Author's personal copy - Florida Atlantic Universitybiology.fau.edu/koch/Documents/Publication PDFs...Author's personal copySynergistic effects of high temperature and sulfide on tropical

This article was originally published in a journal published byElsevier, and the attached copy is provided by Elsevier for the

author’s benefit and for the benefit of the author’s institution, fornon-commercial research and educational use including without

limitation use in instruction at your institution, sending it to specificcolleagues that you know, and providing a copy to your institution’s

administrator.

All other uses, reproduction and distribution, including withoutlimitation commercial reprints, selling or licensing copies or access,

or posting on open internet sites, your personal or institution’swebsite or repository, are prohibited. For exceptions, permission

may be sought for such use through Elsevier’s permissions site at:

http://www.elsevier.com/locate/permissionusematerial

http://www.elsevier.com/locate/permissionusematerial

Auth

or's

pe

rson

al

copy

Synergistic effects of high temperatureand sulfide on tropical seagrass

M.S. Koch a,⁎, S. Schopmeyer a, C. Kyhn-Hansen a, C.J. Madden b

a Aquatic Plant Ecology Laboratory, Biological Sciences Department, Florida Atlantic University,777 Glades Road, Boca Raton, Florida 33431, USA

b South Florida Water Management District, Everglades Division, 3301 Gun Club Rd., West Palm Beach, Florida, 33406, USA

Received 17 August 2006; received in revised form 31 August 2006; accepted 9 October 2006

Abstract

To examine the synergism of high temperature and sulfide on two dominant tropical seagrass species, a large-scale mesocosmexperiment was conducted in which sulfide accumulation rates (SAR) were increased by adding labile carbon (glucose) to intactseagrass sediment cores across a range of temperatures. During the initial 10 d of the 38 d experiment, porewater SAR in coresincreased 2- to 3-fold from 44 and 136 μmol L−1 d−1 at 28–29 °C to 80 and 308 μmol L−1 d−1 at 34–35 °C in Halodule wrightiiand Thalassia testudinum cores, respectively. Labile C additions to the sediment resulted in SAR of 443 and 601 μmol L−1 d−1 at28–29 °C and 758 to 1,557 μmol L−1 d−1 at 34–35 °C in H. wrightii and T. testudinum cores, respectively. Both T. testudinum andH. wrightii were highly thermal tolerant, demonstrating their tropical affinities and potential to adapt to high temperatures. Whileplants survived the 38 d temperature treatments, there was a clear thermal threshold above 33 °C where T. testudinum growthdeclined and leaf quantum efficiencies (Fv/Fm) fell below 0.7. At this threshold temperature, H. wrightii maintained shootdensities and leaf quantum efficiencies. Although H. wrightii showed a greater tolerance to high temperature, T. testudinum had agreater capacity to sustain biomass and short shoots under thermal stress with labile C enrichment, regardless of the fact that sulfidelevels in the T. testudinum cores were 2 times higher than in the H. wrightii cores. Tropical seagrass tolerance to elevatedtemperatures, predicted in the future with global warming, should be considered in the context of the sediment-plant complexwhich incorporates the synergism of plant physiological responses and shifts in sulfur biogeochemistry leading to increased plantexposure to sulfides, a known toxin.© 2006 Elsevier B.V. All rights reserved.

Keywords: Dissolved organic carbon; Florida Bay; Glucose; Halodule wrightii; Thalassia testudinum

1. Introduction

Sediments in seagrass and other shallow marine soft-bottom communities are characterized as highly organic

(Hemminga and Duarte, 2000). The decomposition oforganic matter in these coastal marine sediments ismicrobially mediated through the dissimilatory reduc-tion of sulfate to sulfide (H2S, HS

−) (Sørensen et al.,1979; Howarth and Hobbie, 1981; Jørgensen, 1982;Canfield, 1993; Holmer and Kristensen, 1996; Holmeret al., 2003). Consequently, without reoxidation, pore-water sulfides build up in coastal marine sediments andcan lead to a chronic exposure of seagrass belowground

Journal of Experimental Marine Biology and Ecology 341 (2007) 91–101www.elsevier.com/locate/jembe

⁎ Corresponding author. Tel.: +1 561 297 3325; fax: +1 561 2972749.

E-mail addresses: [email protected] (M.S. Koch),[email protected] (C.J. Madden).

0022-0981/$ - see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.jembe.2006.10.004

Auth

or's

pe

rson

al

copy

tissues to high concentrations of sulfide, a known phy-totoxin (Linthurst, 1979; Havill et al., 1985; Koch et al.,1990; Goodman et al., 1995; Raven and Scrimgeour,1997; Holmer and Bondagarrd, 2001; Barber andCarlson, 1993; Carlson et al., 1994). The degree towhich seagrass below ground tissues are exposed tosulfides is determined by multiple factors: microbialrespiration rates, the ability of plants to oxidize theirinternal lacunae tissue (aerenchyma), the extent of theoxidized microzone of sediment surrounding the roots(rhizosphere), and local sediment biogeochemistry(Eldridge et al., 2004).

Temperature and the availability of labile organicsubstrates are primary factors controlling microbialsulfate reduction rates (SRR) (Holmer and Kristensen,1996; Blaabjerb et al., 1998; Cotner et al., 2004; Palludand Van Cappellen, 2006; Marbà et al., 2006). Labilecarbon produced by high rates of seagrass photosynthe-sis provides low molecular weight carbon substrates tofuel the sulfate reducing microbial community (Blaab-jerb et al., 1998; Hines et al., 1999; Cotner et al., 2004).Thus high photosynthetic rates of tropical seagrasses,such as Thalassia testudinum Banks ex Köing (Four-qurean et al., 2001), and warm sub-tropical temperatures(N30 °C) probably account for the high millimolarporewater sulfide levels found in the sediments ofFlorida Bay (Barber and Carlson, 1993; Carlson et al.,1994), a large semi-enclosed subtropical lagoon at theterminus of the Florida peninsula. The dominantseagrass species in the Bay, T. testudinum, appears tobe well adapted to reduced sediment conditions. It hasextensive aerenchyma (Tomlinson, 1969) and a highcapacity to oxidize its rhizosphere (Lee and Dunton,2000; Borum et al., 2005), characteristics well estab-lished in the wetland plant literature as adaptations tosediment hypoxia (Iizumi et al., 1980; Sand-Jensenet al., 1982, Smith et al., 1984, Caffrey and Kemp, 1991;Lee and Dunton, 2000).

While T. testudinum and other seagrass species have ahigh capacity to oxidize their rhizophere, microbial SRRin the tropics may overwhelm the ability of plants tocompletely reoxidize reduced sulfur compounds in pore-waters. Confounding this problem is the fact that tropicalcarbonate sediments low in iron are inefficient at bindingsulfides into solid-phase forms, contrasting temperatemarine sediments dominated by pyrite and iron sulfidecompounds (Berner, 1984; Chambers et al., 2001).Seagrass systems in carbonate sediments and with highorganic loads, either from high internal organic produc-tion and/or undergoing anthropogenic eutrophication,tend to be exposed to high levels of porewater sulfides(Azzoni et al., 2001; Holmer et al., 2003).

Sulfides have been shown to be phytotoxic to a rangeof aquatic plant species, including seagrasses (Ingoldand Havill, 1984; Havill et al., 1985; Koch andMendelssohn, 1989; Koch et al., 1990; Goodmanet al., 1995). In fact, sulfide toxicity has been proposedas an important factor in promoting “die-off” events ofT. testudinum in Florida Bay (Robblee et al., 1991)and other seagrass species worldwide (Seddon et al.,2000; Azzoni et al., 2001; Holmer and Bondagarrd,2001; Plus et al., 2003; Holmer et al., 2005). Althoughthese studies implicate sediment reducing conditions,and specifically porewater sulfide, as an agent causingmortality in seagrass, there is conflicting evidence onthe direct role of sediment sulfide in causing seagrassmortality (Terrados et al., 1999), and few experimentalstudies have been conducted to determine the upperthreshold levels of sulfide or compare different speciestolerances. Further, sulfide accumulation is highestduring warm summer months when plants may alsoexperience high thermal stress, particularly in tropicalclimes, therefore this potential interaction requiresevaluation.

Herein we present results of a large-scale mesocosmexperiment where sulfide accumulation was stimulatedby adding labile carbon (glucose) to intact sedimentcores and the seagrass response determined. We alsoexamined the effects of increased temperature on sulfideaccumulation rates (SAR) and the synergism of thesetwo stressors (high temperature and sulfide) on plantgrowth and physiological response in two dominanttropical seagrass species T. testudinum and HalodulewrightiiAschers. We hypothesized that high temperaturemay increase sulfide accumulation causing a cumulativeimpact on tropical seagrasses.

2. Materials and methods

2.1. Plant collection and experimental setup

Intact plant cores were collected from Florida Bay, atthe southern terminus of the Florida (U.S.A.) peninsulaMay 21–29th, 2004. Intact cores of H. wrightii (15 cmdiameter×20 cm depth) were collected from Porjoe Key(25°13′41″N/80°47′37″W) and cores of T. testudinumwere collected from Green Mangrove Key (24°55′20″N/80°47′33″W) and transported to the Florida AtlanticUniversity Marine Lab (Boca Raton, FL) in coolers.Upon arrival, intact cores were immediately placed intomesocosm tanks with ambient coastal Atlantic seawater(36 psu), put on a 12:12 hr light–dark cycle and allowedto equilibrate for 4 weeks. The mesocosm setup in-cluded sixteen 500 L (3 m diameter×3 m height)

92 M.S. Koch et al. / Journal of Experimental Marine Biology and Ecology 341 (2007) 91–101

Auth

or's

pe

rson

al

copy

fiberglass tanks equipped with 2 powersweeps, one forcirculation at canopy height, and the other for surface tobottom circulation and continuous aeration (detailed inKoch et al., in press). In summary, each tank had a1000 W metal halide light delivering PAR light levels of864±34 μmol photons m−2 s−1 just below the watersurface and 582±56 μmol photons m−2 s−1 at thecanopy. The entire mesocosm experiment was run as aclosed systemwith deionized water amended to the tanksas needed to compensate for evaporation and to maintainambient salinities (35–37 psu). Coastal seawater fromthe flow through system on each tank was added weeklyto maintain nutrient levels in the tanks.

Temperatures were raised at a rate of 1 °C d−1 (June21, 2004) to allow for slow thermal acclimation. Therewere four replicate tanks for each temperaturetreatment (28–29 °C [ambient], 30–31 °C, 32–33 °Cand 34–35 °C). The elevated temperatures in thetreatment tanks were achieved by using two 300 Waquarium heaters in each tank for the 30–31 °C and32–33 °C treatments, and one 1000 W titanium heaterwith digital controls in each of the four tanks for the34–35 °C treatment. Once tanks reached temperaturetreatment levels (June 30, 2004) they were maintainedfor 38 d.

To stimulate sulfide production, glucose wasinjected through two vertical sippers permanentlyinstalled in the intact plant cores. Cores were injectedwith 5 mL of deoxygenated artificial seawater (InstantOcean 35 psu) alone (controls) or with 3.2 mol L−1

glucose dissolved in artificial seawater calculated toyield a pore water molarity of approximately 10 mmolL− 1 (Carlson, personal communication). Glucoseinjections were conducted on July 1 (day 1), 3 (day3), 5 (day 5), 19 (day 19), and 26 (day 26). Onceinjections were made, sippers were immediately closedoff with three-way valves to prevent porewaterexchange. Preliminary experiments showed this methodof adding glucose stimulated sulfide production in theporewater to the desired upper mM range found in thefield.

2.2. Plant response measurements

Weekly, leaf elongation rates were determined forT. testudinum using the leaf marking technique (Zieman,1974) and live shoots were counted (number core−1) ineach core (surface area=201 cm−2) for both species.Net shoot change (%) was calculated relative to theinitial short shoot number in the cores (initial mean±S.E.; Thalassia: 22.8±0.7 control and 21.8±0.5 glucose;Halodule: 92.3±9.0 control and 89±8.3 glucose) after

temperature treatment was attained. Quantum efficiencyof photosystem II or chlorophyll fluorescence (Fv/Fm)was measured on dark adapted (5 min) leaves eachweek using a Diving PAM (Pulse Amplitude Modu-lation; Walz, Germany). Leaf Fv/Fm ratios have beenfound to be an excellent indicator of stress in terrestrialand submerged aquatic plants with non-stressed sea-grass ratios in the range of 0.7 to 0.8 (Björkman andDemmig, 1987; Ralph, 1999; Durako et al., 2002). Inthe presentation of our results, we use 0.7 as a thresh-old value below which the plants are assumed to bestressed.

After 38 d of treatments T. testudinum plant tissuewas harvested followed by H. wrightii. Plant tissue wasseparated (leaf, root and rhizome) and immediatelyfrozen in liquid N2, freeze dried, weighed, ground withliquid N2, and stored in a desiccator for carbohydrateanalysis. Total soluble carbohydrates and starch weremeasured in leaf and rhizome tissue using methodsdescribed in Erskine and Koch (2000) modified fromYemm and Willis (1954) according to Zimmerman et al.(1989).

2.3. Physicochemical measurements

Pore water was sampled every 2–5 d by extracting5 mL from each sipper (10 mL) while simultaneouslyadding 5 mL of artificial seawater (35 psu) in order tomaintain porewater pressure. One subsample (5 mL) wasused to measure salinity (refractometer) and pH (Orion420A pH meter and Orion Triode pH electrode); thesecond 5 mL subsample was transferred into vialscontaining 5 mL of a highly alkaline sulfide buffer andimmediately measured with a sulfide ion electrode (mV;Orion 420A pH meter and an Orion Model 9616 Sure-Flow Combination Silver/ Sulfide Electrode). Theresulting total porewater sulfide pool measured is ex-pressed asΣTSpwwith the H2S (pKa1=7; pKa2=19) andHS−speciation being defined by pH. The ratio of H2S:HS− is approximately 50% at pH 7. Tank salinity andtemperature were monitored daily (YSI 85) and light (Li-Cor 1400 Data Logger and Li-Cor spherical sensor) anddissolved oxygen (YSI 85) measured weekly.

During the initial 10 d following the first glucoseamendment, sulfide rapidly accumulated in the pore-waters. We calculated the rate of porewater sulfideaccumulation (SAR) in the cores over time (d) using alinear model (n=5) with a high degree of fit for glucose(R2 =0.92–0.99) and control (R2 =0.75–0.99) treat-ments. We used SAR as an approximation of a minimumsulfate reduction rate (SRR). The control SAR calcu-lated closely approximated the SRR we measured in

93M.S. Koch et al. / Journal of Experimental Marine Biology and Ecology 341 (2007) 91–101

Auth

or's

pe

rson

al

copy

Florida Bay sediments using 35SO42− (Koch and Jensen,

unpublished data).

3. Results

3.1. Porewater sulfide and pH

Sediment SAR was significantly influenced by labileC and high temperature treatments and was species and/or sampling location specific. The addition of glucose toT. testudinum intact cores from Green Mangrove Keyunder ambient temperatures (28–29 °C) resulted inaverage ΣTSpw concentrations of 2.3 mM over the 38 dexperiment (Fig. 1). With labile C added, the SAR was 5

times higher (502 μmol L− 1 d− 1) than controls(114 μmol L−1 d−1, Table 1) and increased as a functionof temperature to levels as high as 1300 μmol L−1 d−1

at 32–35 °C during the first 10 d of the experiment(Table 1). The high temperatures resulted in maximumSAR and ΣTSpw of 14 mM when sulfides peaked after10 d, but over the 38 d experiment averaged 5.5 mM(Fig. 1). In contrast, without labile C additions, ΣTSpwpeaked to 3.8 mM at 34–35 °C after 10 d and averagedb2 mM over the 38 d experiment (Fig. 1).

Glucose amendments also stimulated SAR in H.wrightii cores, but the rates calculated and resultingΣTSpw concentrations were much lower than those in T.testudinum cores with and without glucose amendments(Fig. 1, Table 1). H. wrightii cores from Porjoe Key hadapproximately 50% lower maximum ΣTSpw concentra-tions (6.5 mM) than T. testudinum at day 10 following Cenrichment. The SAR in H. wrightii cores were 33 to74% of those calculated for T. testudinum cores withglucose amendments, and 32 to 51% compared toT. testudinum controls (Table 1). The SAR and ΣTSpwresponse to temperature and glucose treatments was alsonot as pronounced as was found for T. testudinum,suggesting a lower saturating rate (Fig. 1, Table 1).Further, ΣTSpw concentrations in H. wrightii controlsnever reached 1 mM, even in the highest temperaturetreatment.

A general lowering of pH in porewaters with labilecarbon amendments (Table 2) was coincident with highsulfide accumulation rates (Table 1). This pH shift mayindicate sulfide oxidation and/or release of organic acids.If sulfide oxidation occurred, our results in Table 1probably represent a minimum estimate of sulfide pro-duction. A shift from approximately pH 7 to 6 underglucose amendments would have shifted the sulfidespeciation (H2S:HS

−) ∼10% in favor of H2S.

3.2. Shoot mortality, growth and physiological response

Net shoot loss in T. testudinum was less than 5%across all temperature treatments in the control cores, but

Fig. 1. Average porewater total sulfide (ΣTSpw=H2S+HS−) concen-

trations from intact sediment cores of Thalassia testudinum and Ha-lodule wrightii extracted fromWestern (Eagle Key) and Eastern (PorjoeKey) Florida Bay, respectively, with glucose additions and artificialseawater controls over the 38 d experiment. Means±S.E. (n=10).

Table 1Sediment sulfide (ΣTSpw=H2S+HS

−) accumulation rates (SAR) inThalassia testudinum andHalodule wrightii intact cores from the eastern(Porjoe Key) andwestern (GreenMangrove Key) Bay sites, respectively,with andwithout glucose amendments and across a range of temperatures

Thalassia Halodule

ΣTSpw(μmol L−1 d−1) R2

ΣTSpw(μmol L−1 d−1) R2

Glucose28–29 °C 502 0.95 370 0.9930–31 °C 790 0.99 531 0.9832–33 °C 1285 0.98 429 0.9934–35 °C 1300 0.98 633 0.92Controls28–29 °C 114 0.98 37 0.7530–31 °C 88 0.82 33 0.9932–33 °C 116 0.86 59 0.9534–35 °C 257 0.96 67 0.93

Rates were calculated based on linear fits of increasing porewatersulfide concentrations over the first 10 d of the experiment aftermesocosm tanks attained incubation temperatures and glucoseamendments were applied to treatment cores.

Table 2Porewater pH in Thalassia testudinum and Halodule wrightii intactcores after 38 d exposure to temperature and glucose treatments

Temperature(°C)

Thalassia Halodule

Control Glucose Control Glucose

28–29 7.08 (0.11) 6.82 (0.10) 7.40 (0.12) 6.65 (0.42)30–31 7.13 (0.12) 6.85 (0.09) 7.27 (0.06) 6.74 (0.20)32–33 7.29 (0.17) 6.69 (0.26) 7.32 (0.14) 7.10 (0.39)34–35 6.92 (0.18) 6.35 (0.08) 7.33 (0.06) 6.49 (0.08)

Means given with standard deviation in parentheses (n=4).


Auth

or's

pe

rson

al

copy

with the combination of high temperature (34–35 °C)and glucose amendments, net shoot loss was 65% overthe course of the 38 d experiment (Fig. 2). Growth rateswere a sensitive indicator of stress with a highly signifi-cant (Pb0.01) affect of temperature and glucose treat-ments on mature leaf elongation rates (2-way ANOVA;Fig. 3). These trends in net production, as indexed by leafelongation, were also apparent in new leaf allocation.New leaf emergence declined significantly beyond thethreshold temperature of 33 °C, in both control andglucose amended cores (Fig. 3), and no new leafemergence was found in plants at the highest tempera-tures with glucose. Leaf quantum efficiencies paralleledgrowth responses in T. testudinum, with Fv/Fm ratiossignificantly different between all temperatures (Pb0.01;2-way ANOVA) with the exception of 28–29 °C and30–31 °C, indicating a stepwise response to temper-ature (Fig. 4). Although leaf florescence was not sig-nificantly different with glucose amendments, quantumyields tended to decline more steeply in the glucosetreatment as a function of increasing temperature rela-tive to controls, particularly beyond 31 °C (Fig. 4).

While T. testudinum was sensitive to increasing tem-peratures,H. wrightiiwas quite remarkable in its thermal

tolerance. Across all temperatures, H. wrightii sustainedshoot numbers greater than initial densities (N100%;Fig. 2) and maintained leaf quantum efficienciesN0.7 intreatments without glucose (Fig. 4). AlthoughH. wrightii exhibited a high thermal tolerance, it wasapparently highly sensitive to sediment sulfide stimulat-ed by glucose amendments. This was the case eventhough porewater sulfide levels were lower than inT. testudinum cores (Fig 1). Percent short shoot survivalwas 50% less with glucose amendments (64%) relativeto controls (121%). It appears that below groundexposure to sulfides affected individual shoot mortality.This is evidenced by the fact that the short shoots whichsurvived maintained high leaf quantum efficiencies≥0.70 (Fig. 4), even though the ratio of Fv/Fm signifi-cantly declined in the glucose treatment (0.715) relativeto controls (0.727; Pb0.05, 2-way ANOVA).

3.3. Plant biomass allocation

T. testudinum final leaf biomass was moderately af-fected at the highest temperature (0.94 g) relative tolower temperature treatments (1.00–1.19 g), but

Fig. 2. Thalassia testudinum and Halodule wrightii net percent changein short shoots over 38 d at temperature treatments with and withoutglucose amendments. Means±S.E. (n=4). See text for initial shootdensities.

Fig. 3. Growth rates of Thalassia testudinum mature and newlyemergent leaves after 38 d of exposure to high temperature treatmentswith and without glucose amendments. Means±S.E. (n=4). New leafemergence may have occurred during the 5 d growth interval, thus newleaf rates are considered a minimum.


Auth

or's

pe

rson

al

copy

consistent with results on net shoot loss, differenceswere only significant with glucose amendments (Fig. 5).Root biomass was not affected by temperature, but onaverage a 50% reduction in root biomass (2.21 to 1.38 g)was observed under glucose amendments. Rhizome bio-mass did not change significantly under either treatment,but total plant biomass tended (P=0.08, 2-way ANOVA)to be lower in the glucose (8.9 g) versus control treatment(10.8), particularly above 29 °C (Fig. 5).

The moderate biomass response observed in T.testudinum is sharply contrasted by the highly significantchange in H. wrightii biomass in response to bothtemperature and glucose treatments (Fig. 5). H. wrightiiincreased biomass in the leaves and roots up to 33 °C,beyond which leaf, root, and total biomass declined(Fig. 5). Total H. wrightii biomass significantly declinedwith glucose amendments (Pb0.01, 2-way ANOVA),but the pattern of increasing leaf and root biomass atelevated temperatures was sustained. It is interesting tonote however that the threshold beyond which totalbiomass declined was 2 °C lower under glucose treat-ments (30 °C) compared to the controls (Fig. 5). At34–35 °C and glucose amendments,H. wrightii biomasswas b1 g in the entire core, 5-times lower than in the

control treatment at this temperature, indicating the highsensitivity of H. wrightii to the interaction of sedimentreducing conditions and high temperature.

3.4. Leaf and rhizome carbohydrates

Leaf tissue of both T. testudinum and H. wrightii hadsignificantly higher soluble and insoluble (starch) carbo-hydrates with increasing temperature (28 to 34 °C,Fig. 6). A general pattern of carbohydrate reduction inthe rhizomes paralleled this increase in leaf carbohy-drates (Fig. 6), but this reduction was not statisticallysignificant. Although an increase in carbohydrates wasobserved in T. testudinum and H. wrightii leaves withincreasing temperatures, leaf carbohydrate (soluble andstarch) contents were only significantly different(Pb0.01, 2-way ANOVA) at the highest temperature(34–35 °C), and there was no significant change withthe addition of glucose in either species. As a result ofhigh variance, there was no significant reduction foundin T. testudinum rhizome starch with increasing

Fig. 4. Thalassia testudinum and Halodule wrightii leaf quantumefficiencies after 38 d at temperature treatments with and withoutglucose amended to intact cores. Means±S.E. (n=4).

Fig. 5. Thalassia testudinum and Halodule wrightii final biomass andpartitioning after 38 d at temperature treatments with and without(control) glucose amended to intact cores (means, n=4).


Auth

or's

pe

rson

al

copy

temperatures; however, there was a clear tendency forrhizome starch to decline at 34–35 °C in both species,most prominently under glucose treatments (Fig. 6).

4. Discussion

High temperature can cause direct physiologicalstress in tropical marine plants living at their thermaltolerance limits (Zieman, 1970, Thorhaug, 1974). Inaddition, high temperatures stimulate microbial sulfatereduction rates in marine sediments increasing sedimenthypoxia and exposure of below-ground tissues tosulfide, a known phytotoxin (Linthurst, 1979; Havill

et al., 1985; Koch et al., 1990; Goodman et al., 1995;Raven and Scrimgeour, 1997; Holmer and Bondagaard,2001; Barber and Carlson, 1993; Carlson et al., 1994;Azzoni et al., 2001). SAR in the porewater of intactseagrass cores from Florida Bay increased 2- to 3-foldwith a temperature increase from 28–29 °C to 33–34 °C. Upper temperature treatments in our studysimulated maximum temperatures (35–36 °C) recordedin Florida Bay (Boyer et al., 1997; Koch in preparation)which approach the thermal optima for bacterial SRR.Sulfate reduction rates by sulfate reducing bacteria havebeen shown to have an optimum temperature of approxi-mately 35–40 °C (Wieland and Kühl, 2000; Rabus et al.,

Fig. 6. Leaf and rhizome soluble carbohydrate and starch for Thalassia testudinum and Halodule wrightii after 38 d of exposure to high temperaturetreatments with (glucose) and without (control) glucose amendments. Means±S.E. (n=4).


Auth

or's

pe

rson

al

copy

2002), which can be several degrees (10 °C) higher thanthe growth optima of psychrotolerant species (Rabuset al., 2002); although the upper thermal range forbacterial sulfate reduction is 60 to 80 °C with hyper-thermophilic groups found at N110 °C (Machel, 2001).The majority of studies on SRR and sulfide productionin the tropics and warm water environments havebeen conducted on microbial mats where it has beenshown that sulfur cycling (SRR and sulfide oxidation) isstrongly regulated by temperature (Wieland and Kühl,2000). Thus, tropical seagrass tolerance to hightemperature needs to be considered in the context ofsediment-plant interactions in marine sediments. This isprincipally the case when the stress response promotesthe release of plant organic exudates which can betranslocated to the sediment microbial community(Rooney-Varga et al., 1997; Blaabjerb et al., 1998;Hines et al., 1999), thereby stimulating SRR (Blaabjerbet al., 1998; Hines et al., 1999;Cotner et al., 2004; Palludand Van Cappellen, 2006) and plant exposure toporewater sulfides.

By increasing labile C to the sediment of intact cores,we effectively stimulated the potential sulfide produc-tion rates 5-fold and shifted the sulfide equilibriaslightly in favor of H2S, the most toxic form of theΣTSpw pool (Bagarinao 1992). The addition of glucoseas well as other labile substrates (lactate) has beenshown in freshwater, brackish, and marine sediments tostimulate SRR with greatest stimulation at marine sites(Holmer et al., 2005; Pallud and Van Cappellen, 2006).We observed regional differences in SAR, a proxy forSRR, in Florida Bay sediment in response to labile Camendments, but these differences were not accountedfor by salinity, as was found in the aforementionedstudies, because in our mesocosm experiment all tankswere maintained at 35 psu. H. wrightii sediments fromeastern Florida Bay (Porjoe Key) never attained ΣTSpwlevels of 1 mM, while T. testudinum controls collectedfrom the western Bay site (Green Mangrove Key) fre-quently had maximum ΣTSpw levels just below 4 mM.Seagrass species may explain this discrepancy; howev-er, the same pattern was seen with the addition ofglucose. Adding labile C resulted in T. testudinumΣTSpw levels 2-fold higher than the levels recorded inH. wrightii cores. This difference in ΣTSpw accumula-tion could have been influenced by sulfide oxidation orsequestration of sulfide by Fe. Fe gradients in the Bayshow 2–3 times lower reactive Fe in western (1–1.5 μmol g−1) versus eastern (3–3.5 μmol g−1) Baysediments (Zhang et al., 2004). While sulfide oxidationand sequestration by reactive Fe could contribute to thedifferences found in ΣTSpw in our study, we have

conducted subsequent experiments examining potentialSAR (with and without glucose amendments) across theBay and find significantly higher potential rates inwestern versus eastern Bay sites (Koch et al., inpreparation) and measured higher SRR in westernversus eastern Bay sites using 35SO4

2− (Koch andJensen, unpublished data). This in situ east–west SRRgradient, probably accounted for the higher ΣTSpwlevels in T. testudinum cores versus H. wrightii cores inthis study.

Remarkably, without glucose amendments, both T.testudinum and H. wrightii were highly thermal tolerant,demonstrating their tropical affinities and potential tothermally adapt to high temperatures. Both species sus-tained short shoot densities across all temperature treat-ments with no mass mortality of shoots found in anycore with temperature treatments up to 35 °C. While theplants survived temperature treatments, there was a clearthermal threshold above 33 °C where T. testudinumgrowth, including new allocation of leaf tissue, declinedand leaf quantum efficiencies fell below 0.7, indicativeof thermal breakdown of photosynthetic function(Campbell et al., 2006). Above this same thresholdtemperature, H. wrightii maintained shoot densities andhigh leaf quantum efficiencies (N0.7), but had an overallreduction in biomass. Although few studies have exam-ined tropical seagrass tolerance to long-term high tem-perature exposure, our thermal thresholds agree quiteremarkably with field observations. Zieman (1970)found T. testudinum in South Florida to drop out ofthe benthic community at thermal effluent sites withtemperatures sustained above 36 °C. Further, Zieman(1975) measured maximum growth rates at all sites withtemperatures in the range of 28–31 °C and lowest attemperatures of 34–35 °C when salinities were alsodepressed (13–15 psu), a potential interaction. Inanother mesocosm experiment (salinity× temperature;Koch et al., unpublished data), we found T. testudinumgrowth to decline linearly with time (1–39 d) at 36 °Cand precipitously at 40 °C, consistent with thetemperature threshold for this species established inthe present study and reported by Zieman (1970, 1975).

While the thermal threshold for T. testudinum undersustained high temperature exposure is in the region of33 °C, H. wrightii was shown in this study to increaseshoots and sustain relatively high quantum efficienciesat 34–35 °C, albeit with lower overall biomass.McMillan (1984) also determined that H. wrightii wasmore thermally tolerant than T. testudinum, with all H.wrightii shoots surviving 4 weeks at 36 °C and 2 weeksat 37 °C, while some shoots of T. testudinum did notsurvive the first treatment and all shoots were lost in the


Auth

or's

pe

rson

al

copy

second treatment, respectively. Halodule is also fre-quently reported to occupy shallow waters in the tropics(Fiji) and subtropics (Florida) that can reach tempera-tures of 40 °C during some period of the day (seeMcMillan, 1984).

Tropical seagrass species thermal thresholds for highsustained temperatures (this study; Zieman, 1970) areslightly lower than those established using short-termpulsed (hourly) exposures to high temperature. A recentstudy by Campbell et al. (2006) confirmed themaintenance of leaf photosynthetic yield in Thalassia(hemprichii) and Halodule (uninervis) exposed to 1–4 hpulses of 35–40 °C, with Halodule exhibiting a greaterthermal tolerance than Thalassia at 45 °C. This short-term adaptation to temperaturesN40 °C may account forthe presence of H. wrightii and T. testudinum in the fieldwith diurnal maximum temperatures above thresholdsestablished under long-term exposure. While short-termintermittent exposures to extreme temperatures appeartolerable, this type of rapid exposure would not lead tosignificant changes in the sediment biogeochemistryfound in this study with sustained elevated tempera-tures. Under the latter scenario, and increased organicmatter loading to estuaries and coastal lagoons witheutrophication (Holmer et al., 2003), the competitionamong seagrass species in terms of their thermaltolerance might shift.

We found that T. testudinum had a greater capacity tosustain biomass and short shoots with labile C enrich-ment under high thermal stress compared to H. wrightii.This was the case regardless of the fact that sulfideexposure to T. testudinum was 2 times greater than forH. wrightii. Thus, while H. wrightii is very thermallytolerant (McMillan, 1984), its low capacity to storeoxygen and its very small and delicate root-andrhizome-system, may make it an inferior competitor toT. testudinum under extreme hypoxia and high sulfideexposure at moderately high temperatures common inFlorida Bay. However, if both species were to succumbto hypoxia, H. wrightii, an early successional species,would have the capacity to more rapidly recolonize inthe field. It is interesting to note that while H. wrightiiincreased its overall biomass (primarily as leaf biomass)up to 32–33 °C, under glucose treatments biomass wasmaximum at 30–32 °C. Beyond this temperature asignificant amount of below-ground tissue was lost andat 34–35 °C biomass was only 20% of controls. Whilemore tolerant of hypoxia and high temperature, T.testudinum had lower leaf growth and quantum effi-ciencies with increasing temperature in glucose treat-ments and at the highest temperature had no new leafproduction. Even though total mortality of shoots was

not observed in either species under high sulfide andtemperature interactions, and probably cannot solelyaccount for sudden seagrass die-off events in the Bay, itis evident that these interactive stressors individuallyand synergistically affect net carbon allocation inseagrass (Touchette and Burkholder, 2000). This wasevident in the reduction of total biomass at hightemperatures and further with glucose amendments,and is reflected in shifts in tissue carbohydrate levels.

Elevated temperatures increased leaf soluble andinsoluble carbohydrate levels in T. testudinum and H.wrightii. These data suggest that as temperature in-creased both species were attempting to keep pace withrespiratory carbon demands for sucrose, the major formof soluble carbohydrates in seagrass (Touchette andBurkholder, 2000). A stimulation in sucrolysis inresponse to stress has been observed in seagrass speciesexposed to elevated temperature and salinity determinedby sucrose-P-synthase activity (reviewed in Touchetteand Burkholder, 2000). Increases in leaf solublecarbohydrates with temperature did not correspond to adecline in leaf or rhizome starch. These data suggest newcarbohydrate synthesis in the leaves or simply accumu-lation. The latter explanation is supported in T.testudinum which had lower leaf elongation rates andquantum efficiencies at high temperatures, but H.wrightii sustained leaf quantum efficiencies and in-creased biomass, at least up to 33 °C, suggesting astimulation in carbon production. With glucose amend-ments and at high temperature, T. testudinum tended tohave lower rhizome carbohydrate levels. A carbon drainmay have resulted from anaerobic respiration, metabolicbreakdown of belowground tissues, and/or disruption oftranslocation due to hypoxia (Alcoverro et al., 1999)caused by extremes of high temperature and porewatersulfides. These data indicate that high temperatures andthe interactive stressor sulfide have the potential todisrupt carbon metabolism in tropical seagrasses.

Based on this mesocosm study we conclude thefollowing: (1) Sediment sulfides have the potential toaccumulate to very high levels in carbonate-dominatedseagrass sediments at seasonal high temperatureextremes. (2) Porewater sulfide accumulation is relatedto a sediment biogeochemical response to labile C. (3)Based on the species examined, tropical seagrasses havea high thermal tolerance, but are living very close totheir thermal limits of sustained high temperatures, eventhough short durations at temperature extremes appeartolerable. (4) The competitive dominance of individualseagrass species to high temperature may shift withthe added stress of increased sediment hypoxia and/orsulfide accumulation associated with eutrophication. (5)


Auth

or's

pe

rson

al

copy

Tropical seagrass tolerance to elevated temperatures,predicted in the future with global warming, should beconsidered in the context of the sediment-plant complexwhich incorporates the synergism between plant phys-iological response to temperature and changes in sedi-ment biogeochemical sulfur cycling.

Acknowledgements

We acknowledge the South Florida Water Manage-ment District (West Palm Beach, FL) for funding thisresearch and Everglades National Park (Homestead, FL)for their logistical support through the InteragencyScience Center at Key Largo, FL. The FL Institute ofOceanography's Keys Marine Laboratory facility alsoprovided field support. We thank numerous graduateand undergraduate students from our lab that spentcountless hours setting up and running the experimentand Neal Tempel for his assistance in designing themesocosms. We are grateful to the anonymous review-ers that provided insightful comments which improvedthis manuscript. [SS]

References

Alcoverro, T., Zimmerman, R.C., Kohrs, D.G., Alberte, R.S., 1999.Resource allocation and sucrose mobilization in light-limitedeelgrass Zostera marina. Mar. Ecol. Prog. Ser. 187, 121–131.

Azzoni, R., Giordani, G., Bartoli, M., Welsh, D.T., Viaroli, P., 2001.Iron, sulphur and phosphorus cycling in the rhizosphere sedimentsof a eutrophic Ruppia cirrhosa meadow (Volle Smarlacca, Italy).J. Sea Res. 45, 15–26.

Bagarinao, T., 1992. Sulfide as an environmental factor and toxicant:tolerance and adaptations in aquatic organisms. Aquat. Toxicol. 24,21–62.

Barber, T.R., Carlson, P.R., 1993. Effects of seagrass die-off on benthicfluxes and porewater concentrations of ΣCO2, ΣH2S, and CH4 inFlorida Bay Sediments. In: Oremland, R.S. (Ed.), Biogeochemistryof Global Change: Radiatively Active Trace Gases. Chapman andHall, New York, pp. 530–550.

Berner, R.A., 1984. Sedimentary pyrite formation: an update.Geochim. Cosmochim. Acta 48, 605–615.

Björkman, O., Demmig, B., 1987. Photon yield of O2 evolution andchlorophyll fluorescence characteristics at 77 K among vascularplants of diverse origins. Planta 170, 489–504.

Blaabjerb, V., Mouritsen, K.N., Finster, K., 1998. Diel cycles ofsulphate reduction rates in sediments of a Zostera marina bed(Denmark). Aquat. Microb. Ecol. 15, 97–102.

Borum, J., Pedersen, O., Greve, T.M., Frankovich, T.A., Zieman, J.C.,Fourqurean, J.W., Madden, C.J., 2005. The potential role of plantoxygen and sulphide dynamics in die-off events of the tropicalseagrass, Thalassia testudinum. J. Ecol. 93, 148–158.

Boyer, J.N., Fourqurean, J.W., Jones, R.D., 1997. Spatial characteristicsof water quality in Florida Bay and Whitewater Bay by multivariateanalysis: zones of similar influence (ZSI). Estuaries 20, 743–758.

Caffrey, J.M., Kemp, W.M., 1991. Seasonal and spatial patterns ofoxygen production, respiration and root-rhizome release in

Potamogeton perfoliatus L. and Zostera marina L. Aquat. Bot.40, 109–128.

Campbell, S.J., McKenzie, L.J., Kerville, S.P., 2006. Photosyntheticresponses of seven tropical seagrasses to elevated seawatertemperature. J. Exp. Mar. Biol. Ecol. 330, 455–468.

Canfield, D.E., 1993. Organic matter oxidation in marine sediments.In: Wollast, R., Chou, L., Mackenzie, F. (Eds.), Interactions of C,N, P and S Biogeochemical Cycles. Springer, Berlin, pp. 333–363.

Carlson, P.R., Yarbro, L.A., Barber, T.R., 1994. Relationship ofsediment sulfide to mortality of Thalassia testudinum in FloridaBay. Bull. Mar. Sci. 54, 733–746.

Chambers, R.M., Fourqurean, J.W., Macko, S.A., Hoppenot, R., 2001.Biogeochemical effects of iron availability on primary producers ina shallow marine carbonate environment. Limnol. Oceanogr. 46,1278–1286.

Cotner, J.B., Suplee, M.W., Chen, N.W., Shormann, D.E., 2004.Nutrient, sulfur, and carbon dynamics in a hypersaline lagoon.Estuar. Coast. Shelf Sci. 59, 639–652.

Durako, M.J., Hall, M.O., Merello, M., 2002. Patterns of change in theseagrass dominated Florida Bay hydroscape. In: Porter, J.W.,Porter, K.G. (Eds.), The Everglades, Florida Bay, and the CoralReefs of the Florida Keys. CRC Press, Boca Raton, pp. 523–538.

Eldridge, P.M., Kaldy, J.E., Burd, A.B., 2004. Stress response modelfor the tropical seagrass Thalassia testudinum: the interaction oflight, temperature, sedimentation, and geochemistry. Estuaries 27,923–937.

Erskine, J.M., Koch, M.S., 2000. Sulfide effects on Thalassiatestudinum carbon balance and adenylate energy charge. Aquat.Bot. 67, 275–285.

Fourqurean, J.W., Willsie, A., Rose, C.D., Rutten, L.M., 2001. Spatialand temporal pattern in seagrass community composition andproductivity in South Florida. Mar. Biol. 138, 341–354.

Goodman, J.L., Moore, K.A., Dennison, W.C., 1995. Photosyntheticresponses of eelgrass (Zostera marina L.) to light and sedimentsulfide in a shallow barrier island lagoon. Aquat. Bot. 50, 37–47.

Havill, D.C., Ingold, A., Pearson, J., 1985. Sulphide tolerance incoastal macrophytes. Vegetatio 62, 279–285.

Hemminga, M.A., Duarte, C.M., 2000. Seagrass Ecology. CambridgeUniversity Press, Cambridge.

Hines, M.E., Evans, R.S., Genthner, B.R., Sharak, W., Friedman, S.,Rooney-Varga, J.N., 1999. Molecular phylogenetic and biogeo-chemical studies of sulfate-reducing bacteria in the rhizosphere ofSpartina alterniflora. Appl. Environ. Microbiol. 65, 2209–2216.

Holmer, M., Bondagarrd, E.J., 2001. Photosynthetic and growthresponse of eelgrass to low oxygen and high sulfide concentrationsduring hypoxic events. Aquat. Bot. 70, 29–38.

Holmer, M., Kristensen, E., 1996. Seasonality of sulfate reduction andpore water solutes in marine fish farm sediment: the importance oftemperature and sedimentary organic matter. Biogeochemistry 32,15–39.

Holmer, M., Duarte, C.M., Marbá, N., 2003. Sulfur cycling andseagrass (Posidonia oceanica) status in carbonate sediments.Biogeochemistry 66, 223–239.

Holmer, M., Frederiksen, M.S., Mollegaard, H., 2005. Sulfuraccumulation in eelgrass (Zostera marina) and effect of sulfur oneelgrass growth. Aquat. Bot. 81, 367–379.

Howarth, R.W., Hobbie, J.E., 1981. The regulation of decompositionand heterotrophic microbial activity in salt marsh soils. Estuaries 4,289.

Iizumi, H., Hattori, A., McRoy, C.P., 1980. Nitrate and nitrite ininterstitial waters of eelgrass beds in relation to the rhizosphere.J. Exp. Mar. Biol. Ecol. 47, 191–201.


Auth

or's

pe

rson

al

copy

Ingold, A., Havill, D.C., 1984. The influence of sulphide on thedistribution of higher plants in salt marshes. J. Ecol. 72, 1043–1054.

Jørgensen, B.B., 1982. Mineralization of organic matter in the seabed — the role of sulphate reduction. Nature 296, 643–645.

Koch, M.S., Mendelssohn, I.A., 1989. Sulphide as a soil phytotoxin:differential responses in two marsh species. J. Ecol. 77, 565–578.

Koch, M.S., Mendelssohn, I.A., McKee, K.L., 1990. Mechanism forthe hydrogen sulfide-induced growth limitation in wetlandmacrophytes. Limnol. Oceanogr. 35, 399–408.

Koch, M.S., Schopmeyer, S., Kyhn-Hansen, C., Nielsen, O., Madden,C.J., in preparation. Seagrass die-off events in Florida Bay: links tosediment biogeochemical processes.

Koch, M.S., Schopmeyer, S.A., Kyhn-Hansen, C., Madden, C.J.,Peters, J.S., in press. Tropical seagrass species tolerance tohypersalinity stress. Aquat. Bot.

Lee, K.S., Dunton, K.H., 2000. Diurnal changes in pore water sulfideconcentrations in the seagrass Thalassia testudinum beds: theeffects of seagrasses on sulfide dynamics. J. Exp. Mar. Biol. Ecol.225, 201–214.

Linthurst, R.A., 1979. The effect of aeration on the growth of Spartinaalterniflora Loisel. Am. J. Bot. 66, 685–691.

Machel, H.G., 2001. Bacterial and thermochemical sulfate reduction indiagenetic settings—old and new insights. Sediment. Geol. 140,143–175.

Marbà, N., Santiago, R., Diaz-Almela, E., Alvarez, E., Duarte, C.M.,2006. Seagrass (Posidonia oceanica) vertical growth as an earlyindicator of fish farm-derived stress. Estuar. Coast. Shelf Sci. 67,475–483.

McMillan, C., 1984. The distribution of tropical seagrass with relationto their tolerance of high temperatures. Aquat. Bot. 19, 369–379.

Pallud, C., Van Cappellen, P., 2006. Kinetics of microbial sulfatereduction in estuarine sediments. Geochim. Cosmochim. Acta 70,1148–1162.

Plus, M., Deslous-Paoli, J.M., Dagault, F., 2003. Seagrass (Zosteramarina L.) bed recolonisation after anoxia-induced full mortality.Aquat. Bot. 77, 121–134.

Rabus, R., Brüchert, V., Amann, J., Köonneke, M., 2002. Physiolog-ical response to temperature changes of the marine, sulfate-reducing bacterium Desulfobacterium autotrophicum. FEMSMicrobiol. Ecol. 42, 409–417.

Ralph, P.J., 1999. Photosynthetic response of Halophila ovalis (R. Br.)Hook. f. to combined environmental stress. Aquat. Bot. 65, 83–96.

Raven, J.A., Scrimgeour, C.M., 1997. The influence of anoxia onplants of saline habitats with special reference to the sulphur cycle.Ann. Bot. 79, 79–86.

Robblee, M.B., Barber, T.R., Carlson, P.R., Durako, M.J., Four-qurean, J.W., Muehlstein, L.K., Porter, D., Yarbro, L.A., Zieman,R.T., Zieman, J.C., 1991. Mass mortality of the tropical seagrassThalassia testudinum in Florida Bay. Mar. Ecol. Prog. Ser. 71,297–299.

Rooney-Varga, J.N., Devereux, R., Evans, R.S., Hines, M.E., 1997.Seasonal changes in the relative abundance of uncultivated sulfate-reducing bacteria in a salt marsh sediment and in the rhizosphere ofSpartina alterniflora. Appl. Environ. Microbiol. 63, 3895–3901.

Sand-Jensen, K., Prahl, C., Stokholm, H., 1982. Oxygen release fromroots of submerged aquatic macrophytes. Oikos 38, 349–354.

Seddon, S., Connolly, R.M., Edyvane, K.S., 2000. Large-scaleseagrass dieback in northern Spencer Gulf, South Australia.Aquat. Bot. 66, 297–310.

Smith, R.D., Dennison, W.C., Alberte, R.S., 1984. Role of seagrassphotosynthesis in root aerobic processes. Plant Physiol. 74,1055–1058.

Sørensen, J., Jørgensen, B.B., Revsbech, N.P., 1979. A comparison ofoxygen, nitrate and sulfate respiration in coastal marine sediments.Microb. Ecol. 5, 105–115.

Terrados, J., Duarte, C.M., Kamp-Nielsen, L., Agawin, N.S.R., Gacia,E., Lancap, D., Fortes, M.D., Borum, J., Lubanski, M., Greve, T.,1999. Are seagrass growth and survival constrained by reducingconditions of the sediment? Aquat. Bot. 65, 175–197.

Thorhaug, A., 1974. Effects of thermal effluents on the marine biologyof southeastern Florida. In: Gibbons, J.W., Sharitz, R.R. (Eds.),Thermal Ecology. Technical Information Center, US AtomicEnergy Commission, pp. 518–531.

Tomlinson, P.B., 1969. On the morphology and anatomy of turtlegrass, Thalassia testudinum (Hydrocharitaceae). III. Floral mor-phology and anatomy. Bull. Mar. Sci. 19,, 286–305.

Touchette, B.W., Burkholder, J.M., 2000. Overview of the physiolog-ical ecology of carbon metabolism in seagrasses. J. Exp. Mar. Biol.Ecol. 250, 169–205.

Wieland, A., Kühl, M., 2000. Short-term temperature effects onoxygen and sulfide cycling in a hypersaline cyanobacterial mat(Solar Lake, Egypt). Mar. Ecol. Prog. Ser. 196, 87–102.

Yemm, E.W., Willis, A.J., 1954. The estimation of carbohydrates inplant extracts by anthrone. J. Biochem. 57, 508–514.

Zhang, J.Z., Fischer, C.J., Ortner, P.B., 2004. Potential availability ofsedimentary phosphorus to sediment resuspension in Florida Bay.Glob.Biogeochem.Cycles 18,GB4008. doi:10.1029/2004GB002255.

Zieman, J.C., 1970. The effects of a thermal effluent stress on the sea-grasses andmacroalgae in the vicinity ofTurkey Point, BiscayneBay,Florida. PhD thesis, University of Miami, Coral Gables, Florida.

Zieman, J.C., 1974. Methods for the study of growth and production ofturtle grass, Thalassia testudinumKönig. Aquaculture 4, 139–143.

Zieman, J.C., 1975. Seasonal variation of turtle grass, Thalassiatestudinum Köing, with reference to temperature and salinityeffects. Aquat. Bot. 1, 107–123.

Zimmerman, R.C., Smith, R.D., Alberte, R.S., 1989. Thermalacclimation and whole-plant carbon balance in Zostera marinaL. (eelgrass). J. Exp. Mar. Biol. Ecol. 130, 93–109.
