The effects of environmental history and thermal stress on coral physiology and immunity

Rising ocean temperatures can induce the breakdown of the symbiosis between reef building corals and Symbiodinium in the phenomenon known as coral bleaching. Environmental history may, however, influence the response of corals to stress and affect bleaching outcomes. A suite of physiological and immunological traits was evaluated to test the effect of environmental history (low vs. high variable pCO2) on the response of the reef coral Montipora capitata to elevated temperature (24.5 °C vs. thermal ramping to 30.5 °C). Heating reduced maximum photochemical efficiency (Fv/Fm) and chlorophyll a but increased tissue melanin in corals relative to the ambient treatment, indicating a role of the melanin synthesis pathway in the early stages of thermal stress. However, interactions of environmental history and temperature treatment were not observed. Rather, parallel reaction norms were the primary response pattern documented across the two temperature treatments with respect to reef environmental history. Corals with a history of greater pCO2 variability had higher constitutive antioxidative and immune activity (i.e., catalase, superoxide dismutase, prophenoloxidase) and Fv/Fm, but lower melanin and chlorophyll a, relative to corals with a history of lower pCO2 variability. This suggests that reef environments with high magnitude pCO2 variability promote greater antioxidant and immune activity in resident corals. These results demonstrate coral physiology and immunity reflect environmental attributes that vary over short distances, and that these differences may buffer the magnitude of thermal stress effects on coral phenotypes.


Introduction
The mutualistic symbiosis between scleractinian corals and dinoflagellates of the genus Symbiodinium underpins the function of hermatypic corals and their capacity to engineer tropical reef ecosystems . Environmental disturbances destabilize this symbiosis and reduce the abundance of Symbiodinium cells and/or their photosynthetic pigments within coral tissues; a stress response called coral bleaching (Coles and Jokiel 1978). Elevated seawater temperatures have driven three global coral bleaching events to date (Hoegh-Guldberg et al. 2017), and ocean warming and the frequency of bleaching-level stress are predicted to increase as climate change intensifies Hughes et al. 2017). While corals have persisted through considerable environmental change in the geologic record (Pandolfi and Kiessling 2014), the magnitude and rate of change in the thermal and chemical properties of seawater during the Anthropocene is unprecedented (Zeebe 2012;IPCC 2014;Hubbard 2015).
The response of corals to thermal stress is influenced by physical conditions that precede (Brown et al. 2002a;Middlebrook et al. 2008;Carilli et al. 2012;Guest et al. 2012;Ainsworth et al. 2016) and/or co-occur with elevated temperatures (Coles and Jokiel 1978;Dunne and Brown 2001;Nakamura and van Woesik 2001;Jokiel and Brown 2004;Anthony et al. 2008;Wiedenmann et al. 2012). It is recognized that organisms are equipped with diverse biochemical mechanisms to acclimate and adapt to physiological stress (Hochachka and Somero 2002), and in corals, evidence supports the role of the coral animal (Kenkel et al. 2013a) and Symbiodinium (Levin et al. 2016) in confronting environmental challenges (Edmunds and Gates 2008;Hume et al. 2016;Palumbi et al. 2014). For instance, corals experiencing thermal (Lesser 2004;Fitt et al. 2009;Kenkel et al. 2011Kenkel et al. , 2013b and/or photo-stress (Brown et al. 2002b) can mitigate cellular damage by up-regulating stress proteins (i.e., fluorescent and heat-shock proteins) (Lesser 2004;Palmer et al. 2009;Louis et al. 2017). Coral immunity and pathogen defense mechanisms (e.g., melanin synthesis pathway) (Söderhäll and Cerenius 1998) are also dynamically regulated in response to bleaching stress ). Indeed, maintaining high baseline immunity and tissue-protective properties (e.g., antioxidative enzymes) may represent a conserved mechanism of coral physiological resilience to both disease and thermal stress (Weis 2008;Palmer et al. 2010;Louis et al. 2017).
Rising concentrations of carbon dioxide (pCO 2 ) and other greenhouse gases are driving global climate change by increasing air and ocean temperatures (IPCC 2014). In addition, the dissolution of atmospheric CO 2 in the upper ocean is disrupting seawater carbonate chemistry and causing ocean acidification (OA), which threatens the net calcification of coral reef ecosystems (Andersson and Gledhill 2013). Exposure to elevated pCO 2 also has the potential to influence coral immunity and the response of corals to warming temperatures Kaniewska et al. 2012). For instance, pCO 2 can exacerbate thermal stress effects and cause bleaching in some corals ; but see Wall et al. 2014;Noonan and Fabricius 2016). Additionally, corals in experimentally elevated pCO 2 conditions or from naturally high pCO 2 -seeps show an upregulation of genes involved in oxidative stress and innate immune pathways (Kaniewska et al. 2012;Kenkel et al. 2017).
Natural field settings where elevated pCO 2 conditions are persistent (Fabricius et al. 2011;Albright et al. 2015;Padilla-Gamiño et al. 2016;Kenkel et al. 2017), or dynamic in nature (Drupp et al. 2013), can provide insight into the consequences of high pCO 2 /low pH on marine taxa not apparent in short-term laboratory experiments (Calosi et al. 2013;Noonan and Fabricius 2016). Leveraging natural field settings with unique seawater properties can, therefore, clarify the influence of pCO 2 history on coral physiology and thermal stress responses (Noonan and Fabricius 2016). Within Kāne'ohe Bay (windward O'ahu, Hawai'i) a combination of factors (e.g., physical forcing, seawater residence time, watershed and oceanic biogeochemistry) (Lowe et al. 2009;Drupp et al. 2011Drupp et al. , 2013Shamberger et al. 2011) has created regions where corals are exposed to seawater pCO 2 projected to occur under end-of-the-century climate change scenarios (van Vuuren et al. 2011). As such, Kāne'ohe Bay provides an ideal natural setting to address the hypothesis that environmental history-specifically, regimes of contrasting pCO 2 variability (Drupp et al. , 2013alters the biology of reef corals and their response to stress. The dynamic interplay between multiple stressors, environmental history and physiological acclimatization shapes reef resilience in varying ways. The goal of this study was to test the interaction of environmental history in the context of pCO 2 variability and short-term thermal stress on the physiological, photochemical, and immunological responses of corals from two Kāne'ohe Bay reefs with contrasting pCO 2 conditions. Considering the potential for elevated pCO 2 to negatively influence coral performance and cause bleaching, we tested the hypothesis that corals from environments with a history of high variable pCO 2 would show greater sensitivity to thermal stress by exhibiting greater declines in photochemical efficiency, photopigment concentrations and Symbiodinium densities relative to corals from environments with a history of low variable pCO 2 . We also expected corals from high variable pCO 2 environments to display increased antioxidative activity as a mechanism to mitigate cellular damage (Weis 2008), as well as greater immune activity measured by elevated melanin synthesis pathway activity (prophenoloxidase and melanin).

Environmental monitoring
Seawater pCO 2 conditions proximate to each reef (LV-Lilipuna, HV-Reef 14) were sourced from quality-controlled, publically available NOAA PMEL-CRIMP CO 2 -Platform moored buoys (https ://www.pmel.noaa.gov/co2/story / Coral +Reef+Moori ngs) (Sabine et al. 2012;Sutton et al. 2016 (Long et al. 2012). While collections of environmental data (i.e., pCO 2 , temperature, PAR) are temporally distinct and reefs may not experience seawater with identical carbonate chemistry as measured at buoys, collectively, these data are useful in describing trends in environmental characteristics among the two reef locations and their relationship to coral performance.

Coral collection and laboratory treatments
Fifty M. capitata (Dana 1846) branch tips (ca. 4 cm in length) were collected from each reef on 5 February 2014 (State of Hawai'i Department of Land and Natural Resources, Special Activity Permit 2013-47); accordingly, holobiont biomass should be considered seasonally acclimated to Kāne'ohe Bay winter conditions (Fitt et al. 2000). Fragments were transported in seawater to HIMB and epoxied to plastic bases using Z-spar A788 splash zone compound in a flow-through water table. One day after collection, corals were transferred into two custom-built experimental flow-through aquaria (50 L; Aqualogic, Inc., North Haven, Connecticut) receiving sand-filtered seawater from Kāne'ohe Bay at a rate of ca. 0.2 L min −1 and maintained at ambient conditions of 36 salinity and ca. 24.5 °C.
After 1 week of acclimation to laboratory conditions, corals (N = 100) were randomly allocated to four flow-through aquaria (50 L) (two replicate tanks treatment −1 ) at a density of 25 fragments tank −1 .
Seawater temperatures in each tank were independently regulated using a combination of 100 W submersible heaters and a programmable solenoid controller that independently regulated the delivery of chilled water through an in-line mixing column (Multi Temp MT-1 Model #2TTB3024A1000AA, Aqualogic) receiving tank seawater. Temperature treatments represented ambient temperature conditions (24.5 °C: Ambient) for January-February 2014 (NOAA 2017) and a heated treatment gradually ramped to elevated temperatures (30.5 °C: Heated) (Fig. 3). Temperature ramping lasted 7 days and increased at a rate of ca. 0.75 °C day −1 . Corals were maintained at 30.5 °C for 2 days, which is near the upper thermal limit of Hawaiian reef corals (Coles et al. 1976;Coles and Jokiel 1978). The ramping regime was comparable to other studies (Middlebrook et al. 2010) and was implemented to avoid acute heat shock, ensuring observation of progressive heating effects on coral performance. Corals were exposed to treatments from 11 to 19 February 2014. ANOVA confirmed the establishment of two separate temperature treatments (F 1,239 = 231.300, P < 0.001); temperatures did not differ among replicate heated tanks (F 1,130 = 0.018, P = 0.893) but replicate ambient tanks differed by 0.24 °C (F 1,107 = 71.578, P < 0.001).

Physiological metrics
Pulse amplitude modulation (PAM) fluorometry was used to measure temperature effects on the photochemical performance of Symbiodinium spp. in hospite for all corals on the 8th day of treatment exposure using a Diving-PAM (Waltz, GmbH, Effeltrich, Germany). The Diving-PAM was operated at a gain of 7, saturation intensity of 8, an electronic signal damping of 2; under these conditions, the signal to noise ratio was optimized and the minimum fluorescence was stabilized at ca. 400-700 (arbitrary units). The minimum (F o ) and maximum (F m ) fluorescence yield and the maximum photochemical efficiency (F v /F m ) of darkadapted PSII reaction centers were measured during the day at 14:00 h (which coincided with the period of peak PAR exposure (Jones and Hoegh-Guldberg 2001) following a 20-min dark acclimation period. Measurements were obtained using the 5-mm diameter fiber-optic probe positioned 5 mm above the surface of the coral tissue following F o stabilization. Following 9 days of exposure, all corals were immediately snap-frozen in liquid nitrogen and stored at -80 °C. A subset of corals were used for physiological assays (Symbiodinium density, chlorophyll a extraction, and total protein; n = 10-13 treatment −1 ) and analyzed at HIMB. Another subset of corals were used in immunological assays (protein, melanin, prophenoloxidase, catalase, and superoxide dismutase; n = 11-12 treatment −1 ) and analyzed at the University of Texas at Arlington. Corals remained at -80 °C and were not thawed prior to analysis. For physiological analyses, coral tissue was removed from the skeleton using an airbrush and filtered seawater (0.7 μm). The resulting coral tissue slurry was briefly homogenized and aliquots taken for each response variable, following Wall et al. (2017). The coral skeleton was placed in 10% bleach solution and allowed to dry at 60 °C before measuring the surface area of the skeleton by the paraffin wax-dipping technique (Stimson and Kinzie 1991). Symbiodinium densities were quantified by repeated cell counts (n = 6 sample −1 ) using a haemocytometer, and cell densities were standardized to coral surface area (cells cm −2 ). Chlorophyll a was extracted by centrifuging the tissue slurry (13,000 rpm × 3 min) and isolating the alga pellet, followed by adding 100% acetone and extracting at − 20 °C in darkness for 36 h. The pigment extract was measured spectrophotometrically (λ = 630 and 663 nm) and chlorophyll a concentrations were determined using equations for dinoflagellates (Jeffrey and Humphrey 1975). Chlorophyll a was standardized to surface area (μg cm −2 ) and to the density of Symbiodinium cells (pg cell −1 ). Total protein (soluble and insoluble) in the tissue slurry was measured using the Pierce BCA (bicinchoninic acid) Protein Assay Kit (Pierce Biotechnology, Waltham, Massachusetts). Solubilization of protein was achieved by adding 1 M NaOH to the tissue slurry, heating at 90 °C for 1 h, and neutralizing to pH ca. 7.5 using 1 N HCl. The total protein in three technical replicates sample −1 was measured in a 96-well microtiter plate (λ = 562 nm) against a bovine serum albumin standard curve and standardized to coral surface area (mg protein cm −2 ).

Immunological assays
Coral immunology was assessed following previously established protocols for protein extractions and enzymatic assays (Mydlarz et al. , 2010Palmer et al. 2010Palmer et al. , 2011aMydlarz and Palmer 2011). Briefly, 3-4 mL of coral tissue slurry was obtained by airbrushing with coral extraction buffer (100 mM TRIS buffer + 0.05 mM dithiothreitol). The resulting slurry was homogenized for 1 min on ice using a hand-held tissue homogenizer (Powergen 125, Fisher Scientific, Waltham, Massachusetts). For melanin concentration estimates, 1 mL of the tissue slurry was freeze-dried for 24 h using a VirTis BTK freeze-dryer (SP Scientific, Warminster, Pennsylvania). The remaining slurry was centrifuged at 4 °C at 2500×g (Eppendorf 5810 R centrifuge, Hamburg, Germany) for 5 min to remove cellular debris, and enzymatic assays were performed on aliquots of the supernatant, representing a cell-free extract or soluble protein extract of the host coral. All assays were run in duplicates on separate 96-well microtiter plates using a Synergy HT multidetection microplate reader using Gen5 software (Biotek Instruments, Winooski, Vermont). Protein concentrations were estimated using the RED660 protein assay (G Biosciences, Saint Louis, Missouri) against a bovine serum albumin standard curve.

Antioxidant profile
Antioxidant enzymes catalase (CAT) and superoxide dismutase (SOD) were measured. CAT is monitored as a change in absorbance after 25 mM hydrogen peroxide is added to crude protein extract and 50 μL of 10 mM PBS (pH 6.0). CAT activity was estimated as the mM H 2 O 2 scavenged min −1 mg protein −1 . SOD activity was analyzed using a commercially available kit (SOD determination kit #19160; Sigma-Aldrich, St. Louis, Missouri) following manufacturer's instructions and expressed as SOD activity mg protein −1 . SOD activity was estimated by comparing the absorbance of samples at 450 nm to a positive and negative standard after incubating 10 μL of crude protein extract with manufacturer-provided reagents.

Melanin synthesis pathway
Prophenoloxidase (PPO) activity and melanin (MEL) concentration per sample were used to study the melanin synthesis pathways. PPO activity was determined by incubating 20 μL of protein extract and 50 μL of 10 mM phosphate buffered saline (PBS) (pH 7.0) at room temperature with 20 μL of trypsin (0.2 mg mL −1 concentration) for 30 min. 20 μL of 25 mM l-DOPA (Sigma-Aldrich) was then added as a substrate. PPO activity was estimated as change in absorbance min −1 mg protein −1 . MEL concentration was estimated using a weighed freeze-dried portion of initial tissue slurry. Melanin was allowed to extract for 48 h in 400 μL of 10 M NaOH after a brief period of beadbeating with 1-mm glass beads. 65 μL of extracted melanin was used to determine endpoint absorbance at 495 nm and resulting values were standardized to a standard curve of commercial melanin (Sigma-Aldrich) and calibrated to μg melanin mg tissue −1 .

Statistical analysis
Dependent variables were analyzed using a two-way linear model using lme with temperature treatment (Ambient vs. Heated) and reef environmental history (LV-Lilipuna vs. HV-Reef 14) as fixed effects. Environmental data (light and temperature) were analyzed using a linear mixed effect model in the package lme4 (Douglas et al. 2015) with site as a fixed factor and the repeated measure (sampling time) as a random factor. Analysis of variance tables (linear models) and analysis of deviance tables (linear mixed effect models) were calculated using Type-II sum of squares with Satterthwaite approximation of degrees of freedom using lmerTest (Kuznetsova et al. 2016). The assumptions of analysis of variance were confirmed by graphical inspection of residuals combined with Shapiro-Wilk's test and Levene's test and transformed where assumptions of ANOVA were not met. Transformations were selected using a Box-Cox power transformation using the package MASS (Box and Cox 1964;Venables and Ripley 2002). All analyses were performed in R, version 3.3.0 (R Development Core Team 2016). Experimental data and R code to reproduce figures and analyses are accessible on Zenodo (https ://zenod o.org/recor d/11750 34).

Discussion
Thermal stress and bleaching can suppress coral immunity (Couch et al. 2008), leaving corals vulnerable to opportunistic infections and disease (Miller et al. 2009). The initial ability of corals to avoid thermal stress and to resist pathogenesis (i.e., constitutive antioxidative and immune activity) is an important driver of coral fate during environmental perturbation (Mydlarz et al. 2010;Palmer and Traylor-Knowles 2012). Therefore, the effect of environmental history on coral antioxidant profiles/immune activity has important implications for organismal performance, disease susceptibility, and responses to local and climate stressors (Couch et al. 2008). Environmental history is an important factor influencing the response of corals to physiological stress (Brown et al. 2000(Brown et al. , 2002aAinsworth et al. 2016) and the capacity of corals to acclimatize and/or adapt to climate change (Palumbi et al. 2014;Dixon et al. 2015;Torda et al. 2017). As such, studying the ability of coral populations and Symbiodinium (Mayfield et al. 2012) to tolerate temperature variability (Maynard et al. 2008; Barshis et al. 2010), persistent high pCO 2 (Fabricius et al. 2011), and variable pCO 2 ) is critical to understanding the Montipora capitata from both reefs maintained their Symbiodinium densities, chlorophyll a per Symbiodinium cell, and total protein content when exposed to short-term thermal stress. However, chlorophyll a cm −2 and photochemical performance (i.e., F m and F v /F m ) declined in corals from both reefs when exposed to the heated treatment. The discrepancy in heating effects reducing chlorophyll a cm −2 but not affecting symbiont densities or chlorophyll a cell −1 may in part be explained by methodology (i.e., sample variability, statistical power), in addition to biological processes (i.e., photoacclimation), differences in tissue and skeletal optical properties (Wangpraseurt et al. 2012), and the internal light environments where Symbiodinium reside. F v /F m differed according to reef environmental history, with higher F v /F m at HV-Reef 14 under ambient and heated conditions, suggesting environmental history affected properties of Symbiodinium photomachinery and rates of electron transport (Warner et al. 2010). The heated treatment reduced F m and F v /F m , and indicates temperature-mediated damage to the photosynthetic machinery (Lesser 1997;Jones et al. 1998;Warner et al. 1999) and/or the activation of photoprotective mechanisms (Hoegh-Guldberg and Jones 1999;Osmond et al. 1999). The decreased photochemical performance, along with visible tissue paling and reduced chlorophyll a cm −2 , confirms M. capitata were experiencing stress commensurate with the onset of bleaching prior to the appreciable loss of symbiont cells. Declines in F v /F m often precede reductions in Symbiodinium or photopigment densities (Fitt et al. 2001;Rodrigues and Grottoli 2007). Indeed, short-term laboratory experiments have shown M. capitata maintains high symbiont densities despite loss of pigmentation from thermal (Rodrigues and Grottoli 2007) and ultraviolet (UV) Env. history environmental history of low variable pCO 2 (LV-Lilipuna) or high variable pCO 2 (HV-Reef 14), Treatment ambient (24.5 °C) or heated (30.5 °C), F o (minimum fluorescence yield), F m (maximum fluorescence yield), and F v /F m (maximum photochemical efficiency) of dark-adapted Symbiodinium photosystem II reaction centers, SS sum of squares, df degrees of freedom Bold P values represent significant effects (P < 0.05)   radiation stress (Grottoli-Everett and Kuffner 1995), suggesting M. capitata and its Symbiodinium possess an especially robust capacity for photoacclimation. In all corals sampled, levels of the antioxidative enzymes superoxide dismutase (i.e., SOD) and catalase (i.e., CAT) were not significantly affected by temperature treatments. Rather, significant differences were only observed as a function of reef environmental history. The lack of change in antioxidative activity was an unexpected observation, and we present three possible explanations for this result: (1) constitutive levels of antioxidants provided adequate protection during short-term thermal challenge; (2) antioxidative responses are a secondary form of defense not employed in the early stages of thermal stress, as such enzymes are energetically costly to produce (Palmer et al. 2011a); or (3) other compounds such as melanin (i.e., MEL) have dual function and exhibit some antioxidant activity (Nappi and Christensen 2005) providing sufficient protection against cellular damage during the onset of thermal stress. In other studies, the production of specialized antioxidative enzymes (e.g., SOD and CAT) is only induced after prolonged exposure (Downs et al. 2002). Melanisation may also function as a general acclimatization response to environmental perturbation. In this way, the melanin synthesis pathway may be an important immune parameter activated in corals exposed to periodic environmental stressors such as elevated irradiance and temperatures. Regardless, the observation of site-specific antioxidant profiles and melanin synthesis activity indicate phenotypic differences in corals at these two locations, potentially as a result of distinct pCO 2 histories at these locations. It is not known if the reef-specific differences in coral phenotypes observed here reflect mechanisms of acclimatization or local adaptation to a history of distinct environmental conditions; however, such lines of inquiry should be advanced to further our understanding of environmental history effects on corals.
The melanin synthesis pathway was responsive to heated treatments in corals from both reefs regardless of environmental history. This pathway begins with the proteolytic cleavage of inactive PPO to the active phenoloxidase (i.e., PO), and through a series of intermediate reactions ultimately leads to the production and deposition of melanin into coral tissues (Mydlarz and Palmer 2011;Nappi and Christensen 2005). PPO levels typically drop upon induction of melanin production as reserves of the latent enzyme are converted to their active form and consumed Palmer et al. 2011a, b). This agrees with results observed in the present study, where thermal stress caused reductions in coral tissue PPO and simultaneous increases in melanin production. However, other enzymes are also capable of completing the melanin synthesis pathway. For example, peroxidases can compete with PO for the hydroxylation of tyrosine and subsequent melanin deposition (Nappi Fig. 6 Immunological responses of Montipora capitata from two Kāne'ohe Bay reefs (LV-Lillipuna vs. HV-Reef 14) exposed to two temperature treatments (Ambient vs. Heated). a Superoxide dismutase (SOD) concentration, b catalase (CAT) activity, c prophenoloxidase (PPO) activity, d melanin (MEL) concentration. Values represent mean ± SE (n = 11-12), and symbols indicate significant effects (P < 0.05) of temperature treatment (*) or reef environmental history ( †) and Vass 1993; Nappi and Christensen 2005), although these enzymes were not measured here.
Melanin is a multifunctional compound that serves many roles. It is important for both wound healing (Palmer et al. 2011b) and pathogen encapsulation (Ellner et al. 2007;Mydlarz et al. 2008). It is also implicated in Symbiodinium photoprotection (Palmer et al. 2010(Palmer et al. , 2011a, as it is a known UV-absorbing molecule in mammals (Ortonne 2002;Sugumaran 2002). Similarly, invertebrate photoprotection is demonstrated in the water flea, Daphnia spp., where melanisation is positively correlated to UV exposure (Rautio and Korhola 2002). This photoprotective function was also recently confirmed in sponges, where melanin produced by symbiotic bacteria was protective against UV-induced intracellular reactive oxygen species (Vijayan et al. 2017). The sea fan, Gorgonia ventalina, displayed melanisation in response to elevated temperatures , and higher constitutive levels of melanin and melanin-containing granular cells have also been documented in coral species considered resistant to thermal bleaching (Palmer et al. 2010). The exact role of coral melanisation in response to increased temperature has yet to be elucidated, however, and the causes and consequences of increased melanin synthesis activity under short-term and prolonged thermal stress have interesting implications for cellular adaptive mechanisms.
In the present study, we found that corals exhibit constitutive differences in photobiology, chlorophyll a, antioxidative enzymes, and immunity. However, environmental history effects did not interact with temperature treatments to alter thermal stress response trajectories. Therefore, while environmental history can shape the response of corals to bleaching stress (Brown et al. 2002a), the specific environmental conditions at the two reefs in the present study did not influence the biological response of corals to short-term thermal stress. Other physical or biological factors in addition to pCO 2 history may also be responsible for influencing the reef-specific responses observed here. Such factors may include pathogen infections or immune response elicitors (Palmer et al. 2011a) and their present and historical distribution within Kāne'ohe Bay (Aeby et al. 2010), as well as low coral/high bare substrate cover and dissolved inorganic nitrogen concentrations (Couch et al. 2008). However, such factors do not appear to have played a significant role in the present study. First, previous exposure to disease and physiological stress can elevate coral immune activity Palmer et al. 2011a), and historically, coral disease (i.e., Montipora white syndrome) prevalence is greater in southern Kāne'ohe Bay reefs proximate to LV-Lilipuna, relative to central and northern reefs (Aeby et al. 2010). However, we observed greater antioxidative (SOD and CAT) and immune (PPO) activity at HV-Reef 14 in central Kāne'ohe Bay. Therefore, historical disease prevalence does not explain greater antioxidant or immune activity at HV-Reef 14. Alternatively, it is possible immune activity in LV-Lilipuna corals surviving historically high disease pressure (southern Kāne'ohe Bay) is a consequence of resistance/immunity to immune activity elicitors. Second, coral cover at the two reefs are comparable (ca. 75%) and inorganic nutrients within Kāne'ohe Bay are not different from those measured on offshore reefs (Cox et al. 2006), suggesting the influence of coral cover and dissolved nutrients in explaining differences among corals in the present study may be minimal. Seawater temperature (Coles and Jokiel 1978) and flow speed influence coral performance (Dennison and Barnes 1988), and it is possible slightly cooler daily minimum temperature at LV-Lilipuna (0.12 °C) or other properties of seawater associated with residence time/flow (Lowe et al. 2009) exerted influence here. In addition, differences in holobiont traits due to seasonality (Fitt et al. 2000) or symbiont abundance (Cunning and Baker 2014) at the time when stress is applied can influence stress outcomes, and winteracclimation may have attenuated heating effects in corals in the present study. Therefore, while pCO 2 history remains the most salient difference between LV-Lilipuna and HV-Reef 14 (Drupp et al. , 2013 best explaining the distinct responses of corals to short-term heating, the influence of other physical factors should not be wholly dismissed. Differences in symbiont communities among M. capitata colonies (Stat et al. 2011) can also influence physiological responses and stress outcomes (Sampayo et al. 2008;Cunning et al. 2016). Montipora capitata in the Main Hawaiian Islands are known to associate with both clade C and/or D Symbiodinium (Stat et al. 2013), namely C31 and D1-4-6 (S. glynnii) (Cunning et al. 2016;Wham et al. 2017). The latter are often found in corals from reefs with a history of thermal stress and/or variance and degraded water quality, such as Kāne'ohe Bay (Stat et al. 2013(Stat et al. , 2015. Thus, the reefspecific effects reported here may result from a combination of several non-mutually exclusive factors including environmental history (Brown et al. 2002a), host genotypes (Barshis et al. 2010;Bongaerts et al. 2010), symbiont community (Sampayo et al. 2008), and microbial consortia (Morrow et al. 2015), as well as unidentified genetic mechanisms (i.e., gene expression plasticity, DNA methylation) (Kenkel and Matz 2016;Putnam et al. 2016).
The role of environmental history in shaping coral physiology remains an important and burgeoning field of inquiry (Brown et al. 2002a;Middlebrook et al. 2008;Kenkel et al. 2013a, b;Ainsworth et al. 2016;Kenkel and Matz 2016), especially in the context of thermal and pCO 2 stress (Fabricius et al. 2011;Noonan and Fabricius 2016;Gibbin et al. 2017;Kenkel et al. 2017). Environmental history and phenotypic plasticity are important considerations for predictions in the biology, ecology, and evolution of marine organisms (Gaylord et al. 2015;Torda et al. 2017). Here, distinct environmental histories of pCO 2 variability did not interact with thermal stress to shape the suite of host and symbiont responses. Nevertheless, environmental history exerted strong influence over coral and Symbiodinium at both ambient and elevated temperatures, emphasizing differences among local reef environments even at small spatial scales are important in determining coral holobiont performance under favorable and challenging conditions. Finally, the melanin synthesis pathway was significantly upregulated during the early stages of thermal stress, and provides further evidence that melanisation is an important general stress response in corals exposed to warming seawater preceding the onset of symbiont losses.