MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog Ser

Vol. 313: 1–12, 2006 Published May 11

INTRODUCTION

Increasing irradiance leads to a saturation of photo-synthesis (Falkowski & Raven 1997), and at highirradiances, excessive light absorption can damagethe photosynthetic apparatus or, alternatively, inducephysiological processes that dissipate energy beforedamage can occur (Demmig-Adams & Adams 1992).This phenomenon, photoinhibition, results in a 6 to25% decline in daily carbon gain in higher plants andphytoplankton (Ogren & Sjostrom 1990, Pahl-Wostl

© Inter-Research 2006 · www.int-res.com*Email: mia.hoogenboom@jcu.edu.au

FEATURE ARTICLE

Energetic cost of photoinhibition in corals

Mia O. Hoogenboom*, Kenneth R. N. Anthony, Sean R. Connolly

ARC Centre of Excellence for Coral Reef Studies, and School of Marine Biology and Aquaculture, James Cook University, Townsville, Queensland 4811, Australia

ABSTRACT: Photoinhibition may constitute an ener-getic cost for photosynthetic organisms through dam-age to the photosynthetic apparatus, or by increasedmetabolism due to damage avoidance or repair. Forseveral species of scleractinian corals, fluorescencetechniques have revealed a significant reduction inphotochemical efficiency of symbiotic dinoflagellateswithin coral tissue in response to excess light ab-sorption. To date, it has been unclear whether or notphotoinhibition has a negative impact on energybudgets in corals. We simultaneously quantified theeffect of exposure to excessive light on net rates ofphotosynthesis and on fluorescence-derived photo-chemistry. We acclimated colonies of the reef-buildingcoral Turbinaria mesenterina to 3 different irradianceregimes in the laboratory. The corals were thenexposed to light levels up to 10 times higher thantheir acclimation irradiance and assayed for rates ofphotosynthesis and photochemical yields. Resultsindicated that daily costs of photoinhibition are negli-gible. Reduced net rates of photosynthesis in theafternoon, compared to the morning, were predomi-nantly due to enhanced afternoon rates of dark respi-ration. However, photoacclimation to high light levelsreduces daily energy acquisition in the long term,primarily due to decreased chlorophyll concentra-tions. Therefore, although changes in the photosyn-thetic activity of symbiotic dinoflagellates over adiurnal irradiance cycle do not cause a measurabledecline in net oxygen evolution for coral colonies,repeated exposure to excessive irradiance canreduce energy acquisition per unit surface area, andhence influence the upper limit of the depth distribu-tion of scleractinian corals.

KEY WORDS: Photoinhibition · Respirometry · PAMfluorescence · Scleractinian coral · Energy budget ·Photosynthetic efficiency · Turbinaria mesenterina

Resale or republication not permitted without written consent of the publisher

The scleractinian coral Turbinaria mesenterina is abundant oninshore reefs, where light regimes are highly variable. Con-spicuous pale patches may develop on the colony surfacefollowing repeated exposure to very high light levels. Thisrepresents an energetic cost of excessive light exposure(photoinhibition) that becomes apparent over a time scale ofseveral days.

Photo: Dr. Andrew Baird, ARC Centre of Excellence for Coral Reef Studies,

Townsville, Australia

OPENPEN ACCESSCCESS

Mar Ecol Prog Ser 313: 1–12, 2006

1992, Werner et al. 2001), leading to reduced growthrates (Laing et al. 1995). For numerous scleractiniancoral species, a marked reduction in the efficiency oflight use for photochemistry by zooxanthellae, thesymbiotic dinoflagellates within coral tissue, is appar-ent under high light conditions (e.g. Brown et al. 1999,Jones & Hoegh-Guldberg 2001, Winters et al. 2003).Furthermore, when isolated from their coral host, zoo-xanthellae exhibit reduced rates of photosynthesis athigh light levels (Shick et al. 1995, Goiran et al. 1996;but see Iglesias-Prieto & Trench 1994). Correspond-ingly, avoidance of excessive light levels is a determi-nant of morphology in coral colonies, and many speciesgenerate self-shading morphologies in high-light habi-tats (Oliver et al. 1983, Titlyanov 1991a, Muko et al.2000). Despite this, the impact of photoinhibition onnet rates of photosynthetic energy acquisition of thecoral symbiosis is not well understood. If photoinhibi-tion incurs significant costs for corals, it may excludelight-sensitive species from shallow habitats. Even inthe presence of morphological strategies that reducelight absorption, or potential changes to the composi-tion of the symbiont population toward light-tolerantclades (e.g. Iglesias-Prieto et al. 2004), costs of photoin-hibition potentially demarcate an upper bound of thedepth distribution of reef-building corals.

Using O2 or CO2 flux measurements (respirometry),photoinhibition has traditionally been detected as adecline in the light-saturated photosynthetic rate (e.g.Platt et al. 1980). An alternative method for assayingphotosynthesis is fluorometry, which allows estimatesof photosynthetic activity from the fluorescent proper-ties of chlorophyll in vivo (Maxwell & Johnson 2000).In the fluorometry method, photoinhibition is typicallyinferred from a decline in photochemical efficiencyduring and/or after exposure to high irradiance. Sucha decline may be due to reversible processes suchas increased dissipation of absorbed light as heat(‘dynamic’ photoinhibition), or to damage to photosyn-thetic units requiring de novo synthesis of protein forrepair ('chronic' photoinhibition: Hoegh-Guldberg &Jones 1999, Gorbunov et al. 2001). In corals, changes inphotochemical yield of zooxanthellae over a diurnalirradiance cycle are generally attributed to dynamicphotoinhibition (Hoegh-Guldberg & Jones 1999, Lesser& Gorbunov 2001, Winters et al. 2003). However,chronic photoinhibition has also been reported, withup to 30% of photosynthetic reaction centers damagedby exposure to full sunlight in shallow waters (see Gor-bunov et al. 2001). Nevertheless, there is no evidenceto indicate whether such changes in photochemistry ofzooxanthellae lead to reduced photosynthetic energyacquisition of the coral symbiosis. In some corals, ratesof photosynthetic oxygen evolution are higher in theafternoon than in the morning (Levy et al. 2004), even

though photoinhibition should be greater in the after-noon following exposure to high irradiance at midday.To determine the ecological significance of photo-inhibition for reef-building corals, we must resolvewhether excess light absorption entails an energeticcost, and, if so, whether this cost constitutes a sig-nificant proportion of daily photosynthetic energyacquisition.

Photoinhibition, whether dynamic or chronic, maycomprise an energetic cost by reducing the maximumrate of photosynthesis (PMAX) at supra-saturating lightlevels (e.g. Platt et al. 1980). Alternatively, costs mayarise from increased sub-saturation irradiance (EK)following exposure to high irradiance, correspondingto lower efficiency of light utilization for photosynthe-sis (e.g. Kana et al. 2002). Finally, photoinhibition maycause elevated rates of respiration (RDARK) due toincreased biosynthesis for damage repair.

The principal aim of this study was to resolvewhether photoinhibition comprises a significant ener-getic cost for corals. Our focus was upon costs incurredby the coral/zooxanthellae symbiosis, the ecologicallyrelevant unit of study for analyses of energy budgetsfor coral colonies. We quantified the cost of photo-inhibition through changes in the parameters of thediurnal photosynthesis–irradiance (PI) relationship ofcorals. Given the influence of photoacclimation on theshape of the PI curve (e.g. Anthony & Hoegh-Guldberg2003), we calculated costs of photoinhibition for coralsacclimated to 3 different light regimes and exposed to2 different diurnal irradiance cycles. In all cases, weexpected that changes in photosynthesis parameterswould be most pronounced for corals acclimatedto lower light levels.

Our second aim was to evaluate the functional rela-tionship between fluorescence and oxygen respirome-try as assays of photosynthesis in reef-building corals.Respirometry and fluorometry measure different as-pects of the PI relationship: the former captures netphotosynthetic gas exchange averaged over a photo-synthetic surface and the latter indicates gross photo-synthetic electron transport from a small area of thatsurface (e.g. Maxwell & Johnson 2000). One of the keyvalidations for the use of fluorescence as a measure ofphotosynthesis is proportionality between the quan-tum yields (photosynthesis per unit light absorbed) ofoxygen evolution and photochemistry (e.g. Brown etal. 1999). Although a linear relationship between thesevariables has been established experimentally forsome plants (Genty et al. 1989, Maxwell et al. 1998),field-based measurements often reveal a non-linearrelationship (e.g. Seaton & Walker 1990, Fryer et al.1998). This non-linearity means that gross photo-chemical activity may vary considerably withouthaving any effect on net rates of photosynthesis. For

2

Hoogenboom et al.: Energetic cost of photoinhibition in corals

corals, the functional relationship between Photosys-tem II (PSII) photochemistry and oxygen evolution isunknown. This is significant, because properties of thecoral/zooxanthella symbiosis may cause this relation-ship to diverge considerably from that demonstratedfor higher plants (e.g. due to respiration of coraltissue independently of photosynthetic production bysymbionts).

MATERIALS AND METHODS

We used the foliose coral Turbinaria mesenterina(Dendrophyllidae) as our study species, because itgenerates self-shading colony morphologies in highlight habitats, suggesting sensitivity to excess lightabsorption (Willis 1985, Anthony et al. 2005). Wecollected 36 flat fragments (measuring approximately10 × 10 cm) from deep (6 m depth) and shallow (1 mdepth) sites at Nelly Bay, Magnetic Island (19° 09’ S,146° 53’ E) on 15 April 2004. Colonies were transportedto aquarium facilities at James Cook University, di-vided into groups and allowed to photoacclimate to3 different irradiance regimes—‘High’, ‘Medium’ and‘Low’—which corresponded to maximum daily irradi-ances of 570, 270 and 120 μmol photons m–2 s–1 andrepresent 4, 7 and 10 m depth at the collection site,respectively (M. O. Hoogenboom unpubl. data). Weused a minimum experimental depth of 4 m, as previ-ous attempts to photoacclimate flat fragments of thestudy species to higher light levels resulted in highmortality (K. R. N. Anthony unpubl. data). This makesour results conservative with respect to the prevalenceof photoinhibition under field conditions, as all col-onies were acclimated to relatively low light levels.Colonies were allowed 6 wk for recovery and photo-acclimation, which is ample for this species (Anthony &Hoegh-Guldberg 2003). To avoid potential changes tophotosynthetic properties of our corals due to nutrientlimitation, colonies were fed live rotifers (50 rotifersml–1). Water temperature was maintained between26.5 and 28°C, approximating the modal temperatureat the collection site (≈ 27°C, AIMS Cleveland BayWeather Station Data).

Experimental setting. An array of 6 closed, clear-perspex incubation chambers (2.7 l) coupled with cali-brated Clark-type oxygen electrodes (Cheshire Sys-tems) was used for the oxygen respirometry assays. Thechambers were designed as flumes, equipped withpumps to maintain continuous water circulation at 5 to6 cm s–1 (laminar flow), and were flushed for 4 min inevery 20 min measuring period to prevent oxygen su-per-saturation. To control for photosynthesis and respi-ration of microorganisms within the water, we regularlyleft 1 chamber empty during respirometry runs. In

addition, chambers were cleaned at the end of each runto prevent biofilm formation. Oxygen concentrationswere recorded every 20 s using a data logger (CR10X,Campbell Scientific). We suspended 400 W metalhalide lamps (Eye) above both the chambers andaquaria to provide the light source for both diurnalphotosynthesis measurements and the light acclimationtreatments. These lamps have a spectrum resemblingnatural sunlight with the ultraviolet component pre-sent. For the photosynthesis measurements, we variedirradiance over the course of the day by changing theelevation of the lamps every 20 min, during the flushingperiod. Light levels (downwelling PAR) were measuredusing a cosine-corrected Licor probe (Li-192S) con-nected to a Li-1000 data logger (Licor). The chamberswere submerged in a 1000 l water jacket with runningseawater connected to a temperature control unit(C023, Carrier Systems) to prevent changes in watertemperature during the measurement period.

On each day of data collection, 5 to 6 colonies from aphotoacclimation treatment were selected for oxygenevolution measurements. Photosynthesis versus irradi-ance curves (PI curves) were constructed over 10 h di-urnal cycles, with irradiance at time t, E(t), varying ap-proximately according to a cubic sine function: E(t) =EMAX sin3(πt/d), where t is time (h after dawn), d is daylength (10 h) and EMAX is maximum daily irradiance.This function closely approximates diurnal irradiancevariation under natural aquatic conditions (Marra1978). Dark respiration was measured twice each day,at the beginning and end of each respirometry run. PIcurves were measured on consecutive days with EMAX =600 μmol photons m–2 s–1 on Day 1 and 1200 μmol pho-tons m–2 s–1 on successive days. We used different diur-nal irradiance cycles to assess whether costs of photoin-hibition were influenced by the degree to whichexperimental irradiance exceeded the growth (acclima-tion) irradiance. For the corals acclimated to low light,we repeated measurements at EMAX = 1200 μmol pho-tons m–2 s–1 over 3 d to determine how photosyntheticproperties change through time under potentially pho-toinhibitory conditions. To relate oxygen flux to colonytissue surface area, colonies were photographed to-gether with a ruler using a digital camera (Canon G3).Surface areas were later determined using Image Tools(University of Texas Health Science Centre).

At the same time as colonies were selected for O2

evolution measurements, an identical set of colonieswas selected for fluorescence measurements, and posi-tioned outside the respirometry chambers. Irradianceat the surface of colonies inside and outside the cham-bers varied <5%. Fluorescence assays were carriedout using a pulse-amplitude-modulated fluorometer(Mini-PAM, Walz) fitted with a 5 mm fiber optic probe.At the end of each 20 min illumination period, fluo-

3

Mar Ecol Prog Ser 313: 1–12, 2006

rescence was measured at the centre of each colony.Apparent quantum yield of photochemistry (ΔF/Fm’,dimensionless) was calculated as ΔF/Fm’ = (Fm’–F)/Fm’,where F is steady state fluorescence in the light andFm’ is maximal fluorescence in the light (Genty et al.1989). Similarly, maximum quatum yield (Fv/Fm, dimen-sionless) was calculated as Fv/Fm = (Fm–Fo)/Fm where Fo

and Fm are steady state and maximal fluorescence iindarkness, respectively. We also calculated a non-photo-chemical quenching coefficient as NPQ = (Fm–Fm’)/Fm’(after Maxwell & Johnson 2000). Yield measurementswere converted to relative electron transport rates(rETR) using the formula rETR = ΔF/Fm’ × E × 0.5,where E is irradiance and 0.5 is a factor that accountsfor the distribution of electrons between Photosystem Iand PSII (e.g. Hoegh-Guldberg & Jones 1999). Weused this relative measure of the rate of electron trans-port, as the light absorption characteristics of tissue areunknown for this species.

Chlorophyll concentration. Immediately after photo-synthesis assays were completed, colonies were frozenat –40°C and later used to determine chlorophyll con-tent. Colonies were broken into 2 fragments and pho-tographed for surface area measurements (see above).Fragments were ground to a fine paste, and chloro-phyll was extracted in the dark using cold acetone(100%). To ensure that all chlorophyll was extractedfrom each coral fragment, the initial overnight extrac-tion (12 h) was followed by two 1 h extractions. Thecombined volume of all 3 extractions was measured,and the extract centrifuged; 10 replicate absorbancereadings at 630 and 663 nm were carried out for eachfragment, and concentration of chl a and chl c2 wasdetermined using the equations of Jeffrey & Humphrey(1975). Chlorophyll concentrations were then nor-malised to fragment surface area.

Data analysis. We fitted a hyperbolic tangent func-tion (Eq. 1) to PI data in order to estimate PMAX (maxi-mum rate of photosynthesis, μmol O2 cm–2 h–1) and EK

(sub-saturation irradiance, μmol photons m–2 s–1) usingmeasured values of RDARK (rate of respiration in dark-ness, μmol O2 cm–2 h–1):

(1)

We fitted this function to data using a Levenberg-Marquardt non-linear estimation routine in STATIS-TICA (StatSoft). We first estimated PMAX and EK for themorning part of the diurnal photosynthesis curve. Asthere was no evidence in the data of a change in (net)PMAX in the afternoon compared with the morning, wesubsequently used the fitted PMAX value from the morn-ing part of the PI curve to obtain an estimate of EK

during the afternoon. This allowed us to detect whetherexposure to light over the course of the morning altered

the shape of the PI curve during the afternoon. We usedrepeated-measures ANOVA to investigate the effect ofphotoacclimation state and diurnal irradiance cycle onchanges in photosynthesis parameters over the day. Wethen used the fitted parameters to calculate photosyn-thetic oxygen evolution during mornings (5 h) and af-ternoons (5 h), based on the integral of Eq. (1). We useda numeric integration routine in MATLAB (MathWorks)to evaluate this integral for each colony. Subsequently,we calculated the energetic cost of photoinhibition asthe proportional difference between total photosynthe-sis summed over the morning, compared with the after-noon, and tested statistical significance of the cost bypaired-samples t-tests. The cost therefore representsthe reduction in net carbon acquisition due to exposureto excessive irradiance over a daily timescale. We chosethis method in place of commonly used photosynthe-sis/respiration ratios (e.g. Muscatine et al. 1981), asrates of respiration over the day are not constant, as wasassumed under the latter method (e.g. Kuhl et al. 1995,Al-Horani et al. 2003). To ascertain whether repeatedexposure to excessive irradiance alters the photosyn-thetic properties of Turbinaria mesenterina, we usedpaired-samples t-tests to compare mean values of 5 keyphotosynthetic properties between Days 1 and 3 ofexposure.

Quantum yield of oxygen evolution, in mol oxygen(mol photons)–1, was obtained by dividing rates of photo-synthesis (converted to appropriate units of μmol O2 m–2

s–1 and corrected for dark respiration) by incident irra-diance (μmol photons m–2 s–1) (Seaton & Walker 1990).Because dark respiration in corals is enhanced by lightexposure (e.g. Edmunds & Spencer-Davies 1988), weused linear interpolation between estimates of RDARK

for the morning and afternoon to convert measurednet photosynthesis to estimates of gross photosynthesis.For comparison between fluorescence and respirometryassays of photosynthesis, we normalized our data tochlorophyll concentration. We chose this normalizationto avoid confounding differences in photochemistry be-tween light acclimation treatments with any differencesin chlorophyll concentration.

RESULTS

Net photosynthesis and rETR

Photoinhibitory responses varied greatly, dependingon whether respirometry or fluorescence assays ofphotosynthesis were used. In general, although rETRdecreased under high irradiance in response to declin-ing apparent photochemical yield (Fig. 1B,C), light-saturated rates of photosynthesis remained constantfor corals acclimated to all the 3 light regimes, regard-

P E PE

ERMAX

KDARK( ) tanh= ⎛

⎝⎞⎠ −

4

Hoogenboom et al.: Energetic cost of photoinhibition in corals

less of whether they were exposed to a daily irradiancecycle of EMAX = 600 or 1200 μmol photons m–2 s–1

(Fig. 1D–F). For all of our light-acclimation treatments,NPQ was highest at midday, mirroring the decrease inphotochemical yield (results not shown). The magni-tude of NPQ was also related to photoacclimation state,with corals acclimated to high light showing greaterquenching capacity. Compared with rates of oxygenevolution, rETR approached saturation more slowlyand was generally higher in the morning than in theafternoon at the same irradiance. This effect was mostapparent at mid-afternoon under both diurnal irradi-

ance cycles, but rETR returned to morning levels bylate afternoon. In other words, effects of photoinhibi-tion on rETR were short-lived, and photochemicalactivity returned to initial values by the end of the day,despite exposure to light levels well above those towhich corals were acclimated. Overall, we found noevidence that low photochemical efficiency at middayhad any effect on net photosynthesis, i.e. althoughrETR via photochemistry decreased under high irradi-ance, rates of photosynthesis remained saturated evenwhen corals were exposed to light levels much higherthan their acclimation irradiance.

5

Fig. 1. Turbinaria mesenterina. Relative electron transport (rETR) and diurnal net photosynthesis versus irradiance in corals accli-mated to 3 light regimes and subjected to 2 daily irradiance cycles (d: EMAX = 600 μmol photons m–2 s–1; n: EMAX = 1200 μmol photonsm–2 s–1), measured at 20 min intervals over the day; (A,D) high light; (B,E) medium light; (C,F) low light (mean ± SE; n = 5 to 8)

Mar Ecol Prog Ser 313: 1–12, 2006

As expected, the shape of the PI curve for colonies ofTurbinaria mesenterina varied with photoacclimationstate. PMAX was significantly lower for corals accli-mated to low light levels compared with the other 2growth irradiances (main effect of State forPMAX, Table 1). PMAX was also higher underexposure to more intense light for coralsacclimated to all growth irradiances (maineffect of Treatment, Table 1). This latter find-ing indicates that photosynthesis did notreach complete saturation under the lowerdiurnal irradiance cycle (also evident fromthe filled points in Fig. 1D–F). EK increasedwith acclimation irradiance when corals wereexposed to low light levels, but not underexposure to high light levels (Fig. 2A,B).This difference was apparent as a significantState × Treatment interaction (Table 1). Toour knowledge, this finding represents thefirst evidence that corals dynamically down-regulate net oxygen evolution due to exces-sive light absorption. We found no evidenceof any variation in EK between morning andafternoon (all effects involving Time werenon-significant for EK, Table 1; filled pointscompared to triangles, Fig. 2A,B).

RDARK was markedly influenced by lightintensity, both in terms of acclimation irradi-ance (significant main effect of State, Table 1)and diurnal irradiance treatment. Under bothdiurnal irradiance cycles, RDARK increasedwith acclimation irradiance (Fig. 2C,D). Fur-

thermore, RDARK was distinctly higher in the afternoonthan in the morning under the high diurnal irradiancetreatment (open triangles compared to filled points inFig. 2D), but not under exposure to lower light levels

6

Table 1. ANOVA for the effect of photoacclimation state (State), diurnal irradiance cycle (Treatment) and time of day (Time) onphotosynthesis–irradiance curve parameters, and results of post hoc tests (Tukey’s Unequal N HSD). PMAX: maximum rate ofphotosynthesis; EK: sub-saturation irradiance; RDARK: rate of respiration in darkness. L, M, H: corals acclimated to low, mediumand high irradiance regimes, respectively; 600/1200: estimates under daily irradiance cycles with EMAX = 600 and 1200 μmol pho-tons m–2 s–1, respectively. AM, PM: estimates from the morning and afternoon half of the diurnal photosynthesis curve

Units Effect F (df) p Post hoc

PMAX (μmol O2 cm–2 h–1) State 11 (2,29) <0.01 L < M = HTreatment 6.9 (1,29) <0.05 600 < 1200 for L, M, HState × Treatment 1.8 (2,29) 0.19 –

EK (μmol photons m–2 s–1) State 0.24 (2,29) 0.79 –Treatment 23 (1,29) <0.01 naState × Treatment 5.7 (2,29) <0.01 L & M, 1200 > 600Time 0.03 (1,29) 0.86 –State × Time 0.33 (2,29) 0.72 –Treatment × Time 0.02 (1,29) 0.90 –State × Treatment × Time 0.03 (2,29) 0.97 –

RDARK (μmol O2 cm–2 h–1) State 22 (2,29) <0.01 L < M = HTreatment 2.6 (1,29) 0.12 –State × Treatment 0.71 (2,29) 0.50 –Time 19 (1,29) <0.01 naState × Time 0.53 (2,29) 0.59 –Treatment × Time 4.7 (1,29) <0.05 AM < PM when EMAX = 1200State × Treatment × Time 0.21 (2,29) 0.81 –

Fig. 2. Turbinaria mesenterina. (A,B) Sub-saturation irradiance (EK), and(C,D) respiration (RDARK) during morning (AM) and afternoon (PM) incorals acclimated to 3 light regimes (low, medium, high) and exposed to 2

irradiance cycles (mean ± 95% confidence intervals; n = 5 to 8)

Hoogenboom et al.: Energetic cost of photoinhibition in corals

(Fig. 2C). Consequently, we found a significant Treat-ment × Time interaction for RDARK (Table 1). Therefore,differences between rates of net photosynthesis in theearly morning compared with the late afternoon at thesame irradiance (Fig. 1E,F) are predominantly due tochanges in rates of respiration. The post-illuminationincrease in rates of dark respiration was consistentacross all photoacclimation groups (i.e. none of theinteraction effects involving State were significant forRDARK, see Table 1). This suggests that increased meta-bolic activity following light exposure may be a gen-eral response of the coral–zooxanthellae symbiosis,and is not influenced by the degree to which exposureirradiance exceeds the irradiance to which corals areacclimated.

Cost of photoinhibition

Total net oxygen evolution (and equivalently, carbonfixation) was significantly lower during the afternoonthan in the morning for corals acclimated to all 3 irradi-ance regimes when exposed to the higher daily irradi-ance cycle. Specifically, integrated rates of net photo-synthesis under exposure to high light levels showed a15 to 17% decline in the afternoon, compared with themorning (Fig. 3A; Table 2). Moreover, corals accli-mated to the lowest light levels showed a 6% reductionin integrated net rates of photosynthesis during theafternoon when exposed to the lower daily irradianceregime. Although this reduction constitutes a consider-able proportion of daily photosynthetic activity, grossphotosynthetic energy acquisition (disregarding oxy-gen consumption through respiration) summed overthe morning and the afternoon did not differ signifi-cantly for any of the comparisons (Fig. 3B; Table 2). Inother words, photoinhibition only represents a sig-nificant energetic cost for corals if metabolic activityassociated with repair of damaged components of the

photosynthetic apparatus is the major cause of post-illumination enhancement of respiration.

Although daily costs of photoinhibition appear to benegligible for Turbinaria mesenterina, photoacclima-tion to high light levels does not maximise integrateddaily photosynthesis. In fact, total daily energy acquisi-tion per unit surface area was lower on average for thehigh light acclimated corals than for corals acclimatedto intermediate light levels when exposed to the samedaily irradiance cycle (Fig. 4). This is primarily due to alower chlorophyll concentration, higher EK, and ahigher RDARK for corals at high light levels, comparedto those acclimated to lower light levels. When inte-grated photosynthesis was calculated for each photo-acclimation treatment over a daily irradiance cyclewith EMAX equal to the growth irradiance, resultsindicated that higher light availability does not trans-late to higher daily carbon acquisition for T. mesente-rina. In other words, photoacclimation causes carbonacquisition to remain constant across an approximately5-fold light gradient.

7

* * * *

-30

-20

-10

0

10

20

% d

iffer

ence

pho

tosy

nthe

sis

AM

:PM

B

A

Photoacclimatory state/EMAX

-30

-20

-10

0

10

20

H60

0

M60

0

L600

H12

00

M12

00

L120

0

Table 2. Turbinaria mesenterina. Proportional difference be-tween integrated net and gross photosynthesis over thecourse of the afternoon compared to the morning for coralsacclimated to 3 different light regimes (State), and exposed to2 diurnal irradiance cycles (maximum daily irradiance, EMAX,

μmol photons m–2 s–1). Results of paired samples t-test

EMAX State df Net photosyn. Gross photosyn.t p t p

600 High 5 –0.07 <0.95 –0.06 0.96Medium 4 –1.99 <0.12 –0.51 0.64Low 5 –3.37 <0.05 –0.15 0.89

1200 High 4 –4.4 <0.05 –0.09 0.93Medium 7 –4.5 <0.05 –0.71 0.50Low 4 –7.6 <0.05 –0.31 0.77

Fig. 3. Turbinaria mesenterina. Percentage difference in (A)net and (B) gross photosynthesis between morning and after-noon in corals acclimated to 3 different light regimes (H: high;M: medium; L: low) and exposed to 2 diurnal irradiance cycles(EMAX, 600 and 1200 μmol photons m–2 s–1) (mean ± SE; n = 5to 8). *Statistically significant difference (dependent samples

t-test, see Table 3)

Mar Ecol Prog Ser 313: 1–12, 2006

To determine the medium-term consequences ofphotoinhibition and changes in photoacclimation stateon carbon gain, we repeated measurements of photo-synthesis for the low light acclimated corals over 3 d ofexposure to a daily irradiance cycle with EMAX =1200 μmol photons m–2 s–1. Although all of the measuredphotosynthetic properties varied over this time frame(Table 3), changes were only statistically significant forchlorophyll concentration and light-saturated PMAX. Theincrease in (fitted) PMAX was partially due to a corre-sponding increase in RDARK. The net effect of this wasthat although PMAX was higher after repeated exposureto high light levels, integrated daily photosynthesis wasactually lower, although not significantly so. In general,these changes in photosynthetic properties were consis-tent with photoacclimation to high light levels (increasedPMAX, EK, and RDARK, lower chlorophyll concentrationand maximum photochemical efficiency). Collectively,our results indicate that while gradual loss of functionof individual symbionts over the course of the day hasa negligible (short-term) impact on energy budgets,repeated exposure leads to lower concentration of chlo-rophyll per unit surface, and consequently lower dailyenergy acquisition in the long term.

Fluorescence versus respirometry

In contrast to the linear relationship between fluores-cence and gas exchange measures of photosynthetic ac-tivity typically assumed for corals (e.g. Brown et al. 1999,Hoegh-Guldberg & Jones 1999), our results demonstratea curvilinear relationship (Fig. 5). Furthermore, therewas considerable variability in the relationship betweenthe quantum yields of oxygen evolution and photochem-istry (Fig. 5A). At high light levels, ETR may decline by>50% without any measurable effect on the net rate ofphotosynthesis. Similarly, a 3-fold variation in rETR wasevident at PMAX. On the other hand, rETR was directlyproportional to rates of oxygen evolution at low light lev-els (i.e. low rates of photosynthesis). The breakdown ofthe linear relationship between these 2 measures at highirradiance is driven by the substantially greater degreeof hysteresis exhibited by rETR versus irradiance com-pared with photosynthesis versus irradiance when mea-sured over a diurnal cycle. In addition to the hysteresiseffect, lower correlation at high irradiance was related tovariation in PMAX (oxygen evolution) due to photo-acclimation. For corals acclimated to low light levels,oxygen evolution per electron transported was lower(lower rate of photosynthesis at the same rETR, Fig. 5B).Collectively, these results demonstrate that for Turbi-naria mesenterina, the relationship between biochemicaland energetic assays of photosynthesis is influenced bythe photoacclimation state of individual colonies, evenwhen variation in chlorophyll concentration betweencolonies is taken into account.

DISCUSSION

Our results demonstrate that exposure of coralsto high light levels only leads to a depression of netphotosynthesis when average daily irradiances aremuch higher than growth (acclimation) irradiances.Moreover, reduced rates of net photosynthesis in theafternoon are primarily associated with increased res-piration rates. Previous studies of light-enhanced res-piration in corals have demonstrated a 6- to 12-fold

8

10

15

20

25

H60

0

M60

0

L600

H12

00

M12

00

L120

0

Photoacclimatory state/EMAX

Pho

tosy

nthe

sis

(µm

ol O

2 cm

–2 d

–1)

Fig. 4. Turbinaria mesenterina. Daily integrated photo-synthetic oxygen evolution in corals acclimated to 3 lightregimes (H: high; M: medium; L: low) and exposed to 2 diurnalirradiance cycles (EMAX, 600 and 1200 μmol photons m–2 s–1)

(mean ± SD)

Table 3. Turbinaria mesenterina. Change in photosynthetic properties (mean ± SE) in corals acclimated to low light followingrepeated exposure to excessive irradiance (EMAX = 1200 μmol photons m–2 s–1), and results of dependent samples t-test (df = 3).EK: sub-saturation irradiance; PMAX: maximum rate of photosynthesis; RDARK: rate of respiration in darkness; Fv/Fm: maximum

quantum yield of photochemistry

Chlorophyll EK PMAX RDARK Fv/Fm

(μg cm–2) (μmol photons m–2 s–1) (μmol O2 cm–2 h–1) (μmol O2 cm–2 h–1)

Day 1 32 ± 6 229 ± 36 2.3 ± 0.06 0.29 ± 0.17 0.66 ± 0.03Day 3 23 ± 8 345 ± 54 2.6 ± 0.06 0.40 ± 0.10 0.52 ± 0.06t 3.3 –2.1 –7.8 0.63 2.4p <0.05 0.13 <0.01 0.57 0.10

Hoogenboom et al.: Energetic cost of photoinhibition in corals

increase in oxygen consumption following light expo-sure (Kuhl et al. 1995, Al-Horani et al. 2003). A propor-tion of this increase may reflect energy expenditure forrepair of damage to the photosynthetic apparatus ofzooxanthellae, and/or coral tissue damage. However,light-enhanced respiration in corals is more likely dueto enhanced metabolic activity related to e.g.increased skeletal growth (e.g. Reynaud-Vaganay etal. 2001) and increased photosynthetic activity, as sig-nificant light enhancement of respiration has been

measured in corals exposed to irradiances as low as140 μmol photons m–2 s–1 (Al-Horani et al. 2003). Ifincreased rates of respiration following exposure tohigh light levels were predominantly due to costs ofrepair of the photosynthetic apparatus, the magnitudeof the change in respiration between morning andafternoon would depend upon the degree to whichexposure irradiance exceeded acclimation irradiance(damage being greater for corals acclimated to lowerlight levels). As we observed the same post-illumina-tion increase in rates of respiration for all acclimationtreatments following exposure to the same diurnalirradiance cycle, we conclude that daily energetic costsof photoinhibition in corals are negligible.

Dissipation of light through non-photochemical path-ways is recognised as an effective photoprotectivemechanism in corals (e.g. Hoegh-Guldberg & Jones1999, Gorbunov et al. 2001), and the same mechanismis present in other photosynthetic organisms for whichenergetic costs of photoinhibition are apparent (e.g.higher plants and phytoplankton; see Pahl-Wostl 1992,Werner et al. 2001). Our results raise the question as tohow corals avoid these costs. Coral tissue containsamino acids that absorb, reflect or fluoresce ultravioletlight (mycosporine-like amino acids or MAAs; Jokiel &York 1982), and pigments that absorb light over photo-synthetic wavelengths (Salih et al. 2000, Dove 2004).Therefore, the coral tissue layer may act as a protectivescreen for the zooxanthellae. Indeed, there is some evi-dence that host tissue reduces the light levels reachingsymbionts by >50% in some marine organisms (e.g.hydroids; Fitt & Cook 2001). Alternatively, zooxanthel-lae may shade each other, with symbionts in the uppertissue layers shielding those in lower layers. Zooxan-thellae in different parts of a coral colony experiencedifferent light environments (Jones et al. 2000), andvary in their chlorophyll fluorescence characteristics(Hill et al. 2004). Moreover, histology of bleachedcorals has revealed a greater loss of zooxanthellae fromupper tissue layers as opposed to deeper tissues (Brownet al. 1995). Therefore, although zooxanthellae in uppertissue layers may suffer reduced rates of photosynthe-sis due to photoinhibition, those in lower layers arelikely to remain protected and potentially compensatefor the reduced photosynthesis of the upper layers.

If zooxanthellae do indeed shade each other, thenthe use of fluorometry to measure photosynthesis incorals may return biased results. For instance, inmicrophytobenthic assemblages, the thickness of thefluorescing layer has a pronounced impact on mea-sured photochemical efficiency, with values over-estimated by up to 60% in thick biofilms (Forster &Kromkamp 2004). An equivalent phenomenon mayexplain our findings of lower oxygen evolution perelectron transported for corals acclimated to low light

9

Fig. 5. Turbinaria mesenterina. Comparison between respiro-metry and fluorescence assays of photosynthetic activity.(A) Quantum yield of oxygen evolution (normalized to chloro-phyll concentration) versus photochemical yield, and (B) rateof photosynthesis (corrected for dark respiration) versusrelative electron transport rate, in corals acclimated to 3 lightregimes (high, medium, low). Data are means; error bars

omitted for clarity (n = 5 to 8)

Mar Ecol Prog Ser 313: 1–12, 2006

levels compared with medium- and high-light accli-mated colonies. Self-shading of zooxanthellae wouldresult in overestimation of apparent photochemicalefficiency as the measured efficiency of the upperlayer would be augmented by higher efficiency of sym-bionts exposed to lower light levels deeper in the tis-sue. This would in turn lead to overestimation of rETR,and an apparently lower rate of oxygen evolution perelectron transported through PSII. Overall, our find-ings indicate that future studies should take account oftissue properties of corals, such as symbiont density,tissue thickness and presence of other light-absorbingpigments, when interpreting fluorescence assays ofphotochemistry.

Due to the insensitivity of rates of net photosynthesis tosupra-saturating light levels, we find that photochemicalelectron transport (in zooxanthellae) and photosyntheticoxygen evolution (in the coral–zooxanthellae symbiosis)correlate poorly at high light levels; under midday irradi-ances, even very low photochemical efficiencies (andhigh NPQ) do not necessarily result in reduced rates ofphotosynthesis. Because the fluorescence behavior ofTurbinaria mesenterina is consistent with that of severalother species (e.g. Goniastrea aspera: Brown et al. 1999;Stylophora pistillata: Jones & Hoegh-Guldberg 2001,Winters et al. 2003), our findings are likely to reflectgeneral properties of the coral–zooxanthellae symbiosis.Moreover, our results demonstrate that the influence ofnon-assimilatory electron flow at high irradiances on therelationship between fluorescence and respirometryassays of photosynthesis may represent a general trendfor photosynthetic organisms. We find that for corals,oxygen evolution saturates prior to rETR, as has beenshown for cyano-lichens, macroalgae and higher plants(Sundberg et al. 1997, Fryer et al. 1998, Figueroa et al.2003). This indicates that once the assimilatory reactionsof photosynthesis become saturated, electrons may stillbe transported through PSII. The importance of electronsinks in addition to CO2 reduction for avoidance ofphotoinhibition has been noted previously (Krall &Edwards 1992, Hoegh-Guldberg & Jones 1999), withphotorespiration and the Mehler cycle as the mostobvious candidates for non-photosynthetic electrontransport (see Fryer et al. 1998, Figueroa et al. 2003). Itis clear that for a range of photosynthetic organisms,including corals, a decline in rETR at high light levelsis not representative of a decline in photosyntheticoxygen evolution.

Despite the negligible costs associated with photoin-hibition on a daily basis, repeated exposure to highirradiance does have a negative impact on photosyn-thetic energy acquisition in corals. Cell damage causedby prolonged exposure to ultraviolet light representsan additional factor that may inhibit coral growth inshallow (high light) habitats (e.g. Jokiel & York 1982),

although there is some evidence that higher concen-trations of MAAs in corals from shallow water may besufficient to mitigate the effects of ultraviolet light onrates of photosynthesis (Shick et al. 1995). Clearly,colonies of Turbinaria mesenterina acclimated to highlight levels would have greater daily energy acquisi-tion if they had higher symbiont population densities,lower EK and higher PMAX. Although our resultsdemonstrate that there are negligible energetic costsassociated with short-term (1 d) exposure to excessiveirradiance, that the abovementioned combination ofphotosynthetic properties does not occur is evidencethat avoiding damage to the photosynthetic apparatusis a fundamental component of acclimation to highlight environments. This conclusion is supported byour observation of a reduction in chlorophyll concen-tration for low light acclimated corals followingrepeated exposure to excessive irradiance. Continuedreduction of chlorophyll concentration through timemust eventually lead to a reduced daily energy acqui-sition per unit area of colony (e.g. daily integrated pho-tosynthesis of high light acclimated corals comparedwith medium light acclimated corals in this study). Ourresults indicate that costs of photoinhibition in coralsare manifest over time scales of days to weeks, ratherthan being apparent over a diurnal irradiance cycle asobserved in other taxa (e.g. Platt et al. 1980, Ogren &Sjostrom 1990). These findings are consistent withrecent observations of seasonal fluctuations in photo-synthetic activity of several Caribbean coral species(Warner et al. 2002), and also explain why a moderatedecrease in light availability within habitats either hasno effect on photosynthetic energy acquisition forcorals, or leads to higher photosynthesis for coloniesfrom shaded habitats (Titlyanov 1991b). The trade-offbetween efficient utilization of light for photosynthesisand avoidance of cumulative damage to the photosyn-thetic apparatus due to repeated exposure to excessiveirradiance means that higher light availability does notequate to higher energy acquisition.

Overall, changes in the photochemical activity ofzooxanthellae over a diurnal irradiance cycle do notcause a reduction in photosynthetic energy acquisitionfor coral colonies. Photoinhibition in corals is mani-fested through a different mechanism than in otherphotosynthetic organisms. For some species of micro-algae, energetic costs of photoinhibition become appar-ent through decreased PMAX under exposure to lightlevels comparable to those used in this study (e.g. Plattet al. 1980, Pahl-Wostl et al. 1992), or through reducedphotosynthetic efficiency following high light exposurethat causes a reduction in rates of photosynthesis (e.g.Kana et al. 2002). In contrast, energetic costs of photoin-hibition in corals become apparent as a gradual reduc-tion in photosynthetic capacity over time (several days)

10

Hoogenboom et al.: Energetic cost of photoinhibition in corals

following repeated exposure to excessive irradiance.These results indicate that, in corals, long-term ratherthan diurnal changes in photosynthetic properties arethe key to understanding ecological impacts of environ-mental gradients in light intensity.

Acknowledgements. This study was funded by the AustralianResearch Council and James Cook University. We thank M.Dornelas, D. Abrego and J. Cox for their assistance with fieldand laboratory work. This study is a contribution from theARC Centre of Excellence for Coral Reef Studies.

LITERATURE CITED

Al-Horani FA, Al-Moghrabi SM, de Beer D (2003) Micro-sensor study of photosynthesis and calcification in thescleractinian coral Galaxea fascicularis: active internalcarbon cycle. J Exp Mar Biol Ecol 288:1–15

Anthony KRN, Hoegh-Guldberg O (2003) Kinetics of photo-acclimation in corals. Oecologia 134:23–31

Anthony KRN, Hoogenboom MO, Connolly SR (2005) Adap-tive variation in coral geometry and the optimisation ofinternal colony light climates. Funct Ecol 19:17–26

Brown BE, Ambarsari I, Warner ME, Fitt WK, Dunne RP, GibbSW, Cummings DG (1999) Diurnal changes in photochem-ical efficiency and xanthophyll concentrations in shallowwater reef corals: evidence for photoinhibition and photo-protection. Coral Reefs 18:99–105

Brown BE, Le Tissier MDA, Bythell JC (1995) Mechanisms ofbleaching deduced from histological studies of reef coralssampled during a natural bleaching event. Mar Biol 122:655–663

Demmig-Adams B, Adams WW (1992) Photoprotection andother responses of plants to high light stress. Annu RevPlant Physiol Plant Mol Biol 43:599–626

Dove S (2004) Scleractinian corals with photoprotective hostpigments are hypersensitive to thermal bleaching. MarEcol Prog Ser 272:99–116

Edmunds PJ, Spencer-Davies P (1988) Post-illumination stim-ulation of respiration rate in the coral Porites porites. CoralReefs 7:7–9

Falkowski PG, Raven JA (1997) Aquatic photosynthesis.Blackwell, Malden, MA

Figueroa FL, Conde-Alvarez R, Gomen I (2003) Relationsbetween electron transport rates determined by pulseamplitude modulated chlorophyll fluorescence and oxy-gen evolution in macroalgae under different light condi-tions. Photosynth Res 75:259–275

Fitt WK, Cook CB (2001) Photoacclimation and the effect ofthe symbiotic environment on the photosynthetic responseof symbiotic dinoflagellates in the tropical marine hydroidMyrionema amboinense. J Exp Mar Biol Ecol 256:15–31

Forster RM, Kromkamp JC (2004) Modelling the effects ofchlorophyll fluorescence from subsurface layers on photo-synthetic efficiency measurements in microphytobenthicalgae. Mar Ecol Prog Ser 284: 9–22

Fryer MJ, Andrews JR, Oxborough K, Blowers DA, Baker NR(1998) Relationship between CO2 assimilation, photosyn-thetic electron transport, and active O2 metabolism inleaves of maize in the field during periods of low tempera-ture. Plant Physiol 116:571–580

Genty B, Briantais JM, Baker NR (1989) The relationship be-tween the quantum yield of photosynthetic electron trans-port and quenching of chlorophyll fluorescence. Biochim

Biophys Acta 990:87–92Goiran C, Al-Moghrabi SM, Allemand D, Jaubert J (1996)

Inorganic carbon uptake for photosynthesis by the sym-biotic coral/dinoflagellate association. I. Photosyntheticperformance of symbionts and dependence on sea waterbicarbonate. J Exp Mar Biol Ecol 199:207–225

Gorbunov MY, Kolber ZS, Lesser MP, Falkowski PG (2001)Photosynthesis and photoprotection in symbiotic corals.Limnol Oceanogr 46:75–85

Hill R, Schreiber U, Gademann R, Larkum AWD, Kuhl M,Ralph PJ (2004) Spatial heterogenetity of photosynthesisand the effect of temperature-induced bleaching condi-tions in 3 species of corals. Mar Biol 144:633–640

Hoegh-Guldberg O, Jones RJ (1999) Photoinhibition andphotoprotection in symbiotic dinoflagellates from reef-building corals. Mar Ecol Prog Ser 183:73–86

Iglesias-Prieto R, Trench RK (1994) Acclimation and adapta-tion to irradiance in symbiotic dinoflagellates. I. Responsesof the photosynthetic unit to changes in photon fluxdensity. Mar Ecol Prog Ser 113:163–175

Iglesias-Prieto R, Beltran VH, LaJeunesse TC, Reyes-BonillaH, Thome PE (2004) Different algal symbionts explain thevertical distribution of dominant reef corals in the easternPacific. Proc R Soc Lond Ser B 271:1757–1763

Jeffrey SW, Humphrey GF (1975) New spectrophotometricequations for determining chlorophylls a, b, c and c2 inhigher plants, algae and natural phytoplankton. BiochimBiophys Acta 167:191–194

Jokiel PL, York RH Jr (1982) Solar ultraviolet photobiologyof the reef coral Pocillopora damicornis and symbioticzooxanthellae. Bull Mar Sci 32:301–315

Jones RJ, Hoegh-Guldberg O (2001) Diurnal changes in thephotochemical efficiency of the symbiotic dinoflagellates(Dinophyceae) of corals: photoprotection, photoinactiva-tion and the relationship to coral bleaching. Plant CellEnviron 24:89–99

Jones RJ, Ward S, Yang Amri A, Hoegh-Guldberg O (2000)Changes in quantum efficiency of Photosystem II of sym-biotic dinoflagellates of corals after heat stress, and ofbleached corals sampled after the 1998 Great Barrier Reefmass bleaching event. Mar Freshw Res 51:63–71

Kana R, Lazar D, Prasil O, Naus J (2002) Experimental andtheoretical studies on the excess capacity of PhotosystemII. Photosynth Res 72:271–284

Krall JP, Edwards GE (1992) Relationship between photo-system II activity and CO2 fixation in leaves. Physiol Plant86:180–187

Kuhl M, Cohen Y, Dalsgaard T (1995) Microenvironment andphotosynthesis of zooxanthellae in scleractinian coralsstudied with microsensors for O2, pH and light. Mar EcolProg Ser 117:159–172

Laing WA, Greer DH, Schnell TA (1995) Photoinhibition ofphotosynthesis causes a reduction in vegetative growthrates of dwarf bean (Phaseolus vulgaris) plants. Aust JPlant Physiol 22:511–520

Lesser MP, Gorbunov MY (2001) Diurnal and bathymetricchanges in chlorophyll fluorescence yields of reef coralsmeasured in situ with a fast repetition rate fluorometer.Mar Ecol Prog Ser 212:69–77

Levy O, Dubinsky Z, Schneider K, Achituv Y, Zakai D,Gorbunov MY (2004) Diurnal hysteresis in coral photo-synthesis. Mar Ecol Prog Ser 268:105–117

Marra J (1978) Effect of short-term variations in light intensityon photosynthesis of a marine phytoplankter: a laboratorysimulation study. Mar Biol 46:191–202

Maxwell K, Badger MR, Osmond CB (1998) A comparison ofCO2 and O2 exchange patterns and the relationship with

11

Mar Ecol Prog Ser 313: 1–12, 2006

chlorophyll fluorescence during photosynthesis in C3 andCAM plants. Aust J Plant Physiol 25:45–52

Maxwell K, Johnson GN (2000) Chlorophyll fluorescence—a practical guide. J Exp Bot 51:659–668

Muko S, Kawasaki K, Sakai K (2000) Morphological plasticityin the coral Porites sillimaniani and its adaptive signifi-cance. Bull Mar Sci 66:225–239

Muscatine L, McCloskey LR, Marian RE (1981) Estimating thedaily contribution of carbon from zooxanthellae to coralanimal respiration. Limnol Oceanogr 26(4):601–611

Ogren E, Sjostrom M (1990) Estimation of the effect of photo-inhibition on the carbon gain in leaves of a willow canopy.Planta 181:560–567

Oliver JK, Chalker BE, Dunlap WC (1983) Bathymetric adap-tations of reef-building corals at Davies Reef, GreatBarrier Reef, Australia. I. Long-term growth responses ofAcropora formosa (Dana 1846). J Exp Mar Biol Ecol 73:11–35

Pahl-Wostl C (1992) Dynamic versus static models for photo-synthesis. Hydrobiologia 238:189–196

Platt T, Gallegos CL, Harrison WG (1980) Photoinhibition ofphotosynthesis in natural assemblages of marine phyto-plankton. J Mar Res 38:687–701

Reynaud-Vaganay S, Juillet-Leclerc A, Jaubert J, Gatusso JP(2001) Effect of light on skeletal 13C and 18O and inter-action with photosynthesis, respiration and calcification in2 zooxanthellate scleractinian corals. Paleogeogr Palaeo-climatol Palaeoecol 175:393–404

Salih A, Larkum A, Cox G (2000) Fluorescent pigments incorals are photoprotective. Nature 408:850–853

Seaton GGR, Walker DA (1990) Chlorophyll fluorescence as a

measure of photosynthetic carbon assimilation. Proc R SocLond Ser B 242:29–35

Shick JM, Lesser MP, Dunlap WC, Stochaj WR, ChalkerBE, Won JW (1995) Depth-dependent responses to solarultraviolet radiation and oxidative stress in the zoo-xanthellate coral Acropora microphthalma. Mar Biol122:41–51

Sundberg B, Campbell D, Palmqvist K (1997) Predicting CO2

gain and photosynthetic light acclimation from fluores-cence yield and quenching in cyano-lichens. Planta 201:138–145

Titlyanov EA (1991a) Light adaptation and production char-acteristics of branches differing by age and illumination ofthe hermatypic coral Pocillopora verrucosa. Symbiosis 10:249–261

Titlyanov EA (1991b) The stable level of coral primary pro-duction in a wide light range. Hydrobiologia 216/217:383–387

Warner ME, Chilcoat GC, McFarland FK, Fitt WK (2002) Sea-sonal fluctuations in the photosynthetic capacity of photo-system II in symbiotic dinoflagellates in the Caribbeanreef-building coral Montastrea. Mar Biol 141:31–38

Werner C, Ryel RJ, Correia O, Beyschlag W (2001) Effects ofphotoinhibition on whole-plant carbon gain assessed witha photosynthesis model. Plant Cell Environ 24:27–40

Willis BL (1985) Phenotypic plasticity versus phenotypicstability in the reef corals Turbinaria mesenterina andPavona cactus. Proc 5th Int Coral Reef Symp 4:107–112

Winters G, Loya Y, Roettgers R, Beer S (2003) Photoinhibitionin shallow-water colonies of the coral Stylophora pistillataas measured in situ. Limnol Oceanogr 48:1388–1393

12

Editorial responsibility: Howard I. Browman (AssociateEditor-in-Chief), Storebø, Norway

Submitted: September 16, 2005; Accepted: December 27, 2005Proofs received from author(s): March 29, 2006