at SciVerse ScienceDirect

Estuarine, Coastal and Shelf Science 99 (2012) 42e49

Contents lists available

Estuarine, Coastal and Shelf Science

journal homepage: www.elsevier .com/locate/ecss

Invasive algal mats degrade coral reef physical habitat quality

Jonathan A. Martinez a,b,c,*, Celia M. Smith b, Robert H. Richmond a

aKewalo Marine Laboratory, University of Hawai‘i at M�anoa, 41 Ahui St, Honolulu, HI 96813, USAbBotany Department, University of Hawai‘i at M�anoa, 3190 Maile Way, Honolulu, HI 96822, USAcNational Oceanic and Atmospheric Administration ONMS PIR 6600 Kalanianaole Hwy, Suite 302, Honolulu, HI 96825, USA

a r t i c l e i n f o

Article history:Received 10 November 2011Accepted 10 December 2011Available online 30 December 2011

Keywords:algaeGracilaria salicorniacoral reefdiurnal respirationacidificationpH

* Corresponding author.E-mail address: jmartin1@hawaii.edu (J.A. Martine

0272-7714/$ e see front matter Published by Elseviedoi:10.1016/j.ecss.2011.12.022

a b s t r a c t

Invasive species alter the ecology of marine ecosystems through a variety of mechanisms or combinationof mechanisms. This study documented critical physical parameters altered by the invasive red macro-alga Gracilaria salicornia in situ, including: reduced irradiance, increased sedimentation, and markedvariation in diurnal dissolved oxygen and pH cycles in K�ane‘ohe Bay, O‘ahu, Hawai‘i. Paired studiesshowed that algal mats reduced irradiance by 99% and doubled sediment accumulation. Several matsdeveloped hypoxia and hyperoxia in the extreme minima and maxima, though there was no statisticaldifference detected in the mean or the variability of dissolved oxygen between different 30 min timepoints of 24 h cycles between algal mat-open reef pairs. The algal mat significantly acidified the waterunder the algal mat by decreasing pH by 0.10e0.13 pH units below open reef pH. A minimum of pH 7.47occurred between 14 and 19 h after sunrise. Our combined results suggest that mats of G. salicornia canalter various physical parameters on a fine scale and time course not commonly detected. These changesin parameters give insight into the underlying basis for negative impact, and suggest new ways in whichthe presence of invasive species leads to decline of coral reef ecosystems.

Published by Elsevier Ltd.

1. Introduction

Worldwide, algae compete with coral through a variety ofmechanisms (McCook et al., 2001; Jompa andMcCook, 2003; Birrellet al., 2008) and algal dominance is maintained by typically intenseherbivory and low nutrient additions (Littler and Littler, 1984;Smith et al., 2001; Sotka and Hay, 2009). In coastal O‘ahu waters,there are several species of non-indigenous macroalgae includingGracilaria salicornia, which spread across the benthos, overgrownative organisms (Smith et al., 2002) and are not palatable (Stimsonet al., 2007). G. salicorniawas introduced to O‘ahu in 1971 and 1978,and to Moloka‘i in the late 1980s from samples that originated fromHawai‘i Island (Smith et al., 2004).

The field growth rate of Gracilaria salicornia under ambientconditions has been estimated at 0.037 g d�1; this alga can increaseits biomass by as much as 10% d�1, based on field growth experi-ments (Smith et al., 2004). G. salicornia is found growing in veryhigh sediment areas on soft bottom and hard bottom substratesfrom intertidal to 4 m deep (Abbott, 1999) and can accumulatebiomass of 5.193 kg m�2 wet weight (Smith et al., 2004) whichoften grows in mounds and mats as thick as 8.9 cm (Beach et al.,

z).

r Ltd.

1997). This proliferative alga has become well established andinvasive, in some cases estimated to cover the reef at a rate of280 m y�1 (Rogers and Cox, 1999). Increases in algal cover havebeen associated with decreases in coral cover and hard substrate(Conklin, 2007), as well as native algal species (Stimson et al.,2001). The mound morphology of G. salicornia creates an unusualthree-dimensional open mesh of fused branches with space fora variety of residential organismsmost commonly filter feeders anddetritivores (Rodolf Pan pers. comm.) as well as other epiphytes andepifauna. These algal mats effectively smother benthic sessileorganisms, particularly the two most common coral species inK�ane‘ohe Bay: the branching coral (Porites compressa) and plating/foliose coral (Montipora capitata).

Given the complex but open mesh network of thalli that createthe mound form of Gracilaria salicornia, accumulation of suspendedsediment seems likely (Wolanski et al., 2009). To support this idea,evidence for hypoxic and hyperoxic oxygen flux can be observed insitu. For example, when this alga grows over sediment, the sedi-ment underneath is frequently black (possibly hydrogen sulfideassociatedwith sulfur reducing bacteria living in anoxic conditions)and yet, near the end of a day, mats can be found with bubblesloosely associated with the surface (supersaturation from thermalregime change, photosynthesis by canopy portions of G. salicornia,or both).

Fig. 1. Map of the study area at Moku o Lo‘e, K�ane‘ohe Bay, O‘ahu, Hawai‘i. Replicatepaired study sites were along a 100 m transect in the lagoonal portion of southernportion of the windward side of the island.

J.A. Martinez et al. / Estuarine, Coastal and Shelf Science 99 (2012) 42e49 43

Beyond the well studied factors of nutrient enrichment andherbivory that regulate algal dominance on coral reef regions,a variety of more subtle mechanisms are also implicated incontrolling algal-coral competition. These include active interac-tions such as pre-empting space, shading, smothering, abrasion,allelopathy (reviewed in McCook et al. (2001)), by the alga, as wellas passive interactions such as increased sedimentation (Birkeland,1977). Further, an alga could have complex longer-term impacts bypossibly promoting disease (Smith et al., 2006), as well as pre-venting coral recruitment (Birrell et al., 2005; Birrell et al., 2008).

Little attention has been paid to the role of respiratory gases inaffecting biogeochemistry at microscales and rapid time frames.Specifically, dissolved oxygen (DO), carbon dioxide (CO2) and pHhave physiochemical relationships, which allow these factors tovary substantially in reef settings. Biologically, DO is consumed andCO2 is produced via respiration by marine organisms. When CO2concentration increases, CO2 combines with H2O to produceH2CO

2�3 (carbonic acid), which exists in equilibrium with Hþ and

HCO�3 . The increased availability of protons decreases the pH,

thereby acidifying the water. Following this acidification, Hþ reactwith CO3 (carbonate) molecules thereby reducing carbonate, whichin turn decreases the saturation states for aragonite and calcite inseawater (Kleypas et al., 1999; Doney, 2010). At a low saturationstate, calcification rates can decrease and CaCO3 can dissolve. Hauriet al. (2010) found lower DO under algal mats than in referencesettings above themat for research sites on the Great Barrier Reef ata fixed time point in the day. However, DO and pH levels may differbetween the benthos and several cm above an algal mat due to‘background’ benthic respiration under conditions without mac-roalgal mats and with varying hydrodynamic processes. Further,very little is known about how the processes of photosynthesis andrespiration affect the diurnal flux of DO and pH under algal matsand how these changes may compare with ‘natural’ benthicprocesses. The persistence of Gracilaria salicornia allows it toimpose any one or combination of mechanisms of impact on otherbenthic organisms, especially the obligately sessile life forme coral.

The goal of this study was to understand howmats of Gracilariasalicornia and the associated biological communities can affectcoral reef habitat quality at fine scales and in rapid time courses.Wecompared physical parameters under the algal mats with open reefbenthic reference sites. Our goals were to assess the quantity ofirradiance attenuated by algal mats, the quantity of sedimenttrapped by an algal mat and to explore the relationship betweenthese phenomena with algal mat morphometrics. To frame theseinterests, we tested a series of hypotheses: a) Does photosynthesisfrom the algal mats increase DO and pH under the algal mats tolevels above the open benthic reef; b) Does night time respiration ofthe algal mat and associated species reduce DO and pH under thealgal mats to levels below the open benthic reef; c) Do extrememaxima, minima and variability imposed by these processes differfrom the open benthic reef.

2. Materials and methods

2.1. Site description and experimental design

The study site was the lagoonal portion of southern portion ofthe windward side of Moku o Lo‘e, K�ane‘ohe Bay, O‘ahu, Hawai‘i(21� 25.8740 N, 157� 47.2570 W; Fig. 1). This locationwas selected forthe ubiquitous mats of Gracilaria salicornia growing over live coralalong the reef at a similar low tide depth (0.4e1.8 m).

Ten replicate sites were chosen along a 100 m transect that metthe following criteria: within 2 m depth interval, presence of analgal mat growing over carbonate reef areas with live coral, spacedat least 10 m from other research sites, and the presence of

a suitable reference pair (without algae) at least 1 m from algal mat.A paired sampling designwas used to test mean values for specifiedphysical parameters between algal mat and reference sites.

2.2. Irradiance measurements

Measurements of photosynthetically active radiation (PAR)were taken in water, on November 15, 2010 at 1145 HST (HawaiianStandard Time) with direct sunlight, directly under the center ofthe algal mat and then just above the algal mat for referencesamples. Measurements were taken with a LICOR LI-250Aphotometer with a flat plate cosine quantum sensor and cor-rected for underwater attenuation.

2.3. Sediment measurements

Tomeasure the amount of sediment accumulated under an algalmat and compare with reference values, sediment traps were

J.A. Martinez et al. / Estuarine, Coastal and Shelf Science 99 (2012) 42e4944

deployed (Storlazzi et al., 2011). Two plastic 50 ml TB Falcon tubeswere stacked with the bottom of one tube cut (to allow sediment tosink) and zip tied to small wooden stakes which were securedinside crevices or sand pockets on the reef (Fig. 2a). The total heightof the trap was 20.5 cm and the surface area of the opening was6.6 cm2. One trap was placed under the horizontal center of an algalmat and oriented vertically at 90�. A reference trap was placed 1 mfrom that algal mat, for a total of 10 pairs of sediment traps. Thetraps were collected after eight days. Care was taken to cap thetubes underwater to not lose sediment. The trap contents werevacuum filtered through pre-weighed Millipore 0.2 mm filter anddried in an oven for 2 d at 40 �C. The filters were weighed and thedifference was used to calculate the mass of dry sediment as gsediment d�1.

2.4. Respiration, pH & Salinity measurements

Paired replicates in this study were sampled February 4, 11,2010; July 7, 15, 2010; and August 6, 18, 2010 to evaluate short andlong day variation. We placed deployable data logging sondes(YSI 6600 EDS V2, Yellow Springs, OH) with conductivity, temper-ature, turbidity, depth, dissolved oxygen and pH probes in an algalmat and 1 m away (reference pair), each fixed the same distancefrom the bottom which was approximately 5 cm. Each algal matwas carefully lifted and the sonde was placed underneath (Fig. 2b).Data were logged every 5 min for several days. Six replicatedobservations using different algal mats and reference pairs weremade over three to five days.

2.5. Algal mat Size

After other measurements were taken, a six by six cm section ofeach algal mat was collected, the mat height was measured with

Fig. 2. a) Sediment traps were placed under algal mats and 1 m away for a reference site. b)away for a reference site.

a standard metric measure, the epifauna and flora were removed,the mat was rinsed to remove salt, dried at 80 �C in the laboratoryfor two days and weighed for computation of biomassmeasurements.

2.6. Statistical analyses

Minitab 16 statistical software was used to calculate tests ofhomogeneity of variances. PRISM GraphPad v. 5.0b was used toproduce all plots. All remaining statistical analyses were performedusing the SPSS statistical software 17.0.

2.7. Sizes of algal mats

Sizes of algal mat were calculated using several metrics: dryweight per area covered by mat (kg m�2), thickness (cm) anddensity (gm�3). Relationship between algal mat thickness and algalmat density, as well as algal mat biomass and algal mat densitywere analyzed with linear regressions.

2.8. Irradiance

A Student’s paired t-test was used to compare the mean quantaof PAR transmitted on top of the algal mat and underneath it.

2.9. Sediment

The sediment trap rate (Storlazzi et al., 2011) was calculated asthemean sediment per day per area (g d�1 cm�2) for each trap. Datadid not meet the assumptions of a parametric paired test of means;non-parametric sign test was used to compare the mean sedimenttrap rates between the algal mat and reference group. Regressionanalyses were performed to: find the best metric of algal mat size

Probes from data logging sondes were placed into mats of Gracilaria salicornia and 1 m

Table 1Dissolved oxygen and pH extreme maxima and minima, mean and standard devi-ation data from six algal mats and six reference pairs over a 24 h diurnal cycle.

Descriptive statistics

Maximum Minimum Mean Std. deviation

DO algae (mg/l) 11.78 1.04 5.00 2.64pH algae 8.2 7.47 7.84 0.159DO reference (mg/l) 8 2.12 4.78 1.2pH reference 8.23 7.61 7.93 0.133

J.A. Martinez et al. / Estuarine, Coastal and Shelf Science 99 (2012) 42e49 45

and quantify the relationship between algal size and the sedimenttrap rate.

2.10. DO & pH

Data for analyses were selected from time points at least 24 hafter placement of sensors to allow the mound dynamics attainsteady state rates after sensor deployment. 24 h cycles from sixreplicate paired sites with data from 30 min intervals werenormalized to time from sunrise, with astronomical data from theUS Naval Observatory Astronomical Data Services. This wasnecessary as the data varied over time of day associated withphotosynthesis by Gracilaria salicornia. For data that were found tomeet parametric statistical assumptions, the mean DO and pHlevels for time intervals of 30 min throughout a 24 h cycle weretested between the algal mat and reference groups (48 tests perparameter) using a Student’s t-test. For data that did not meet theparametric statistical assumptions, the non-parametric sign testwas used to test for differences with ranked medians.

Spearman rho correlation analyses were conducted to examinethe relationship between DO and pH for algal mats and controlswith pooled data from a 24 h cycle of five replicate pairs.

3. Results

3.1. Size of algal mats

The biomass of algal mats sampled ranged between7.27 kge20.20 kg dry weight m�2. Linear regression analyses testedthe relationship between the algal size morphometrics of biomass(g m�2) and height of mat (cm) with density (g m�3). Bothparameters were found to predict density, with biomass explaining72% of the variation in density with a positive association andheight explaining 42% of the variation with a negative association.Biomass, y¼ 22.989xþ 44.273, R2¼ 0.71, p< 0.05, DF¼ 9. Height ofmat, y ¼ �3455.625x þ 29,375.761, R2 ¼ 0.42, p < 0.05, DF ¼ 9.

3.2. Irradiance

The Students paired t-test revealed significantly less photosyn-thetically active irradiance transmitted to the benthos under algalmats than for the reference group (p < 0.05, DF ¼ 9). The meanquantity of photosynthetically active irradiance transmittedunder algal mats (n ¼ 10) was 1.17 � 2.31 SE mmol photons m�2 s�1

and the mean for the reference group (n ¼ 10) was1191.20 � 60.201 SE mmol photons m�2 s�1. The mean ratio of lightattenuated by algal mats was 99.3% � 0.17 SE.

3.3. Sediment

The non-parametric sign test revealed a significantly highersediment trap rate in the algal mat when compared with thereference group (p< 0.05, DF¼ 8). The mean sediment trap rate forthe algal mat was 25.28 � 2.371 SE g m�2 d�1; the mean for thereference was about 50% lower, at 12.71 � 2.435 SE g m�2 d�1. Themajority of sediment collected appeared to be fine particulate.The sediment collected fromwithin themats was dark brown/blackand had a hydrogen sulfide-like odor where as the sediment incontrols was light brown or white.

Linear regression analyses tested the relationship between thealgal biomass (kg m�2), and morphometrics of height of mat (cm),and density (kg m�2) as well as the rate of trapping sediment(kgm�2 d�1). Algal biomass andmorphometric data explained littleof the variation and did not predict the rate at which sedimentswere trapped (data not shown).

3.4. Dissolved oxygen

Mean dissolved oxygen for five replicate algal mat pairs, changedwith time from sunrise. With appropriate paired sample statisticaltests, no significant differences in mean DO levels were detected atparticular 30 min intervals for algae and control groups, (p > 0.05,n¼4) for all timepoints (Table 1). However, night timeminimumDOvalues (Table 1) in algalmatswere found to behalf those of referencevalues. Values of DO during daytimemaximumswere 47.25% higherunder the algal mats than at the reference sites. Two mats hadminimal DO values of 1.04e1.99 mg l�1 and developed hypoxia(<2 mg l�1 or <30% saturation) between time intervals 12 h 30 mfrom sunrise to 1 h 30m after sunrise the next day or approximately1830-0730 Hawaiian Standard Time (HST). Each instance of hypoxialasted between 30 and 120 m. There were only near hypoxic valuesfound for two reference sites with a minimum value of2.12e2.99 mg l�1 between 18 h 30 m from sunrise to 1 h 30 m aftersunrise the next day or approximately 0030-0730 HST. Themaximumvalue for the algal mat group reached 11.78 mg l�1 (180%saturation), and three mats maintained hyperoxia (>100% satura-tion) at various times of the day. Themaximal value for the referencegroup was 8.00 mg l�1 (108.9% saturation), and one reference groupsite had hyperoxic DO levels between 2 h 30 m to 12 h from sunriseor approximately 0830-1800 HST. Though the overall standarddeviation of the mean dissolved oxygen in the algal mat group wastwice that of the reference group, the test of homogeneity of vari-ances revealed no significant differences between the variabilityimposed on the oxygen levels at a 30 min interval (p > 0.05).

3.5. pH

Significant differences in pH were detected (0.1e0.13 pH units)in 14 of the 30 min intervals (just under 5% of all observations)when examining the algal mats against the reference groups at14 h,15 h 30min and between 19 h from sunrise to 2 h after sunrisethe next day or approximately 2000, 2130, 2300 - 0200 h HSTrespectively. Extremely low pH values of 7.61 in the reference and7.47 in the algal mats were also detected. Maximal values were pH8.23 for the reference and pH 8.20 for the algal mats.

3.6. DO & pH

To test relationships between DO and pH, we performedSpearman rho correlation analyses for data points from the refer-ence group and the algal mats. Tests found a moderately positivecorrelation between DO and pH in the reference group (0.674,p < 0.000) and a stronger correlation in the algal mat treatment(0.749, p < 0.000). Mats had a more diffuse relationship betweenDO and pH than that observed at the reference site (Fig. 3). DO andpH levels for one of the sites that experienced hypoxia in a 24 hcycle normalized for time from sunrise for the algal mat andreference pair respectively are shown in Fig. 4a and b. Whennormalized for time from sunrise for the algal mat and referencegroups respectively (error bars removed for visualization, similar

a b

Fig. 3. The relationship between dissolved oxygen and pH in a diurnal cycle for (a) six algal mats and (b) six reference pairs. Horizontal bars indicate the time from sunrise thatthese values occurred.

J.A. Martinez et al. / Estuarine, Coastal and Shelf Science 99 (2012) 42e4946

patterns were revealed (Fig. 4c and d)). For the reference group, DOand pH began to increase 2 h after sunrise (0800 HST). DO increasedto its maximummean value (6.00 mg l�1), 7 h 30 me9 h 30 m aftersunrise (approximately 1330-1530 h HST). pH reached a maximummean value of 8.00 at 4 h 30 me11 h 30 m after sunrise (1030-1730HST). Near the end of daylight, DO and pH declined until theyreached night time minimum mean values of 3.52 mg l�1 and pH7.80 at 22 h 30 m from sunrise (approximately 0430 HST)respectively.

For algal mats, DO began to increase 1 h 30 m after sunrise, withpH increasing 2 h after sunrise. DO increased to its maximum levels(6.80 mg l�1) from 7 h 30 me9 h 30 m after sunrise (approximately1330-1530 HST), while pH increased to its maximummean value of7.95 between 7 h 30 me10 h 30 m after sunrise (1300-1800 HST).Near the end of daylight, DO and pH declined until they reachednight time minimum mean values of 3.40 mg l�1 and pH 7.71 at23 h from sunrise (approximately 0500 HST) respectively.

a

c d

Fig. 4. Dissolved oxygen (solid line) and mean pH (dashed line) over a 24 h diurnal cycle indissolved oxygen (solid line) and mean pH (dashed line) over a 24 h cycle in time from su

4. Discussion

This study has demonstrated several important modifications ofthe benthic physical environment caused by algal mats of Gracilariasalicornia. Decreased irradiance, sediment trapping, and periods ofhyperoxia, hypoxia and acidification are all stressful conditionscaused by the algal mats. We found significant changes in thetimescale of 30 m, which underscore the need to take environ-mental measurements over a comprehensive timescale, as detailedin next sections.

4.1. Algal size

Algal biomass was found to be a strong predictor of algal density.Algal mat height was found to also predict algal density but witha negative association with thicker mats being less dense. Thisfinding was surprising, as mat forms in situ appear homogeneous as

b

time from sunrise at site 372 under an algal mat (a) and the reference pair (b). Meannrise for (c) six algal mats and (d) six reference pairs.

J.A. Martinez et al. / Estuarine, Coastal and Shelf Science 99 (2012) 42e49 47

presented in a 2D surface, suggesting from casual observation thatdensity of the mats did not vary. Though it is possible that watermotion or herbivory could crop the mats or pull them loose.

4.2. Irradiance

The findings in this study corroborate with other studies (Beachet al., 1997; Dailer, 2006), which have reported substantial atten-uation of PAR by algal mats.

Several physiological features of Gracilaria salicornia likely allowit to persist in such a high biomass state in both high light and lowlight environments. G. salicornia mats are known to self shade,acclimating with a gradient of accessory pigments increasing fromoutside the mat to within (Beach et al., 1997). In addition, the alga’sphotosynthetic saturating irradiance (Ik) can range from 134 to633 mmol photons m�2 s�1 and compensating can range from 6 to37 mmol photons m�2 s�1 in different habitats (Phooprong et al.,2007), which allows this alga to maintain high productivity andmeet its basic metabolic needs under relatively low lightconditions.

The consequences of reduced irradiance to coral have clearphysiological implications. The near maximal levels of irradiancefound under these algal mats (1.17 mmol photons m�2 s�1 � 2.31 SE,n ¼ 10) are analogous to irradiance levels at greater than 45 mdepths (Rivero-Calle et al., 2008) and are insufficient to achievemaximal photosynthesis in shallow reef corals living undereither: attenuated full sun (w231e586 mmol photons m�2 s�1

Langdon and Atkinson, 2005) or deep shade (w31e89 mmolphotons m�2 s�1 Porter et al., 1984). Further, these under matirradiances may not allow for either physiological group of corals tocompensate for respiration (w150e350 mmol photons m�2 s�1

Langdon and Atkinson, 2005; Levy et al., 2004; Falkowski andDubinsky, 1981); or deep shade-adapted coral (w21e50 mmolphotons m�2 s�1, Porter et al., 1984; Falkowski and Dubinsky,1981). While mesophotic communities can exist in extremely lowirradiance, the taxa in those deep reefs are not common to shallowcoastal zones (Rooney et al., 2010). A point of final concern is e

under low irradiances, coral rates of calcification fall to low levels(Marubini et al., 2001).

4.3. Sediment

There was a significantly higher sediment trap rate for finesediments in the algal mat treatment than for the reference group.As sediment fell or moved with water motion, the three-dimensional structure of algal mats may allow multiple surfacesto baffle water flow, allowing sediment to fall out. It is also possiblethat waste, particulate organic matter or decomposing tissue of thebiological community inhabiting the algal mat contributes partic-ulate material to the benthos below. Other studies have foundeither correlations between algae and sediment accumulation fromfield studies (Purcell, 2000) or measured sediment previouslyaccumulated on algae in the field (Stamski and Field, 2006), but notactually transported to the benthos underneath. This studydemonstrates enhanced sediment accumulation under algal mats.Sediment has several biological effects including preventing algal(Umar et al., 1998) and invertebrate larvae from recruiting (Birrellet al., 2005), inducing coral disease (Garrison et al., 2005), smoth-ering and directly killing benthic organisms as well as increasingnutrients (Fabricius, 2005). Algal mats like those of Gracilaria sali-cornia may further trap sediment, which may support algal growthand thus contribute to a positive feedback loop. Data from thisstudy suggests that algal removal efforts, which have a secondarygoal to reduce accumulated sediment, may be successful for bothbiomass removal and reduction in secondary impacts from

sediment accumulation. Further, it might be expected that filter-feeding organisms might benefit from inhabiting the algal mat asthere may be an increased opportunity for catching food and thiscorroborates with previous observations (Rodolf Pan pers. comm.).The macroalga Dictyosphaeria cavernosa can take advantage ofnutrients from sediment and waste from inhabiting animals whenwater column nutrient conditions are insufficient (Stimson et al.,1996) and this may be true for G. salicornia as well.

4.4. DO

There was a great deal of variability in the data from the algalmat and the reference groups; we did not detect a statisticallysignificant difference for DO in mean or variances. Yet, water underalgal mats reached extreme minimum hypoxic and maximumhyperoxic conditions in this study. There are several biologicalimplications for an organism experiencing these phenomena. Lowoxygen availability reduces the aerobic respiration capacity of anorganism to lethal levels and can drive mobile species to evacuatean area (Grantham et al., 2004). At the other end of the spectrum,high concentrations of oxygen in water can produce reactiveoxygen species (ROS). ROS such as superoxide, hydrogen peroxide,hydroxyl and peroxyl molecules are all known to induce oxidativedamage to biological organisms (Lesser, 2006).

Extreme fluctuations themselves can impose a great deal ofstress into a habitat, which could drive the community compositionbased on organisms that can tolerate living in these conditions.Crustaceans are known to have significantly higher lethal hypoxiaoxygen thresholds than other groups (making them the mostsensitive), though fish have the highest sublethal hypoxia oxygenthreshold with Crustaceans and Molluscs next and Polychaetes,Echinoderms and Cnidarians with the lowest (Vaquer-Sunyer andDuarte, 2008). If hypoxia induced by the algal mats was an issueon the reef, it might be expected thatmobile organisms such as fish,echinoderms, crustaceans and gastropods may be able to tempo-rarily inhabit these algal mats in the daytime and then move as DOconditions go hypoxic. In contrast it might be expected that benthicsessile organisms such as coral, algae, tubeworms, snails andsponge may be most susceptible to the extreme DO minima.

Kuhl et al. (1995) measure DO in coral tissue under differentflow and light regimes and found that % saturation of DO fall from100% to <2% in stagnant and 5e20% in low flow conditions withthick boundary layers. Under moderate irradiance, photosynthesiscan increase DO to as much as 240% saturation under high flowconditions with a thin boundary layer. These ranges are well belowand above the minima and maxima found in this study. Themeasurements taken in this study were near the center of themicrohabitat under the algal mat and not at the coral surface.The levels of oxygen that corals experienced under these algal matsare unknown but it is possible that DO levels are much lower thanwhat we measured in this study. While some algae and coral copewith daily hypoxic and hyperoxic conditions with enzymatic andnon-enzymatic antioxidant defenses (Lesser, 2006), this tolerancemight allow them to survive under oxidative damaging conditionsat least temporarily. Gracilaria salicornia imposes these conditionson itself when it grows in large biomass. Portions of the algae arefrequently buried in mud and are still pigmented and alive and soone of the keys to this alga’s success may likely be awell-developedand effective oxidative damage response physiology.

4.5. Respiration & pH

Levels of DO can be used as a proxy measure for respiration(Clavier et al., 2008). As the algal mat community respires, O2 isconsumed and CO2 is released. The variability imposed on the

J.A. Martinez et al. / Estuarine, Coastal and Shelf Science 99 (2012) 42e4948

relationship between DO and pH in the algal mat group could likelybe from respiratory acidification in the daytime while photosyn-thetic basification (O2 production) was occurring. The relationshipbetween DO and pH can be further interpreted from their corre-sponding diurnal cyclicity as was seen in our data. As O2 productionincreased, there was a point at which pH peaks in both data sets.This may be a reflection of community respiration in the day, whereCO2 and O2 production are both occurring. It is also interesting tonote that it takes about 1.5 h until production meets the compen-sation point for respiration (where DO begins to increase).

Themean andminimum pH under the algal mats, were found tobe less than that for the reference groups, though themaximum pHwas similar. The daily increase in pH may be explained by theconsumption of CO2 in photosynthesis and the production ofhyperoxic water with ROS, which could accept hydrogen ions andthus increase the pH. A corresponding pH maximum for the algalgroup was not detected above the reference group but insteadstayed maximally at 8.2. It is possible that CO2 produced fromdiurnal metabolic respiration and photorespiration (from the algalmat and epiphytes) was concurrently consumed by carbon fixationreactions of photosynthesis.

The impacts of decreased pH to coral and coral reef inverte-brates are now just beginning to become understood. Minimal pHlevels found in this study under the algae are comparable to pub-lished values that showed impacts to coral health such as skeletondissolution, decreased settlement, fertilization success, and growth.pH ranges of 7.3e7.6 have been found to dissolve coral skeletons(Fine and Tchernov 2007). pH conditions of 7.7e7.9 decreaseseveral biological processes in coral such as fertilization success,gamete viability, settlement of larvae, growth in juvenile recruits(Albright et al., 2008, 2010). pH conditions of 7.3e7.6 delay the algalinfection rate of recruits (Suwa et al., 2010). The abundance of redcalcareous algae has also been found to decrease under low pHconditions with shifts to increased abundance of non-calcareousalgae (Kuffner et al., 2007; Fabricius et al., 2011).

It is still uncertain how the global process of ocean acidificationwill affect these kinds of microscale processes. With estimates ofglobal surface acidification of 0.14e0.35 pH units by the 21stcentury (IPCC, 2007), microscale processes like nocturnal benthicrespiration could affect or exacerbate detrimental effects on corals.While DO and pH can vary in these algal microhabitats, theimplications to organisms like coral when considered in synergywith other factors remain important but uncertain. Such concernsbecome more important as non-palatable invasive species likeGracilaria salicornia spread at w280 m y�1 in K�ane‘ohe Bay, O‘ahu(Rogers and Cox, 1999), as well as outside (Abbott, 1999) and if itremains unchecked will impact macro-scale classes of benthichabitat on at least two islands.

At present conditions, coral reefs can undergo a small dissolu-tion at night (Suzuki et al., 1995) but light and dark calcification canmore than compensate for that (Yates and Halley, 2006). It ispossible that the lower pH found under invasive algal mats maydecrease the aragonite saturation states at night and acceleratedissolution of CaCO3 reef material, biological skeletons and reducethe ‘dark calcification rate’ of calcifying organisms. For confirma-tion, calcification, alkalinity and carbonate chemistry measure-ments should be taken diurnally under the algal mats tounderstand the direct influence to the processes of calcification anddissolution.

Because DO and pH minima in the algal mats are lower thanthose observed in open reef conditions, flow in the algal mat islikely to be reduced and the boundary layer above coral overgrownby the algal mats would be expected to be much larger than openreef regions. Several studies have shown that flow (Carpenter andWilliams, 1993; Escartin and Aubrey, 1995) as well as advection

(Worcester, 1995) can be reduced by macrophytes. Water flow isimportant for aerobic benthic heterotrophic organisms for food andto reduce the boundary layer thus increasing diffusion of respira-tory gases. Additionally for autotrophic benthic organisms flowincreases the diffusion of nutrients, and can reduce photorespira-tion and photoinhibition (Nakamura, 2010).

5. Conclusions

Mats of Gracilaria salicornia impose microscale changes inwaterquality to the living community under these mats. These modifi-cations include changes in physical processes of sediment accu-mulation and attenuation of irradiance. Additionally, as a functionof their own respiration with nightly periods of hypoxia andaccelerated acidification, mats impose single and synergisticcombinations of stress on organisms living under mats. These datasuggest that the presence of large biomass stands of G. salicorniamats have the capability to drastically change benthic habitatquality, that several mat morphometrics be considered whenmeasuring physical parameters in and under the alga, thatcomprehensive diurnal sampling be performed when trying tounderstand microscale biogeochemical processes, and that oceanacidification models at the appropriate scale include these vari-ables, especially when invasive algal biomass reaches the largescale of reefs.

Acknowledgments

The authors would like to thank Carey Morishige for fieldassistance, advice and artwork as well as: Jorg Anson, Leinson Neth,Brandi Sasaki, Sean Macduff, Chloé Brahmi, for field assistance; andDrs. Flo Thomas, Megan Donahue, Olga Cordero-Brana and EricWolanski for advice. Funding was provided by the University ofHawai’i Graduate Professional Access program, the NationalOceanographic and Atmospheric Administration Educational Part-nership Program Graduate Science Program and the Hawai’i CoralReef Initiative (HCRI). The conclusions presented from this researchare not necessarily the opinion of the National Oceanic andAtmospheric Administration or the US government.

References

Abbott, I., 1999. Marine Red Algae of the Hawaiian Islands. Bishop Museum, Hon-olulu, p. 214.

Albright, R., Mason, B., Miller, M., Langdon, C., 2008. Effect of aragonite saturationstate on settlement and post-settlement growth of Porites astreoides larvae.Coral Reefs 27, 485e490.

Albright, R., Mason, B., Miller, M., Langdon, C., 2010. Ocean acidification compro-mises recruitment success of the threatened Caribbean coral Acropora palmata.Proceedings of the National Academy of Science 107, 20400e20404.

Beach, K.S., Borgeas, H.B., Nishimura, N.J., Smith, C.M., 1997. In vivo absorbancespectra and the ecophysiology of reef macroalgae. Coral Reefs 16, 21e28.

Birkeland, C., 1977. The importance of rate of biomass accumulation in earlysuccessional stages of benthic communities to the survival coral recruits. In:Proceedings of the 3rd International Coral Reef Symposium Publ 1, pp. 15e21.

Birrell, C.L., McCook, L.J., Willis, B.L., 2005. Effects of algal turfs and sediment oncoral settlement. Marine Pollution Bulletin 51, 408e414.

Birrell, C.L., McCook, L.J., Willis, B.L., Diaz-Pulido, A., 2008. Effects of benthic algaeon the replenishment of corals and the implications for the resilience of coralreefs. Oceanography and Marine Biology: An Annual Review 46, 25e63.

Carpenter, R.C., Williams, S.L., 1993. Effects of algal turf canopy height and micro-scale substratum topography on profiles of flow speed in a coral fore-reefenvironment. Limnology and Oceanography 38, 687e694.

Clavier, J., Chauvaud, L., Cuet, P., Esbelin, C., Frouin, P., Taddei, D., Thouzeau, G., 2008.Diel variation of benthic respiration in a coral reef sediment (Reunion Island,Indian Ocean). Estuarine, Coastal and Shelf Science 76, 369e377.

Conklin, E.J., 2007. The influence of preferential foraging, alien algal species, andpredation risk on the interaction between herbivorous fishes and reef macro-algae. Ph.D. thesis. The University of Hawai‘i at M�anoa, Honolulu, Hawai’i,United States of America. p. 257.

Dailer, 2006. Photoecological strategies influencing the invasive success of theinvasive marine macrophyte Eucheuma denticulatum on Hawaiian coral reefs.

J.A. Martinez et al. / Estuarine, Coastal and Shelf Science 99 (2012) 42e49 49

MS thesis. University of Hawai‘i at M�anoa, Honolulu, Hawai’i, United States ofAmerica. p. 149.

Doney, S.C., 2010. The growing human footprint on coastal and open-oceanbiogeochemistry. Science 328, 1512e1516.

Escartin, J., Aubrey, D.G., 1995. Flow structure and dispersion within algal mats.Estuarine, Coastal and Shelf Science 40, 451e472.

Fabricius, K.E., 2005. Effects of terrestrial runoff on the ecology of corals and coralreefs: review and synthesis. Marine Pollution Bulletin 50, 125e146.

Fabricius, K.E., Langdon, C., Uthicke, S., Humphrey, C., Noonan, N., De’ath, G.,Okazaki, R., Muehllehner, N., Glas, M.S., Lough, J.M., 2011. Losers and winners incoral reefs acclimatized to elevated carbon dioxide concentrations. NatureClimate Change 1, 165e169.

Falkowski, P.G., Dubinsky, Z., 1981. Light-shade adaptation of Stylophora pistillata,a hermatypic coral from the Gulf of Eilat. Nature 189, 172e174.

Fine, M., Tchernov, D., 2007. Scleractinian coral species survive and recover fromdecalcification. Science 315, 1811.

Garrison, V.H., Shinn, E.A., Foreman, W.T., Griffin, D.W., Holmes, C.W., Kellogg, C.A.,Majewski, M.S., Richardson, L.L., Ritchie, K.B., Smith, G.W., 2005. African andAsian dust from desert soils to coral reefs. Bioscience 53, 469e480.

Grantham, B.A., Chan, F., Nielsen, K.J., Fox, D.S., Barth, J.A., Huyer, A., Lubchenco, J.,Menge, B.A., 2004. Upwelling-driven nearshore hypoxia signals ecosystem andoceanographic changes in the northeast Pacific. Nature 429, 749e754.

Hauri, C., Fabrcius, K.E., Schaffelke, B., Humphrey, C., 2010. Chemical and physicalenvironmental conditions underneath mat- and canopy-forming macroalgae,and their effects on understorey corals. Plos One 5, 1e9.

IPCC, 2007. Climate Change 2007: Synthesis Report. Contribution of WorkingGroups I, II and III to the Fourth Assessment Report of the IntergovernmentalPanel on Climate Change. In: Core Writing Team, Pachauri, R.K., Reisinger, A.(Eds.). IPCC, Geneva, Switzerland, 104 pgs. p. 37.

Jompa, J., McCook, L.J., 2003. Coral-algal competition: macroalgae with differentproperties have different effects on corals. Marine Ecology Progress Series 258,87e95.

Kleypas, J.A., Buddemeier, R.W., Archer, D., Gattuso, J.P., Langdon, C., Opdyke, B.N.,1999. Geochemical consequences of increased atmospheric carbon dioxide oncoral reefs. Science 284, 118e120.

Kuffner, I.B., Andersson, A.J., Jokiel, P.L., Rodgers, K.S., Mackenzie, F.T., 2007.Decreased abundance of crustose coralline algae due to ocean acidification.Nature Geoscience 1, 114e117.

Kuhl, M., Cohen, Y., Dalsgaard, T., Barker Jorgensen, B., Revsbech, N.P., 1995.Microenvironment and photosynthesis of zooxanthellae in scleractinian coralsstudied with microsensors for O2, pH and light. Marine Ecology Progress Series117, 159e172.

Langdon, C., Atkinson, M.J., 2005. Effect of elevated p CO2 on photosynthesis andcalcification of corals and interactions with seasonal change in temperature/irradiance and nutrient enrichment. Journal of Geophysical Research 110, 1e16.

Lesser, M.P., 2006. Oxidative stress in marine environments: biochemistry andphysiological ecology. Annual Review of Physiology 36, 253e278.

Levy, O., Dubinsky, Z., Schneider, K., Achituv, Y., Zakai, D., Gorbunov, M.Y., 2004.Diurnal hysteresis in coral photosynthesis. Marine Ecology Progress Series 268,105e117.

Littler, M.M., Littler, D.S., 1984. Models of tropical reef biogenesis: the contributionof algae. Progress in Phycological Research 3, 323e364.

Marubini, F., Barnett, H., Langdon, C., Atkinson, M.J., 2001. Dependence of calcifi-cation on light and carbonate ion concentration for the hermatypic coral Poritescompressa. Marine Ecology Progress Series 220, 153e162.

McCook, L.J., Jompa, J., Diaz-Pulido, G., 2001. Competition between corals and algaeon coral reefs: a review of evidence and mechanisms. Coral Reefs 19, 400e417.

Nakamura, T., 2010. Importance of water-flow on the physiological responses ofreef-building corals. Galaxea. Journal of Reef Studies 12, 1e14.

Phooprong, S., Ogawa, H., Hayashizaki, K., 2007. Photosynthetic and respiratoryresponses of Gracilaria salicornia (C. Ag.) Dawson (Gracilariales, Rhodophyta)from Thailand and Japan. Journal of Applied Phycology 19, 795e801.

Porter, J.W., Muscatine, L., Dubinsky, Z., Falkowski, P.G., 1984. Primary productionand photoadaptation in light and shade adapted colonies of the symbiotic coral,Stylophora pistillata. Proceedings of the Royal Society of London B 222, 161e180.

Purcell, S.W., 2000. Association of epilithic algae with sediment distribution ona windward reef in the northern Great Barrier Reef, Australia. Bulletin of MarineScience 66, 199e214.

Rivero-Calle, S., Armstrong, R.A., Soto-Santiago, F.J., 2008. Biological and physicalcharacteristics of a mesophotic coral reef: Black Jack reef, Vieques, Puerto Rico.In: Proceedings of the 11th International Coral Reef Symposium, Ft. Lauderdale,Florida, 1, pp. 567e571.

Rogers, K., Cox, E., 1999. Rate of spread of introduced rhodophytesKappaphycus alvarezii, Kappaphycus striatum, and Gracilaria salicornia and theircurrent distributions in Kane‘ohe Bay, O‘ahu, Hawai‘i. Pacific Science 53,232e241.

Rooney, J., Donham, E., Montgomery, A., Spalding, H., Parrish, F., Boland, R.,Fenner, D., Gove, J., Vetter, O., 2010. Mesophotic coral ecosystems in theHawaiian Archipelago. Coral Reefs 29, 361e367.

Smith, J.E., Smith, C.M., Hunter, C.L., 2001. An experimental analysis of the effects ofherbivory and nutrient enrichment on benthic community dynamics ona Hawaiian reef. Coral Reefs 19, 332e342.

Smith, J.E., Hunter, C.L., Smith, C.M., 2002. Distribution and reproductive charac-teristics of nonindigenous and invasive marine algae in the Hawaiian Islands.Pacific Science 56, 299e315.

Smith, J., Hunter, C., Conklin, E., Most, R., Sauvage, T., Squair, C., Smith, C., 2004.Ecology of the invasive red alga Gracilaria salicornia (Rhodophyta) on O‘ahu,Hawai‘i. Pacific Science 58, 25e34.

Smith, J.E., Shaw, M., Edwards, R.A., Obura, D., Pantos, O., Sala, E., Sandin, S.A.,Smriga, S., Hatay, M., Rohwer, F.L., 2006. Indirect effects of algae on coral: algae-mediated, microbe-induced coral mortality. Ecology Letters 9, 835e845.

Sotka, E.E., Hay, M.E., 2009. Effects of herbivores, nutrient enrichment, and theirinteractions on macroalgal proliferation and coral growth. Coral Reefs 28,555e568.

Stamski, R.E., Field, M.E., 2006. Characterization of sediment trapped by macroalgaeon a Hawaiian reef flat. Estuarine, Coastal and Shelf Science 66, 211e216.

Stimson, J., Larned, S.T., McDermid, K., 1996. Seasonal growth of the coral reefmacroalga Dictyosphaeria cavernosa (Forskål) Børgesen and the effects ofnutrient availability, temperature and herbivory on growth rate. Journal ofExperimental Marine Biology and Ecology 196, 53e77.

Stimson, J., Larned, S.T., Conklin, E., 2001. Effects of herbivory, nutrient levels, andintroduced algae on the distribution and abundance of the invasive macro-alga Dictyosphaeria cavernosa in Kaneohe Bay, Hawaii. Coral Reefs 19,343e357.

Stimson, J., Cunha, T., Philippoff, J., 2007. Food preferences and related behavior ofthe browsing sea urchin Tripneustes gratilla (Linnaeus) and its potential for useas a biological control agent. Marine Biology 151, 1761e1772.

Storlazzi, C.D., Field, M.E., Bothner, M.H., 2011. The use (and misuse) of sedimenttraps in coral reef environments: theory, observations, and suggested protocols.Coral Reefs 30, 23e38.

Suwa, R., Nakamura, M., Morita, M., Shimada, K., Iguchi, A., Sakai, K., Suzuki, Atsushi,2010. Effects of acidified seawater on early life stages of scleractinian corals(Genus Acropora). Fisheries Science 76, 93e99.

Suzuki, A., Nakamori, T., Kayanne, H., 1995. The mechanism of productionenhancement in coral reef carbonate systems: model and empirical results.Sedimentary Geology 99, 259e280.

Umar, M.J., McCook, L.J., Price, I.R., 1998. Effects of sediment deposition on theseaweed Sargassum on a fringing coral reef. Coral Reefs 17, 169e177.

US Naval Observatory http://www.usno.navy.mil/USNO/astronomical-applications/data-services/data-services.

Vaquer-Sunyer, R., Duarte, C.M., 2008. Thresholds of hypoxia for marine biodiver-sity. Proceedings of the National Academy of Science 105, 15452e15457.

Wolanski, E., Martinez, J.A., Richmond, R.H., 2009. Quantifying the impact ofwatershed urbanization on a coral reef: Maunalua Bay, Hawaii. Estuarine,Coastal and Shelf Science 84, 259e268.

Worcester, S.E., 1995. Effects of eelgrass beds on advection and turbulent mixing inlow current and low shoot density environments. Marine Ecology ProgressSeries 126, 223e232.

Yates, K.K., Halley, R.B., 2006. Diurnal variation in rates of calcification andcarbonate sediment dissolution in Florida Bay. Estuaries and Coasts 29, 24e39.