Short periods of decreased water flow may modulate long-term ocean acidification in reef-building corals

17 Ocean acidification (OA) poses a major threat to reef -building corals. Although water flow 18 variability is common in coral reefs and modulates coral physiology, the interactive effects of 19 flow and OA on corals remain poorly understood. Therefore, we performed a three-month OA 20 experiment investigating the effect of changes in flow on coral physiology. We exposed the 21 reef-building corals Acropora cytherea, Pocillopora verrucosa, and Porites cylindrica to control 22 (pH 8.0) and OA (pH 7.8) conditions at moderate flow (6 cm s-1) and monitored OA effects on 23 growth. Throughout the experiment, we intermittently exposed all corals to low flow (2 cm s-1) 24 for 1.5 h and measured their photosynthesis :photosynthesis (P:R) ratio under low and 25 moderate flow. On average, corals under OA calcified 18 % less and grew 23 % less in surface 26 area than those at ambient pH. We observed species-specific interactive effects of OA and 27 flow on coral physiology. P:R ratios decreased after 12 weeks of OA in A. cytherea (22 %) and 28 P. cylindrica (28 %) under moderate flow , but were unaffected by OA under low flow. P:R 29 ratios were stable in P. verrucosa. These results suggest that short periods of decreased water 30 flow may modulate OA effects on some coral species, indicating that flow variability is a factor 31 to consider when assessing long-term effects of climate change. 32

33 ocean acidification, reef-building corals, water flow, coral physiology 34 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 3 1. INTRODUCTION 35 Ocean acidification (OA) constitutes a main future threat to reef-building corals (Doney et al. 36 2009), typically reducing their calcification rate (Kroeker et al. 2013) . However, the effect of 37 OA on respiration and photosynthe sis of reef -building corals, i.e., key processes of coral 38 physiology, may vary depending on environmental conditions and species . Although meta-39 analyses show that net photosynthesis generally remains stable under OA conditions (Kroeker 40 et al. 2013; Godefroid et al. 2022), it has also been reported to decrease (e.g., Reynaud et al. 41 2003; Bedwell-Ivers et al. 2017) or increase (e.g., Comeau et al. 2018; Biscéré et al. 2019) . 42 Coral reefs are highly dynamic ecosystems where local environmental variability modulates 43 coral physiology (McLachlan et al. 2021). For instance, natural diel oscillation of seawater pH 44 in reefs (Hannan et al. 2020) may cause differences in the physiological response of coral 45 species to OA and further complicate the assessment of its effects (Comeau et al. 2014a; 46 Bedwell-Ivers et al. 2017; Enochs et al. 2018). 47 Additional physical factors such as water flow, which vary on short time scales, may underlie 48 the complex responses of corals to OA. Water flow in coral reefs –typically high -energy 49 ecosystems exposed to currents, waves, and tides (Sheppard et al. 2018) –can vary greatly 50 within and among reefs (Lowe and Falter 2015). Flow velocities vary even within single reef 51 locations (e.g., 0–30 cm s-1; Roik et al. 2016) , with temporal decreases during the diel cycle 52 associated with tides (Green et al. 2018; Lindhart et al. 2021). Flow also differs between reef 53 environments, with back -reef environments typically experiencing lower flow than reef-crest 54 environments (Madin et al. 2006). Periods of low flow may be relatively common in some reefs 55 (e.g., Hench et al. 2008) , and flow immediately adjacent to coral colonies is further reduced 56 due to the formation of recirculation zones in complex reef topographies (Hench and Rosman 57 2013). 58 In reef-building corals, water flow modulates respiration and photosynthesis (Dennison and 59 Barnes 1988; Patterson et al. 1991) with species -specific effects (e.g., Rex et al. 1995; 60 Schutter et al. 2010) . Differences in physiological responses between species to w ater flow 61 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 4 may be associated with its effect on the coral boundary layer. This is the layer of seawater 62 bordering the coral surface that controls mass transfer between the coral and bulk seawater 63 (Atkinson and Bilger 1992; Shashar et al. 1993) . Water flow variability also has implications 64 for coral physiology. For instance, photochemical efficiency may be higher under alternating 65 high-to-low flow conditions than under constant flow (Smith and Birkeland 2007). Thus, speed 66 and short-term variability of water flow elicit complex patterns in coral physiology. 67 Although water flow is a prevailing characteristic of all coral reefs, knowledge of the interactive 68 effects of different flow regimes and OA on reef-building corals remains limited (Noisette et al. 69 2022). While high-flow environments have been proposed as refuges from the effects of ocean 70 warming (Fifer et al. 2021), low-flow environments are currently considered potential refuges 71 from OA for calcifying organisms (Hurd 2015). Effects of flow and OA on coral physiology may 72 be complex, potentially interactive, and differ by exposure time. For instance, during 1-h short-73 term OA exposure, net photosynthesis was similar between low (1 cm s-1) and moderate water 74 flow (4–13 cm s -1; Osinga et al. 2017) . Similarly, a fter two -day exposure to OA, coral 75 communities briefly exposed to high (35 cm s-1) and moderate flow (8 cm s-1) also had similar 76 net photosynthesis under OA (Anthony et al. 2013) . Whereas, after two-month exposure to 77 OA and different flow regimes (2.5 or 8 cm s -1), net photosynthesis under low flow was 78 decreased in Acropora yongei and increased in Plesiastrea versipora (Comeau et al. 2019a). 79 Similarly, d ecreased calcification of reef -building corals under OA may be alleviated by 80 temporary 24-h exposure to moderate water flow (5 and 10 cm s -1) compared to low flow 81 (2 cm s-1; Comeau et al. 2014c) . These results suggest that acclimatisation to OA and flow 82 regimes may occur on different time scales. Thus, a systematic investigation of long-term OA 83 effects on coral physiology combined with short -term fluctuations of water flow , as occur in 84 coral reefs, is needed. 85 The overarching aim of this study was to assess the physiological response of three reef-86 building coral species, Acropora cytherea (Dana, 1846) , Pocillopora verrucosa (Ellis & 87 Solander, 1786), and Porites cylindrica Dana, 1846, to changes in water flow under control 88 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 5 and OA conditions. Specifically, we tested how OA conditions (pH 7.8) maintained for three 89 months at constant moderate flow (6 cm s-1) affected I) coral calcification and surface growth 90 compared to control conditions (pH 8.0) and II) coral photosynthesis:respiration (P:R) ratios 91 under short periods of low flow (2 cm s-1) compared to responses under moderate flow. This 92 study will help to disentangle the complex effects of OA on coral physiology and better 93 understand the role of hydrodynamics in the response of reef-building corals to OA. 94 2. MATERIALS AND METHODS 95 2.1. Study species and experimental design 96 A three-month experiment was conducted to investigate the physiological response of three 97 scleractinian coral species, Acropora cytherea, Pocillopora verrucosa, and Porites cylindrica, 98 to OA together with the effect of changes in water flow conditions using respirometry assays. 99 Coral colonies (Table S1) were maintained at the ‘Ocean2100’ long -term coral experimental 100 facility (8,000 L closed recirculating system of artificial seawater , Table S2) at Justus Liebig 101 University Giessen, Germany, for at least six months before the experiment. Conditions in the 102 long-term culturing tanks (256 L) were 11:13 h light:dark photoperiod with a light intensity of 103 230 µmol m-2 s-1, temperature of 26.0 ± 0.5 °C, and daily feeding of a mix of frozen Artemia sp., 104 Mysis sp., and copepods. For the experiment, a total of four colonies per species were used 105 and cut into eight fragments using a small angle grinder (Dremel Multitool 3000 -15, The 106 Netherlands). Fragments were attached to tiles with two-component glue (CoraFix SuperFast, 107 Grotech, Germany) and transferred to the experimental setup. Corals were acclimated to the 108 experimental setup for five weeks before the start of the experiment. During the first 10 days 109 of the acclimation period, corals were administered the long-term culturing feed, after which 110 the culturing feed was provided with lower frequency (2.7 mg L-1 of frozen copepods every two 111 days) due to the lower density of coral fragments within experimental tanks. C orals also 112 received constant dissolved nutrient conditions via the connected water system. 113 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 6 Experimental t reatments consisted of two pH levels , with the control treatment mimicking 114 present-day atmospheric pCO2 concentration on some reefs (~500 µatm pCO2; Ziegler et al. 115 2021), and the OA treatment with values projected in the long term (2081 –2100) for surface 116 ocean pH in coral reefs (UNEP-WCMC et al. 2021) under SSP2-4.5 (0.20 pH units lower 117 relative to 1961 –1990; Iturbide et al. 2022; IPCC 2023) . The physiological response of the 118 corals was monitored throughout both the acclimation and experimental period, which were 119 conducted from 15 October 2019 to 28 February 2020 . Buoyant weight was measured three 120 weeks before the start of gradual pH decrease (t-3) and after 13 weeks under OA (t13), while 121 surface area was documented four weeks before the start of gradual pH decrease (t -4) and 122 after 14 weeks under OA (t14). The time points of these two parameters differed by one week 123 due to the duration of measurements ( buoyant weight measurement of all coral fragments 124 took three days and surface area six days). P:R ratios were measured in low (2 cm s-1) and 125 moderate flow (6 cm s -1) four times on all fragments (32 fragments per species per fl ow 126 condition = 192 fragments per time point). Measurements of all fragments took six days to 127 complete per time point and were performed once at the end of the acclimation period (t-1, one 128 week before the start of gradual pH decrease) and three times during the experimental period 129 (t3, after three weeks under OA; t7, after seven weeks; t12, after 12 weeks). A complete timeline 130 of all physiological measurements can be found in Fig. 1A and details of the physiological 131 assessments are outlined below. 132 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 7 133 Fig. 1 Experimental design and setup. (A) Timeline of physiological measurements performed 134 throughout the acclimation and experimental periods. Different colours indicate different types 135 of physiological measurements (calcification, dark brown; surface growth, light brown; 136 photosynthesis:respiration ratio, orange). The white diamond indicates the start of the 137 experiment. Dates are provided as DD/MM/YYYY. (B) Overview of the experimental setup. 138 See Fig. S1 for a detailed close-up of an individual experimental tank. (C) Snapshot of diel pH 139 oscillation during the experiment. Shaded grey areas indicate nighttime. OA, ocean 140 acidification; t-1, one week before the start of gradual pH decrease; t-3, three weeks before; t-4, 141 four weeks before; t3, after three weeks under OA conditions, including two weeks of gradual 142 pH decrease; t7, after seven weeks; t12, after 12 weeks; t13, after 13 weeks; t14, after 14 weeks; 143 pHT, pH on the total scale 144 2.2. Experimental setup and treatment conditions 145 The experiment al setup consisted of eight 120 L tanks divided into two experimental pH 146 treatments (four tanks per treatment, 16 fragments per species per treatment) (Fig. 1B). Each 147 tank housed one fragment per colony (total of four fragments per species in each tank) with 148 15 cm spacing between them in the direction of flow. In addition, experimental tanks contained 149 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 8 other scleractinian and octocorals with the same number of individuals per tank (Fig. S1A). 150 The experimental tanks were supplied with water from the 8,000 L closed recirculating system 151 of artificial seawater (calcium concentration: 396 ± 6 mg L-1, phosphate: < 0.02 mg L-1, nitrate: 152 < 0.02 mg L -1, nitrite: < 0.01 mg L -1) with an inflow rate of 20–40 L h-1 (corresponding to a 153 100 % tank volume turnover every 3–6 h). In addition, the large water system received weekly 154 water changes of ~10 % of the water volume. Temperature was maintained at 26 °C through 155 a feedback -controlled heater ( 300 W; 548, Schego, Germany). Water flow conditions , 156 consisting of a flow velocity of 6 cm s-1 (measured at the position of coral fragments; OTT MF 157 pro, OTT Hydromet GmbH, Germany) and a standing wave with an amplitude of 5 mm, were 158 generated with two circulating pumps (ES-28, Aqualight, Germany) and one wave generator 159 (6208, Tunze, Germany). Salinity in the tanks was monitored daily using a conductivity sensor 160 (TetraCon 925, WTW, Germany) and maintained at 35. Light was provided by two T5 bulbs 161 (54 W, Aqua-Science, Germany), producing a light intensity of 176 ± 31 µmol m-2 s-1 in an 162 11:13 h light:dark photoperiod. Light intensity in the experimental tanks differed slightly from 163 conditions in the culturing tanks due to technical reasons. 164 Seawater pH was constantly monitored using a digital controller (Profilux 3, GHL, Germany) 165 attached to pH electrodes in each tank (GHL, Germany) , which were calibrated every two 166 weeks using NBS buffers. Values of pHNBS were converted to total scale (pHT) using equations 167 from Millero (2013) and Takahashi et al. (1982). pH values are expressed in total scale 168 throughout the text. OA conditions in each treatment tank were generated individually via pH-169 controlled CO2 dosing (bubbling) using solenoid valves, which controlled the release of CO2. 170 Pumping was done through one of the circulating pumps to aid CO2 dissolution and dispersion 171 in seawater. pH was gradually decreased in OA treatment tanks and lowered by 0.01–0.02 172 units every day over two weeks until target values were reached ( Fig. S2A). Target OA 173 conditions were maintained for 8 6 days, including diel oscillation of pH mimicking naturally 174 occurring variability (Hannan et al. 2020; Ziegler et al. 2021) . Total alkalinity (TA) was 175 measured in each experimental tank by open -cell potentiometric titration using a titrator 176 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 9 (TitroLine 7000, SI Analytics, Germany ) equipped with a glass pH -combination electrode 177 (A 162 2M -DIN-ID, SI Analytics, Germany ). Measurem ents were made following SOP3b 178 (Dickson et al. 2007) on 50 g samples with 0.1N HCl (Titrisol, Merk, Germany) in 35 g L-1 NaCl 179 and corrected using certified reference materials (Batch 183, A.G. Dickson Laboratory, 180 Scripps Institution of Oceanography, UCSD, USA; Dickson et al. 2003). Measurements were 181 performed every 2–4 days during the first two weeks of the experiment and then every 1 –2 182 weeks. TA was calculated using a modified Gran approach (Millero 2013). Alkalinity was also 183 monitored daily and maintained with two automatic in-house constructed calcium reactors (pH 184 6.2–6.4, coral rubble) and dosing of NaHCO3 in a common reservoir tank. The calcium reactor 185 was feedback controlled by an alkalinity controller (Alkatronic, Focustronic, Hong Kong) based 186 on 3-hourly automatic titrations. Seawater carbonate chemistry was calculated from days with 187 TA measurements using pHNBS and temperature values for a whole day and the corresponding 188 value of TA and salinity. TA and salinity values were assumed to be representative of the 189 conditions of the entire day on which they were measured. Calculations were performed in the 190 program CO2SYS (v25; Pelletier et al. 2007) with carbonic acid dissociation constants from 191 Mehrbach et al. (1973) refit by Dickson and Millero (1987). This approach is suitable for 192 biological OA experiments with treatments that have differences larger than 100 µatm pCO2 193 (Watson et al. 2017) and allowed us to account for the diel oscillation of pH in the tanks. 194 2.3. Coral calcification and surface growth determination 195 Calcification was determined from measurements of buoyant weight (Jokiel et al. 1978) and 196 calculated as the difference between final (t13) and initial (t-3) buoyant weight ( accuracy 197 0.001 g; KB 360 -3N, KERN & Sohn GmbH, Germany), converted to dry weight using an 198 aragonite density of 2.93 g cm-3 (Spencer Davies 1989), and normalised by initial surface area 199 (t-4) and time. Surface growth was determined using 3D scanning (Artec Spider 3D with Artec 200 Studio 9, Artec 3D, Luxembourg) by documenting coral surface area, following Reichert et al. 201 (2016). Briefly, corals were placed on a rotating plate and scans were captured in air within 202 60–90 seconds. 3D models were calculated by performing fine serial registration, global 203 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 10 registration (minimal distance: 10, iterations: 2,000, based on texture and geometry), sharp 204 fusion (resolution : 0.2, fill holes by radius, max. hole radius: 5) , and o utlier removal 205 (A. cytherea and P. cylindrica, SD: 3; P. verrucosa, SD: 4) . Artefact objects were removed 206 (small objects filter). To determine coral surface areas, meshes were trimmed manually at the 207 tissue border. All meshes were exported as Wavefront “.obj” files to MeshLab Visual 208 Computing Lab-ISTI-CNR (v1.3.4, BETA, 2014; Cignoni et al. 2008) , and surface area was 209 calculated using the “compute geometric measures” tool. Surface growth r ates were 210 determined as the difference between final (t14) and initial (t-4) surface area and normalised by 211 initial surface area and time. 212 2.4. Measurement of photosynthesis:respiration ratios 213 P:R ratios were derived from oxygen production and consumption rates at low and moderate 214 flow conditions measured at the same time of day to avoid bias due to diurnal variation. Coral 215 fragments were incubated individually in sealed 1 L glass chambers for 90 min at 191 ± 23 216 µmol m-2 s-1 to measure oxygen production, followed by 90 min in darkness to measure oxygen 217 consumption. Chambers were filled with seawater from the corresponding treatment and 218 maintained at 26 ºC. Low and moderate water flow conditions in the incubation chambers were 219 generated with a magnetic stirring bar (Fig. S1B) (Rades et al. 2022) and measured by visual 220 tracking of small, neutrally buoyant plastic beads . Dissolved oxygen concentration was 221 measured in each chamber at the start and end of each incubation using an optical oxygen 222 sensor (FDO 925, WTW, Germany). Four empty chambers were included in every incubation 223 run to control for background biological activity. Rates of oxygen production and consumption 224 were calculated as the change in dissolved oxygen per incubation volume (calculated as the 225 difference between water volume and volume of the coral fragment with its tile) and normalised 226 to incubation time. P:R ratios were calculated as the ratio of gross oxygen production (i.e., the 227 sum of net oxygen production and consumption) to oxygen consumption with an 11:24 h 228 metabolic cycle (i.e., the ratio of total oxygen produced during daylight hours to that consumed 229 during a 24 h period) to estimate daily autotrophic capability (McCloskey et al. 1978). Rates 230 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 11 of respiration and photosynthesis were not analysed in this study due to the presence of 231 inconclusive patterns between rates measured during the acclimation period (t-1) and at the 232 end of the experimental period (t12), but are available for inspection in Fig. S3 and Table S3. 233 Additional details are provided as Supplementary Text. 234 2.5. Statistical analysis 235 All statistical analyses were performed in R (v.4.1.0; R Core Team 2021) using RStudio 236 (v1.4.1103; RStudio Team 2021) . All plots were produced using the R package ggplot2 237 (Wickham 2016). Changes in the physiological parameters of the three studied corals were 238 investigated using linear mixed -effects models (LMMs). Differences between species in 239 calcification and surface growth rates were assessed using LMMs with species (3 levels: 240 A. cytherea, P. verrucosa, and P. cylindrica) as a fixed factor, and colony and treatment as 241 random factors, while differences in P:R ratios were analysed using the same model structure 242 but with the addition of coral fragment identity (ID) as a random effect. To test the effect of OA 243 on calcification and surface growth , we used LMMs constructed for each species with 244 treatment (2 levels: control and OA) as a fixed factor and colony as a random factor. The P:R 245 ratio response to OA and flow over time was assessed using LMMs constructed for each 246 species with treatment (2 levels: control and OA), flow (2 levels: low and moderate), and time 247 (3 levels: t3, t7, and t12) as fixed factors in a fully crossed design, and ID, colony, and tank as 248 random factors. LMMs were performed using the R package lme4 (Bates et al. 2015). Model 249 validation was performed by graphically assessing homogeneity and normality assumptions , 250 and models were inspected for any influential observations using the R package performance 251 (Lüdecke et al. 2021) . The numerical output of LMMs was extracted using the R package 252 sjPlot (Lüdecke 2021) and is provided with model formulas in Tables S4 and S5. We then 253 computed type -II ANOVA tables of the fixed effects of LMMs using the Kenward-Roger 254 approximation for the degrees of freedom in the R package car (Fox & Weisberg 2019. Type 255 II sums of squares was selected to compute ANOVAs, following previously recommended 256 protocol for assessing main effects individually in the absence of interactions (Langsrud 2003; 257 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 12 Hector et al. 2010). Post hoc analyses were performed using the R package emmeans (Lenth 258 2021) with Bonferroni adjustment of p-values. 259 Differences in seawater chemistry between treatments were tested using daily mean values 260 from days with TA measurements and the same approach as above (LMM-ANOVA) with 261 treatment as a fixed factor (2 levels: control and OA) and tank and date as random factors. 262 All fragments of A. cytherea in one control tank bleached and subsequently died eight weeks 263 into the experiment. The fragments of P. verrucosa from the same tank also showed signs of 264 bleaching after 11 weeks into the experiment and were driving the response patterns in the 265 data analyses (as revealed by correlation analysis, Fig. S4). Therefore, all fragments from this 266 tank were excluded from the analyses. Based on the regular monitoring of water parameters, 267 the underlying reason for the affected fragments could not be identified. The other fragments 268 of these species appeared healthy. In addition, the coral fragments did not show visual signs 269 of bleaching or necrosis in response to the OA treatment. 270 3. RESULTS 271 3.1. Ocean acidification conditions 272 During the three months of the experiment , pH of the control was significantly higher at 273 7.98 ± 0.13 (mean ± SD; daily range: 7.79–8.19) than in the OA treatment at 7.78 ± 0.13 274 (range: 7.60–8.01; LMM-ANOVA, F = 298, p < 0.001) . The control and OA treatment s had 275 similar diel pH oscillation s (Fig. 1C), with a diel range of 0.4 pH units in both treatments 276 throughout the experiment (Fig. S2B; Table S6). pCO2 values were significantly lower in the 277 control at 480 ± 171 µatm (range: 244–769 µatm) than in the OA treatment at 813 ± 286 µatm 278 (range: 416–1,262 µatm; LMM-ANOVA, F = 273, p 0.05; Table 1). Seawater 280 chemistry per tank and a summary of the full recording of pH and temperature values are 281 provided in the Supplementary Material (Fig. S2B,C; Tables S6,S7). 282 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 13 Table 1 Seawater chemistry during a three-month ocean acidification experiment. Values are 283 expressed as mean ± SD with measurement replication (n). pH T, pH on the total scale; TA, 284 total alkalinity; pCO2, partial pressure of CO 2; DIC, dissolved inorganic carbon; Ω ca, calcite 285 saturation; Ωar, aragonite saturation 286 Control Ocean Acidification Salinity 34.6 ± 0.4 (10) 34.7 ± 0.4 (10) Temperature (ºC) 25.9 ± 0.3 (1,736) 25.9 ± 0.2 (1,503) pHT 7.98 ± 0.13 (1,736) 7.78 ± 0.13 (1,503) Daily Minimum pHT 7.79 ± 0.12 (10) 7.60 ± 0.12 (10) Daily Maximum pHT 8.19 ± 0.04 (10) 8.01 ± 0.06 (10) TA (µmol kg-1) 2,155 ± 52 (47) 2,155 ± 60 (44) pCO2 (µatm) 480 ± 171 (1,736) 813 ± 286 (1,503) Daily Minimum pCO2 (µatm) 244 ± 31 (10) 413 ± 67 (10) Daily Maximum pCO2 (µatm) 769 ± 195 (10) 1,262 ± 330 (10) DIC (µmol kg-1) 1,906 ± 76 (1,736) 2,009 ± 74 (1,503) CO2 (µmol kg-1) 13 ± 5 (1,736) 23 ± 8 (1,503) HCO3 - (µmol kg-1) 1,709 ± 109 (1,736) 1,857 ± 94 (1,503) CO3 !- (µmol kg-1) 184 ± 43 (1,736) 129 ± 33 (1,503) Ωca 4.45 ± 1.05 (1,736) 3.13 ± 0.81 (1,503) Ωar 2.94 ± 0.69 (1,736) 2.07 ± 0.53 (1,503) 287 3.2. Ocean acidification decreased coral calcification and surface growth 288 During the experimental period, all coral fragments grew in weight and surface area, but 289 increases differed between the three investigated species (LMM-ANOVA, F = 17.0/17.1, 290 p < 0.01; Fig. 2). Acropora cytherea had a 16 % weight gain over the experiment (pooled over 291 treatments) and a significantly lower overall calcification rate than Pocillopora verrucosa and 292 Porites cylindrica, which presented a weight gain of 42 and 30 %, respectively (Tables S8,S9). 293 Also, while the surface area of A. cytherea increased by 58 % during the experiment across 294 treatments, it increased by 163 and 140 % in P. verrucosa and P. cylindrica, respectively, and 295 was significantly higher than surface area growth in A. cytherea (Tables S8,S9). 296 The OA treatment decreased calcification and surface growth rates in all species (Fig. 2B,C). 297 Acropora cytherea showed a 24 and 30 % reduction in calcification and surface growth, 298 respectively (LMM-ANOVA, F = 14.1/11.1, p < 0.01) (Table S8). In P. verrucosa, calcification 299 and surface growth decreased by 20 and 23 % (LMM-ANOVA, F = 6.8/13.8, p < 0.05/0.01), 300 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 14 and in P. cylindrica by 13 and 19 %, respectively (LMM-ANOVA, F = 7.3/5.8, p < 0.05) (Table 301 S8). 302 303 Fig. 2 Growth effects of ocean acidification (OA) on three reef -building coral species. (A) 304 Photographs of the investigated reef-building coral species taken at the end of the acclimation 305 period: (1) Acropora cytherea, (2) Pocillopora verrucosa, (3) Porites cylindrica. Scale bar = 5 306 mm. (B) Calcification and (C) surface growth of A. cytherea, P. verrucosa, and P. cylindrica 307 during three months in a control and OA treatment. Boxes represent the first and third quartiles 308 with lines as medians and whiskers as the minimum and maximum values or up to the 1.5 ´ 309 interquartile range (IQR), whichever is reached first. Stars indicate significant differences 310 between the control and OA treatment (p < 0.01**, p < 0.05*, from linear mixed-effects models 311 with ANOVA). 312 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 15 3.3. Ocean acidification effects on photosynthesis:respiration ratios increased over 313 time 314 The three investigated coral species displayed time -delayed physiological responses to OA 315 treatments with complex interactions with water flow ( Fig. 3). P:R ratios were overall similar 316 between species during the experiment (pooled over treatment s and time, LMM -ANOVA, 317 F = 3.9, p > 0.05). They increased after the pre-OA phase (t-1; Fig. 3; Table S10) and differed 318 between time points during the OA-phase for all species (pooled over treatment, LMM-319 ANOVA, A. cytherea, F = 11.1, p < 0.001; P. verrucosa, F = 11. 3, p < 0.001; P. cylindrica, 320 F = 63.0, p < 0.001; Table S11). The P:R ratio was reduced in the OA treatment compared to 321 the control with interactive effects with time (LMM -ANOVA, Treatment-Time interaction) in 322 A. cytherea (F = 13.9, p < 0.001) and P. cylindrica (F = 3.4, p 0.05). OA effects occurred from three 324 weeks onward in P. cylindrica and after 12 weeks in A. cytherea (Table S11). P:R ratio was 325 on average lower under low flow than moderate flow conditions (LMM-ANOVA, A. cytherea, 326 F = 6.0, p < 0.05; P. cylindrica, F = 20.9, p < 0.0 01), except in P. verrucosa, for which this 327 effect was reversed (LMM-ANOVA, F = 150.9, p < 0.001) (Fig. 3). 328 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 16 329 Fig. 3 Photosynthesis:Respiration (P:R) ratio of Acropora cytherea, Pocillopora verrucosa, 330 and Porites cylindrica in a control and ocean acidification (OA) treatment and measured under 331 low flow (LF, 2 cm s -1) and moderate flow (MF, 6 cm s -1) conditions during the acclimation 332 period (measured one week before the start of gradual pH decrease) and experimental period 333 (after three, seven, and 12 weeks under OA, including two weeks of gradual pH decrease). 334 Data from the acclimation period are presented separated by the respective treatment applied 335 during the OA phase. Boxes represent the first and third quartiles with lines as medians and 336 whiskers as the minimum and maximum values or up to the 1.5 ´ interquartile range (IQR), 337 whichever is reached first. Stars indicate significant differences between the control and OA 338 treatment within each time point and flow condition (p < 0.001***, p < 0.01**, p < 0.05*, from 339 linear mixed-effects models with ANOVA). 340 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 17 4. DISCUSSION 341 Our study shows that the physiological response of corals to future OA conditions varies with 342 water flow conditions. In this study, all coral species showed reduced calcification and surface 343 area growth rates under near-natural OA conditions with diel pH oscillation. In addition, short 344 periods of low water flow modulated P:R ratios, which developed with OA-exposure time and 345 differed among species with potential downstream effects on their calcification rates. 346 4.1. Decreased growth under ocean acidification with diel pH oscillation 347 In our study, calcification and surface growth decreased in all species under OA, which is 348 consistent with previous findings (e.g., Sekizawa et al. 2017) and the generalised effect of OA 349 on coral calcification (Kornder et al. 2018) . Still, calcification i n Acropora cytherea may be 350 unaffected by OA at stable and even higher pCO2 than tested here (1,000 µatm pCO2; 351 Godefroid et al. 2021) . Likewise, Pocillopora verrucosa maintained stable calcification rates 352 with moderately and highly elevated pCO2 under stable pH conditions (700 and 1,000 µatm 353 pCO2; Comeau et al. 2019b). In coral reefs, pH oscillates naturally with diel ranges, which vary 354 between and within reefs (Hannan et al. 2020; Cyronak et al. 2020) and are expected to 355 increase under future OA conditions (Shaw et al. 2013) . pH variability is thus an important 356 element of biological OA manipulation experiments that is essential for extrapolating 357 responses observed in the laboratory to coral reefs (Ziegler et al. 2021). In our study, pH was 358 not stable but had a simulated diel oscillation of 0.4 pH units, which is representative of ranges 359 present in some reefs (Cyronak et al. 2020). Since pH oscillation may exacerbate OA-induced 360 calcification decreases (Comeau et al. 2014a) , the calcification responses observed in our 361 study are thus potentially associated with diel pH oscillation as observed in natural reef 362 ecosystems. 363 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 18 4.2. Physiological response to ocean acidification dependent on exposure time and 364 water flow 365 OA is a slowly developing press disturbance. Therefore, short-term acute assays will fail to 366 capture the breadth of organismal responses to naturally -occurring acidification. Here, we 367 assessed the progression of OA effects by monitoring P:R ratios over time. P:R ratios were in 368 line with daily P:R ratios reported previously (~1.1; Jacquemont et al. 2022) and also when 369 calculated as a direct P:R ratio (an estimate of autotrophic capability during daylight hours; 370 ~2.3; Biscéré et al. 2019) . Under OA, P:R ratios decreased at moderate flow , with time -371 delayed differential responses, except in P. verrucosa, where they remained stable. These 372

could suggest an increasing reliance on heterotrophically fixed energy under OA, 373 consistent with the reported alleviation of OA effects on calcification with increased feeding 374 (Towle et al. 2015) . These data thus indicate different vulnerabilities among species with 375 variable trophic dependence. 376 Specifically, the response of A. cytherea , which showed significant differences between 377 treatments only after 12 weeks of exposure to OA, might indicate an initial compensation for 378 adverse effects that the coral was then unable to sustain over time . Acroporids are highly 379 autotrophic, which is a trophic mode associated with high susceptibility to environmental 380 stress, such as high temperature (Conti-Jerpe et al. 2020) . However, some Acropora spp. 381 have the capacity to increase feeding rates under OA conditions and heterotrophically 382 compensate for OA effects (Towle et al. 2015) , which might explain the delayed response 383 observed in A. cytherea. Therefore, our results support the classification of A. cytherea with 384 other acroporids as OA-susceptible in the long term (Kornder et al. 2018), despite being able 385 to potentially compensate for short-term OA challenges. 386 With changes in P:R ratios after just three weeks of exposure to OA, the response of Porites 387 cylindrica to OA was the most immediate in our study . While massive Porites spp. are 388 consistently observed to be OA-resilient (Fabricius et al. 2011; Comeau et al. 2019b), poritids 389 with branching growth forms , such as P. cylindrica, appear more vulnerable (Comeau et al. 390 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 19 2014b). However, its overall milder response compared to A. cytherea may be due to the 391 mixotrophic strategy of poritids, i.e., their more balanced contributions from autotrophy and 392 heterotrophy (Conti-Jerpe et al. 2020) . In contrast to A. cytherea and P. cylindrica, P. 393 verrucosa maintained stable P:R ratios across treatments throughout the experiment and was 394 thus more resilient to OA, supporting the current view that pocilloporids are highly resilient 395 (Kornder et al. 2018). A higher contribution of heterotrophic feeding and physiological plasticity 396 in Pocillopora than in Acropora or Porites spp. may underlie these differences (Hoogenboom 397 et al. 2015; Radice et al. 2019). 398 Furthermore, our results from A. cytherea and P. cylindrica confirm the expectation of stronger 399 effects under moderate flow than under low flow. Ecologically this means that some species 400 threatened by OA on a global scale might benefit from local environmental variation and offset 401 OA effects on local scales (e.g., in environments with large short-term OA fluctuations and/or 402 low flow such as reef flats under calm weather; Lowe et al. 2009; Shaw et al. 2012) . 403 Accordingly, low-flow environments are considered refuge environments from OA for many 404 calcifying organisms (Hurd 2015) . Notably, the effects of heat stress, which is a pulse 405 disturbance, may be stronger for Acropora spp. under low flow than high flow conditions 406 (Nakamura and Van Woesik 2001; Page et al. 2021), and recovery from bleaching events may 407 also be faster under high flow conditions (Fifer et al. 2021). Yet, field studies have found lower 408 bleaching intensity at lagoon sites than high -flow environments (McClanahan et al. 2007; 409 Hoogenboom et al. 2017), which may not be the case for all reef habitats (e.g., Ainsworth et 410 al. 2021). Therefore, future coral reef conservation strategies to address climate change will 411 likely benefit from a diverse portfolio of refug ia to balance trade -offs associated with reef-412 specific characteristics and timescales of stressors. 413 The low-flow effects reported in this study might be underestimated when compared to laminar 414 low-flow conditions. Given our respirometry setup, the flow regimes tested were likely relatively 415 turbulent with lower intensity in the low flow condition due to its lower velocity. Under turbulent 416 flow conditions, boundary layer behaviour may be limited (Reidenbach et al. 2006), reducing 417 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 20 the effect of flow changes on coral physiology (Lesser et al. 1994) . Therefore, our results 418 support the notion of milder OA effects with reduced flow, even when flow is relatively 419 turbulent, as may occur on coral reefs. The flow conditions simulated in this study could be 420 representative of conditions in reef areas where oscillatory flow may limit the development of 421 boundary layers, such as reef flats (Davis et al. 2021). Still, in situ flow regimes are influenced 422 by various factors (Monismith 2007; Davis et al. 2021) and are thus generally more complex 423 than the flow conditions in our study . Also, coral colonies larger than the coral fragments in 424 our study may have a larger effect on the flow patterns around them (e.g., Hench and Rosman 425 2013; Hossain and Staples 2020). Therefore, future studies incorporating ecologically relevant 426 flow dynamics and a range of colony sizes and shapes will be important to disentangle OA 427 and flow effects. 428 5. CONCLUSIONS 429 Our study indicates that short periods of low flow modulate the physiological response of 430 corals to OA. We show that OA conditions of moderately elevated pCO2 and naturally 431 oscillating pH I) reduce coral calcification and surface growth rates and that II) coral species 432 display differential time-delayed P:R ratio responses to OA, which may be mitigated by 433 temporarily reduced flow conditions. This differential progression of OA responses over time 434 may be related to differences in trophic strategy and explain the variable susceptibility to long-435 term OA among coral species, which should be subject of follow-up work. Future research on 436 this topic could potentially inform the design and management of coral nurseries. Overall, our 437

highlight that the combination of long-term OA exposure in the variable hydrodynamic 438 conditions of coral reefs may lead to complex biological outcomes that require consideration 439 of the spatial and temporal scales at which they occur . Finally, in situ flow regimes are 440 generally more complex than the low flow conditions in our study . Therefore, future studies 441 incorporating ecologically relevant flow regimes and dynamics will be important to disentangle 442 OA and flow effects and better understand the potential of low-flow environments as refugia. 443 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 21 DATA AVAILABILITY 444 The datasets and code for the analyses presented in this study can be 445 found in the online Figshare repository and are accessible at 446 https://doi.org/10.6084/m9.figshare.23538474. 447

448 We thank Giulia Puntin (JLU, Germany) for help during respirometry measurements. This 449 study was conducted as part of the ‘Ocean2100’ global change simulation project of the 450 Colombian-German Center of Excellence in Marine Sciences (CEMarin) , funded by the 451 German Academic Exchange Service (DAAD, project number 57480468). 452 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 22

453 Ainsworth TD, Leggat W, Silliman BR, Lantz CA, Bergman JL, Fordyce AJ, Page CE, Renzi 454 JJ, Morton J, Eakin CM, Heron SF (2021) Rebuilding relationships on coral reefs: Coral 455 bleaching knowledge-sharing to aid adaptation planning for reef users. BioEssays. doi: 456 10.1002/bies.202100048 457 Anthony KRN, Diaz-Pulido G, Verlinden N, Tilbrook B, Andersson AJ (2013) Benthic buffers 458 and boosters of ocean acidification on coral reefs. Biogeosciences 10:4897–4909. doi: 459 10.5194/bg-10-4897-2013 460 Atkinson MJ, Bilger RW (1992) Effects of water velocity on phosphate uptake in coral reef-flat 461 communities. Limnol Oceanogr 37:273–279. doi: 10.4319/lo.1992.37.2.0273 462 Bates D, Mächler M, Bolker B, Walker S (2015) Fitting linear mixed-effects models using lme4. 463 J Stat Softw 67:1–48. doi: 10.18637/jss.v067.i01 464 Bedwell-Ivers HE, Koch MS, Peach KE, Joles L, Dutra E, Manfrino C (2017) The role of in 465 hospite zooxanthellae photophysiology and reef chemistry on elevated pCO2 effects in 466 two branching Caribbean corals: Acropora cervicornis and Porites divaricata. ICES J 467 Mar Sci 74:1103–1112. doi: 10.1093/icesjms/fsw026 468 Biscéré T, Zampighi M, Lorrain A, Jurriaans S, Foggo A, Houlbrèque F, Rodolfo -Metalpa R 469 (2019) High pCO2 promotes coral primary production. Biol Lett 15:20180777. doi: 470 10.1098/rsbl.2018.0777 471 Cignoni P, Callieri M, Corsini M, Dellepiane M, Ganovelli F, Ranzuglia G (2008) MeshLab: An 472 Open-Source Mesh Processing Tool. Sixth Eurographics Italian Chapter Conference 473 129–136. doi: 10.2312/LocalChapterEvents/ItalChap/ItalianChapConf2008/129-136 474 Comeau S, Edmunds PJ, Spindel NB, Carpenter RC (2014a) Diel pCO2 oscillations modulate 475 the response of the coral Acropora hyacinthus to ocean acidification. Mar Ecol Prog 476 Ser 501:99–111. doi: 10.3354/meps10690 477 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 23 Comeau S, Edmunds PJ, Spindel NB, Carpenter RC (2014b) Fast coral reef calcifiers are 478 more sensitive to ocean acidification in short -term laboratory incubations. Limnol 479 Oceanogr 59:1081–1091. doi: 10.4319/lo.2014.59.3.1081 480 Comeau S, Edmunds PJ, Lantz CA, Carpenter RC (2014c) Water flow modulates the response 481 of coral reef communities to ocean acidification. Sci Rep 4:1–6. doi: 10.1038/srep06681 482 Comeau S, Cornwall CE, DeCarlo TM, Krieger E, McCulloch MT (2018) Similar controls on 483 calcification under ocean acidification across unrelated coral reef taxa. Glob Chang Biol 484 24:4857–4868. doi: 10.1111/gcb.14379 485 Comeau S, Cornwall CE, Pupier CA, DeCarlo TM, Alessi C, Trehern R, McCulloch MT (2019a) 486 Flow-driven micro-scale pH variability affects the physiology of corals and coralline 487 algae under ocean acidification. Sci Rep 9:12829. doi: 10.1038/s41598-019-49044-w 488 Comeau S, Cornwall CE, DeCarlo TM, Doo SS, Carpenter RC, McCulloch MT (2019b) 489 Resistance to ocean acidification in coral reef taxa is not gained by acclimatization. Nat 490 Clim Chang 9:477–483. doi: 10.1038/s41558-019-0486-9 491 Conti-Jerpe IE, Thompson PD, Wong CWM, Oliveira NL, Duprey NN, Moynihan MA, Baker 492 DM (2020) Trophic strategy and bleaching resistance in reef -building corals. Sci Adv. 493 doi: 10.1126/sciadv.aaz5443 494 Cyronak T, Takeshita Y, Courtney TA, DeCarlo EH, Eyre BD, Kline DI, Martz T, Page H, Price 495 NN, Smith J, Stoltenberg L, Tresguerres M, Andersson AJ (2020) Diel temperature and 496 pH variability scale with depth across diverse coral reef habitats. Limnol Oceanogr Lett 497 5:193–203. doi: 10.1002/lol2.10129 498 Davis KA, Pawlak G, Monismith SG (2021) Turbulence and coral reefs. Ann Rev Mar Sci 499 13:343–373. doi: 10.1146/annurev-marine-042120-071823 500 Dennison WC, Barnes DJ (1988) Effect of water motion on coral photosynthesis and 501 calcification. J Exp Mar Biol Ecol 115:67–77. doi: 10.1016/0022-0981(88)90190-6 502 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 24 Dickson AG, Millero FJ (1987) A comparison of the equilibrium constants for the dissociation 503 of carbonic acid in seawater media. Deep Sea Research Part A Oceanographic 504 Research Papers 34:1733–1743. 505 Dickson AG, Afghan JD, Anderson GC (2003) Reference materials for oceanic CO2 analysis: 506 A method for the certification of total alkalinity. Mar Chem 80:185 –197. doi: 507 10.1016/S0304-4203(02)00133-0 508 Dickson AG, Sabine CL, Christian JR (2007) Guide to best practices for ocean CO 2 509 measurements. PICES Special Publication 3, CA 510 Doney SC, Fabry VJ, Feely RA, Kleypas JA (2009) Ocean acidification: The other CO 2 511 problem. Ann Rev Mar Sci 1:169–192. doi: 10.1146/annurev.marine.010908.163834 512 Enochs IC, Manzello DP, Jones PJ, Aguilar C, Cohen K, Valentino L, Schopmeyer S, Kolodziej 513 G, Jankulak M, Lirman D (2018) The influence of diel carbonate chemistry fluctuations 514 on the calcification rate of Acropora cervicornis under present day and future 515 acidification conditions. J Exp Mar Biol Ecol 506:135 –143. doi: 516 10.1016/j.jembe.2018.06.007 517 Fabricius KE, Langdon C, Uthicke S, Humphrey C, Noonan S, De’ath G, Okazaki R, 518 Muehllehner N, Glas MS, Lough JM (2011) Losers and winners in coral reefs 519 acclimatized to elevated carbon dioxide concentrations. Nat Clim Chang 1:165 –169. 520 doi: 10.1038/nclimate1122 521 Fifer J, Bentlage B, Lemer S, Fujimura AG, Sweet M, Raymundo LJ (2021) Going with the 522 flow: How corals in high-flow environments can beat the heat. Mol Ecol 30:2009–2024. 523 doi: 10.1111/mec.15869 524 Fox J, Weisberg S (2019) An R companion to applied regression, Third. Sage, Thousand 525 Oaks, CA 526 Godefroid M, Arçuby R, Lacube Y, Espiau B, Dupont S, Gazeau F, Metian M, Hédouin L (2021) 527 More than local adaptation: High diversity of response to seawater acidification in 528 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 25 seven coral species from the same assemblage in French Polynesia. J Mar Biol Assoc 529 United Kingdom 101:675–683. doi: 10.1017/S0025315421000618 530 Godefroid M, Dupont S, Metian M, Hédouin L (2022) Two decades of seawater acidification 531 experiments on tropical scleractinian corals: Overview, meta -analysis and 532 perspectives. Mar Pollut Bull 178:113552. doi: 10.1016/j.marpolbul.2022.113552 533 Green RH, Lowe RJ, Buckley ML (2018) Hydrodynamics of a tidally forced coral reef atoll. J 534 Geophys Res Oceans 123:7084–7101. doi: 10.1029/2018JC013946 535 Hannan KD, Miller GM, Watson S-A, Rummer JL, Fabricius K, Munday PL (2020) Diel pCO2 536 variation among coral reefs and microhabitats at Lizard Island, Great Barrier Reef. 537 Coral Reefs 39:1391–1406. doi: 10.1007/s00338-020-01973-z 538 Hector A, von Felten S, Schmid B (2010) Analysis of variance with unbalanced data: An update 539 for ecology & evolution. J Anim Ecol 79:308–316. doi: 10.1111/j.1365 -540 2656.2009.01634.x 541 Hench JL, Rosman JH (2013) Observations of spatial flow patterns at the coral colony scale 542 on a shallow reef flat. J Geophys Res Oceans 118:1142–1156. doi: 10.1002/jgrc.20105 543 Hench JL, Leichter JJ, Monismith SG (2008) Episodic circulation and exchange in a wave -544 driven coral reef and lagoon system. Limnol Oceanogr 53:2681 –2694. doi: 545 10.4319/lo.2008.53.6.2681 546 Hoogenboom M, Rottier C, Sikorski S, Ferrier-Pagès C (2015) Among-species variation in the 547 energy budgets of reef-building corals: Scaling from coral polyps to communities. J Exp 548 Biol 218:3866–3877. doi: 10.1242/jeb.124396 549 Hoogenboom MO, Frank GE, Chase TJ, Jurriaans S, Álvarez-Noriega M, Peterson K, Critchell 550 K, Berry KLE, Nicolet KJ, Ramsby B, Paley AS (2017) Environmental drivers of 551 variation in bleaching severity of Acropora species during an extreme thermal anomaly. 552 Front Mar Sci. doi: 10.3389/fmars.2017.00376 553 Hossain MM, Staples AE (2020) Effects of coral colony morphology on turbulent flow 554 dynamics. PLoS One 15:1–25. doi: 10.1371/journal.pone.0225676 555 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 26 Hurd CL (2015) Slow-flow habitats as refugia for coastal calcifiers from ocean acidification. J 556 Phycol 51:599–605. doi: 10.1111/jpy.12307 557 IPCC (2023) Atlas. In: Masson-Delmotte V, Zhai P, Pirani A, Connors SL, Péan C, Berger S, 558 Caud N, Chen Y, Goldfarb L, Gomis MI, Huang M, Leitzell K, Lonnoy E, Matthews JBR, 559 Maycock TK, Waterfield T, Yelekçi O, Yu R, Zhou B (eds) Climate Change 2021 – The 560 Physical Science Basis: Working Group I Contribution to the Sixth Assessment Report 561 of the Intergovernmental Panel on Climate Change. Cambridge University Press, 562 Cambridge, UK and New York, NY, USA, pp 1927–2058 563 Iturbide M, Fernández J, Gutiérrez JM, Pirani A, Huard D, Al Khourdajie A, Baño -Medina J, 564 Bedia J, Casanueva A, Cimadevilla E, Cofiño AS, De Felice M, Diez-Sierra J, García-565 Díez M, Goldie J, Herrera DA, Herrera S, Manzanas R, Milovac J, Radhakrishnan A, 566 San-Martín D, Spinuso A, Thyng KM, Trenham C, Yelekçi Ö (2022) Implementation of 567 FAIR principles in the IPCC: The WGI AR6 Atlas repository. Sci Data 9:629. doi: 568 10.1038/s41597-022-01739-y 569 Jacquemont J, Houlbrèque F, Tanvet C, Rodolfo-Metalpa R (2022) Long-term exposure to an 570 extreme environment induces species-specific responses in corals’ photosynthesis and 571 respiration rates. Mar Biol 169:82. doi: 10.1007/s00227-022-04063-6 572 Jokiel PL, Maragos JE, Franzisket L (1978) Coral growth: Buoyant weight technique. In: 573 Stoddart DR, Johannes RE (eds) Coral reefs: research methods. UNESCO, pp 529 –574 542 575 Kornder NA, Riegl BM, Figueiredo J (2018) Thresholds and drivers of coral calcification 576 responses to climate change. Glob Chang Biol 24:5084–5095. doi: 10.1111/gcb.14431 577 Kroeker KJ, Kordas RL, Crim R, Hendriks IE, Ramajo L, Singh GS, Duarte CM, Gattuso J 578 (2013) Impacts of ocean acidification on marine organisms: Quantifying sensitivities 579 and interaction with warming. Glob Chang Biol 19:1884–1896. doi: 10.1111/gcb.12179 580 Langsrud Ø (2003) ANOVA for unbalanced data: Use Type II instead of Type III sums of 581 squares. Stat Comput 13:163–167. doi: 10.1023/A:1023260610025 582 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 27 Lenth R V. (2021) emmeans: Estimated Marginal Means, aka Least -Squares Means. R 583 package version 1.6.2-1. https://cran.r-project.org/package=emmeans. 584 Lesser MP, Weis VM, Patterson MR, Jokiel PL (1994) Effects of morphology and water motion 585 on carbon delivery and productivity in the reef coral, Pocillopora damicornis (Linnaeus): 586 Diffusion barriers, inorganic carbon limitation, and biochemical plasticity. J Exp Mar Biol 587 Ecol 178:153–179. doi: 10.1016/0022-0981(94)90034-5 588 Lindhart M, Rogers JS, Maticka SA, Woodson CB, Monismith SG (2021) Wave modulation of 589 flows on open and closed reefs. J Geophys Res Oceans 126:e2020JC016645. doi: 590 10.1029/2020JC016645 591 Lowe RJ, Falter JL (2015) Oceanic forcing of coral reefs. Ann Rev Mar Sci 7:43 –66. doi: 592 10.1146/annurev-marine-010814-015834 593 Lowe RJ, Falter JL, Monismith SG, Atkinson MJ (2009) A numerical study of circulation in a 594 coastal reef-lagoon system. J Geophys Res 114:C06022. doi: 10.1029/2008JC005081 595 Lüdecke D (2021) sjPlot: Data visualization for statistics in social science. R package version 596 2.8.9. https://CRAN.R-project.org/package=sjPlot. 597 Lüdecke D, Ben -Shachar M, Patil I, Waggoner P, Makowski D (2021) performance: An R 598 package for assessment, comparison and testing of statistical models. J Open Source 599 Softw 6:3139. doi: 10.21105/joss.03139 600 Madin JS, Black KP, Connolly SR (2006) Scaling water motion on coral reefs: From regional 601 to organismal scales. Coral Reefs 25:635–644. doi: 10.1007/s00338-006-0137-2 602 McClanahan TR, Ateweberhan M, Muhando CA, Maina J, Mohammed MS (2007) Effects of 603 climate and seawater temperature variation on coral bleaching and mortality. Ecol 604 Monogr 77:503–525. doi: 10.1890/06-1182.1 605 McCloskey LR, Wethey DS, Porter JW (1978) Measurement and interpretation of 606 photosynthesis and respiration in reef corals. In: Stoddart DR, Johannes RE (eds) 607 Coral reefs: research methods. UNESCO, pp 379–396 608 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 28 McLachlan RH, Price JT, Muñoz -Garcia A, Weisleder NL, Jury CP, Toonen RJ, Grottoli AG 609 (2021) Environmental gradients drive physiological variation in Hawaiian corals. Coral 610 Reefs 40:1505–1523. doi: 10.1007/s00338-021-02140-8 611 Mehrbach C, Culberson CH, Hawley JE, Pytkowicx RM (1973) Measurement of the apparent 612 dissociation constants of carbonic acid in seawater at atmospheric pressure. Limnol 613 Oceanogr 18:897–907. 614 Millero FJ (2013) Chemical oceanography, 4th edn. CRC Press, Boca Raton, Florida 615 Monismith SG (2007) Hydrodynamics of coral reefs. Annu Rev Fluid Mech 39:37 –55. doi: 616 10.1146/annurev.fluid.38.050304.092125 617 Nakamura T, Van Woesik R (2001) Water -flow rates and passive diffusion partially explain 618 differential survival of corals during the 1998 bleaching event. Mar Ecol Prog Ser 619 212:301–304. doi: 10.3354/meps212301 620 Noisette F, Pansch C, Wall M, Wahl M, Hurd CL (2022) Role of hydrodynamics in shaping 621 chemical habitats and modulating the responses of coastal benthic systems to ocean 622 global change. Glob Chang Biol 28:3812–3829. doi: 10.1111/gcb.16165 623 Osinga R, Derksen-Hooijberg M, Wijgerde T, Verreth JAJ (2017) Interactive effects of oxygen, 624 carbon dioxide and flow on photosynthesis and respiration in the scleractinian coral 625 Galaxea fascicularis. J Exp Biol 220:2236–2242. doi: 10.1242/jeb.140509 626 Page CE, Leggat W, Heron SF, Fordyce AJ, Ainsworth TD (2021) High flow conditions mediate 627 damaging impacts of sub-lethal thermal stress on corals’ endosymbiotic algae. Conserv 628 Physiol 9:1–19. doi: 10.1093/conphys/coab046 629 Patterson MR, Sebens KP, Olson RR (1991) In situ measurements of flow effects on primary 630 production and dark respiration in reef corals. Limnol Oceanogr 36:936 –948. doi: 631 10.4319/lo.1991.36.5.0936 632 Pelletier G, Lewis E, Wallace D (2007) CO2Sys. xls: A calculator for the CO 2 system in 633 seawater for Microsoft Excel/VBA. Washington State Department of 634 Ecology/Brookhaven National Laboratory, Olympia, WA/Upton, NY, USA 635 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 29 R Core Team (2021) R: A language and environment for statistical computing. R Foundation 636 for Statistical Computing, Vienna, Austria 637 Rades M, Schubert P, Ziegler M, Kröckel M, Reichert J (2022) Building plan for a temperature-638 controlled multi -point stirring incubator. protocols.io. doi: 639 10.17504/protocols.io.dm6gpb34dlzp/v1 640 Radice VZ, Brett MT, Fry B, Fox MD, Hoegh -Guldberg O, Dove SG (2019) Evaluating coral 641 trophic strategies using fatty acid composition and indices. PLoS One 14:e0222327. 642 doi: 10.1371/journal.pone.0222327 643 Reichert J, Schellenberg J, Schubert P, Wilke T (2016) 3D scanning as a highly precise, 644 reproducible, and minimally invasive method for surface area and volume 645 measurements of scleractinian corals. Limnol Oceanogr Methods 14:518 –526. doi: 646 10.1002/lom3.10109 647 Reidenbach MA, Koseff JR, Monismith SG, Steinbuckc J V., Genin A (2006) The effects of 648 waves and morphology on mass transfer within branched reef corals. Limnol Oceanogr 649 51:1134–1141. doi: 10.4319/lo.2006.51.2.1134 650 Rex A, Montebon F, Yap HT (1995) Metabolic responses of the scleractinian coral Porites 651 cylindrica Dana to water motion. I. Oxygen flux studies. J Exp Mar Biol Ecol 186:33 –652 52. doi: 10.1016/0022-0981(95)00150-P 653 Reynaud S, Leclercq N, Romaine -Lioud S, Ferrier -Pagés C, Jaubert J, Gattuso J (2003) 654 Interacting effects of CO 2 partial pressure and temperature on photosynthesis and 655 calcification in a scleractinian coral. Glob Chang Biol 9:1660 –1668. doi: 656 10.1046/j.1529-8817.2003.00678.x 657 Roik A, Röthig T, Roder C, Ziegler M, Kremb SG, Voolstra CR (2016) Year-long monitoring of 658 physico-chemical and biological variables provide a comparative baseline of coral reef 659 functioning in the central Red Sea. PLoS One 11:e0163939. doi: 660 10.1371/journal.pone.0163939 661 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 30 RStudio Team (2021) RStudio: Integrated development environment for R. RStudio, PBC, 662 Boston, MA 663 Schutter M, Crocker J, Paijmans A, Janse M, Osinga R, Verreth AJ, Wijffels RH (2010) The 664 effect of different flow regimes on the growth and metabolic rates of the scleractinian 665 coral Galaxea fascicularis. Coral Reefs 29:737–748. doi: 10.1007/s00338-010-0617-2 666 Sekizawa A, Uechi H, Iguchi A, Nakamura T, Kumagai NH, Suzuki A, Sakai K, Nojiri Y (2017) 667 Intraspecific variations in responses to ocean acidification in two branching coral 668 species. Mar Pollut Bull 122:282–287. doi: 10.1016/j.marpolbul.2017.06.061 669 Shashar N, Cohen Y, Loya Y (1993) Extreme diel fluctuations of oxygen in diffusive boundary 670 layers surrounding stony corals. Biol Bull 185:455–461. doi: 10.2307/1542485 671 Shaw EC, McNeil BI, Tilbrook B (2012) Impacts of ocean acidification in naturally variable coral 672 reef flat ecosystems. J Geophys Res Oceans 117:C03038. doi: 673 10.1029/2011JC007655 674 Shaw EC, McNeil BI, Tilbrook B, Matear R, Bates ML (2013) Anthropogenic changes to 675 seawater buffer capacity combined with natural reef metabolism induce extreme future 676 coral reef CO2 conditions. Glob Chang Biol 19:1632–1641. doi: 10.1111/gcb.12154 677 Sheppard CRC, Davy SK, Pilling GM, Graham NAJ (2018) The abiotic environment. In: 678 Sheppard C, Davy S, Pilling G, Graham N (eds) The Biology of Coral Reefs, 2nd edn. 679 Oxford University Press, Oxford, UK, pp 68–99 680 Smith LW, Birkeland C (2007) Effects of intermittent flow and irradiance level on back reef 681 Porites corals at elevated seawater temperatures. J Exp Mar Biol Ecol 341:282 –294. 682 doi: 10.1016/j.jembe.2006.10.053 683 Spencer Davies P (1989) Short -term growth measurements of corals using an accurate 684 buoyant weighing technique. Mar Biol 101:389–395. doi: 10.1007/BF00428135 685 Takahashi T (1982) Carbonate chemistry. In: GEOSECS Pacific expedition, Volume 3, 686 Hydrographic Data 1973-1974. National Science Foundation, Washington, DC, pp 77–687 83 688 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint 31 Towle EK, Enochs IC, Langdon C (2015) Threatened Caribbean coral is able to mitigate the 689 adverse effects of ocean acidification on calcification by increasing feeding rate. PLoS 690 One 10:e0123394. doi: 10.1371/journal.pone.0123394 691 UNEP-WCMC, WorldFish Centre, WRI, TNC (2021) Global distribution of warm -water coral 692 reefs, compiled from multiple sources including the Millennium Coral Reef Mapping 693 Project. Version 4.1. Includes contributions from IMaRS-USF and IRD (2005), IMaRS-694 USF (2005) and Spalding et al. (2001). In: UN Environment World Conservation 695 Monitoring Centre. Cambridge, UK. https://doi.org/10.34892/t2wk -5t34. Accessed 22 696 Jul 2023 697 Watson S -A, Fabricius KE, Munday PL (2017) Quantifying pCO2 in biological ocean 698 acidification experiments: A comparison of four methods. PLoS One 12:e0185469. doi: 699 10.1371/journal.pone.0185469 700 Wickham H (2016) ggplot2: Elegant graphics for data analysis. Springer-Verlag, New York 701 Ziegler M, Anton A, Klein SG, Rädecker N, Geraldi NR, Schmidt-Roach S, Saderne V, Mumby 702 PJ, Cziesielski MJ, Martin C, Frölicher TL, Pandolfi JM, Suggett DJ, Aranda M, Duarte 703 CM, Voolstra CR (2021) Integrating environmental variability to broaden the research 704 on coral responses to future ocean conditions. Glob Chang Biol 27:5532 –5546. doi: 705 10.1111/gcb.15840 706 707 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 28, 2024. ; https://doi.org/10.1101/2024.02.23.581783doi: bioRxiv preprint