glud-biogeochemical responses to mass coral spawning at the great barrier reef- effects on...

7/27/2019 Glud-Biogeochemical Responses to Mass Coral Spawning at the Great Barrier Reef- Effects on Respiration and Pri

1/11

Biogeochemical responses to mass coral spawning at the Great Barrier Reef: Effects on

respiration and primary production

Ronnie N. Glud1

Marine Biological Laboratory, University of Copenhagen, Strandpromenaden 5, DK-3000 Helsingr, Denmark

Bradley D. Eyre and Nicole PattenCenter for Coastal Biogeochemistry, Southern Cross University, P.O. Box 157, Lismore NSW 2480, Australia

Abstract

Coral mass-spawning represents a spectacular annual, short-term, fertilization event of many oligotrophic reefcommunities. The spawning event in 2005 at Heron Island, Great Barrier Reef, was followed by an intense bloomof benthic dinoflagellates. Within a day from the first observed spawning, the primary production of the watercolumn and the benthic compartment increased by factors of 4 and 2.5, respectively. However, the phototrophiccommunities were intensively grazed by macrozoans, and after 45 d the net photosynthesis (P) returned to thepre-spawning background level. The heterotrophic activity (R) mirrored the phototrophic response: a short termof elevated activity was followed by a rapid decline. However, the net autotrophic microbial communitiesexhibited a marked increase in the P : R ratio just after coral mass-spawning, indicating a preferential

phototrophic recycling of nutrients rather than a microbial exploitation of the release of labile organic carbon.The heterotrophic and phototrophic activity of the benthic community exceeded the pelagic activity by ,2- and,5-fold, respectively, underlining the importance of benthic activity for coral reef ecosystem function. Massbalance calculations indicated an efficient recycling of spawn-derived nitrogen (N) and carbon (C) within thebenthic reef community. This was presumably facilitated by advective solute transport within the coarse,permeable, carbonate sand.

Introduction

During a few days every year, most corals in the GreatBarrier Reef (GBR) synchronically release their gametesinto the ambient water. The event is mainly triggered by thelunar phase, although a number of other stimuli may affect

the exact timing (Harrison et al. 1984). The intensity of themass-spawning event shows interannual variations, buttypically .100 coral species participate in a concertedbroadcast release of egg and sperm bundles that initiallyfloat to the water surface. The bundles soon disintegrate,and fertilized eggs develop into planula larvae that settle tothe substrata after a few days. The mass-spawning strategyincreases the probability of fertilization within species andprobably lowers the overall predation level by swampingthe predators in food (Hughes et al. 2000). Several factorssuch as turbidity, temperature, salinity, water depth, and

local hydrodynamics affect the fertilization and settlingsuccess, but the event represents a short-term nutrient andorganic carbon enrichment of the entire reef community.

Coral reefs are highly productive ecosystems despitetheir oligotrophic setting. Reef waters of the GBRgenerally have dissolved inorganic nitrogen (DIN) and

dissolved inorganic phosphorus (DIP) concentrations,,1.0 mmol L21 (Furnas et al. 1990; Ayukai 1993), andthe low pelagic primary production is generally nutrient-limited (e.g., Furnas et al. 2005). The ecosystem productionis typically dominated by benthic primary producers suchas microphytic communities that take advantage ofnutrients released during benthic mineralization, and thesandy sediments of reef flats can contribute a substantialproportion of total ecosystem primary production (John-stone et al. 1990). Studies have indicated that nutrientaddition can further stimulate benthic productivity (Heil etal. 2004), and convincing documentation shows thatmacrofauna-regenerated nutrients stimulate the benthicprimary production (Uthicke and Klumpp 1998).

Coral mass-spawning represents a massive input oflabile, energy-rich material to the reef, and spawn materialforms slicks on the water and sediment surface and alongshorelines of reef islands (Simpson et al. 1993; Wolanski etal. 1989). The mass-spawning event is a feast for much ofthe reef fauna that consume the spawning material(Prachett et al. 2001; Westneat and Resing 1988), andoxygen (O2) depletion events associated with coral mass-spawning have been ascribed to intensified microbialmetabolic activity (Simpson et al. 1993). The benthic O2uptake at the Heron Island reef flat increased by 2.5-fold(from a background level of,90 mmol m22 d21) just after

1 Corresponding author ([email protected]). Presentaddress: Scottish Association for Marine Science (SAMS),Dustaffnage Marine Laboratory, Dunbeg, Oban, Argyll PA371QA, Scotland, United Kingdom.

AcknowledgmentsWe thank Ian Alexander and Anni Glud for laboratory

assistance and sample analysis. We also thank Christian Wildfor his comments on an earlier version of the manuscript and RalfHaese and one anonymous reviewer for constructive reviews thathelped improve the manuscript. The study was supportedfinancially by the Danish Research Agency and a Southern CrossUniversity Visiting Researcher Fellowship (RG), an AustralianResearch Council Discovery Grant (DP0342956) (BE), and theSouthern Cross Postgraduate funding program (NP). Researchpermit G06/18413.

Limnol. Oceanogr., 53(3), 2008, 10141024

E 2008, by the American Society of Limnology and Oceanography, Inc.

1014


2/11

the mass-spawning event in 2001 (Wild et al. 2004a).Subsequently the benthic O2 uptake quickly declined andreached the background level within a period of 810 days.These observations indicated an extremely rapid microbial-

mediated benthic degradation of the spawn material (Wildet al. 2004a). However, the biogeochemical responsetoward this annual fertilization of mass-spawning hasonly been sporadically studied. Through a number of insitu incubations, the current study evaluates, quantifies,and discusses the heterotrophic and autotrophic responseof the pelagic and benthic communities during a mass-spawning event at Heron Island, GBR. Further, weevaluate the extent to which the released material,representing an important nutrition source, is recycledwithin the reef community or exported to the ambient sea.The present manuscript compliments another manuscriptfocusing on the nutrient dynamic during the same coral

mass-spawning event (Eyre et al. 2008).

Materials and methods

Study site and sediment characteristicsAll samplingand in situ measurements were performed during 20November 200528 November 2005 at the inner reef flatoff Heron Island at a distance of ,100 m from theshoreline (Fig. 1A). A full moon was registered on 16November 2005. Heron Island is situated at the southernend of the Great Barrier Reef, 72 km off Gladstone,Queensland, Australia, and the island covers roughly 1.5%of the 26.4 km2 large local reef flat. The inner and central

reef flats are a mosaic of coral clusters separated by patchesof coarse carbonate sands. On average, ,85% of thesurface area at the study site was covered by sand. Thesediment porosity was determined at a depth resolution of

1 cm from the density and the weight loss after 24 h at 75uCof the respective sediment slices. The value at the sedimensurface was 0.57 6 0.02 (vol : vol) (n 5 3) but graduallydeclined to 0.52 6 0.02 (vol : vol) (n 5 3) at a sedimentdepth of 10 cm. The sediment permeability at the onset ofthe study was measured by a constant head permeameter(Klute and Dirksen 1986) for the depth intervals of 05 cmand 510 cm and equaled 6.0 6 0.73 10211 m22 and 1.660.6 3 10211 m22 (n 5 4), respectively. The water depthfollowed the tidal cycle and varied between ,0.3 m and,1.8 m. Due to a broken light sensor no light recordingswere obtained, but the entire period was characterized by aclear sky without any cloud cover.

Chamber measurements and sample analysisTo measure the in situ benthic exchange rate, we deployed threeparallel transparent custom-made Plexiglas chambers. Thecircular chambers had an inner diameter of 190 mm, a heighof 330 mm, were stirred by a rotating disc with a diameter o150 mm and a thickness of 10 mm, and were placed 5 cmbelow the chamber lid (Huettel and Gust 1992). The internawater height during the respective incubations rangedbetween 19 cm and 24 cm (Fig. 1B). Chambers wereinserted at the sampling locations with lifted lids to ensureeffective exchange with ambient water and were continuously stirred for 3 h before the incubations were initiated by

Fig. 1. (A) Heron Island, GBR, with the surrounding reef at low tide (photo by Christian Wild). The black rectangle roughlyencloses the present study site and SB indicates the location of Shark Bay, where investigations on benthic exchange rates have beenperformed previously (Rasheed et al. 2004, Wild et al. 2004b). (B) The applied benthic chambers during incubations at the inner reef flat(C) Slick of coral spawning material washing ashore after the mass-spawning event in 2005. (D) Holothurians grazing on a benthic bloomof dinoflagellates a few days after the coral mass-spawning.

Effects of mass coral spawning 1015


3/11

lid closure at 21:00 h each day. The chambers were thensampled 57 times during the next 20 h, ensuring that soluteconcentration changes during both night and day werefollowed. Thus, the O2 concentrations within the chambertypically varied between 90450 mmol L21 on a diurnalscale. These changes were similar to changes that were

recorded in the ambient water column (see below). After each20-h diurnal incubation the chambers were moved to a newlocation, and the incubation procedure was repeated. Intotal, measurements were performed for nine consecutivedays, and all measurements were confined within an area of,500 m2 (see Fig. 1A).

Most parallel deployments were performed with twochambers stirred at 40 rotations per minute (RPM) whileone chamber was run in diffusive mode, i.e., reversingthe stirring direction at a time interval of 20 s ensuring awell-mixed water phase without the development of a stablepressure gradient (Janssen et al. 2005). On a few occasionsthe three chambers were incubated at the diffusive mode,40 RPM and 80 RPM, respectively, to study the effect ofincreased sediment percolation for the benthic exchangerates. The continuous stirring of the chambers at the givenwater height induced a stable pressure gradient of0.2 Pa dm21 and 1.0 Pa dm21, respectively (Janssen et al2005). Given the initial sediment permeability, temperature,and salinity, the sediment percolation rates could thus becalculated at 43 L m22 d21 and 213 L m22 d21, at 40RPM and 80 RPM, respectively, using the proceduredescribed by Janssen et al. (2005) and Cook et al. (2007).

At each sampling event 100 mL of water was withdrawnfrom each chamber using two 50-mL presoaked (in situwater) plastic syringes. The water was replaced by ambientwater through a valve in the chamber lid, diluting theincubating water by ,2%. The samples were used todetermine the concentrations of O2, dissolved inorganiccarbon (DIC), and nutrients. Only the O2 and DIC data arereported here, while the handling, treatment, and results ofthe nutrient analysis are presented in the companionmanuscript (Eyre et al. 2008). At the end of mostincubations, three samples (,2 cm2) of surface sediment(0.01.0 cm) were collected using cut-off plastic syringes ineach chamber to determine the benthic Chl a concentra-tion. However, the flocculent microalgae cover wasextremely difficult to sample without inducing resuspensionand subsequent loss of phototrophic biomass, and thevalues must thus be regarded as minimum estimates.

During the first few sampling campaigns the O2

concentration was determined in parallel by Winklertitration (Strickland and Parson 1972) and by the use ofan optode-based O2 sensor (Hach, LDO, HQ10). The datasets were inseparable (Fig. 2), and thus the O2 concentra-tions were only measured by the optode in the last half ofthe measuring campaign. Samples for DIC were stored ingastight Exitainers (Labco High Whycombe, UK) spikedwith mercuric chloride (HgCl2) until analysis. Duringanalysis the samples were acidified and the carbon dioxide(CO2) was carried by nitrogen (N2) through an inferred gasanalyzer (ADZ, Analytical Development Company, Hod-desdon, UK, 225, MK3, LTD). The Chl a samples werefrozen at 220uC in acid-washed 10-mL polyethylene vials

until extractions in 90% acetone and spectrophotometricdeterminations of the Chl a concentrations could beperformed (Strickland and Parson 1972).

Benthic exchange rates were calculated from themeasured concentration changes, accounting for thesample dilution, the enclosed water volume, and thesediment area. Nighttime incubations integrated theactivity from 21:00 h to 05:00 h, while daytime incubationintegrated the period from 05:00 h to 18:00 h. Incubationsof ambient water (see below) revealed that the activity ofthe overlying water during chamber incubations, onaverage, only accounted for 6% of the concentrationchanges observed during the chamber incubations.

Water column measurementsWater samples for nutri-ents and Chl a were collected 25 times each day in two 1-Lacid-washed and -rinsed polyethylene containers. Treat-ment of nutrient samples and nutrient data are presentedelsewhere (Eyre et al. in press). The sampling volume wasfiltered through a glass microfiber filter (Whatmann GF/F), and the filter was frozen for later determinations of Chla concentrations. In addition to water sampling, temper-ature, salinity, and pH were measured in situ by acalibrated Q-10 multiprobe (Hydrolab), and the O2

concentrations were measured by the handheld O2 optode.To measure pelagic net photosynthesis and respiration,34 250-mL glass bottles of seawater were collected eachday at sunrise, and the bottles were left suspended near themiddle of the water column at the sampling site to incubateuntil sunset. Net photosynthesis was calculated from themeasured increase in O2 concentration. Parallel bottleswrapped in light-proof calico bags were used to quantifythe respiration. Nighttime respiration was determined in asimilar manner by initiating bottle incubations at sunset.Bottle breakage caused by swift tidal currents forcing thebottles against corals limited the number of incubationsand replicates that could be performed.

Fig. 2. The O2 concentration in water samples from selectedchamber incubations as measured with a macro-optode plotted

against O2 determinations made by parallel Winkler titrations.The best linear regression fitted to the data points is included.

1016 Glud et al.


4/11

Results

Spawning and water-column observationsThe first coralspawning in 2005 was observed during the night of 20November. From direct in situ observation it was estimatedthat ,20% of the dominant acroporid corals (Montipora

spp. and Acropora spp.) spawned that night, while theremaining ,80% of the specimens spawned the followingnight. Most of the massive corals (e.g., Platygyra daedalea)spawned the night of 22 November 2005, and a fewspecimens spawned the following night. The spawningevent thus lasted 4 d, but it was most intense on the nightsof 21 November 2005 and 22 November 2005. Initially, thereleased eggsperm bundles floated to the water surface.Following the major nights of spawning, spawn materiallocally aggregated into massive slicks (Fig. 1C). In someareas slicks covered the sediment, and it was observed thatsinking spawn material percolated into the coarse sand dueto tide and wind-induced forcing.

The shallow-water column did not express any stratifi-cation and was always well-mixed. Salinity remained at aconstant value of 37.4 6 0.1 (n 5 35) during the entirestudy period, while the pH showed peak values of 8.4around noon and minimum values of 7.8 at sunrise (datanot shown), reflecting the net fixation and net release ofinorganic carbon during daytime and nighttime, respec-tively. Water temperature at the inner reef expressed adiurnal variation of up to 5uC, regulated by the solarirradiance and the tidal exchange (Fig. 3). Overall, therewas a trend of increasing temperatures until 25 November2005, where after temperatures stabilized. The O2 concen-trations exhibited an extensive diurnal variation primarilydriven by the respiration and photosynthesis at the reef

(Fig. 3C). The O2 concentration reached a minimum of52 mmol L21 (


5/11

explain ,20% of the observed O2 accumulation rate asobserved during 22 November 2005. A significant fractionof the ecosystem-based primary production thus occurredwithin the benthic community. Similarly, for the hetero-trophic activity, only on the night of 22 November 2005,the night of the most intense spawning, did the integratedpelagic O2-consumption rate exceed the benthic O2-consumption rate (see below).

Benthic O2 and DIC exchange during and after coral mass-spawningThe benthic net photosynthesis as measured bythe chambers stirred at 40 RPM on 20 November 2005 and

21 November 2005 was 144 6 11 mmol O2 m22 d21 (n 54), while the benthic O2 consumption during nighttimeequaled 72 6 5 mmol O2 m22 d21 (n 5 4) (Fig. 5A). Thenet photosynthesis increased by a factor of ,2 on thesecond day after the spawning event, a level that wasmaintained for four consecutive days before it declined tothe original values (Fig. 5A). The nighttime O2-consump-tion rate mirrored this development, but O2 uptake was notstimulated to the same extent as the daytime netphotosynthesis rate (Fig. 5B). The benthic communitywas net autotrophic during the entire study period, andthe net photosynthesis : net respiration ratio (P : R) was ,2

Fig. 4. (A) Net O2 production, (B) net O2 consumption and (C) Chl a content of the watercolumn at the inner reef flat of Heron Island around the coral mass-spawning. Error barsrepresent the standard deviation of replicate measurements for the respective days ( n 5 36).Respiration rates were measured at day and nighttime; the latter was measured in a bottlewrapped in light-proof calico bags (no replication). The grey bars indicate the spawning intensityas described in Fig 3. NM means no measurements. Due to breakage of bottles, no measurementswere performed after 26 November 2005.

1018 Glud et al.


6/11

at the beginning and end of the study, but ranged between2.5 and 3.5 during the five central days of the study(Fig. 5B).

The stimulated phototropic activity was associated witha dense olive-green cover on the sediment surface that wasintensively grazed by fish and holothurians (see Fig. 1D).On-site microscopic investigations revealed that the benthicbloom was dominated by a dinoflagellate, presumably aProrocentrum sp., even though some pennate diatoms alsowere encountered. Unfortunately, the benthic Chl a wasfirst quantified for 22 November 2005 (i.e., two days afterthe initial spawning), but the maximum value of 15.2 mg

Chl a m22 was reached on 25 November 2005. Thereafterthe values gradually declined to a minimum of 11.1 mg Chla m22 on the last sampling day (data not shown), a trendthat probably proceeded after the measuring campaignThe values for benthic microphytic biomass must beregarded as minimum estimates because the loose floccu-

lent dinoflagellate cover was very difficult to samplewithout any resuspension and subsequent loss.The chambers stirred at 40 RPM consistently gave

higher benthic O2-exchange rates than the chambers stirredin the diffusive mode (Fig. 5A). This was most explicitduring daytime and in the beginning and end of the studyperiod when the benthic net-photosynthesis was at itminimum (Fig. 6). During 26 November 2005 the sedimentwas still covered by a layer of benthic dinoflagellates, andincreased stirring only marginally stimulated the nephotosynthesis during the day; changing the diffusivestirring mode to 80 RPM only enhanced the net photo-synthesis rate by 24% (246305 mmol m22 d21). The

Fig. 6. The O2 exchange rates measured at three differenstirring modes during daytime and nighttime (A) 26 November2005 and (B) 27 November 2005, respectively (no replicationwas performed).

Fig. 5. (A) The benthic net-photosynthesis and O2 consump-tion measured during daytime (open symbols) and nighttime(closed symbols), respectively. Two chambers were stirred at 40RPM while one chamber was stirred in diffusive mode (on twooccasions on 26 November 2005 and 27 November 2005 only one40-RPM chamber was deployed). Negative values indicate uptakeand error bars represent the range of the two parallel chambersand were in many instances smaller than the symbol size. (B) Theratio between the average benthic net-photosynthesis (12 h) andthe average O2 consumption measured during nighttime (12 h) atthe respective dates. The grey bars indicate the spawning intensityas in Fig. 3.



7/11

corresponding value for the nighttime respiration was 29%(86113 mmol m22 d21). However, on 27 November 2005the visual dinoflagellate cover had gone, and the absoluteO2 exchange rates were lower but were markedly stimulatedby increased stirring especially during daytime (Fig. 6B).Here the net photosynthesis rate increased by 160% (80

209 mmol m22 d21) when the stirring was shifted fromdiffusive to 80 RPM. The nighttime O2 consumptionincreased by 85% (61110 mmol m22 d21) following thesame shift in stirring mode.

In parallel to the O2 exchange rates we also determinedthe DIC exchange rates. In general the temporal dynamicsmirrored the O2 fluxes, but the scatter was higher (Fig. 7A).On average the DIC uptake was 1.30 6 0.11 times largerthan the O2 release during daytime, while the DIC releaseduring nighttime was 1.57 6 0.37 times larger than the O2uptake. Deviation from a 1.0 ratio presumably reflects acombination of incomplete reoxidation of reduced inor-ganic components arising during anaerobic mineralizationand carbonate dissolution contribution to the DICexchange (see below). The mismatch was largest when theabsolute exchange rates were high (Fig. 7B).

Discussion

Microbial heterotrophic activity during coral spawningBenthic microbially-mediated heterotrophic activity is animportant component in the recycling of material on coralreefs (e.g., Clavier and Garrigue 1999; Rasheed et al. 2004).Most reef flats are dominated by coarse, permeable,carbonate sand, and advection induced by wave action,tidal currents, or differences in water level between reeflagoons and surrounding channels ensures benthic entrap-

ment of particulate material and efficient supply of oxygenfor benthic degradation processes (Huettel et al. 1996;Werner et al. 2006; Huettel et al. 2006). Mucus released bycorals represents a significant fraction of the benthiccarbon input in reef communities, and it acts as animportant energy carrier by scavenging the water columnfor particulate material before being filtered into the seabed(Wild et al. 2004b,c; Huettel et al. 2006). The total benthicO2 uptake (TOU) expresses a proxy for the integratedbenthic mineralization rate (e.g., Canfield et al. 1993), andrates measured on coral reef flats are often similar or evenhigher (Johnstone et al. 1990; Clavier and Garrigue 1999;Wild et al. 2005) than values from similar water depths ineutrophicated systems (Wild et al. 2005; Cook et al. 2007,

and references therein). Measurements from darkenedchambers deployed in the daytime in Shark Bay at HeronIsland (Fig. 1A) gave TOU rates ranging between25 mmol m22 d21 and 86 mmol m22 d21 (Rasheed et al.2004; Wild et al. 2004b). Another study inferring TOUfrom O2 microprofiles at four locations around HeronIsland reflected considerable local variations with mini-mum median values in Shark Bay of 57 mmol m22 d21 andmaximum values of 197 mmol m22 d21 in channels closeto the shoreline (Werner et al. 2006). Further, this studyconcluded that aerobic respiration dominated the benthicmineralization and was the most important O2 consumingprocess in the reef sediments (Werner et al. 2006).

In the present study, the nighttime background respira-tion prior to spawning at moderate stirring (40 RPM) wasintermediate to the values measured in Shark Bay [72 65 mmol m22 d21 (n 5 4)] and confirms that the benthic O2uptake is substantial despite the oligotrophic setting.However, the TOU increased after the mass-spawningand remained at a plateau of 106 6 6 mmol m22 d21 (n 58) for 45 consecutive nights (22 Nov 0526 Nov 05) beforereturning to the base level. Integrating the elevated benthic

Fig. 7. (A) The DIC exchange measured during daytime(open symbols) and nighttime (closed symbols), respectively. Allchambers were stirred at 40 RPM (no replication on 26 Nov 05and 27 Nov 05). Negative values indicate uptake and error barsrepresent the range for the two chambers and were, in oneinstance, smaller than the symbol size. The grey bars indicate thespawning intensity as described in Fig. 3. (B) The mean O2 net-production plotted against the mean DIC uptake for the stirred

benthic chambers incubated during daytime for the respectivedays. The best linear regression fitted to the data is included.

1020 Glud et al.


8/11

O2 consumption (TOU background level) and assumingthat this increase reflects stimulated heterotrophic activity,that the respiratory quotient (RQ) is 1.0, and thatrespiration during the daytime and nighttime are similar,

the elevated benthic O2-consumption rate corresponds tothe mineralization of 2.0 g C m22. Using the TOU ratherthan just the fraction exceeding the background level, thetotal benthic mineralization for this period amounts to6.4 g C m22. Despite careful inspection, no macrofaunawas observed during any chamber incubations, and thebenthic activity is mainly ascribed to microbial carbonturnover even though meiofauna could have contributedslightly. Sediment-trap data from the inner reef flatresolved an average vertical POC transport rate of 7.8 gC m22 (Table 1) obtained during the same period(calculated from Wild et al. 2008), which corresponds tothe total benthic heterotrophic activity. In general, paralleltrap deployments suggested a relatively even distribution ofsediment material at the reef flat (Wild et al. 2008).

In principle, the DIC exchange rate should reflect the netmineralization, including equivalents from the anaerobicdegradation not being oxidized. Previous studies of coralmucus degradation at Heron Island reflected benthicDIC:O2 exchange ratios of 1.51.8, and the discrepancyfrom a 1.0 ratio was ascribed to a transient accumulation ofreduced sulfur compounds leaving little room for contri-butions from concurrent carbonate dissolution (Wild et al2004b). We attempted to quantify the calcium carbonate(Ca2+) exchange rate during our chamber incubations inorder to asses any potential contribution from carbonatedissolution, but on a variable background .1 mmol L21

we where not able to resolve any net exchange of Ca2+

(datanot shown). Although we cannot exclude any carbonatedissolution, the net DIC and TOU exchange rates willbracket the benthic carbon turnover. Applying the DICexchange rates, the elevated activity corresponded to themineralization of 4.4 g C m22, and the total DIC exchangein the equivalent period was 10.8 g C m22. Taken together,the benthic exchange rates suggest a very efficient recyclingof spawning or spawning-derived material during the studyperiod.

The heterotrophic activity in the water column peakedduring the first night of intense spawning and exhibited agradual decline during the study period. The background

level of respiration in the water column was not wellconstrained, but it appeared to have been very low(,0.4 mmol L21 h21). Integrating the pelagic O2 consumption rate from the 22 November 2005 to 26 November 2005(accounting for the water depth and rates measured duringboth daytime and nighttime, Fig. 4B), it corresponded to

the mineralization of 3.0 g C m22 (equivalent to 2847% ofthe total benthic activity of 6.410.8 g C m22).

During the mass-spawning event at the Heron Island reefin 2001 it was estimated that a total of 310.0 kg of C and18.0 kg of N were released into the ambient water in theform of coral eggs (Wild et al. 2004a). If all of this remainson the reef, it roughly corresponds to an average of 11.7 gC m22 and 0.7 g N m22 for the entire reef area of 26.4 km2

The sum of the entire elevated heterotrophic activity asinferred from the O2 consumption rates in the watercolumn and the benthic DIC-O2 exchange during theperiod from 22 November 2005 to 26 November 2005corresponds to the mineralization of 5.07.4 g C m22

Assuming that the spawning event in 2005 occurred on thesame scale as the spawning event in 2001, the heterotrophicactivity of the water column and the sediment thucorresponds to the degradation of 26% and 1738% (atotal of 4364%) of the carbon released as coral eggsrespectively (Table 1).

The results from our chamber incubations performedduring the nighttime align with the observations of Wild etal. (2004a). During a study in Shark Bay (Fig. 1A) in 2001they observed a rapid 2.5-fold increase in the benthic O2consumption rate that gradually approached the background level ,9 d after the first spawning was observedThey concluded that the elevated benthic activity corresponded to 490 mmol O2 m22, equivalent to 5.9 g C m22 or50% of the carbon released as coral eggs. (Note that theauthors in the original manuscript did a miscalculation andthus concluded that the benthic activity only correspondedto 0.2% of the carbon release as coral eggs [Wild et al2004a].) No independent assessment of the water-columnactivity was performed during this study, and the darkened-chamber incubations were performed during the daytime.

To the extent that the two study sites are representativefor the Heron Island reef flat in general, the twoinvestigations imply an efficient microbially-mediatedbenthic and pelagic recycling of spawning-related OC athe reef flat. The microbial diagenetic activity is abruptlyand markedly stimulated just after coral spawning, and theelevated metabolism corresponds to a significant fraction

of the estimated release of spawning material.The simplified approaches described above entirelyascribe the elevated heterotrophic activity to degradationof spawning material and thereby ignore a potential foincreased respiration following stimulated photosynthesisIn fact, both the sediment and the water column remainednet-autotrophic during the entire study period and thebenthic P : R ratio reached an elevated plateau of 2.53.0during 22 November 200526 November 2005. Thissuggests that a fraction of the elevated heterotrophicactivity measured during nighttime (or in darkenedchambers) can be ascribed to enhanced respiration of theaccumulation phototrophic biomass or a stimulated

Table 1. Heterotrophic activity, average coral egg release,and particulate organic carbon (POC) sedimentation at the innerreef flat (22 Nov 0526 Nov 05).

Compartment Parameter Rate (g C m22)

Sediment Total activity 6.410.8 Elevated activity 2.04.4Water column Total activity 3.1 Elevated activity 3.0*Sedimentation of POC** 7.8Total estimated coral egg

release***11.7

* Assuming a background respiration of 0.4 mmol L21 h21** Calculated from Wild et al. (2008)*** In 2001 (Wild et al. 2004a).



9/11

prokaryotic respiration following increased leakage ofphotosynthesates. This would imply a preferential photo-trophically-mediated recycling of nutrients rather thanmicrobial exploitation of the bulk of OC released duringcoral spawning.

Stimulation of phototrophic activityThe benthic net-photosynthesis rates measured prior to coral spawningwere in the high end of reports from coastal settings (e.g.,Cahoon 1999; Glud et al. 2002; Jahnke et al. 2000) butcomparable to the relatively few available measurementsfrom coral reefs (Table 2). This activity was, however,markedly stimulated after the spawning event. During fiveconsecutive days after 22 November 2005 the net photo-

synthesis rate was 281 6 15 mmol m22 d21 (n 5 7),ranking among the highest measurements performed in reefsettings (Table 2). The daytime, benthic, net-photosynthet-ic activity for the period 22 November 200526 November2005 integrates to 8.4 g C m22, while the correspondingvalue for the water column is 1.7 g C m22. The total netphotosynthesis for the two compartments is 10.1 g C m22.In addition, due to occasional formation of gas bubbles insome chambers, the benthic rates must be regarded asminimum values because the O2 present in bubbles was notaccounted for in the exchange measurements. The relativelygood balance of 1.30 6 0.11 between DIC and O2 exchangerates, an expression of the photosynthetic quotient (PQ),

does, however, suggest that this problem was of minorimportance.These values underline the importance of benthic

primary production in reef communities. Note that theseare net rates and thus include the daytime respiration of thetwo compartments. The gross rates may have beensubstantially higher, but are difficult to access. Eventhough it is a common procedure to estimate gross primaryproduction by simply adding the activity of incubationsperformed in darkened chambers and bottles to incuba-tions performed in light, the approach severely underesti-mates gross production by not accounting for thelight-stimulated respiration (Epping and Jrgensen 1996;

Fenchel and Glud 2000). Microsensor studies have revealedthat respiration in light-exposed benthic phototrophiccommunities exceed nighttime respiration by a factor ofat least 1.41.8 (Fenchel and Glud 2000 and referencestherein). Elevated O2 consumption during the daytime maybe even more pronounced in permeable systems where

advection can transport the O2 produced along thesediment surface deep into otherwise anoxic sedimentlayers (Precht et al. 2004). The benthic exchange ratesmay also be affected by the short- and long-term variationsin the in situ temperature to some extent. Temperatureaffects many of the factors regulating the benthic O2-exchange rate (i.e., water viscosity, solubility, biologicalactivity). In general, increasing temperatures shift benthicphototrophic communities toward a stronger heterotrophyresponse as O2-consuming processes are stimulated morethan photosynthesis with increasing temperature (Hanckeand Glud 2004). This indicates that the stimulatedautotrophic response following spawning potentially couldhave been even higher had it not been for the concurrenttemperature increase (Fig. 3B). However, it is a complicat-ed task to ascertain the temperature effect from the in situbenthic metabolic response.

Accounting for the nighttime respiration, the photo-trophic activity corresponded to a net accumulation ofbenthic phototrophic biomass on the order of 5.2 g C m22

during the period from 22 November 2005 to 26 November2005, and the corresponding value for the water columnwas 1.3 g C m22. Indeed the pelagic phototrophic biomassincreased 26 fold during the study, as facilitated by theincrease in Chl a. However, due to a lack of pre-spawn datawe have no measurement of the increase in benthicphototrophic biomass. Nevertheless, the benthic biomasswas also intensively grazed by fish and holothurians, aheterotrophic activity not included in our chamber (orbottle) incubations. We have no means to discriminatebetween the O2 consumption related to mineralization offreshly produced phototrophic biomass and the degrada-tion of spawn-related material, but benthic dinoflagellatesdo appear to be an important component in the nutrientrecycling following coral mass-spawning. Some insight can,however, be gained by evaluating the N-mineralization fluxand the C : N ratio of the potential sources of organicmaterial (see Eyre et al. 2008).

Coral eggs have a relatively high C : N ratio of 1621(Wild et al. 2004a), but the mass-spawning event still musthave enhanced the pool of available nutrients which

presumably induced the benthic bloom of dinoflagellatedeveloping just after the event began. Phosphorus wasapparently not a limiting factor at any time due to a largebioavailable benthic phosphorus pool (Eyre et al. 2008).The bloom was most likely dominated by a Prorocentrumsp. (as evaluated by on-site microscopic investigations) thatapparently took advantage of nutrients release by miner-alization of the spawning material. However, mixotrophy iscommon among dinoflagellates, including Prorocentrumspp. that can complement their nutritional needs byassimilating of organic material (Jacobson and Anderson1996; Stoecker et al. 1997). In Prorocentrum minimum thefeeding behavior is seen as a mechanism for obtaining

Table 2. Benthic O2 production rates in shallow, coral-sandreef flats as derived from chamber incubations. Values arerecalculated to standardized units from the respective sourcesand represent exchanges rates measured in light withoutsubtraction of nighttime or daytime respiration.

SourceWater

depth (m)Day time photosynthesis

(mmol m22 d21)

Sournia (1976) 0.51.0 86265Uthicke & Klumpp

(1998)0.53.5 1332

Clavier & Garrigue(1999)

14.0 510

Rasheed et al. (2004) 0.22.5 61137Johnstone et al. (1990) 1.07.0 95285Present study prior to

spawning0.31.8 141148

Present study just afterspawning

0.31.8 232317

1022 Glud et al.


10/11

limiting nutrients rather than a supplementation of theircarbon nutrition (Stoecker et al. 1997). Thus initialbacterial degradation of organic carbon might not havebeen necessary to facilitate the N demand of the dominantbenthic phototroph. We speculate that the bloomingorganism rapidly exploited the available PON and DON

pools and that the net surplus of N facilitated enhancedgrowth and photosynthesis.

Importance of advective porewater transportDespitehigh nutrient and DIC levels in sediments, these com-pounds may become limiting in dense microphytobenthiccommunities due to diffusion limitations (Glud et al. 1992;Dalsgaard 2003). However, in permeable sediment thesolute transport rate may be enhanced by orders ofmagnitude due to advection (Precht and Huettel 2003).Recent investigations have documented that advection canrelieve DIC limitation of benthic microphytes in permeablesilicate sand (Cook and Ry 2006). In the present studysediment percolation consistently resulted in higher O2exchange rates. Without a dense microphytobenthic cover,the net photosynthesis rate gradually increased up to 2.6fold as the stirring was shifted from diffusive topercolation rates of 43 L m22 d21 and 213 L m22 d21,respectively. This clearly indicated that continuous advec-tion was required to maintain the high photosyntheticactivity. Given that the sediment present consisted ofcarbonate sand, the stimulation was most likely associatedwith an enhanced N supply. The nighttime respiration wasalso enhanced by sediment percolation, and this can beascribed to stimulated O2 transport to the deeper sedimentlayers. Increased stirring also enhanced the O2 exchangerates during the benthic dinoflagellate bloom wherecompetition for nutrients was presumably higher; theresponse was, however, markedly lower. This was mostlikely caused by reduced sediment permeability during thebloom impeding the advective porewater transport. How-ever, despite several attempts, we were not able to samplesediment cores with an intact microphytobenthic cover,and thus no trustworthy permeability measurements couldbe made during this phase of the study. Nevertheless,advection appears to be important in sustaining the highbenthic activity of the reef flat, but without concurrentincrease in grazing pressure the process may to some extentbe transiently counteracted by the accumulation ofmicrobial biomass clogging up the filter capabilities of thesand.

Like most studies on permeable sand, it is difficult toevaluate the in situ porewater percolation rates at the reefflat and compare them to the percolation rates induced bythe chamber incubations especially because the permeabil-ity of the reef flat, wind, tide, and wave forcing changedduring the study. We have, however, applied the samechamber configurations as most reliable benthic exchangestudies conducted in silicate and carbonate sands (Cook etal 2007 and references therein). The initial percolation ratesranged between 43 L m22 d21 and 213 L m22 d21 forchambers stirred at 40 RPM and 213 RPM, respectively.This range is similar to those of the few existing estimateson in situ flushing rates in permeable sediments: 5

130 L m22 d21 (Cook et al. 2007), 60130 m22 d21 (Prechet al. 2004), and 50100 m22 d21 (Precht and Huette2003). Future investigations should be required to assess insitu porewater percolation especially at coral reefs.

Coral reefs represent highly productive ecosystems inoligotrophic settings and thus rely on an efficient nutrient

recycling. Several nutrient addition experiments on corareefs have been performed, but opposing conclusions onthe microphytic response towards fertilization have beenmade (e.g., Koop et al. 2001; Furnas et al. 2005). Coralmass-spawning on the Great Barrier Reef represents ashort-term natural fertilization event, and the present studyconcludes that this markedly stimulated the primaryproductivity. Benthic exploitation of the released nutrientsled to a distinct bloom of dinoflagellate that was grazedintensively by the macrozoans. The heterotrophic activitywas stimulated partly by the intensified primary productionand partly by the labile carbon released during the mass-spawning event. Advection apparently facilitated thebenthic turnover of both carbon and nutrients, and asignificant fraction of the released spawning products wererecycled within the reef flat area.

References

AYUKAI, T. 1993. Temporal variability of the nutrient environ-ment on Davis Reef in the central Great Barrier ReefAustralia. Pacific Science 47: 171179.

CAHOON, L. B. 1999. The role of benthic microalgae inneritic ecosystems. Oceanogr. Mar. Biol. Annu. Rev. 374786.

CANFIELD, D. E., AND oTHERS. 1993. Pathways of organic carbonoxidation in three continental margin sediments. Mar. Geol113:

2740.CLAVIER, J., AND C. GARRIGUE. 1999. Annual sediment primaryproduction and respiration in a large coral reef lagoon (SWNew Caledonia). Mar. Ecol. Prog. Ser. 191: 7989.

COOK, P. L. M., AND H. RY. 2006. Advective relief of CO2limitation in microphytobenthos in highly productive sandysediments. Limnol. Oceanogr. 51: 15941601.

, F. WENZHOEFER, R . N . GLUD, F . JANSSEN, AND MHUETTEL. 2007. Benthic solute exchange and carbon mineralization in two shallow subtidal sandy sediments: Effect ofadvective porewater exchange. Limnol. Oceanogr. 5219431963.

DALSGAARD, T. 2003. Benthic primary production and nutriencycling in sediments with benthic microalgae and transienaccumulation of macroalgae. Limnol. Oceanogr. 48

21382150.EPPING, E. H. G., AND B. B. JRGENSEN. 1996. Light-enhancedoxygen respiration in benthic phototrophic communitiesMar. Ecol. Prog. Ser. 139: 193203.

EYRE, B. D., R. N. GLUD, AND N. PATTEN. 2008. Mass coraspawninga natural large-scale nutrient addition experimentLimnol. Oceanogr. 53: 9971013.

FENCHEL, T., AND R. N. GLUD. 2000. Benthic primary productionand O2CO2 dynamics in a shallow-water sediment: Spatiaand temporal heterogeneity. Ophelia 53: 159171.

FURNAS, M . J . , A . W . MITCHELL, M . GILMARTIN, AND NREVELANTE. 1990. Phytoplankton biomass and primaryproduction in semi-enclosed reef lagoons of the central GreatBarrier Reef, Australia. Coral Reefs 9: 110.



11/11

FURNAS, M., A. MITCHELL, M. SKUZA, AND J. BRODIE. 2005. In theouter 90%: Phytoplankton responses to enhanced nutrientavailability in the Great Barrier Reef Lagoon. Mar. Pollut.Bull. 51: 253263.

GLUD, R . N . , N . B . RAMSING, AND N. P. REVSBECH. 1992.Photosynthesis and photosynthesis-coupled respiration innatural biofilms quantified with microsensors. J. Phycol. 28:5160.

, M. KUHL, F . WENZHOEFER, AND S. RYSGAARD. 2002.Benthic diatoms of a high arctic fjord (Young Sound, NEGreenland): Importance for ecosystem primary production.Mar. Ecol. Prog. Ser. 238: 1529.

HANCKE, K., AND R. N . GLUD. 2004. Temperature effects onrespiration and photosynthesis in three diatom dominatedbenthic communities. Aqua. Microb. Ecol. 37: 265281.

HARRISON, P. L., R. C. BABCOCK, G. D. BULL, J. K. OLIVER, C. C.WALLACE, AND B. L. WILLIS. 1984. Mass-spawning in tropicalreef corals. Science 223: 11861189.

HEIL, C. A., K. CHASTON, A. JONES, P. BIRD, B. LONGSTAFF, S.COSTANZO, AND W. C. DENNISON. 2004. Benthic microalgae incoral reef sediments of the southern Great Barrier Reef,Australia. Coral Reefs 23: 336343.

HUETTEL, M., AND G. GUST. 1992. Solute release mechanisms fromconfined sediment cores in stirred benthic chambers andflume flows. Mar. Ecol. Prog. Ser. 82: 22412249.

, W. Z IEBIS, AND S. FORSTER. 1996. Flow-induced uptake ofparticulate matter in permeable sediments. Limnol. Oceanogr.41: 309322.

, C. WILD, AND S. GONELLI. 2006. Mucus trap in coral reefs:Formation and temporal evolution of particulate aggregatescaused by coral mucus. Mar. Ecol. Prog. Ser. 307: 6984.

HUGHES, T. P., A. H. BAIRD, E. A. DINSDALE, N. A. MOLTSCHA-NIWSKYJ, M. S. PRATCHETT, J. E. TANNER, AND B. L. WILLIS.2000. Supply-side ecology works both ways: The link betweenbenthic adults, fecundity, and larval recruits. Ecology 81:22412249.

JACOBSON, D. M., AND D. M. ANDERSON. 1996. Widespread

phagocytosis of ciliates and other protists by marinemixotrophic and heterotrophic thecate dinoflagellates. J.Phycol. 32: 279285.

JAHNKE, R. A., J. R. NELSON, R. L. MARINELLI, AND J. E. ECKMAN.2000. Benthic flux of biogenic elements on the SoutheasternUS continental shelf: Influence of porewater advectivetransport and benthic microalgae. Con. Shelf Res. 20:109127.

JANSSEN, F., P. FAERBER, M. HUETTEL, V. MEYER, AND U. WITTE.2005. Pore-water advection and solute fluxes in permeablemarine sediments (I): Calibration and performance of thenovel benthic chamber system Sandy. Limnol. Oceanogr. 50:786778,

JOHNSTONE, R. W ., K . KOOP, AND A. W . D . LARKUM. 1990.Physical aspects of coral reef lagoon sediments in relation todetritus processing and primary production. Mar. Ecol. Prog.Ser. 66: 271283.

KLUTE, A., AND C. DIRKSEN. 1986. Hydraulic conductivity anddiffusivity: Laboratory methods, p. 687700. In A. Klute[ed.], Methods of soil analysispart Iphysical and miner-alogical methods. American Society of Agronomy.

KOOP, K., AND oTHERS. 2001. ENCOREThe effect of nutrientenrichment on coral reefs: Synthesis of results and conclusionsMar. Pollut. Bull. 42: 91120.

PRATCHETT, M. S., N. GUST, G. GOBY, AND S. O. KLANTEN. 2001.Consumption of coral propagules represent a significanttrophic link between corals and reef fish. Coral Reefs 20:1317.

PRECHT, E., AND M. HUETTEL. 2003. Advective pore-waterexchange driven by surface gravity waves and its ecologicalimplications. Limnol. Oceanogr. 48: 16741684.

, U. FRANKE, L . POLERECKY, AND M. HUETTEL. 2004.Oxygen dynamics in permeable sediments with wave-drivenporewater exchange. Limnol. Oceanogr. 49: 693705.

RASHEED, M ., C . WILD, U . FRANKE, AND M. HUETTEL. 2004.Benthic photosynthesis and oxygen consumption in perme-able carbonate sediments at Heron Island, Great Barrier Reef,Aust. Estuar. Coast. Shelf Science 59: 139150.

SIMPSON, C. J., J. L. CARY, AND R. J. MASINI. 1993. Destruction ofcorals and other reef animals by coral spawn slicks onNingaloo Reef, Western Australia. Coral Reefs 12: 185191.

SOURNIA, A. 1973. Primary production of sands in the lagoon ofan atoll and the role of benthic foraminifera symbionts. Mar.Biol. 37: 2932.

STOECKER, D. K., A. LI, D. W. COATS, D. E. GUSTAFSON, AND M.K. NANNEN. 1997. Mixotrophy in the dinoflagellate Pro-rocentrum minimum. Mar. Ecol. Prog. Ser. 152: 112.

STRICKLAND, J. D., AND T. R. PARSON. 1972. A practical handbookof seawater analysis, 2nd ed. Bull. Fish. Res. Bor. Can., 167.

UTHICKE, S., AND D. W. KLUMPP. 1998. Microphytobenthoscommunity production at a near-shore coral reef: seasonalvariation and response to ammonium recycled by holothuri-ans. Mar. Ecol. Prog. Ser. 169: 111.

WERNER, U., P. BIRD, C. WILD, T. FERDELMAN, L. POLERECKY, G.EICKERT, R . JOHNSTONE, O . HOEGH-GULDBERG, AND D. DEBEER. 2006. Spatial patterns of aerobic and anaerobicmineralization rates and oxygen penetration dynamics incoral reef sediments. Mar. Ecol. Prog. Ser. 309: 93105.

WESTNEAT, M., AND J. RESING. 1988. Predation on coral spawn byplanktivorous fish. Coral Reefs 7: 8992.

WILD, C., R. TOLLRIAN, AND M. HUETTEL. 2004a. Rapid recyclingof coral mass-spawning products in permeable reef sediments.Mar. Ecol. Prog. Ser. 271: 159166.

, M. RASHEED, U. WERNER, U. FRANKE, R. JOHNSTONE, AND

M. HUETTEL. 2004b. Degradation and mineralization of coralmucus in reef environments. Mar. Ecol. Prog. Ser. 267:159171.

, M. HUETTEL, A. KLUETER, S. G. KREMB, M. RASHEED,AND B. B. JRGENSEN. 2004c. Coral mucus functions as anenergy carrier and particle trap in the reef ecosystem. Nature428: 6670.

, M. RASHEED, C . JANTZEN, P . COOK, U . STRUCK, M.HUETTEL, AND A. BOETIUS. 2005. Benthic metabolism anddegradation of natural particulate organic matter in carbon-ate and silicate reef sands of the Northern Red Sea. Mar.Ecol. Prog. Ser. 298: 6978.

, C. JANTZEN, U. STRUCK, O. HOEGH-GULDBERG, AND M.HUETTEL. 2008. Biogeochemical responses to coral mass-spawning at the Great Barrier Reef: Pelagicbenthic coupling.

Coral Reefs 27: 123132.WOLANSKI, E., D. BURRAGE, AND B. KING. 1989. Trapping anddispersal of coral eggs around Bowden Reef, Great BarrierReef, following mass coral spawning. Cont. Shelf Res. 9:479496.

Received: 3 August 2007Accepted: 22 December 2007

Amended: 11 January 2008

1024 Glud et al.

glud-biogeochemical responses to mass coral spawning at the great barrier reef- effects on...

Documents