Review paper 9 pages

High summer temperatures amplify functional differences between coral- and algae-dominated reef communities

FLORIAN ROTH ,1,2,3,10 NILS RäDECKER ,1,4,5 SUSANA CARVALHO ,1 CARLOS M. DUARTE ,1,6 VINCENT SADERNE ,1 ANDREA ANTON ,1,6 LUIS SILVA ,1 MARIA LI. CALLEJA ,1,7 XOSÉ ANXELU G. MORÁN ,1

CHRISTIAN R. VOOLSTRA ,1,4 BENJAMIN KüRTEN ,1,8 BURTON H. JONES ,1 AND CHRISTIAN WILD 9

1Red Sea Research Center, King Abdullah University of Science and Technology (KAUST), Thuwal 23955 Saudi Arabia 2Baltic Sea Centre, Stockholm University, Stockholm 10691 Sweden

3Faculty of Biological and Environmental Sciences, Tvärminne Zoological Station, University of Helsinki, Helsinki 00014 Finland 4Department of Biology, University of Konstanz, Konstanz 78457 Germany

5Laboratory for Biological Geochemistry, School of Architecture, Civil and Environmental Engineering, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne 1015 Switzerland

6Computational Biology Research Center, King Abdullah University of Science and Technology (KAUST), Thuwal 23955 Saudi Arabia

7Department of Climate Geochemistry, Max Planck Institute for Chemistry (MPIC), Mainz 55128 Germany 8Project Management Jülich, Jülich Research Centre GmbH, Rostock 52425 Germany

9Marine Ecology, Faculty of Biology and Chemistry, University of Bremen, Bremen 28359 Germany

Citation: Roth, F., N. Rädecker, S. Carvalho, C. M. Duarte, V. Saderne, A. Anton, L. Silva, M. L. L. Call- eja, X. A. G. Morán, C. R. Voolstra, B. Kürten, B. H. Jones, and C. Wild. 2021. High summer tempera- tures amplify functional differences between coral- and algae-dominated reef communities. Ecology 102 (2):e03226. 10.1002/ecy.3226

Abstract. Shifts from coral to algal dominance are expected to increase in tropical coral reefs as a result of anthropogenic disturbances. The consequences for key ecosystem functions such as primary productivity, calcification, and nutrient recycling are poorly understood, par- ticularly under changing environmental conditions. We used a novel in situ incubation approach to compare functions of coral- and algae-dominated communities in the central Red Sea bimonthly over an entire year. In situ gross and net community primary productivity, calci- fication, dissolved organic carbon fluxes, dissolved inorganic nitrogen fluxes, and their respec- tive activation energies were quantified to describe the effects of seasonal changes. Overall, coral-dominated communities exhibited 30% lower net productivity and 10 times higher calcifi- cation than algae-dominated communities. Estimated activation energies indicated a higher thermal sensitivity of coral-dominated communities. In these communities, net productivity and calcification were negatively correlated with temperature (>40% and >65% reduction, respectively, with +5°C increase from winter to summer), whereas carbon losses via respiration and dissolved organic carbon release more than doubled at higher temperatures. In contrast, algae-dominated communities doubled net productivity in summer, while calcification and dis- solved organic carbon fluxes were unaffected. These results suggest pronounced changes in community functioning associated with coral-algal phase shifts. Algae-dominated communities may outcompete coral-dominated communities because of their higher productivity and car- bon retention to support fast biomass accumulation while compromising the formation of important reef framework structures. Higher temperatures likely amplify these functional dif- ferences, indicating a high vulnerability of ecosystem functions of coral-dominated communi- ties to temperatures even below coral bleaching thresholds. Our results suggest that ocean warming may not only cause but also amplify coral–algal phase shifts in coral reefs.

Key words: activation energy; biogeochemical cycling; climate change; community budget; ecosystem functioning; regime shifts.

INTRODUCTION

Community shifts and the ongoing loss of biodiversity (Brondizio et al. 2019) are altering the productivity and biogeochemistry of many ecosystems globally

(Middleton and Grace 2004, Hooper et al. 2012, Naeem et al. 2012). These changes compound with local and global environmental perturbations, which can acceler- ate the alteration of essential ecosystem processes (Bal- vanera et al. 2006, Stachowicz et al. 2007). Thermal stress caused by climate change is, thereby, likely to exhi- bit the most substantial impact (Stillman 2019). Tropical coral reefs are hotspots of biodiversity that

provide various ecosystem services that are supported by

Manuscript received 17 March 2020; revised 6 July 2020; accepted 24 August 2020. Corresponding Editor: Richard B. Aronson.

10 E-mail: [email protected]

Article e03226; page 1

Ecology, 102(2), 2021, e03226 © 2020 The Authors. Ecology published by Wiley Periodicals LLC on behalf of Ecological Society of America This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

one or more metabolic or biogeochemical functions (e.g., primary production, calcification, organic matter fluxes, and nutrient cycling; Moberg and Folke 1999). Many of these processes are primarily driven by sclerac- tinian corals, the “ecosystem engineers” of tropical reefs (Wild et al. 2011). However, the combination of global and local anthropogenic stressors has caused extensive coral mortality and subsequent shifts from complex coral-dominated communities to simplified communities with a predominance of filamentous turf- and macroal- gae in many reefs around the world (Done 1992, Bell- wood et al. 2004, Hughes et al. 2007, Graham et al. 2015). Although coral–algal phase shifts are increasingly observed globally, the consequences for reef ecosystem functions such as productivity, calcification, and nutri- ent cycling are poorly understood. Laboratory and mesocosm studies indicate that reef algae, particularly the widespread filamentous turfs, are metabolically very different from corals, and generally display significantly higher primary production rates (Rix et al. 2015, Cardini et al. 2016). At the same time, the fraction of the photo- synthetically fixed carbon (C) being exuded into the environment is generally more labile (Nelson et al. 2013). At the community level, these differences may result in changes in the carbonate chemistry of seawater (McMahon et al. 2013, Bernstein et al. 2016), disrupted trophic structures (Johnson et al. 1995, Hempson et al. 2018), or increased microbial loads on algae-dominated reefs worldwide (Jessen et al. 2013, Haas et al. 2016). Divergent responses to changing environmental condi-

tions may amplify ecosystem functions of corals and algae differently. As such, changing temperature regimes and recurrent heatwaves, which are increasing in fre- quency and magnitude (Frölicher et al. 2018, Oliver et al. 2018), can have detrimental effects on tropical coral reef taxa (Lough et al. 2018, Hughes et al. 2019). In corals, sublethal heat stress during summer can

compromise primary production and calcification (Rey- naud et al. 2003, Anthony et al. 2008), thereby altering the release of organic and inorganic products (Niggl et al. 2009, Piggot et al. 2009). In contrast, benthic turf- and macroalgae may be less sensitive to heat (Koch et al. 2013), showing increased productivity and net growth with rising temperature (Bender et al. 2014). Likewise, temperature-related productivity optima and mortality thresholds of algae are often well above those of corals (Anton et al. 2020). Similarly, the abundance of reef algae can increase seasonally, especially during the sum- mer months (Lirman and Biber 2000, Diaz-Pulido and Garzón-Ferreira 2002, Ateweberhan et al. 2006). However, few studies investigated the effects of coral–-

algal phase shifts on community metabolism, particu- larly in situ. This paucity of information probably reflects the logistical challenges of quantifying the func- tions of structurally complex communities in their natu- ral environment (Roth et al. 2019). Currently, most data describing ecosystem functions are derived from labora- tory (e.g., Cardini et al. 2016) and mesocosm (e.g.,

Langdon et al. 2003, Bellworthy and Fine 2018, Edmunds et al. 2020) studies using either single organ- isms or simplified reconstructed communities to predict in situ changes at the community scale. Although these approaches provide valuable mechanistic insights and permit a tight control of environmental conditions dur- ing the experiments, they can only approximate natural conditions. However, primary production, calcification, and organic matter recycling critically depend on local environmental conditions, biodiversity, and system heterogeneity (Baird et al. 2007). In addition, large parts of the energy and nutrient pool are remineralized by microbial communities or cryptic fauna within the reef matrix (Richter and Wunsch 1999, de Jongh and Van Duyl 2004, Maldonado et al. 2012), all of which are gen- erally not considered in ex situ experimental setups. Concordantly, Roth et al. (2019) highlighted in a com- parison between laboratory-based single-organism and in situ incubations that ex situ measurements that are scaled up to average-constructed communities can over- estimate community-wide net primary production and underestimate respiration and gross photosynthesis by 20–90%. Hence, laboratory experiments can only pro- vide a glimpse of the complex environmental dynamics (e.g., seasonality) that shape the ecological processes of reef communities (Damgaard 2019). To overcome these experimental constraints, we used

a novel in situ approach that allowed the quantification of major metabolic and biogeochemical pathways (Roth et al. 2019) of co-occurring natural coral- and algae- dominated reef communities in the Red Sea. With a total of 112 light and dark in situ incubations, we measured rates of community production (i.e., net community pro- duction [NCP], community respiration [CR], and gross primary production [GPP]), net community calcification (NCC), net dissolved organic carbon (DOC), and dis- solved inorganic nitrogen (DIN) fluxes bimonthly for an entire year. In addition, we quantified the thermal-de- pendence of the functioning of benthic communities by applying principles of the metabolic theory of ecology (MTE; Sibly et al. 2012). We quantified the temperature sensitivity of metabolic processes using the activation energy (Ea), as the slope or rate of change in the rise and falling phases of a thermal performance curve before and after achieving the optimal temperature. Although activation energies are commonly assessed at the organ- ism level (Garcı́a et al. 2018, Savva et al. 2018, Anton et al. 2020), they also provide useful insights regarding the sensitivity of community metabolism to warming (Follstad Shah et al. 2017, Morán et al. 2017, Padfield et al. 2017). Thus, we (1) directly compare the magnitudes and

directions of key functions of coral-dominated and phase-shifted algae-dominated reef communities, (2) derive their functional responses to environmental changes induced by seasonality, and (3) describe their thermal sensitivity to seasonally variable temperature changes.

Article e03226; page 2 FLORIAN ROTH ET AL. Ecology, Vol. 102, No. 2

MATERIALS AND METHODS

Study site and environmental conditions

This experiment was carried out at Abu Shosha reef located in the central Red Sea on the west coast of Saudi Arabia (22°18’16.3’’ N; 39°02’57.7’’ E) from January 2017 until January 2018. Key environmental variables were monitored at the sampling site and were previously reported in Roth et al. (2018). Water temper- ature was measured continuously (logging interval = 30 min) for the whole study period with Onset HOBO temperature/light data loggers (accuracy: �0.2°C) deployed at the seafloor. Salinity was measured at each day of sampling with a WTW TetraCon® conductivity cell (accuracy: �0.5% of value). Light availability was measured continuously (logging interval = 1 min) on three full days per month with the above-mentioned Onset HOBO data logger. Light readings were con- verted from lux to photosynthetically active radiation (PAR; μmol quanta�m−2�s−1; 400–700 nm wavelengths) by intercalibration and conversion as outlined in Roth et al. (2018), and values are presented as daytime means. Seawater samples for the determination of dis- solved nitrate (NO�3 ), nitrite (NO

� 2 ), ammonium

(NHþ4 ), phosphate (PO 3� 4 ), and monomeric silicate (Si

(OH)4) were taken in triplicates each month from 1 m above the seafloor. Details for sampling and analysis can be found in Appendix S1: Section S1. The sum of NO�3 , NO

� 2 , and NH

þ 4 is termed “dissolved inorganic

nitrogen” (DIN) henceforth.

Benthic communities selected for in situ incubations

Abu Shosha reef is characterized by a heterogeneous mosaic of patches of coral- and algae-dominated com- munities. Thus, this site allows for the quantification of the functionality of both communities under identical environmental conditions. Nnatural coral- and algae- dominated reef communities surrounded by sand were haphazardly selected at the study site at 5 m water depth within an area of 50 × 50 m. The communities had to fulfill the following characteristics to qualify as “suitable candidates” for later incubations: (1) Coral-dominated communities were defined by having >40% coral cover but <10% algae cover; (2) algae-dominated communities were defined by having >40% algae cover but <10% coral cover; (3) all communities had to fit into the incu- bation chambers (max. diameter 50 cm, max. height 39 cm). Among all suitable candidates in the study area, four coral-dominated and four algae-dominated com- munities were chosen randomly. These eight communi- ties were revisited each month of sampling. The community composition at the level of major

functional groups was assessed for each community three times during the study period (i.e., in the begin- ning, after 6 months, and at the end of the experiments). Details on the assessment and statistical evaluation of

the community composition can be found in Appendix S1: Section S1. Two-factor permutational multivariate analysis of variance (PERMANOVA) indicated that dif- ferences between communities grouped according to coral and algal dominance were significant (P = 0.001; visualization in Appendix S1: Fig. S1). As no significant changes over time in the relative benthic cover were detected (Appendix S1: Table S2),the benthic commu- nity composition of each treatment was averaged over all replicates and survey points (Fig. 1a). Algae-dominated communities were considered

“phase-shifted,” as complex structures and the occur- rence of coral rubble indicated that branching corals were present previously. The co-occurrence of coral and algal reef communities within small spatial ranges was reported as “mosaic-dynamics” before (e.g., Edmunds 2002, Tkachenko et al. 2007) and may be explained by a combination of local processes and historical effects, such as previous stress events or adaptation (Done et al. 1991, Bythell et al. 2000, Edmunds 2002). Cryptic habitats can encompass about 60–75% of the

total surface area of a reef (Richter and Wunsch 1999, Richter et al. 2001), but organisms living in cracks and crevices within the communities’ matrices could not be assessed by our conventional benthic surveys. These organisms (e.g., sponges, bryozoans, and tunicates), however, metabolize organic matter in the order of 15–30% of the gross production of a coral reef (reviewed in de Jongh and Van Duyl 2004), driving community res- piration and other biogeochemical fluxes assessed in this study. As our benthic incubation chambers jointly cap- tured the metabolism of all members of the communities (i.e., from the visible surface and cryptic habitats), we refrained from assigning measured metabolic activities to functional groups on the visible surface only, as the inferred contribution would be highly biased. Thus, measurements presented in this study represent commu- nity-wide processes that include all compartments of the reef benthos and the surrounding water.

In situ incubations and quantification of community functions

In situ incubations with benthic chambers were per- formed according to the protocol described in Roth et al. (2019). In brief, chambers were constructed from polymethyl methacrylate (PMMA) cylinders with removable gastight lids of the same material. All cham- bers were equipped with individual water circulation pumps with adjustable flow control, autonomous recording dissolved oxygen (DO), and temperature sen- sors (HOBO U26; temperature corrected and salinity adjusted), and two sampling ports for discrete water samples. Incubations were carried out on three consecu- tive days in January, March, May, July, September, and November 2017, and in January 2018. Generally, on day 1, divers deployed four chambers on coral-dominated, and four chambers on algae-dominated reef

February 2021 FUNCTIONING OF CORAL REEF COMMUNITIES Article e03226; page 3

communities (Fig. 1a). The chambers were positioned carefully and left in place with open tops (no lids) until the next morning. On the second day, incubations started at around 09:00 a.m. by tightly securing the lids and closing all sampling ports during natural daylight conditions. The exact incubation start and end time was recorded for each chamber. Incubations ran for approxi- mately 2 h. The chambers were left in place with open tops for a second set of incubation on the following day. On day 3, benthic communities were incubated at “simu- lated” darkness during the same period used for incuba- tions the previous day. The procedure followed the same as on day two; however, all chambers were covered with thick black PVC covers. Any light penetration into the chambers was prevented, as validated by control read- ings of Onset HOBO temperature/light data loggers within chambers. Between deployments, all materials were rinsed with freshwater, washed with 4% hydrochlo- ric acid (HCl), and subsequently rinsed with deionized

water for reliable water chemistry measurements that included sensitive DOC samples. Discrete water samples for dissolved inorganic carbon

(DIC), total alkalinity (TA), DOC, and DIN were with- drawn from the sampling ports with acid-washed syr- inges at the beginning and the end of each incubation (details for the analysis of these samples can be found in Appendix S1: Section S1). Changes in seawater chem- istry between start and end of incubations were used to calculate rates of NCP, CR, GPP (calculated as GPP = NCP + |CR|), NCC, and fluxes of DOC, and DIN. All rates and fluxes were extrapolated to incubation water volume (in L) and normalized to incubation duration (in h) and the planar reef area (in m2) of the enclosed ben- thic community adapted after Roth et al. (2019). Productivity and respiration rates (NCP and CR; in

mmol C�m−2�h−1) were calculated by changes in DIC concentrations, taking into account calcification and dissolution rates according to an adapted protocol by

(a) (b)

(c)

(

(

(%)

FIG 1. Experimental setup, relative benthic cover, and key environmental variables at the study site. (a) Relative benthic cover of functional groups in the studied coral- and algae-dominated reef communities, and exemplary pictures of the incubation chambers on the respective substrates. Details on the community composition of each replicate and time point can be found in Appendix S1: Section S1 and Fig. S1. (b) Photosynthetically active radiation (PAR, in μmol photons�m−2�s−1), dissolved inorganic nitrogen (DIN, in μM), and aragonite saturation state (Ωarag) at the study site from January 2017 until January 2018. Circles represent values of dis- crete samples for DIN and Ωarag, and daytime averages of three separate days per month for PAR; lines represent the smoothed trend through the means. (c) Plot of seawater temperature at experimental site from January 2017 until January 2018. Each dot rep- resents one measurement at 30-min intervals. Dashed horizontal line depicts the mean maximum annual temperature modeled for the region from 1982 to 2015, taken from Chaidez et al. (2017). [Color figure can be viewed at wileyonlinelibrary.com]

Article e03226; page 4 FLORIAN ROTH ET AL. Ecology, Vol. 102, No. 2

Albright et al. (2013). Rates of NCP and CR based on DIC fluxes were compared to rates based on oxygen fluxes from continuous measurements with DO sensors. No discrepancy between C and oxygen measurements was detected (r = 0.99, P < 0.0001, n = 112). The calcu- lated photosynthetic (1.05 � 0.02) and respiratory (0.97 � 0.02) quotients agree with those obtained by various authors elsewhere, typically ∼1 mol of oxygen produced for 1 mol C fixed, and vice versa (e.g., Gattuso et al. 1999b, Atkinson and Falter 2003). NCC (in mmol CaCO3�m−2�h−1) was calculated by

concentration differences in TA, which are primarily caused by calcification and dissolution of CaCO3, whereby TA is reduced (increased) by two molar equiva- lents for every mole of CaCO3 produced (dissolved) (Zeebe and Wolf-Gladrow 2001). Nutrients fluxes (i.e., NO�3 , NH

þ 4 , PO

3� 4 , and SO

2� 4 ) that cause a change in TA

unrelated to calcification and dissolution were accounted for according to Zeebe and Wolf-Gladrow (2001) and Wolf-Gladrow et al. (2007). DOC (in mmol C�m−2�h−1) and DIN (in µmol N�m−2�h−1) fluxes were calculated from concentration differences between start and endpoints. Any temperature corrections that were necessary for seawater chemistry calculations were achieved by temperature readings from individual tem- perature loggers within each chamber. Although measurements presented in this study only

relate to small benthic communities, most studies cur- rently work with individual reef organisms (e.g., Anton et al. 2020) or reconstructed communities (e.g., Edmunds et al. 2020) to derive community functions. Thus, the results presented here are among the best approximations for the quantification of community- wide (biogeochemical) ecosystem functions of untouched, natural coral reef communities in situ (but see Haas et al. 2013, Van Heuven et al. 2018).

Data analysis

Statistical analyses were performed using JMP© Pro14 (SAS Institute) statistic software. Environmental variables and response parameters from incubations were grouped into spring (March–May), summer (June–September), fall (October–November), and winter (December–February) for statistical analysis. For the seasonal comparison and to derive GPP/CR ratios, hourly rates from light and dark incubations were used to calculate daily net fluxes. We acknowledge that there is a chance for a slight over- or underestimation because of associated changes in environmental conditions (e.g., light) during the course of the day. Thus, to minimize the error associated with extrapolating hourly rates, we chose a time window for daylight incubations (from around 09:00 a.m. to 11:00 a.m.) that is closest to day- time average irradiation and excludes the “ramping up” phase in the early morning hours and extreme values that can occur during midday. As all incubations during all sampling periods were conducted at the very same

time of the day, incubations are comparable across com- munity types and time points. The full seawater carbonate system parameters were

derived for each sampling period from measured salinity, temperature, nutrients, TA, and DIC data using the CO2SYS Microsoft Excel Macro by Pierrot et al. (2006) and the R package Seacarb (Lavigne and Gattuso 2013) (Appendix S1: Table S3). Environmental variables were tested for differences with two-tailed t-tests. Response parameters from incubations (GPP, NCP, CR, NCC, DOC, and DIN) were assessed by linear mixed models (LMMs) to test for differences in the respective response parameters with ‘treatment’ (coral- vs. algae-dominated) and ‘season’ (spring, summer, fall, and winter) as fixed factors, and the sampling dates (date) within seasons and the replicates of the communities (community ID) as random factors. Tukey’s Honest Significant Differ- ence (HSD) test was used for pairwise comparisons if significant interactions (treatment * season) were found. Detailed statistical results, including significant post hoc comparisons, are presented in Appendix S1: Table S4. The relationships between response (e.g., metabolic functions) and explanatory variables (e.g., environmen- tal variables) were assessed by linear regression models. The thermal sensitivity of the metabolic processes

(GPP, NCP, CR, NCC, DOC, and DIN) was explored by calculating the activation energy (Ea) based on Arrhenius equations (Sibly et al. 2012) within the sea- sonal thermal regime (25.0–32.8°C). The activation ener- gies (Ea in eV) were estimated by fitting a linear regression equation between the natural logarithm of the metabolic rates and the reciprocal of temperature (1/kT), where k is the Boltzmann’s constant (8.62 × 10−5 eV/K) and T is the water temperature (K). To deal with obser- vations ≤0 on log-transformed data (e.g., NCC, DOC, and DIN rates), a constant was added (i.e., ln(rate + 1 − min value(rate)) to shift all values above zero (Legendre and Legendre 2012). The alternative of excluding ≤0 val- ues was discarded because these measurements are an important part of the biological processes under investi- gation (Canavero et al. 2018).

RESULTS

Environmental conditions

Monthly monitored environmental variables at the study site exhibited strong seasonal patterns (Fig. 1b,c, Appendix S1: Table S2). The average seawater tempera- ture ranged from 25.8 � 0.2°C in winter to 32.3 � 0.1°C in summer (Fig. 1c). Simultaneously, aver- age daytime PAR intensities at 5 m water depth increased from 130 � 2 μmol photons�m−2�s−1 to 465 � 14 μmol photons�m−2�s−1 (Fig. 1b). Seawater DIN concentrations were lowest in spring and winter (0.46 � 0.02 and 0.66 � 0.04 μM DIN, respectively) and significantly higher in summer and fall (1.03 � 0.06 and 1.08 � 0.13 μM DIN, respectively, Fig. 1b).

February 2021 FUNCTIONING OF CORAL REEF COMMUNITIES Article e03226; page 5

Functions of coral- and algae-dominated reef communities

The rates observed along the various deployments (Fig. 2) were used to calculate average values over the whole study period (Table 1). Average NCP was 30% and CR 50% lower in coral- as compared to algae-domi- nated communities (NCP, mean � SE: 26.7 � 1.2 and 36.9 � 1.7 mmol C�m−2�h−1, respectively; CR: −10.9 � 0.7 and −20.9 � 1.4 mmol C�m−2�h−1, respec- tively; Fig. 2a,b). Integrated over 1 d, these differences yielded a 40% lower GPP of coral- compared to algae- dominated communities, with GPP/CR ratios of 2.4 � 0.1 and 2.0 � 0.2, respectively (Table 1). NCC in the light was sixfold higher in coral- com-

pared to algae-dominated communities, averaging 7.9 � 0.7 mmol and 1.3 � 0.2 mmol CaCO3�m−2�h−1, respectively. In the dark, coral-dominated communities displayed an NCC of 3.0 � 0.3 mmol CaCO3�m−2�h−1, whereas algae-dominated communities exhibited an NCC of −0.5 � 0.2 mmol CaCO3�m−2�h−1 (represent- ing net CaCO3 dissolution). Integrating the hourly rates over 24 h, corals showed a 10-fold higher NCC com- pared to algae-dominated reef communities (Table 1). Coral-dominated communities were net sources of

DOC during both light and dark incubations, with aver- age fluxes of 0.57 � 0.07 and 0.62 � 0.11 mmol C�m−2- �h−1, respectively (Table 1). In contrast, algae-dominated communities released similar amounts of DOC as corals in the light (0.70 � 0.08 mmol C�m−2�h−1) but were net sinks of DOC in the dark (−0.37 � 0.05 mmol C�m−2- �h−1). When integrated over 24 h, net DOC fluxes in coral-dominated communities were 3.5-fold higher than those in algae-dominated communities (Table 1). Both coral- and algae-dominated communities were net

sources of DIN (1.62 � 0.21 and 1.66 � 0.23 mmol N�m−2�d−1, respectively; Table 1), with no signifi- cant differences between treatments. There were, however, significant differences between light and dark incubations: algae-dominated communities released three times more DIN in the dark as compared to light conditions (109.9 � 14.1 μmol N�m−2�h−1 and 28.5 � 8.7, respec- tively). In contrast, coral-dominated communities released DIN at consistent rates during light and in dark incuba- tions (65.7 � 11.6 and 69.7 � 8.9 μmol N�m−2�h−1).

Temporal variability of reef functions

Both C and N fluxes showed temporal variations; however, significant differences in the magnitude and directions occurred between coral- and algae-dominated reef communities (Fig. 2, Table 1; detailed statistics in Appendix S1: Table S4). Daily-integrated GPP in coral-dominated communi-

ties remained stable throughout the year; however, CR increased by >60% from winter to summer, resulting in 40% lower NCP (Table 1). GPP in algae-dominated communities doubled from winter to summer, resulting in significantly increased NCP that peaked at

>500 mmol C�m−2�d−1 in summer. In both community types, variations in NCP were significantly correlated with seawater temperature (Fig. 3a). NCP of coral-

Dark incubationsLight incubations

–80

0

80

160

240

–5

0

5

10

Ja nu

ar y

M ar

ch M

ay Ju

ly S

ep te

m be

r N

ov em

be r

Ja nu

ar y

Ja nu

ar y

M ar

ch M

ay Ju

ly S

ep te

m be

r N

ov em

be r

Ja nu

ar y

N C

C ( m

m o

l C a C

O m

· h

) D

O C

( m

m o

l C m

·h )

D IN

( μ m

o l N

m · h

) N

C P

a n d

C R

( m

m o

l C m

· h

)

–20

0

20

40

Coral-dominated Algae-dominated

(b)

(c) (d)

(e) (f)

(g) (h)

(a)

Months

–0.7

0

0.7

1.4

FIG 2. Hourly biogeochemical fluxes during light (left panels) and dark (right panels) incubations of coral- and algae-dominated reef communities of the central Red Sea. Presented are all data from bimonthly incubations of (a, b) community metabolism (di- vided into net community production (NCP) and community res- piration (CR), (c, d) net community calcification (NCC), (e, f) net dissolved organic carbon (DOC) fluxes, and (g, h) net dissolved inorganic nitrogen (DIN) fluxes. Dashed lines connect the bimonthly means. Shaded areas connect the standard error around the means. [Color figure can be viewed at wileyonlinelibrary.com]

Article e03226; page 6 FLORIAN ROTH ET AL. Ecology, Vol. 102, No. 2

dominated communities exhibited a negative (r = −0.81, P < 0.0001, n = 26), and NCP of algae-dominated com- munities exhibited a positive (r = 0.83, P < 0.0001, n = 26) relationship with increasing temperature. Hence, coral-dominated communities showed apparent negative Ea values for GPP and NCP, indicating them to be in the falling phase of the performance curve (i.e., past the opti- mum temperature), while apparent Ea for CR was posi- tive, indicating an opposite trend (Appendix S1: Fig. S3). In contrast, algae-dominated communities had positive Ea values for GPP, NCP, and, CR (Table 2), indicating that these communities remained within the rising phase of the performance curve throughout the study period (including the summer months; Appendix S1: Fig. S3). NCC in coral-dominated communities peaked in

spring, fall, and winter, with no differences between these seasons, but dropped sharply by >60% in summer (Table 1). Contrary, NCC in algae-dominated communi- ties was consistently low throughout the year (Fig. 2c,d; Table 1). NCC in coral-dominated communities corre- lated negatively with increasing water temperatures (r = −0.62, P = 0.0005, n = 26; Fig. 3b), negatively with Ωarag (r = −0.36, P = 0.0070, n = 54) (Fig. 4a), and pos- itively with NCP (r = 0.73, P < 0.0001, n = 54; Fig. 4b). NCC of algae-dominated communities did neither corre- late significantly with Ωarag, NCP, nor water tempera- ture. The corresponding apparent Ea values are presented in Table 2.

Net DOC fluxes per day in coral-dominated commu- nities were lowest in winter and doubled in summer (Table 1), owing to both increases in DOC releases dur- ing dark and light incubations. While generally one order of magnitude lower, net DOC fluxes integrated per day remained stable in algae-dominated communities throughout most of the year but increased towards sum- mer (Table 1). Both community types showed a positive correlation between net DOC fluxes and increasing tem- perature (Fig. 3c), which was also reflected in positive Ea values (Table 2, Appendix S1: Fig. S3). DIN fluxes in both coral- and algae-dominated com-

munities were lowest in spring and twofold higher during the rest of the year (Table 1). DIN fluctuations did not correspond to changes in seawater temperature (Fig. 4d) but showed a significant positive relationship (r = 0.57, P = 0.0017, n = 26 for corals; r = 0.39, P = 0.0385, n = 26 for algae) with increasing DIN concentrations of the ambient seawater (Appendix S1: Fig. S2).

DISCUSSION

Phase shifts from corals to algae may drastically change reef community functions

The carbon and carbonate pathways are well described in the literature for undisturbed coral reef communities, both in direction and magnitude

TABLE 1. Seasonal gross primary production (GPP), net community production (NCP), community respiration (CR), the ratio of GPP and CR (GPP/CR), net community calcification (NCC), net dissolved organic carbon (DOC) fluxes, and net dissolved inorganic nitrogen (DIN) fluxes of coral- and algae-dominated reef communities of the central Red Sea.

GPP (mmol C⋅m⁻2⋅d⁻1)

NCP (mmol C⋅m⁻2⋅d⁻1)

CR (mmol C⋅m⁻2⋅d⁻1) GPP/CR

NCC (mmol CaCO₃⋅m⁻2⋅d⁻1)

DOC (mmol C⋅m⁻2⋅d⁻1)

DIN (mmol N⋅m⁻2⋅d⁻1)

Annual mean Coral 581 � 19 320 � 15 −261 � 17 2.4 � 0.1 131 � 10 14.3 � 1.8 1.6 � 0.2 Algae 946 � 51 443 � 21 −503 � 33 2.0 � 0.2 10 � 3 4.0 � 0.9 1.7 � 0.2 P <0.0001* <0.0001* <0.0001* <0.0001* <0.0001* <0.00001* 0.9351

Spring Coral 638 � 21 353 � 12 −285 � 25 2.3 � 0.2 162 � 11 14.1 � 3.6 0.7 � 0.4 Algae 974 � 41 439 � 22 −535 � 27 1.8 � 0.0 7 � 6 1.6 � 1.6 1.3 � 0.4 ∣t∣ <0.0001* 0.0833 <0.0001* 0.0134* <0.0001* 0.0246* 0.9573

Summer Coral 575 � 45 231 � 24 −344 � 26 1.7 � 0.1 61 � 11 18.9 � 3.9 2.2 � 0.2 Algae 1,224 � 13 553 � 12 −671 � 9 1.8 � 0.0 13 � 3 7.1 � 1.9 1.9 � 0.4 ∣t∣ <0.0001* <0.0001* <0.0001* 0.9310 0.0024 0.0432* 0.9979

Fall Coral 447 � 44 292 � 25 −154 � 20 2.9 � 0.1 176 � 7 16.2 � 1.7 2.3 � 0.7 Algae 1,053 � 94 470 � 68 −583 � 43 1.8 � 0.1 15 � 11 2.3 � 3.0 2.3 � 0.7 ∣t∣ <0.0001* 0.0017* <0.0001* <.00001* <0.0001* 0.0111* 1.0000

Winter Coral 597 � 20 390 � 10 −207 � 15 2.9 � 0.1 147 � 10 9.1 � 3.0 1.6 � 0.3 Algae 586 � 30 324 � 17 −262 � 18 2.3 � 0.1 8 � 4 4.1 � 0.4 1.5 � 0.5 ∣t∣ 1.0000 0.3351 0.6166 0.0005 <0.0001* 0.1388 1.0000

Notes: Values represent averages from all incubations during the respective season � SE. Significant differences between treat- ments were assessed by linear mixed models (LMMs), and the differences between treatments and seasons by Tukey’s Honest Sig- nificant Difference (HSD; Appendix S1: Table S4) test. Asterisks highlight significant P values (P or |t| < 0.05).

February 2021 FUNCTIONING OF CORAL REEF COMMUNITIES Article e03226; page 7

(Appendix S1: Table S5). Concomitant to previous find- ings, coral-dominated communities in our experiments displayed: (1) low net community production despite

high gross productivity, implying that biomass accumu- lates slowly (Gattuso et al. 1998); (2) net DOC fluxes, where the accumulation of exudates outpaces its con- sumption (Nelson et al. 2013, Quinlan et al. 2018), rep- resenting an organic C loss that further reduces C available to support net benthic biomass accretion, and promotes an efficient transfer of C via benthic-pelagic coupling (Wild et al. 2004); and (3) high rates of net community calcification, consistent with the accretion of carbonate structures typical for tropical coral reefs (Gat- tuso et al. 1998, 1999b, Atkinson and Falter 2003). Yet, our seasonal comparative in situ approach revealed that functions of reef ecosystems change with shifts from a coral- to an algae-dominated benthic community. Specifically, algae-dominated communities displayed a

higher organic C metabolism, along with residual net calcification. Significantly higher GPP and >40% ele- vated NCP in algae-dominated communities indicate a greater potential for autotrophic biomass accumulation per planar square meter of reef (Fong and Paul 2011, Kelly et al. 2017), despite a lower GPP/CR ratio (Table 1). The lower GPP/CR ratio in algae communi- ties compared to coral communities seems counterintu- itive considering the dominant organisms (algae as autotrophs compared to mixotrophic corals); however, algae-dominated reef structures host numerous

N C

C ( m

m o

l C a C

O m

· d

) 26 28 30 32

(b)

0

80

160

240 ·d )

26 28 30 32 32

(c)

0

20

40 N

C P

( m

m o

l C m

· d

)

26 28 30 32

(a) ·d )

26 28 30

(d)

0

2

4

Coral-dominated Algae-dominated

200

400

600

FIG 3. Relationship of temperature with (a) net community production (NCP), (b) net community calcification (NCC), (c) dis- solved organic carbon (DOC) fluxes, and (d) dissolved inorganic nitrogen (DIN) fluxes in coral- and algae-dominated reef commu- nities. Rates represent the net flux integrated over a 24-h day (assuming 12 h of light and 12 h of dark). Solid lines represent the linear regressions; shaded areas in transparent colors represent the 95% confidence intervals. [Color figure can be viewed at wileyon linelibrary.com]

TABLE 2. Apparent activation energies (Ea, eV) for coral- and algae-dominated reef communities as the slope of the Arrhenius relationship between the natural logarithm of specific metabolic rates and the inverted temperature (1/kT).

GPP NCP CR NCC DOC DIN

Coral Algae Coral Algae Coral Algae Coral Algae Coral Algae Coral Algae

Ea (eV) −0.09 0.75 −0.58 0.55 0.49 0.93 −0.82 0.65 0.73 0.19 0.12 0.13 r2 0.03 0.71 0.54 0.59 0.26 0.66 0.30 0.10 0.28 0.01 0.02 0.02 P 0.3861 <0.0001 <0.0001 <0.0001 0.0065 <0.0001 0.0023 0.1073 0.0038 0.5619 0.5126 0.4439

Notes: GPP = gross primary production (mmol C�m−2�d−1); NCP = net community production (mmol C�m−2�d−1); CR = com- munity respiration (mmol C�m−2�d−1); NCC = net community calcification = (mmol CaCO3�m−2�d−1); DOC = dissolved organic carbon fluxes (mmol C�m−2�d−1); DIN = dissolved inorganic nitrogen fluxes (mmol N�m−2�d−1); r2 = square of correlation coeffi- cient.

–5

0

5

10

–30 0 30 60

NCP (mmol C m ·h )

N C

C ( m

m o

l C a C

O m

·h )

– 5

0

5

10

3 4 5 6

N C

C ( m

m o

l C a C

O m

·h )

arag

(a) (b)

Coral-dominated Algae-dominated

FIG 4. Relationship between (a) net community calcifica- tion (NCC) and aragonite saturation state (Ωarag), and (b) NCC and net community production (NCP) in coral- and algae-dom- inated reef communities. Closed circles indicate measurements from light incubations, and open circles indicate dark incuba- tions. Solid lines represent the linear regressions; shaded areas in transparent colors represent the 95% confidence intervals. [Color figure can be viewed at wileyonlinelibrary.com]

Article e03226; page 8 FLORIAN ROTH ET AL. Ecology, Vol. 102, No. 2

heterotrophs that rely on the high algal biomass produc- tion, fueling community-wide respiration by direct her- bivory (Klumpp and McKinnon 1989, Russ 2003) and indirect detritivory by invertebrates (Kramer et al. 2013). Likewise, sponges and other filter feeders are commonly associated with degraded reef habitats (Abele and Patton 1976), feeding on DOC or algal debris (Rix et al. 2017, 2018). In addition, heterotrophic bacteria within algal communities remineralize labile DOC released by algae (Nelson et al. 2011, Haas et al. 2013). Thereby, the consumption of algal-derived C (i.e., the C retention within the system) can shorten the average trophic path length, and reduce the average trophic level of the second-order consumers (reviewed in Johnson et al. 1995). In support of these trophic interactions, we observed an apparent consumption of DOC in the dark in algae-dominated communities, limiting the net DOC flux integrated over 24 h (i.e., release in the light bal- anced by consumption in the dark). Our results indicate that algae-associated organisms readily remineralize DOC, concurring with demonstrated DOC depletion in algae-dominated shallow reefs elsewhere (Nelson et al. 2011, Haas et al. 2016). Along with alterations of the organic C cycle, calcifi-

cation was reduced within algae-dominated communi- ties. The slope of the relationship between NCP and NCC is commonly used as an indicator of reef health, indicative for the relative proportion of calcifying to noncalcifying organisms in a benthic community (Alb- right et al. 2013, Takeshita et al. 2016). We observed a slope of 0.18 in coral- and 0.03 in algae-dominated com- munities, and the slope averages 0.22 based on 52 reefs around the world (Gattuso et al. 1999a). The significant difference highlights the shift from calcifying corals to noncalcifying organisms and a decoupling of the organic carbon (production vs. respiration) and carbonate (calci- fication vs. dissolution) cycles in algae-dominated com- munities (McMahon et al. 2019). Unraveling nitrogen pathways in coral reefs is crucial

to understand how high productivity is supported despite low ambient nutrient concentrations (D’Elia and Wiebe 1990, Szmant 2002, Atkinson and Falter 2003). Both coral- and algae-dominated communities were net sources of DIN over the study period with no significant differences between community types. The flux rates generally fell within the published range of in situ mea- surements (−4 to 5 mmol N as NOx�m−2�d−1; reviewed in Atkinson and Falter 2003). They are in stark contrast, however, with the expectation that net autotrophic com- munities would act as sinks for dissolved inorganic nutri- ents, as generally measured in single organism incubations (e.g., Den Haan et al. 2016). Although a the- oretical N requirement of 34–83 mmol N�m−2�d−1 to support NCP can be expected (stoichiometric calcula- tions with NCP ranging from 230 to 550 mmol C�m−2- �d−1; assuming a C/N ratio of 6.6; see Redfield 1958), the effects of assimilation were likely masked by concurrent community-wide processes that produce DIN (Gruber

et al. 2019). For example, cavities within Red Sea reefs can be considerable sources of DIN (>20 mmol N�m−2- �d−1) as sponges and other filter feeders utilize dissolved organic matter (Richter et al. 2001). Also, microbial communities can consume and transform organic N compounds (Yahel et al. 2003, Moulton et al. 2016, Pfis- ter and Altabet 2019), potentially increasing the commu- nity-wide DIN release into the environment. Other pathways, such as N2 fixation (Cardini et al. 2016) or heterotrophic feeding on particulates (Ribes et al. 2003, Houlbrèque and Ferrier-Pagès 2009) are additional N sources that potentially limit/mask the N uptake from DIN. Overall, algae-dominated communities displayed a

higher potential of biomass accumulation or export (i.e., high NCP), associated with a higher total amount of C available (i.e., high GPP) to the ecosystem. The high NCP of algae communities facilitates rapid lateral and vegetative overgrowth of bare substrates (Diaz-Pulido and Garzón-Ferreira 2002, Roth et al. 2018). At the same time, limited reef accretion (i.e., low NCC) within algal habitats may compromise the topographic com- plexity of phase-shifted reef communities (Wild et al. 2011), limit the recruitment of corals (Harrington et al. 2004, Roth et al. 2017, 2018), and increase reef erosion (Adey 1978).

High temperatures during summer amplify functional differences between coral- and algae-dominated

communities

Our data highlight that functions related to the carbon and carbonate cycles of coral- and algae-dominated communities are strongly but inversely affected by tem- perature (summarized in Fig. 5), with implications for their response to warming. The results from activation energies show a higher

sensitivity to thermal stress of coral-dominated com- pared to algae-dominated communities during summer (Table 2, Appendix S1: Fig. S3). Particularly, apparent activation energies for NCP and NCC of coral-domi- nated communities (Ea = −0.58 and −0.82 eV; respec- tively) were in the falling phase of the performance curves, and thus, past the optimum temperature to peak rates. Thereby, the community metabolism of corals is pushed toward carbon losses because of high CR relative to GPP in summer. Likewise, the activation energy of CR was positive, as respiration continued to increase with temperature through the seasonal thermal range. Both the lower ratio of GPP to CR (GPP/CR) and reduced NCP integrated over diel cycles indicate that coral-dominated communities shifted toward a more heterotrophic state with increasing temperature. It is thus apparent that summer temperatures exceeded the metabolic optima for coral-dominated communities, which was previously suggested for individuals of Pocil- lopora verrucosa (Sawall et al. 2015, Roik et al. 2016, Anton et al. 2020) and Stylophora pistillata (Anton et al.

February 2021 FUNCTIONING OF CORAL REEF COMMUNITIES Article e03226; page 9

2020) in the central Red Sea. This contrasts with many reef locations worldwide, where primary production maxima are typically observed during the warmest months of the year (e.g., Scheufen et al. 2017). At the same time, coral-dominated communities displayed enhanced rates of net DOC fluxes with warming (Ea = 0.73 eV), which can be attributed to an increased release of cellular matter and/or mucoid exudates during thermal stress in corals (Niggl et al. 2009, Scheufen et al. 2017). Although mucus released during higher tempera- tures may help to protect corals against pathogens (Glasl et al. 2016) or high UV radiation (Gleason and Welling- ton 1993), it poses an increased loss of organic C that can be used for community growth and/or export at con- stant GPP rates (Fig. 5). Along with this trend, NCC

dropped by >50% from the annual mean during sum- mer, with most of this decline realized as waters warmed from 30 to 32°C. Overall, decreased NCC strongly corre- lated with decreased NCP as temperatures increased (as revealed by high negative apparent activation energies of NCC, Ea = −0.82 eV), with a temperature threshold at around 30.5°C (Appendix S1: Fig. S3). This thermal threshold is near the reported thermal optimum of Pocil- lopora verrucosa and Stylophora pistillata for gross pri- mary production (29.9 and 31.9°C, respectively; Anton et al. 2020) in the central Red Sea, indicating a strong thermal sensitivity of coral-dominated communities soon after the thermal optimum is exceeded. Accord- ingly, in the temperature range above the optimum, the rate of calcification decreased despite increased Ωarag in

FIG 5. In situ community metabolism of natural coral- and algae-dominated reef communities in the central Red Sea, Saudi Arabia. Schematic was derived from all data available in the given lower (blue) and upper (red) temperature ranges. Organic carbon pathways refer to photosynthesis, respiration, and dissolved organic carbon (DOC) fluxes, and the inorganic carbon pathway refers to the formation and dissolution of calcium carbonate. The thicknesses of the bars scale with the actual flux measurements from in situ incubations. [Color figure can be viewed at wileyonlinelibrary.com]

Article e03226; page 10 FLORIAN ROTH ET AL. Ecology, Vol. 102, No. 2

summer (Fig. 4a; Silverman et al. 2007). Physiological factors can also strongly affect the biomineralization process. As calcification mainly depends on the photo- synthetic efficiency of the endosymbionts within corals (Gattuso et al. 1999a, Allemand et al. 2004), a lower NCC may occur for thermally stressed corals, limiting reef accretion and stabilization (Jokiel and Coles 1977, De’ath et al. 2009). In contrast, positive activation energies for algae-

dominated communities (Ea = 0.93, 0.75, and 0.55 eV for CR, GPP, and NCP; respectively) indicate that these communities benefit from higher temperatures and that thermal optima were not reached in summer. In fact, some macroalgae species (Halimeda tuna) from the cen- tral Red Sea have a reported thermal optimum (31.7°C) for gross primary production (Anton et al. 2020) that is close to the maximum summer temperature recorded during our incubations (32.5°C). The increases in GPP and CR along a thermal gradient in algae-dominated communities highlight a higher turnover of organic C. However, increases in C fixation outweighed increases in respiratory C consumption, resulting in higher NCP in the summer. In contrast to previous reports (e.g., Barron et al. 2014), DOC fluxes in algae-dominated communi- ties showed only a weak temperature dependence. How- ever, although the overall net DOC fluxes remained relatively stable, differences in net production during the light and net consumption in the dark amplified with temperature (Fig. 3), limiting losses of organic C through this process. As a result, more organic C was retained within algae-dominated communities in sum- mer, supporting biomass accumulation and export in the community (Fig. 5).

Implications for reef ecosystem functioning under global change

Theoretical studies have provided a sound under- standing of the relationship between biodiversity loss and ecosystem functions in tropical coral reefs (reviewed in Hughes et al. 2017). However, considerable knowledge gaps remain, in particular, on how metabolic and bio- geochemical processes differ between coral- and algae- dominated communities, and how these respond to sea- sonal fluctuations in environmental conditions. As algal cover is expected to increase in coral reefs, our long-term in situ experiments reveal how these novel communities in general, and how thermal stress in particular, may alter pivotal ecological functions of future reefs. Our data show that fundamental metabolic and biogeochem- ical characteristics of coral-dominated communities are disturbed by shifts from coral to algal dominance and may, thereby, compromise the future stability and resili- ence of coral reef biota. These responses may be further compounded by differential thermal responses between coral and algae species (e.g., Anton et al. 2020). The sensitivity of corals and their symbionts to rising

temperatures has been documented extensively (Hoegh-

Guldberg 1999). Thermal anomalies exceeding 1–2°C above the mean summer maximum temperature can compromise the symbiosis (e.g., Weeks et al. 2008), lead- ing to coral bleaching and reduced coral survival (Baird and Marshall 2002, Baker et al. 2008). However, our study did not record temperatures exceeding the local mean summer maxima reported for the region (see Fig. 1c; Chaidez et al. 2017) and, likewise, no apparent signs of coral bleaching were observed. Nevertheless, growth of Pocillopora verrucosa and Stylophora pistillata (Anton et al. 2020) and calcification of Pocillopora verru- cosa (Roik et al. 2018) are already reduced under current summer conditions in the Red Sea, as also highlighted by the present study. For the warmer part of the year, algae-dominated communities have, thus, a metabolic advantage over coral-dominated communities because they maintain high NCP. Importantly, the consequences of global warming may manifest not only in terms of higher-than-normal temperatures but also in a longer- than-normal duration of the seasonal peak temperatures (Fitt et al. 2001). Under such conditions, if coral mortal- ity events occur, algae may quickly spread and shift coral ecosystems more rapidly towards systems that are domi- nated by algae (McManus et al. 2019), as already reported on some reefs in the southern Red Sea follow- ing the coral bleaching event in 2015 (Anton et al. 2020). The frequency and intensity of climate-driven stress

events on coral reefs will inevitably aggravate in the near future. Our results suggest that the anticipated increase in the spatial footprint of algae-dominated communities would exacerbate the magnitude of the functional changes described here. Ocean warming likely enhances the competitive advantage of algae- over coral-domi- nated communities (Anton et al. 2020), thus promoting a positive feedback loop of reef degradation. Similar effects of warming are likely to be operational in other temperature-sensitive and calcifying communities. As these organisms are central to the formation of reef ecosystems, critical changes in the biodiversity and func- tioning may be witnessed. Thus, appropriate manage- ment practices designed to limit the proliferation of algae are needed for maintaining reefs dominated by corals and the important ecosystem services they sup- port.

ACKNOWLEDGMENTS

We are grateful to the personnel from the King Abdullah University of Science and Technology (KAUST) Coastal and Marine Resources Core (CMOR) Laboratory for logistical sup- port. The authors would also like to acknowledge Rodrigo Vil- lalobos and João Cúrdia, who helped during fieldwork. Figure 5 was produced by Xavier Pita, scientific illustrator at KAUST. We would like to thank the two anonymous reviewers and the editor for their helpful suggestions and comments. The research was supported by KAUST baseline funding to BHJ and by grant Wi 2677/9-1 from the German Research Founda- tion (DFG) to CW. Author contributions: FR, CW, and SC conceptualized and designed research. FR, LS, MLC, and VS performed research. FR, NR, VS, AA, LS, BK, and MLC

February 2021 FUNCTIONING OF CORAL REEF COMMUNITIES Article e03226; page 11

analyzed data. CMD, XAGM, CRV, and BHJ contributed to research materials, logistics and to interpreting data. FR wrote original draft of the manuscript with support by CW. All authors read and approved the final manuscript.

LITERATURE CITED

Abele, L. G., and W. K. Patton. 1976. The size of coral heads and the community biology of associated decapod crus- taceans. Journal of Biogeography 3:35.

Adey, W. H. 1978. Coral reef morphogensis: a multidimensional model. Science 202(4370): 831–837.

Albright, R., C. Langdon, and K. R. N. Anthony. 2013. Dynamics of seawater carbonate chemistry, production, and calcification of a coral reef flat, Central Great Barrier Reef. Biogeosciences Discussions 10:7641–7676.

Allemand, D., C. Ferrier-Pagès, P. Furla, F. Houlbrèque, S. Puverel, S. Reynaud, É. Tambutté, S. Tambutté, and D. Zoc- cola. 2004. Biomineralisation in reef-building corals: from molecular mechanisms to environmental control. Comptes Rendus Palevol 3:453–467.

Anthony, K. R. N., D. I. Kline, G. Diaz-Pulido, S. Dove, and O. Hoegh-Guldberg. 2008. Ocean acidification causes bleaching and productivity loss in coral reef builders. Proceedings of the National Academy of Sciences of the United States of Amer- ica 105:17442–17446.

Anton, A., J. L. Randle, F. C. Garcia, S. Rossbach, J. I. Ellis, M. Weinzierl, and C. M. Duarte. 2020. Differential thermal toler- ance between algae and corals may trigger the proliferation of algae in coral reefs. Global Change Biology 26:4316–4327.

Ateweberhan, M., J. H. Bruggemann, and A. M. Breeman. 2006. Effects of extreme seasonality on community structure and functional group dynamics of coral reef algae in the southern Red Sea (Eritrea). Coral Reefs 25:391–406.

Atkinson, M. J., and J. L. Falter. 2003. Coral reefs. in K. Black and G. Shimmieldeditors. Pages 40–64 Biogeochemistry of Marine Systems, Boca Raton, Florida: CRC Press.

Baird, A. H., and P. A. Marshall. 2002. Mortality, growth and reproduction in scleractinian corals following bleaching on the Great Barrier Reef. Marine Ecology Progress Series 237:133–141.

Baird, D., S. Brown, L. Lagadic, M. Liess, L. Maltby, M. Mor- eira-Santos, R. Schulz, and G. Scott. 2007. In situ–based effects measures: determining the ecological relevance of mea- sured responses. Integrated Environmental Assessment and Management 3:259–267.

Baker, A. C., P. W. Glynn, and B. Riegl. 2008. Climate change and coral reef bleaching: an ecological assessment of long- term impacts, recovery trends and future outlook. Estuarine, Coastal and Shelf Science 80:435–471.

Balvanera, P., A. B. Pfisterer, N. Buchmann, J. S. He, T. Naka- shizuka, D. Raffaelli, and B. Schmid. 2006. Quantifying the evidence for biodiversity effects on ecosystem functioning and services. Ecology Letters 9:1146–1156.

Barron, C., E. T. Apostolaki, and C. M. Duarte. 2014. Dis- solved organic carbon fluxes by seagrass meadows and macroalgal beds. Frontiers in Marine Science 1:42.

Bellwood, D. R., T. P. Hughes, C. Folke, and M. Nyström. 2004. Confronting the coral reef crisis. Nature 429:827–833.

Bellworthy, J., and M. Fine. 2018. The Red Sea Simulator: A high-precision climate change mesocosm with automated monitoring for the long-term study of coral reef organisms. Limnology and Oceanography: Methods 16:367–375.

Bender, D., G. Diaz-Pulido, and S. Dove. 2014. Warming and acidification promote cyanobacterial dominance in turf algal assemblages. Marine Ecology Progress Series 517:271–284.

Bernstein, W. N., K. A. Hughen, C. Langdon, D. C. McCorkle, and S. J. Lentz. 2016. Environmental controls on daytime net community calcification on a Red Sea reef flat. Coral Reefs 35:697–711.

Brondizio, E., J. Settele, S. Diaz, and T. Ngo. 2019. Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science—Policy Platform on Biodiver- sity and Ecosystem Services. IPBES Secretariat, Bonn, Ger- many.

Bythell, J. C., Z. M. Hillis-Starr, and C. S. Rogers. 2000. Local variability but landscape stability in coral reef communities following repeated hurricane impacts. Marine Ecology Pro- gress Series 204:93–100.

Canavero, A., M. Arim, F. Pérez, F. M. Jaksic, and P. A. Mar- quet. 2018. A metabolic view of amphibian local community structure: the role of activation energy. Ecography 41:388–400.

Cardini, U., V. N. Bednarz, N. van Hoytema, A. Rovere, M. S. Naumann, M. M. D. Al-Rshaidat, and C. Wild. 2016. Budget of primary production and dinitrogen fixation in a highly sea- sonal Red Sea coral reef. Ecosystems 19:771–785.

Chaidez, V., D. Dreano, S. Agusti, C. M. Duarte, and I. Hoteit. 2017. Decadal trends in Red Sea maximum surface tempera- ture. Scientific Reports 7:1–8.

D’Elia, C. F., and W. J. Wiebe. 1990. Biogeochemical nutrient cycles in coral reef ecosystems. Pages 49–74 in Z. Dubinsky, editor. Ecosystems of the world 25: coral reefs. Elsevier, New York, New York, USA.

Damgaard, C. 2019. A critique of the space-for-time substitu- tion practice in community ecology. Trends in Ecology and Evolution 34:416–421.

de Jongh, F., and F. C. Van Duyl. 2004. The impact of suspen- sion feeders in cryptic habitats on the trophodynamics of coral reefs. Thesis. http://www.science.uva.nl/ZMA/Inverteb rates/Coel/scirep/Cryptic_Reef_Habitats.pdf

De’ath, G., J. M. Lough, and K. E. Fabricius. 2009. Declining coral calcification on the Great Barrier Reef. Science 323:116–119.

Den Haan, J., et al. 2016. Nitrogen and phosphorus uptake rates of different species from a coral reef community after a nutrient pulse. Scientific Reports 6:1–13.

Diaz-Pulido, G., and J. Garzón-Ferreira. 2002. Seasonality in algal assemblages on upwelling-influenced coral reefs in the Colombian Caribbean. Botanica Marina 45:284–292.

Done, T. J. 1992. Phase shifts in coral reef communities and their ecological significance. in V. Jaccarini and E. Martens editors. Pages 121–132 Book: The ecology of mangrove and related ecosystems, Dordrecht, Netherlands: Springer Inter- national Publishing.

Done, T. J., P. K. Dayton, A. E. Dayton, and R. Steger. 1991. Regional and local variability in recovery of shallow coral communities: Moorea, French Polynesia and central Great Barrier Reef. Coral Reefs 9:183–192.

Edmunds, P. J. 2002. Long-term dynamics of coral reefs in St. John, US Virgin Islands. Coral Reefs 21:357–367.

Edmunds, P. J., S. S. Doo, and R. C. Carpenter. 2020. Year-long effects of high pCO2 on the community structure of a tropical fore reef assembled in outdoor flumes. ICES Journal of Mar- ine Science 77:1055–1065.

Fitt, W., B. Brown, M. Warner, and R. Dunne. 2001. Coral bleaching: interpretation of thermal tolerance limits and ther- mal thresholds in tropical corals. Coral Reefs 20:51–65.

Follstad Shah, J. J., et al. 2017. Global synthesis of the tempera- ture sensitivity of leaf litter breakdown in streams and rivers. Global Change Biology 23:3064–3075.

Fong, P., and V. J. Paul. 2011. Coral reef algae. in Z. Dubinsky and N Stambler editors. Pages 241–272 Book: Coral reefs: an

Article e03226; page 12 FLORIAN ROTH ET AL. Ecology, Vol. 102, No. 2

ecosystem in transition, Dordrecht, Netherlands: . Springer International Publishing.

Frölicher, T. L., E. M. Fischer, and N. Gruber. 2018. Marine heatwaves under global warming. Nature 560:360–364.

Garcı́a, F. C., E. Bestion, R. Warfield, and G. Yvon-Durocher. 2018. Changes in temperature alter the relationship between biodiversity and ecosystem functioning. Proceedings of the National Academy of Sciences of the United States of Amer- ica 115:10989–10994.

Gattuso, J. P., D. Allemand, and M. Frankignoulle. 1999a. Pho- tosynthesis and calcification at cellular, organismal and com- munity levels in coral reefs: a review on interactions and control by carbonate chemistry. American Zoologist 39:160–183.

Gattuso, J. P., M. Frankignoulle, and S. V. Smith. 1999b. Mea- surement of community metabolism and significance in the coral reef CO2 source-sink debate. Proceedings of the National Academy of Sciences of the United States of Amer- ica 96:13017–13022.

Gattuso, J. P., M. Frankignoulle, and R. Wollast. 1998. Carbon and carbonate metabolism in coastal aquatic ecosystems. Annual Review of Ecology and Systematics 29:405–434.

Glasl, B., G. J. Herndl, and P. R. Frade. 2016. The microbiome of coral surface mucus has a key role in mediating holobiont health and survival upon disturbance. ISME Journal 10:2280–2292.

Gleason, D. F., and G. M. Wellington. 1993. Ultraviolet radia- tion and coral bleaching. Nature 365:836–838.

Graham, N. A. J., S. Jennings, M. A. MacNeil, D. Mouillot, and S. K. Wilson. 2015. Predicting climate-driven regime shifts versus rebound potential in coral reefs. Nature 518:1–17.

Gruber, R. K., R. J. Lowe, and J. L. Falter. 2019. Tidal and sea- sonal forcing of dissolved nutrient fluxes in reef communities. Biogeosciences 16 1921–1935. https://doi.org/10.5194/bg-16- 1921-2019.

Haas, A. F., et al. 2016. Global microbialization of coral reefs. Nature Microbiology 1: 16042.

Haas, A. F., C. E. Nelson, F. Rohwer, L. Wegley-Kelly, S. D. Quistad, C. A. Carlson, J. J. Leichter, M. Hatay, and J. E. Smith. 2013. Influence of coral and algal exudates on micro- bially mediated reef metabolism. PeerJ 1:e108.

Harrington, L., K. Fabricius, G. D’eath, A. Negri, L. I. H. Arrington, K. A. F. Abricius, and G. L. D. E. Ath. 2004. Recognition and selection of settlement substrata determine post-settlement survival in corals. Ecology 85:3428–3437.

Hempson, T. N., N. A. J. Graham, M. A. Macneil, A. S. Hoey, and S. K. Wilson. 2018. Ecosystem regime shifts disrupt trophic structure. Ecological Applications 28:191–200.

Hoegh-Guldberg, O. 1999. Climate change, coral bleaching and the future of the world’s coral reefs. Marine and Freshwater Research 50:839–866.

Hooper, D. U., E. C. Adair, B. J. Cardinale, J. E. K. Byrnes, B. A. Hungate, K. L. Matulich, A. Gonzalez, J. E. Duffy, L. Gamfeldt, and M. I. Connor. 2012. A global synthesis reveals biodiversity loss as a major driver of ecosystem change. Nat- ure 486:105–108.

Houlbrèque, F., and C. Ferrier-Pagès. 2009. Heterotrophy in tropical scleractinian corals. Biological Reviews 84:1–17.

Hughes, T. P., et al. 2017. Coral reefs in the Anthropocene. Nat- ure 546:82–90.

Hughes, T. P., et al. 2019. Ecological memory modifies the cumulative impact of recurrent climate extremes. Nature Cli- mate Change 9:40–43.

Hughes, T. P., M. J. Rodrigues, D. R. Bellwood, D. Ceccarelli, O. Hoegh-Guldberg, L. McCook, N. Moltschaniwskyj, M. S. Pratchett, R. S. Steneck, and B. Willis. 2007. Phase shifts,

herbivory, and the resilience of coral reefs to climate change. Current Biology 17:360–365.

Jessen, C., J. F. Villa Lizcano, T. Bayer, C. Roder, M. Aranda, C. Wild, and C. R. Voolstra. 2013. In-situ effects of eutrophi- cation and overfishing on physiology and bacterial diversity of the red sea coral Acropora hemprichii. PLoS One 8:e62091.

Johnson, C., D. Klumpp, J. Field, and R. Bradbury. 1995. Car- bon flux on coral reefs: effects of large shifts in community structure. Marine Ecology Progress Series 126:123–143.

Jokiel, P. L., and S. L. Coles. 1977. Effects of temperature on the mortality and growth of Hawaiian reef corals. Marine Biology 43:201–208.

Kelly, E. L. A., Y. Eynaud, I. D. Williams, R. T. Sparks, M. L. Dai- ler, S. A. Sandin, and J. E. Smith. 2017. A budget of algal pro- duction and consumption by herbivorous fish in an herbivore fisheries management area, Maui. Hawaii. Ecosphere 8:e01899.

Klumpp, W., and A. McKinnon. 1989. Community structure, biomass and productivity of epilithic algal communities on the Great Barrier Reef: dynamics at different spatial scales. Marine Ecology Progress Series 86:77–89.

Koch, M., G. Bowes, C. Ross, and X. H. Zhang. 2013. Climate change and ocean acidification effects on seagrasses and mar- ine macroalgae. Global Change Biology 19:103–132.

Kramer, M. J., O. Bellwood, and D. R. Bellwood. 2013. The trophic importance of algal turfs for coral reef fishes: the crustacean link. Coral Reefs 32:575–583.

Langdon, C., W. S. Broecker, D. E. Hammond, E. Glenn, K. Fitzsimmons, S. G. Nelson, T.-H. Peng, I. Hajdas, and G. Bonani. 2003. Effect of elevated CO2 on the community metabolism of an experimental coral reef. Global Biogeo- chemical Cycles 17:1011.

Lavigne, H., and J. P. Gattuso. 2013. Package “seacarb”: seawa- ter carbonate chemistry with R, v. 2.4. 8. http://www2.uaem. mx/r-mirror/web/packages/seacarb/seacarb.pdf

Legendre, P., and L. Legendre. 2012. Numerical ecology. Third edition. Elsevier, Amsterdam, The Netherlands.

Lirman, D., and P. Biber. 2000. Seasonal dynamics of macroal- gal communities of the northern Florida reef tract. Botanica Marina 43:305–314.

Lough, J. M., K. D. Anderson, and T. P. Hughes. 2018. Increas- ing thermal stress for tropical coral reefs: 1871–2017. Scien- tific Reports 8:6079.

Maldonado, M., M. Ribes, and F. C. van Duyl. 2012. Nutrient fluxes through sponges. Biology, budgets, and ecological implications. Advances in Marine Biology 62:113–182.

McMahon, A., I. R. Santos, T. Cyronak, and B. D. Eyre. 2013. Hysteresis between coral reef calcification and the seawater aragonite saturation state. Geophysical Research Letters 40:4675–4679.

McMahon, A., I. R. Santos, K. G. Schulz, A. Scott, J. Silver- man, K. L. Davis, and D. T. Maher. 2019. Coral reef calcifica- tion and production after the 2016 bleaching event at Lizard Island, Great Barrier reef. Journal of Geophysical Research Oceans 124:4003–4016.

McManus, L. C., V. V. Vasconcelos, S. A. Levin, D. M. Thomp- son, J. A. Kleypas, F. S. Castruccio, E. N. Curchitser, and J. R. Watson. 2019. Extreme temperature events will drive coral decline in the Coral Triangle. Global Change Biology 26:2120–2133.

Middleton, B., and J. Grace. 2004. Biodiversity and ecosystem functioning: synthesis and perspectives. Restoration Ecology 12:611–612.

Moberg, F. F., and C. Folke. 1999. Ecological goods and ser- vices of coral reef ecosystems. Ecological Economics 29:215–233.

Morán, X. A. G., J. M. Gasol, M. C. Pernice, J. Mangot, R. Massana, E. Lara, D. Vaqué, and C. M. Duarte. 2017.

February 2021 FUNCTIONING OF CORAL REEF COMMUNITIES Article e03226; page 13

Temperature regulation of marine heterotrophic prokaryotes increases latitudinally as a breach between bottom-up and top-down controls. Global Change Biology 23:3956–3964.

Moulton, O. M., M. A. Altabet, J. M. Beman, L. A. Deegan, J. Lloret, M. K. Lyons, J. A. Nelson, and C. A. Pfister. 2016. Microbial associations with macrobiota in coastal ecosys- tems: patterns and implications for nitrogen cycling. Frontiers in Ecology and the Environment 14:200–208.

Naeem, S., J. E. Duffy, and E. Zavaleta. 2012. The functions of biological diversity in an age of extinction. Science 336:1401–1406.

Nelson, C. E., A. L. Alldredge, E. A. McCliment, L. A. Amaral-Zettler, and C. A. Carlson. 2011. Depleted dissolved organic carbon and distinct bacterial communities in the water column of a rapid-flushing coral reef ecosystem. ISME Journal 5:1374–1387.

Nelson, C. E., S. J. Goldberg, L. Wegley Kelly, A. F. Haas, J. E. Smith, F. Rohwer, and C. A. Carlson. 2013. Coral and macroalgal exudates vary in neutral sugar composition and differentially enrich reef bacterioplankton lineages. ISME Journal 7:962–979.

Niggl, W., M. Glas, C. Laforsch, Mayr, C., and Wild, C. 2009. First evidence of coral bleaching stimulating organic matter release by reef corals. Pages 905–910inB. Riegl, editor. Pro- ceedings of the 11th International Coral Reef Symposium. https://www.researchgate.net/profile/Martin_Gutbrod2/publi cation/257233311_First_evidence_of_coral_bleaching_stimu lating_organic_matter_release_by_reef_corals/links/56938980 08aec14fa55e9224/First-evidence-of-coral-bleaching-stimula ting-organic-matter-release-by-reef-corals.pdf.

Oliver, E. C. J., et al. 2018. Longer and more frequent marine heatwaves over the past century. Nature Communications 9:1–12.

Padfield, D., C. Lowe, A. Buckling, R. Ffrench-Constant, S. Jennings, F. Shelley, J. S. Ólafsson, and G. Yvon-Durocher. 2017. Metabolic compensation constrains the temperature dependence of gross primary production. Ecology Letters 20:1250–1260.

Pfister, C. A., and M. A. Altabet. 2019. Enhanced microbial nitrogen transformations in association with macrobiota from the rocky intertidal. Biogeosciences 16:193–206.

Pierrot, D., E. Lewis, and D. W. R. Wallace. 2006. MS Excel program developed for CO2 system calculations. Oak Ridge, Tennessee, USA. https://www.ncei.noaa.gov/access/ocean-ca rbon-data-system/oceans/CO2SYS/co2rprt.html.

Piggot, A. M., B. W. Fouke, M. Sivaguru, R. A. Sanford, and H. R. Gaskins. 2009. Change in zooxanthellae and mucocyte tissue density as an adaptive response to environmental stress by the coral, Montastraea annularis. Marine Biology 156:2379–2389.

Quinlan, Z. A., K. Remple, M. D. Fox, N. J. Silbiger, T. A. Oli- ver, H. M. Putnam, L. Wegley Kelly, C. A. Carlson, M. J. Donahue, and C. E. Nelson. 2018. Fluorescent organic exu- dates of corals and algae in tropical reefs are compositionally distinct and increase with nutrient enrichment. Limnology and Oceanography Letters 3:331–340.

Redfield, A. C. 1958. The biological control of chemical factors in the environment. American Scientist 46:230A–A221.

Reynaud, S., N. Leclercq, S. Romaine-Lioud, C. Ferrier-Pagès, J. Jaubert, and J. P. Gattuso. 2003. Interacting effects of CO2 partial pressure and temperature on photosynthesis and calci- fication in a scleractinian coral. Global Change Biology 9:1660–1668.

Ribes, M., R. Coma, M. Atkinson, and R. Kinzie. 2003. Parti- cle removal by coral reef communities: picoplankton is a major source of nitrogen. Marine Ecology Progress Series 257:13–23.

Richter, C., and M. Wunsch. 1999. Cavity-dwelling suspension feeders in coral reefs—A new link in reef trophodynamics. Marine Ecology Progress Series 188:105–116.

Richter, C., M. Wunsch, M. Rasheed, I. Kötter, and M. I. Bad- ran. 2001. Endoscopic exploration of Red Sea coral reefs reveals dense populations of cavity-dwelling sponges. Nature 413:726–730.

Rix, L., V. N. Bednarz, U. Cardini, N. Van Hoytema, F. A. Al- Horani, C. Wild, and M. S. Naumann. 2015. Seasonality in dinitrogen fixation and primary productivity by coral reef framework substrates from the northern Red Sea. Marine Ecology Progress Series 533:79–92.

Rix, L., J. M. De Goeij, D. Van Oevelen, U. Struck, F. A. Al- Horani, C. Wild, and M. S. Naumann. 2018. Reef sponges facilitate the transfer of coral-derived organic matter to their associated fauna via the sponge loop. Marine Ecology Pro- gress Series 589:85–96.

Rix, L., et al. 2017. Differential recycling of coral and algal dis- solved organic matter via the sponge loop. Functional Ecol- ogy 31:778–789.

Roik, A., C. Roder, T. Röthig, and C. R. Voolstra. 2016. Spatial and seasonal reef calcification in corals and calcareous crusts in the central Red Sea. Coral Reefs 35:681–693.

Roik, A., T. Röthig, C. Pogoreutz, V. Saderne, and C. R. Vool- stra. 2018. Coral reef carbonate budgets and ecological dri- vers in the central Red Sea—A naturally high temperature and high total alkalinity environment. Biogeosciences 15:6277–6296.

Roth, F., F. Saalmann, T. Thomson, D. J. Coker, R. Villalobos, B. H. Jones, C. Wild, and S. Carvalho. 2018. Coral reef degra- dation affects the potential for reef recovery after disturbance. Marine Environmental Research 142:48–58.

Roth, F., I. Stuhldreier, C. Sánchez-Noguera, S. Carvalho, and C. Wild. 2017. Simulated overfishing and natural eutrophica- tion promote the relative success of a non-indigenous ascid- ian in coral reefs at the pacific coast of Costa Rica. Aquatic Invasions 12:435–446.

Roth, F., C. Wild, S. Carvalho, N. Rädecker, C. R. Voolstra, B. Kürten, H. Anlauf, Y. C. El-Khaled, R. Carolan, and B. H. Jones. 2019. An in situ approach for measuring biogeochemi- cal fluxes in structurally complex benthic communities. Meth- ods in Ecology and Evolution 10:712–725.

Russ, G. R. 2003. Grazer biomass correlates more strongly with production than with biomass of algal turfs on a coral reef. Coral Reefs 22:63–67.

Savva, I., S. Bennett, G. Roca, G. Jordà, and N. Marbà. 2018. Thermal tolerance of Mediterranean marine macrophytes: Vulnerability to global warming. Ecology and Evolution 8:12032–12043.

Sawall, Y., A. Al-Sofyani, S. Hohn, E. Banguera-Hinestroza, C. R. Voolstra, and M. Wahl. 2015. Extensive phenotypic plas- ticity of a Red Sea coral over a strong latitudinal temperature gradient suggests limited acclimatization potential to warm- ing. Scientific Reports 5:8940.

Scheufen, T., W. E. Krämer, R. Iglesias-Prieto, and S. Enrı́quez. 2017. Seasonal variation modulates coral sensibility to heat- stress and explains annual changes in coral productivity. Sci- entific Reports 7:4937.

Sibly, R., J. Brown, and A. Kodric-Brown. 2012. Metabolic ecology: a scaling approach. Wiley-Blackwell: Oxford, UK ISBN: 978-1-119-96851-1.

Silverman, J., B. Lazar, and J. Erez. 2007. Effect of aragonite saturation, temperature, and nutrients on the community cal- cification rate of a coral reef. Journal of Geophysical Research: Oceans 112:1–14.

Stachowicz, J. J., J. F. Bruno, and J. E. Duffy. 2007. Understand- ing the effects of marine biodiversity on communities and

Article e03226; page 14 FLORIAN ROTH ET AL. Ecology, Vol. 102, No. 2

ecosystems. Annual Review of Ecology, Evolution, and Sys- tematics 38:739–766.

Stillman, J. H. 2019. Heat waves, the new normal: summertime temperature extremes will impact animals, ecosystems, and human communities. Physiology 34:86–100.

Szmant, A. M. 2002. Nutrient enrichment on coral reefs: is it a major cause of coral reef decline? Estuaries 25: 743–766.

Takeshita, Y., W. Mcgillis, E. M. Briggs, A. L. Carter, E. M. Donham, T. R. Martz, N. N. Price, and J. E. Smith. 2016. Assessment of net community production and calcification of a coral reef using a boundary layer approach. Journal of Geo- physical Research: Oceans 121:5655–5671.

Tkachenko, K. S., B. J. Wu, L. S. Fang, and T. Y. Fan. 2007. Dynamics of a coral reef community after mass mortality of branching Acropora corals and an outbreak of anemones. Marine Biology 151:185–194.

Van Heuven, S. M. A. C., A. E. Webb, D. M. de Bakker, E. Meesters, F. C. Van Duyl, G.-J.-J. Reichart, and L. J. De Nooijer. 2018. In-situ incubation of a coral patch for commu- nity-scale assessment of metabolic and chemical processes on a reef slope. PeerJ 2018:e5966.

Weeks, S. J., K. R. N. Anthony, A. Bakun, G. C. Feldman, and O.-H. Guldberg. 2008. Improved predictions of coral bleach- ing using seasonal baselines and higher spatial resolution. Limnology and Oceanography 53:1369–1375.

Wild, C., et al. 2011. Climate change impedes scleractinian cor- als as primary reef ecosystem engineers. Marine and Freshwa- ter Research 62:205–215.

Wild, C., M. Huettel, A. Klueter, S. G. Kremb, M. Y. M. Rasheed, and B. B. Jørgensen. 2004. Coral mucus functions as an energy carrier and particle trap in the reef ecosystem. Nature 428:66–70.

Wolf-Gladrow, D. A., R. E. Zeebe, C. Klaas, A. Körtzinger, and A. G. Dickson. 2007. Total alkalinity: The explicit con- servative expression and its application to biogeochemical processes. Marine Chemistry 106:287–300.

Yahel, G., J. H. Sharp, D. Marie, C. Häse, and A. Genin. 2003. In situ feeding and element removal in the symbiont-bearing sponge Theonella swinhoei: Bulk DOC is the major source for carbon. Limnology and Oceanography 48:141–149.

Zeebe, R. E., and D. A. Wolf-Gladrow. 2001. CO2 in seawater : equilibrium, kinetics, isotopes. First editionAmsterdam, Netherlands: Elsevier Science. ISBN: 9780080929903

SUPPORTING INFORMATION

Additional supporting information may be found in the online version of this article at http://onlinelibrary.wiley.com/doi/ 10.1002/ecy.3226/suppinfo

February 2021 FUNCTIONING OF CORAL REEF COMMUNITIES Article e03226; page 15