff of the Lizard Island Research Station (particularly A. Hoggett and L. Vail) for support.

Footnotes

Author contributions

Conceptualization: T.D.C., D.G.R., S.A.B., B.S., J.S.; Methodology: T.D.C., D.G.R., S.A.B., B.S., J.S.; Software: T.D.C., D.G.R.; Validation: T.D.C.; Formal analysis: T.D.C., D.G.R., J.S.; Investigation: T.D.C., D.G.R., S.A.B., B.S., J.S.; Resources: T.D.C., J.S.; Data curation: T.D.C., D.G.R., S.A.B., B.S.; Writing - original draft: T.D.C.; Writing - review & editing: T.D.C., D.G.R., S.A.B., B.S., J.S.; Visualization: T.D.C., D.G.R.; Project administration: T.D.C.; Funding acquisition: T.D.C., J.S.

Funding

This work was funded by Svenska Forskningsrådet Formas (J.S.: 2013-947), Magnus Bergvalls Stiftelse (J.S.: 2014-00620), IRIS stipendiet (J.S.: 2015-0264), Stiftelsen Lars Hiertas Minne (J.S.: FO2014-0659), Wallenbergstiftelsen (J.S.), Inez Johanssons stiftelse (J.S.), Sederholms utrikes stiftelse (J.S.), and the Fonds de Recherche du Québec - Nature et Technologies (S.A.B., D.G.R.).

Data availability

The R script and raw data used for the analyses are publicly archived following best practices (Roche et al., 2015): https://doi.org/10.6084/m9.figshare.3408310.v1.

References

Bay
,
L. K.
,
Crozier
,
R. H.
and
Caley
,
M. J.
(
2006
).
The relationship between population genetic structure and pelagic larval duration in coral reef fishes on the Great Barrier Reef
.
Mar. Biol.
149
,
1247
-
1256
.
Becker
,
C. D.
and
Genoway
,
R. G.
(
1979
).
Evaluation of the critical thermal maximum for determining thermal tolerance of freshwater fish
.
Environ. Biol. Fishes
4
,
245
-
256
.
Bennett
,
W. A.
and
Beitinger
,
T. L.
(
1997
).
Temperature tolerance of the sheepshead minnow, Cyprinodon variegatus
.
Copeia
1997
,
77
-
87
.
Brewer
,
G. D.
(
1976
).
Thermal tolerance and resistance of the northern anchovy Engraulis mordax
.
Fish. Bull.
74
,
433
-
445
.
Brijs
,
J.
,
Jutfelt
,
F.
,
Clark
,
T. D.
,
Gräns
,
A.
,
Ekström
,
A.
and
Sandblom
,
E.
(
2015
).
Experimental manipulations of tissue oxygen supply do not affect warming tolerance of European perch
.
J. Exp. Biol.
218
,
2448
-
2454
.
Brown
,
D. J. A.
and
Sadler
,
K.
(
1989
).
Fish survival in acid waters
. In
Acid Toxicity and Aquatic Animals
, Vol.
34
(ed.
R.
Morris
,
E. W.
Taylor
,
D. J. A.
Brown
and
J. A.
Brown
), pp.
31
-
44
.
Cambridge
:
Cambridge University Press
.
Cameron
,
J. N.
and
Iwama
,
G. K.
(
1987
).
Compensation of progressive hypercapnia in channel catfish and blue crabs
.
J. Exp. Biol.
133
,
183
-
197
.
Cheung
,
W. W. L.
,
Sarmiento
,
J. L.
,
Dunne
,
J.
,
Frölicher
,
T. L.
,
Lam
,
V. W. Y.
,
Deng Palomares
,
M. L.
,
Watson
,
R.
and
Pauly
,
D.
(
2013
).
Shrinking of fishes exacerbates impacts of global ocean changes on marine ecosystems
.
Nat. Clim. Change
3
,
254
-
258
.
Clark
,
T. D.
,
Sandblom
,
E.
and
Jutfelt
,
F.
(
2013a
).
Aerobic scope measurements of fishes in an era of climate change: respirometry, relevance and recommendations
.
J. Exp. Biol.
216
,
2771
-
2782
.
Clark
,
T. D.
,
Sandblom
,
E.
and
Jutfelt
,
F.
(
2013b
).
Response to Farrell and to Pörtner and Giomi
.
J. Exp. Biol.
216
,
4495
-
4497
.
Cornwall
,
C. E.
and
Hurd
,
C. L.
(
2015
).
Experimental design in ocean acidification research: problems and solutions
.
ICES J. Mar. Sci.
73
,
572
-
581
.
Daufresne
,
M.
,
Lengfellner
,
K.
and
Sommer
,
U.
(
2009
).
Global warming benefits the small in aquatic ecosystems
.
Proc. Natl Acad. Sci. USA
106
,
12788
-
12793
.
Doherty
,
P. J.
,
Mather
,
P.
and
Planes
,
S.
(
1994
).
Acanthochromis polyacanthus, a fish lacking larval dispersal, has genetically differentiated populations at local and regional scales on the Great Barrier Reef
.
Mar. Biol.
121
,
11
-
21
.
Eme
,
J.
and
Bennett
,
W. A.
(
2009
).
Critical thermal tolerance polygons of tropical marine fishes from Sulawesi, Indonesia
.
J. Therm. Biol.
34
,
220
-
225
.
Ern
,
R.
,
Norin
,
T.
,
Gamperl
,
A. K.
and
Esbaugh
,
A. J.
(
2016
).
Oxygen dependence of upper thermal limits in fishes
.
J. Exp. Biol.
219
,
3376
-
3383
.
Ern
,
R.
,
Johansen
,
J. L.
,
Rummer
,
J. L.
and
Esbaugh
,
A. J.
(
2017
).
Effects of hypoxia and ocean acidification on the upper thermal niche boundaries of coral reef fishes
.
Biol. Lett.
13
,
20170135
.
Forster
,
J.
,
Hirst
,
A. G.
and
Atkinson
,
D.
(
2012
).
Warming-induced reductions in body size are greater in aquatic than terrestrial species
.
Proc. Natl Acad. Sci. USA
109
,
19310
-
19314
.
Frédérich
,
B.
,
Sorenson
,
L.
,
Santini
,
F.
,
Slater
,
G. J.
and
Alfaro
,
M. E.
(
2013
).
Iterative ecological radiation and convergence during the evolutionary history of damselfishes (Pomacentridae)
.
Am. Nat.
181
,
94
-
113
.
Freitas
,
V.
,
Cardoso
,
J. F. M. F.
,
Lika
,
K.
,
Peck
,
M. A.
,
Campos
,
J.
,
Kooijman
,
S. A. L. M.
and
van der Veer
,
H. W.
(
2010
).
Temperature tolerance and energetics: a dynamic energy budget-based comparison of North Atlantic marine species
.
Philos. Trans. R. Soc. B Biol. Sci.
365
,
3553
-
3565
.
Heuer
,
R. M.
and
Grosell
,
M.
(
2014
).
Physiological impacts of elevated carbon dioxide and ocean acidification on fish
.
Am. J. Physiol. Regul. Integr. Comp. Physiol.
307
,
R1061
-
R1084
.
Hughes
,
T. P.
,
Baird
,
A. H.
,
Bellwood
,
D. R.
,
Card
,
M.
,
Connolly
,
S. R.
,
Folke
,
C.
,
Grosberg
,
R.
,
Hoegh-Guldberg
,
O.
,
Jackson
,
J. B. C.
,
Kleypas
,
J.
, et al. 
(
2003
).
Climate change, human impacts, and the resilience of coral reefs
.
Science
301
,
929
-
933
.
Hughes
,
T. P.
,
Kerry
,
J. T.
,
Álvarez-Noriega
,
M.
,
Álvarez-Romero
,
J. G.
,
Anderson
,
K. D.
,
Baird
,
A. H.
,
Babcock
,
R. C.
,
Beger
,
M.
,
Bellwood
,
D. R.
,
Berkelmans
,
R.
, et al. 
(
2017
).
Global warming and recurrent mass bleaching of corals
.
Nature
543
,
373
-
377
.
Hutchison
,
V. H.
(
1976
).
Factors influencing thermal tolerances of individual organisms
. In
Symposium Series of the National Technical Information Service
(ed.
G. W.
Esch
,
R.
McFarlane
), pp.
10
-
26
.
Springfield, VA
:
National Technical Information Service
.
Iftikar
,
F. I.
and
Hickey
,
A. J. R.
(
2013
).
Do mitochondria limit hot fish hearts? Understanding the role of mitochondrial function with heat stress in Notolabrus celidotus
.
PLoS ONE
8
,
e64120
.
IPCC
. (
2013
).
Climate change 2013: the physical science basis
. In
Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
(ed.
T. F.
Stocker
,
D.
Qin
,
G.-K.
Plattner
,
M.
Tignor
,
S. K.
Allen
,
J.
Boschung
,
A.
Nauels
,
Y.
Xia
,
V.
Bex
and
P. M.
Midgley
), p.
1535
.
Cambridge, NY
,
USA
:
Cambridge University Press
.
Ishimatsu
,
A.
,
Kikkawa
,
T.
,
Hayashi
,
M.
,
Lee
,
K.-S.
and
Kita
,
J.
(
2004
).
Effects of CO2 on marine fish: larvae and adults
.
J. Oceanogr.
60
,
731
-
741
.
Jeffries
,
K. M.
,
Hinch
,
S. G.
,
Martins
,
E. G.
,
Clark
,
T. D.
,
Lotto
,
A. G.
,
Patterson
,
D. A.
,
Cooke
,
S. J.
,
Farrell
,
A. P.
and
Miller
,
K. M.
(
2012
).
Sex and proximity to reproductive maturity influence the survival, final maturation, and blood physiology of Pacific salmon when exposed to high temperature during a simulated migration
.
Physiol. Biochem. Zool.
85
,
62
-
73
.
Jutfelt
,
F.
and
Hedgärde
,
M.
(
2013
).
Atlantic cod actively avoid CO2 and predator odour, even after long-term CO2 exposure
.
Front. Zool.
10
,
81
.
Jutfelt
,
F.
,
Gräns
,
A.
,
Jönsson
,
E.
,
Wiklander
,
K.
,
Seth
,
H.
,
Olsson
,
C.
,
Dupont
,
S.
,
Ortega-Martinez
,
O.
,
Sundell
,
K.
,
Axelsson
,
M.
, et al. 
(
2014
).
Response to ‘How and how not to investigate the oxygen and capacity limitation of thermal tolerance (OCLTT) and aerobic scope – remarks on the article by Gräns et al.
’.
J. Exp. Biol.
217
,
4433
-
4435
.
Komoroske
,
L. M.
,
Connon
,
R. E.
,
Lindberg
,
J.
,
Cheng
,
B. S.
,
Castillo
,
G.
,
Hasenbein
,
M.
and
Fangue
,
N. A.
(
2014
).
Ontogeny influences sensitivity to climate change stressors in an endangered fish
.
Conserv. Physiol.
2
,
cou008
.
Lefevre
,
S.
(
2016
).
Are global warming and ocean acidification conspiring against marine ectotherms? A meta-analysis of the respiratory effects of elevated temperature, high CO2 and their interaction
.
Conserv. Physiol.
4
,
cow009
.
Lough
,
J. M.
(
2012
).
Small change, big difference: sea surface temperature distributions for tropical coral reef ecosystems, 1950-2011
.
J. Geophys. Res.
117
,
C09018
.
Lutterschmidt
,
W. I.
and
Hutchison
,
V. H.
(
1997
).
The critical thermal maximum: history and critique
.
Can. J. Zool.
75
,
1561
-
1574
.
Menasveta
,
P.
(
1981
).
Lethal temperature of marine fishes of the Gulf of Thailand
.
J. Fish Biol.
18
,
603
-
607
.
Miller-Sims
,
V. C.
,
Gerlach
,
G.
,
Kingsford
,
M. J.
and
Atema
,
J.
(
2008
).
Dispersal in the spiny damselfish, Acanthochromis polyacanthus, a coral reef fish species without a larval pelagic stage
.
Mol. Ecol.
17
,
5036
-
5048
.
Mora
,
C.
and
Maya
,
M. F.
(
2006
).
Effect of the rate of temperature increase of the dynamic method on the heat tolerance of fishes
.
J. Therm. Biol.
31
,
337
-
341
.
Mora
,
C.
and
Ospína
,
A. F.
(
2001
).
Tolerance to high temperatures and potential impact of sea warming on reef fishes of Gorgona Island (tropical eastern Pacific)
.
Mar. Biol.
139
,
765
-
769
.
Moran
,
D.
(
2014
).
The importance of accurate CO2 dosing and measurement in ocean acidification studies
.
J. Exp. Biol.
217
,
1827
-
1828
.
Ortubay
,
S. G.
,
Gómez
,
S. E.
and
Cussac
,
V. E.
(
1997
).
Lethal temperatures of a Neotropical fish relic in Patagonia, the scale-less characinid Gymnocharacinus bergi
.
Environ. Biol. Fishes
49
,
341
-
350
.
Ospína
,
A. F.
and
Mora
,
C.
(
2004
).
Effect of body size on reef fish tolerance to extreme low and high temperatures
.
Environ. Biol. Fishes
70
,
339
-
343
.
Paladino
,
F. V.
,
Spotila
,
J. R.
,
Schubauer
,
J. P.
and
Kowalski
,
K. T.
(
1980
).
The critical thermal maximum - a technique used to elucidate physiological stress and adaptation in fishes
.
Rev. Can. Biol.
39
,
115
-
122
.
Payne
,
N. L.
and
Smith
,
J. A.
(
2017
).
An alternative explanation for global trends in thermal tolerance
.
Ecol. Lett.
20
,
70
-
77
.
Pearce
,
A. F.
and
Feng
,
M.
(
2013
).
The rise and fall of the “marine heat wave” off Western Australia during the summer of 2010/2011
.
J. Mar. Syst.
111-112
,
139
-
156
.
Peck
,
L. S.
,
Clark
,
M. S.
,
Morley
,
S. A.
,
Massey
,
A.
and
Rossetti
,
H.
(
2009
).
Animal temperature limits and ecological relevance: effects of size, activity and rates of change
.
Funct. Ecol.
23
,
248
-
256
.
Pimentel
,
M. S.
,
Faleiro
,
F.
,
Dionísio
,
G.
,
Repolho
,
T.
,
Pousão-Ferreira
,
P.
,
Machado
,
J.
and
Rosa
,
R.
(
2014
).
Defective skeletogenesis and oversized otoliths in fish early stages in a changing ocean
.
J. Exp. Biol.
217
,
2062
-
2070
.
Popova
,
E.
,
Yool
,
A.
,
Byfield
,
V.
,
Cochrane
,
K.
,
Coward
,
A. C.
,
Salim
,
S. S.
,
Gasalla
,
M. A.
,
Henson
,
S. A.
,
Hobday
,
A. J.
,
Pecl
,
G. T.
, et al. 
(
2016
).
From global to regional and back again: common climate stressors of marine ecosystems relevant for adaptation across five ocean warming hotspots
.
Glob. Change Biol.
22
,
2038
-
2053
.
Pörtner
,
H. O.
(
2008
).
Ecosystem effects of ocean acidification in times of ocean warming: a physiologist's view
.
Mar. Ecol. Prog. Ser.
373
,
203
-
217
.
Pörtner
,
H. O.
and
Farrell
,
A. P.
(
2008
).
Ecology: physiology and climate change
.
Science
322
,
690
-
692
.
Pörtner
,
H. O.
and
Knust
,
R.
(
2007
).
Climate change affects marine fishes through the oxygen limitation of thermal tolerance
.
Science
315
,
95
-
97
.
Pörtner
,
H. O.
,
Langenbuch
,
M.
and
Michaelidis
,
B.
(
2005
).
Synergistic effects of temperature extremes, hypoxia, and increases in CO2 on marine animals: From Earth history to global change
.
J. Geophys. Res.
110
,
C09S10
.
Pörtner
,
H.-O.
,
Karl
,
D. M.
,
Boyd
,
P. W.
,
Cheung
,
W. W. L.
,
Lluch-Cota
,
S. E.
,
Nojiri
,
Y.
,
Schmidt
,
D. N.
and
Zavialov
,
P. O.
(
2014
).
Ocean systems
. In
Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
(ed.
C. B.
Field
,
V. R.
Barros
,
D. J.
Dokken
,
K. J.
Mach
,
M. D.
Mastrandrea
,
T. E.
Bilir
,
M.
Chatterjee
,
K. L.
Ebi
,
Y. O.
Estrada
,
R. C.
Genova
, et al. 
), pp.
411
-
484
.
Cambridge
,
United Kingdom
and
New York
,
USA
:
Cambridge University Press
.
Riebesell
,
U.
,
Fabry
,
V. J.
,
Hansson
,
L.
and
Gattuso
,
J.-P.
(
2010
).
Guide to Best Practices for Ocean Acidification Research and Data Reporting
.
Luxembourg
:
Publications Office of the European Union
.
Rijnsdorp
,
A. D.
,
Peck
,
M. A.
,
Engelhard
,
G. H.
,
Möllmann
,
C.
and
Pinnegar
,
J. K.
(
2009
).
Resolving the effect of climate change on fish populations
.
ICES J. Mar. Sci.
66
,
1570
-
1583
.
Roche
,
D. G.
,
Kruuk
,
L. E. B.
,
Lanfear
,
R.
and
Binning
,
S. A.
(
2015
).
Public data archiving in ecology and evolution: how well are we doing?
PLoS Biol.
13
,
e1002295
.
Shamberger
,
K. E. F.
,
Feely
,
R. A.
,
Sabine
,
C. L.
,
Atkinson
,
M. J.
,
DeCarlo
,
E. H.
,
Mackenzie
,
F. T.
,
Drupp
,
P. S.
and
Butterfield
,
D. A.
(
2011
).
Calcification and organic production on a Hawaiian coral reef
.
Mar. Chem.
127
,
64
-
75
.
Shaw
,
E. C.
,
McNeil
,
B. I.
and
Tilbrook
,
B.
(
2012
).
Impacts of ocean acidification in naturally variable coral reef flat ecosystems
.
J. Geophys. Res.
117
,
C03038
.
Stewart
,
H. A.
and
Allen
,
P. J.
(
2014
).
Critical thermal maxima of two geographic strains of channel and hybrid catfish
.
N. Am. J. Aquac.
76
,
104
-
111
.
Sunday
,
J. M.
,
Bates
,
A. E.
and
Dulvy
,
N. K.
(
2012
).
Thermal tolerance and the global redistribution of animals
.
Nat. Clim. Change
2
,
686
-
690
.
Sunday
,
J. M.
,
Bates
,
A. E.
,
Kearney
,
M. R.
,
Colwell
,
R. K.
,
Dulvy
,
N. K.
,
Longino
,
J. T.
and
Huey
,
R. B.
(
2014
).
Thermal-safety margins and the necessity of thermoregulatory behavior across latitude and elevation
.
Proc. Natl Acad. Sci. USA
111
,
5610
-
5615
.
Tewksbury
,
J. J.
,
Huey
,
R. B.
and
Deutsch
,
C. A.
(
2008
).
Putting the heat on tropical animals
.
Science
320
,
1296
-
1297
.
Toews
,
D. P.
,
Holeton
,
G. F.
and
Heisler
,
N.
(
1983
).
Regulation of the acid-base status during environmental hypercapnia in the marine teleost fish Conger conger
.
J. Exp. Biol.
107
,
9
-
20
.
Vaughan
,
G. E.
and
Coble
,
D. W.
(
1975
).
Sublethal effects of three ectoparasites on fish
.
J. Fish Biol.
7
,
283
-
294
.
Victor
,
B. C.
(
1987
).
Growth, dispersal, and identification of planktonic labrid and pomacentrid reef-fish larvae in the eastern Pacific Ocean
.
Mar. Biol.
95
,
145
-
152
.
Wang
,
T.
,
Lefevre
,
S.
,
Iversen
,
N. K.
,
Findorf
,
I.
,
Buchanan
,
R.
and
McKenzie
,
D. J.
(
2014
).
Anaemia only causes a small reduction in the upper critical temperature of sea bass: is oxygen delivery the limiting factor for tolerance of acute warming in fishes?
J. Exp. Biol.
217
,
4275
-
4278
.
Wellington
,
G. M.
and
Victor
,
B. C.
(
1989
).
Planktonic larval duration of one hundred species of Pacific and Atlantic damselfishes (Pomacentridae)
.
Mar. Biol.
101
,
557
-
567
.
Wernberg
,
T.
,
Smale
,
D. A.
,
Tuya
,
F.
,
Thomsen
,
M. S.
,
Langlois
,
T. J.
,
de Bettignies
,
T.
,
Bennett
,
S.
and
Rousseaux
,
C. S.
(
2013
).
An extreme climatic event alters marine ecosystem structure in a global biodiversity hotspot
.
Nat. Clim. Change
3
,
78
-
82
.

Competing interests

The authors declare no competing or financial interests.

Supplementary information