Optical Sensor Nanoparticles in Artificial Sediments−A New Tool ToVisualize O2 Dynamics around the Rhizome and Roots of SeagrassesKlaus Koren,†,# Kasper E. Brodersen,‡,# Sofie L. Jakobsen,† and Michael Kuhl*,†,‡,§

†Marine Biological Section, Department of Biology, University of Copenhagen, Helsingør DK-3000, Denmark‡Plant Functional Biology and Climate Change Cluster, University of Technology Sydney, Ultimo, New South Wales 2007, Australia§Singapore Centre on Environmental Life Sciences Engineering, School of Biological Sciences, Nanyang Technological University,Singapore 639798, Singapore

*S Supporting Information

ABSTRACT: Seagrass communities provide important eco-systems services in coastal environments but are threatened byanthropogenic impacts. Especially the ability of seagrasses toaerate their below-ground tissue and immediate rhizosphere toprevent sulfide intrusion from the surrounding sediment iscritical for their resilience to environmental disturbance. Thereis a need for chemical techniques that can map the O2distribution and dynamics in the seagrass rhizosphere uponenvironmental changes and thereby identify critical stressthresholds of e.g. water flow, turbidity, and O2 conditions in thewater phase. In a novel experimental approach, we incorpo-rated optical O2 sensor nanoparticles into a transparentartificial sediment matrix consisting of pH-buffered deoxy-genated sulfidic agar. Seagrass growth and photosynthesis was not inhibited in the experimental setup when the below-groundbiomass was immobilized in the artificial sulfidic sediment with nanoparticles and showed root growth rates (∼5 mm day−1) andphotosynthetic quantum yields (∼0.7) comparable to healthy seagrasses in their natural habitat. We mapped the real-time belowground O2 distribution and dynamics in the whole seagrass rhizosphere during experimental manipulation of light exposure andO2 content in the overlaying water. Those manipulations showed that oxygen release from the belowground tissue is much higherin light as compared to darkness and that water column hypoxia leads to diminished oxygen levels around the rhizome/roots.Oxygen release was visualized and analyzed on a whole rhizosphere level, which is a substantial improvement to existing methodsrelying on point measurements with O2 microsensors or partial mapping of the rhizosphere in close contact with a planar O2optode. The combined use of optical nanoparticle-based sensors with artificial sediments enables imaging of chemicalmicroenvironments in the rhizosphere of aquatic plants at high spatiotemporal resolution with a relatively simple experimentalsetup and thus represents a significant methodological advancement for studies of environmental impacts on aquatic plantecophysiology.

■ INTRODUCTION

Seagrasses are marine flowering plants that provide a range ofessential eco-engineering services, such as facilitating carbonsequestration, improving water clarity, and protecting coastalareas against erosion.1 Despite being considered as a high-valueecosystem, providing nursery areas and feeding grounds tonumerous important commercial marine fish and crustaceanspecies, seagrass meadows are currently declining with analarming rate mainly due to human activity.2,3 Seagrass plantsmostly inhabit shallow coastal sediments, where they form animportant coastal ecosystem with high productivity andbiodiversity.1,2 However, the below-ground biomass ofseagrasses is anchored and grows in organic rich, reduced andoften sulfidic sediments, which present a challenge to the plantsand can potentially be involved in die-off events.4,5 Sulfide, andespecially dissolved H2S, is highly toxic for seagrasses6,7 thathave developed a variety of structural defense mechanisms such

as the presence of an intracellular gas-filled lacunar system(aerenchyma) enabling rapid and low-resistance exchange ofgases between the above- and below-ground tissue andrhizosphere.8 However, our understanding of the function ofthis system and its role for the survival of seagrasses underenvironmental stress is still incomplete. The chemical micro-environment in the seagrass rhizosphere exhibits a highspatiotemporal heterogeneity that remains to be studied indetail. Especially the O2 dynamics in the rhizosphere is ofimportance as radial O2 loss from the below-ground biomasscan act as a microshield against toxic H2S from the surroundingsediment.9−11 Inadequate O2 transport from seagrass leaves to

Received: November 24, 2014Revised: January 15, 2015Accepted: January 22, 2015

Article

pubs.acs.org/est

© XXXX American Chemical Society A DOI: 10.1021/es505734bEnviron. Sci. Technol. XXXX, XXX, XXX−XXX

the below-ground tissue is regarded as a key mechanism inseagrass die-off events12,13 upon anthropogenic impacts such asdredging and eutrophication affecting the transparency and/orO2 level in the overlaying water.14

It has recently been shown that O2 released from themeristematic region of the rhizome of the seagrass speciesZostera muelleri subsp. capricorni alters the local below-groundchemical microenvironment by chemically reoxidizing H2S andthereby detoxifying the surrounding sediment.11 This chemicaldefense mechanism is highly affected by hypoxic water-columnconditions during darkness, as the oxidation capacity of thebelow-ground tissue of seagrasses is completely dependent onpassive diffusion of O2 from the surrounding water-column,across diffusive boundary layers and into the aerenchyma whenno O2 is evolved via photosynthesis.11−13

Most studies of the O2 dynamics in the seagrass rhizospherehave used electrochemical or optical microsensors9−11,15 thatenable measurements with a very high spatiotemporalresolution but only at a limited set of measuring points.Therefore, studying different regions of the rhizome is verytedious, and it is impossible to account for the complete O2distribution around the below-ground biomass. There is thus aneed for techniques that can map the heterogeneous O2microdistribution and dynamics around the rhizome androots of seagrasses over larger spatial scales.Imaging of chemical parameters using planar optical sensor

foils, i.e., planar optodes, is a powerful alternative tomicrosensor measurements.16 While planar optodes areexcellent tools for visualizing dynamic processes in sedi-ments,17−19 this approach is not as straightforward forinvestigations of the rhizosphere, where a good contactbetween the sensor foil and the plant tissue is needed.Achieving such a good contact is not easy and obviously onlypossible for selected parts of the roots at once.20,21 Dependingon the root geometry and the planar sensor layer thickness,diffusive smearing of the true O2 distribution can also beinduced due to the presence of the O2-impermeable sensor foilup against the biomass.22

In the present study we used O2 sensitive optical nano-sensors23 in combination with an artificial, semitransparentsulfidic sediment matrix to simultaneously map the O2dynamics in the whole rhizosphere of the seagrass Zostera

muelleri. We present the new methodology and show itsapplication for mapping responses in the rhizosphere O2microenvironment due to changes in irradiance and O2 contentin the overlaying water.

■ MATERIALS AND METHODSMaterials for the Nanosensors. Platinum(II) meso-

(2,3,4,5,6-pentafluoro)phenyl porphyrin (PtTFPP) was boughtfrom Frontier Scientific (www.frontiersci.com); Macrolexfluorescence yellow 10GN (MY) was obtained from KremerPigments (http://kremerpigments.com). The styrene maleicanhydride copolymer (PSMA with 8% MA, Mw: 250000g*mol−1) XIRAN was generously provided by Polyscope(http://www.polyscope.eu). Tetrahydrofuran (THF) wasobtained by Sigma-Aldrich.

Materials for the Transparent, Artificial Sediment.Agar powder for microbiology (gel point ∼35 °C; gel strength>300 g cm−2), HEPES buffer (N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid); pKa (at 25 °C) = 7.48; useful pHrange = 6.8−8.2), and sodium sulfide nonahydrate (Na2S·9H2O) were purchased from Sigma-Aldrich (www.sigmaaldrich.com).

Nanosensor Preparation. The sensor nanoparticles wereprepared according to the previously described method byMistlberger et al.24 Briefly, 200 mg of PSMA, 3 mg of MY(reference dye), and 3 mg of PtTFPP (O2 indicator) weredissolved in 20 g of THF. This mixture was quickly poured into200 mL of vigorously stirred distilled water. After evaporatingthe THF under an air stream, the particle suspension wasconcentrated at elevated temperature (60 °C) until aconcentration of 5 mg per mL was reached. The concentrationwas checked by drying and subsequent weighing of 1 mL of theparticle suspension. The obtained particles have a size of severalhundred nm and a strongly negative zeta potential of around−30 mV as shown elswhere.24 The particle suspension could bestored over several weeks without any signs of sedimentation,color change, or change in the calibration characteristics.

Seagrass Collection. Seagrass specimens of Zosteramuelleri subsp. capricorni (Asch.) S.W.L.Jacobs were collectedfrom a shallow (<1 m deep) coastal site at Brisbane Waters,NSW, Australia. The plants were transported (in water fromthe sampling site) to a greenhouse facility at the University of

Figure 1. A: Experimental setup. The below-ground tissue of the seagrass is embedded in the artificial sediment containing the O2 sensitivenanoparticles. A SLR camera and LED are mounted perpendicular to the transparent chamber wall. Gas supply and reference optode are immersedin the overlaying water. B: Calibration curve of the sensor nanoparticles in the artificial sediment. Symbols and error bars represent means ± SD (n =3). The red curve shows a fit of an exponential decay function to the calibration data (R2 > 0.999).

Environmental Science & Technology Article

DOI: 10.1021/es505734bEnviron. Sci. Technol. XXXX, XXX, XXX−XXX

B

Technology Sydney, where they were kept under natural lightat a salinity of ∼34 ppt and a temperature of ∼22 °C. Prior tothe experiments, specimens were gently washed free of anyadhering sediment particles before transferring them to theexperimental chamber.Experimental Setup. The experimental chamber (inner

dimensions 10 × 130 × 120 mm) consisted of a custom-madenarrow, transparent acrylic chamber with a removable frontwindow made of polycarbonate for ease of access when castingthe sediment and improved optical properties during imaging,respectively (see Figure 1 and the Supporting Informationvideo). Illumination of the plant leaves was provided by atungsten halogen lamp equipped with a collimating lens (KL-2500, Schott GmbH, Germany). Stirring and aeration of thewater-column was obtained via a submerged Pasteur pipetconnected to an air pump or a gas mixer (Sensorsense, TheNetherlands).Preparation of the Artificial Sediment. The transparent,

artificial sediment consisted of a deoxygenated ∼0.5% (w/w)agar-seawater solution (100 mL), buffered with HEPES (finalconcentration of 10 mM) and amended with O2-sensitivenanoparticles (2% w/w) and Na2S to a final H2S concentrationof 250 μM (at pH 7). Prior to casting the sediment, the agarpowder had been prewashed overnight in cold seawater toimprove clarity. The reduced, artificial sediment was thusconstructed to mimic chemical settings in natural sedimentwhile allowing for direct visual investigation of the below-ground tissue during measurements.The sensor nanoparticles were added to the heated artificial

sediment mixture during the preparation. Oxygen sensornanoparticles could be homogeneously incorporated into theartificial sediment matrix with no visible formation of largersensor particle aggregates in the agar. To ensure this, the timingof the nanoparticle addition to the agar is important, and thisshould be done shortly before the artificial sediment is pouredinto the chamber (i.e., at ∼38 °C). The concentration ofnanoparticles in the agar ensured a good measuring signal,while preserving a good visual transparency. To further avoidpotential limitations of transparency, we placed the seagrassrhizome close to (without touching) the polycarbonate plate ofthe experimental chamber when pouring the agar withnanoparticles into the chamber (at an agar matrix temperatureof ∼36 °C that rapidly cooled to room temperature uponcontact with the experimental chamber wall). The chambercould then be sealed and positioned in front of the imagingsystem. Gas supply from a gas mixer was ensured, and a fiber-optic optode (Pyro-Science GmbH, Germany) was introducedto continuously monitor the O2 concentration in the watercolumn. The leaf canopy was kept in the upper stirred water-phase. A schematic of the setup can be seen in Figure 1. Adetailed video documentation of these preparation steps can befound in the Supporting Information.Imaging Setup. We used a ratiometric RGB camera setup

for O2 imaging.25 The system consisted of a SLR camera (EOS1000D, Canon, Japan) combined with a macro objective(Macro 100 f2,8 D, Tokina, Japan) equipped with a 455 nmlong pass filter (Uqgoptics.com). Excitation of sensor particleswas achieved with a 405 nm multichip LED equipped with abandpass filter (NT43-156, Edmundoptics.com). The LED waspowered by a USB-controlled LED driver unit for fluorescenceimaging applications (available from http://imaging.fish-n-chips.de). Image acquisition control of the SLR and LED was

done with the software look@RGB (http://imaging.fish-n-chips.de).

Image Analysis and Calibration. Acquired images weresplit into red, green, and blue channels and analyzed using thefreely available software ImageJ (http://rsbweb.nih.gov/ij/). Inorder to obtain O2 concentration images the following stepswere performed: First the red channel (O2 sensitive emission ofPtTFPP) and green channel (emission of the reference dyeMY) images were divided using the ImageJ plugin Ratio Plus(http://rsb.info.nih.gov/ij/plugins/ratio-plus.html). Afterward,the obtained ratio-image was fitted with the previously obtainedcalibration curve using the Curve Fitting tool of ImageJ(exponential decay). The calibration curve was generated asfollows. A small piece of nanosensor-containing nonsulfidic agarwas immobilized in the chamber. Oxygen levels of the water ontop of the agar were altered with the help of compressed air andnitrogen, which were mixed by a PC-controlled gas mixer(SensorSense, The Netherlands). Simultaneously, the O2 levelin the water column was monitored by means of a calibrated O2optode system (Oxygen dipping probe connected to Piccolo2m; PyroScience GmbH, Aachen, Germany). To ensure thatequilibrium was reached, each calibration step was held for 60min. The calibration was obtained by linking the measuredimage ratios to the measured O2 level (Figure 1). Avisualization of the calibration process can be found in theSupporting Information (Figure S1).

Seagrass Photosynthetic Performance. We assessed thephotosynthetic competence of various parts of the seagrassesusing a fiber-optic pulse amplitude modulated (PAM)fluorometer (PocketPAM, Walz GmbH, Germany) measuringthe quantum yield of the PSII photosynthetic electron transportin the dark adapted state (Fv/Fm = the maximal quantum yield)and in the presence of actinic light (YII = the effective quantumyield).26

Experimental Treatments. We monitored the O2distribution around the seagrass roots under 2 experimentalmanipulations:i) During a light-dark transition, where the plant leaves were

first illuminated with an incident photon irradiance of ∼500μmol photons m−2 s−1 for 90 min to ensure that equilibriumconditions were reached in the light. Thereafter, the externalillumination was switched off, and the plant was left in the darkfor 3 h. During the dark incubation, the image acquisition couldbe automated, while measurements in light required switchingthe external light source off for a brief period just before andduring image acquisition.ii) During decreasing O2 contents in the water-column, from

100% air saturation down to ∼0% air saturation. The plant wasfirst kept at an irradiance of ∼500 μmol photons m−2 s−1 for 90min in air saturated water to ensure that equilibrium conditionswas reached. Then the external illumination was switched off,and the overlaying water was flushed with N2 gas for 2.5 h tosimulate water column hypoxia. Finally the water-column wasagain bubbled with air, and the plant was still kept in the darkfor another 4.5 h. The O2 concentration in the water-columnwas monitored simultaneously by the above-mentioned fiber-optic O2 optode. Image acquisition was performed as describedabove.

■ RESULTS AND DISCUSSIONIntroduction of optical sensor nanoparticles into the artificialsediment did not affect the sensor performance. The calibrationcurve of the agar-immobilized O2 nanoparticles showed the

Environmental Science & Technology Article

DOI: 10.1021/es505734bEnviron. Sci. Technol. XXXX, XXX, XXX−XXX

C

expected exponential decay typical for optical O2 sensors basedon luminescence quenching27 (Figure 1).Obviously, it is important to assess potential effects of the

artificial sediment with nanosensors on the seagrass health. Inthis study, we evaluated two plants of the species Z. muelleri.Both plant specimens tolerated the artificial sediment withnanoparticles well, and new root growth (at a rate of ∼5 mmd−1) was actually observed in both plants after a few days. Goodplant health was also confirmed by measurements of photo-synthetic performance of the two plants (Table 1). PAM

fluorometry measurements in both plants revealed a highquantum efficiency of photosynthesis in the leaf canopy (shoot1); i.e., maximum and effective quantum yields of PSII ∼0.7and ∼0.6 (at a light intensity of 500 μmol photons m−2 s−1),respectively. As seagrasses are considered healthy whenmaximum PSII quantum yields of the shoot are around 0.7,28

we concluded that the two studied plants cultivated in theartificial sediment were healthy. Besides the shoot, thephotosynthetic performance of the prophyllum, i.e., singleleaves originating from the horizontal rhizome, was alsoevaluated. Whereas plant 1 had photosynthetically activeprophyllums, albeit with a lower photosynthetic quantumefficiency than the fresh leaves, plant 2’s prophyllums showedno photosynthetic activity (Table 1); these observations werealso supported by the respective images of O2 concentrationaround these structures (see Figure 2). The presence ofphotosynthetically active prophyllums was a surprising finding,as these older plant structures are typically considered inactive.A more detailed investigation of this finding was, however,beyond the scope of this study and will be examined in futurework.While the presented O2 nansosensor methodology is not

applicable to natural nontransparent sediments, it is well suitedto investigate the below-ground chemical microenvironment ofseagrasses embedded in artificial transparent sediment matricesthat mimic key aspects of the sediment biogeochemistry such ashigh sulfide contents.11 With the experimental setup, bothstructural images of the two plants as well as images of O2concentration surrounding their below-ground biomass couldbe recorded (Figure 2). Both images were taken after exposingthe plants to an irradiance of 500 μmol photons m−2 s−1 for 90min. The O2 distribution in the below-ground environment wasevidently very different between the two investigated speci-mens. In plant 1, photosynthetically active prophyllums (onemarked as P in each picture) oxidized the sediment, while plant2 mainly showed O2 leakage around the nodiums (marked as

N), the basal meristem (B), and at the root tips. One of thebenefits of the method presented here is that aligned structuraland chemical (O2 concentration) images can easily be obtained.While the chamber is illuminated by an external light source astructural image can be taken and by switching off the externallight source and triggering the LED illumination, aluminescence image can be acquired that leads to an O2concentration image. In this way the position of structuralelements (e.g., roots, rhizome) can be precisely aligned with thechemical information. A slight drawback is that due to theemission filter in front of the SLR camera, the structural pictureappears slightly yellow.The O2 imaging showed that it is possible to simultaneously

map the O2 distribution within the entire seagrass rhizosphere(Figure 2, 3), and the method is well suited to observe changesin sediment oxygenation under different environmentalconditions. Light exposure of the leaf canopy thus dramaticallychanged the O2 status around the roots and rhizome (Figure3). Under high irradiance, seagrass leaf photosynthesisproduced O2 that was transported to the below-groundbiomass of the plant, where it supported aerobic metabolismand leaked into the immediate rhizosphere leading to locallyincreasing O2 levels during illumination (Figure 3A, 3C; regionof interest (ROI), 1−3). During illumination, leakage of O2 wasalso observed at the node region (ROI 2) and the basalmeristem (ROI 3). In contrast to the area on top of one of theprophyllums (ROI 1), this leakage did not originate fromphotosynthesis at the spot but was due to diffusive transportthrough the aerenchyma.In darkness, O2 diffuses into the leaves from the surrounding

water, across the diffusive boundary layer (DBL), and is thentransported to the below-ground tissue. This caused lessoxygenation of the rhizosphere, due to a relatively lower O2supply from the above-ground tissue and owing to O2consumption along the diffusive transport path (Figure 3;

Table 1. Maximum (Fv/Fm) and Effective (Y(II)) QuantumYields of PSII-Related Photosynthestic Electron Transportin Seagrass Leaves of Plants Mounted in the ExperimentalSetup with Artificial Sediment + O2 Nanoparticles

a

shoot prophyllum prophyllum

plant 1 nodium 1 nodium 5 nodium 8

Fv/Fm 0.74 ± 0.01 0.61 ± 0.02 0.59 ± 0.02Y(II) 0.60 ± 0.01 0.48 ± 0.02 0.39 ± 0.01

shoot prophyllum prophyllum

plant 2 nodium 1 nodium 2 nodium 5

Fv/Fm 0.73 ± 0.01 (−) (−)Y(II) 0.58 ± 0.02 (−) (−)

aMean ± SD; n = 4−6. (−) indicates no photosynthetic activity.

Figure 2. Structural images of the seagrass Z. muelleri mounted in theartificial sediment (A, C) and the respective false color images of theO2 concentration distribution (B, D) around plant 1 (top) and plant 2(bottom) recorded after 90 min illumination of the leaves with 500μmol photons m−2 s−1. Several plant structural elements are pointedout: S − shoot, N − nodium, P − prophyllum, B − basal meristem.

Environmental Science & Technology Article

DOI: 10.1021/es505734bEnviron. Sci. Technol. XXXX, XXX, XXX−XXX

D

Figure 3. A: False color image of the O2 concentration around the seagrass roots taken after 90 min illumination of the leaves with 500 μmolphotons m−2 s−1. B: An O2 depletion image visualizing the change in O2 concentration between the end of the light period (i.e., onset of darkening)and 130 min later. C: time profile of the 3 regions of interest (ROIs) over the light-dark exposure experiment. D: line profile (line shown in A)across some small roots at the time points 90 and 240 min.

Figure 4. A: False color image of the O2 concentration around the seagrass roots taken after 90 min illumination of the leaves at 500 μmol photonsm−2 s−1. Oxygen dynamics pictures visualizing the change in oxygenation between the time points 0 min (light) and 135 min (anoxic water) (B) andbetween the time points 135 min (anoxic water) and 405 min (air saturated water) (C). D−E: time profile of the 6 ROIs. F: line profile across somesmall roots at the time points 0, 120, and 240 min.

Environmental Science & Technology Article

DOI: 10.1021/es505734bEnviron. Sci. Technol. XXXX, XXX, XXX−XXX

E

ROI 1−3 and line profile). To visualize the changes betweenhigh light and dark conditions, we calculated an O2concentration difference picture by subtracting the dark O2image (time point 240 min) from the high light O2 image (timepoint 90) (Figure 3B). This showed that the O2 production inthe prophyllum had a high local impact, but clear differenceswere also evident at the root tips, close to the nodium and thebasal meristem. Such O2 leakage into the rhizosphere canprotect the seagrass from H2S intrusion through developmentof an oxic microshield around the below-ground tissue thatprevents H2S from reaching the tissue surface, via chemicaloxidation with O2 (i.e., 2O2 + H2S → 2H+ + SO4

2−).11

Furthermore, the line profile in Figure 3D showed significantO2 leakage around the root tips. This leakage may have aparticular important defense role for the plant as this mayenable intermittent root growth through relatively oxidizedmicrozones.While light-dark shifts appear on a diel basis, other

environmental changes such as O2 depletion in the watercolumn appear more rarely in the natural environment.Hypoxic water-column conditions may be caused by anthro-pogenic impacts such as eutrophication (e.g., due to land runoffand nutrient loadings) and/or by dredging operations in areasclose to the seagrass meadows. During the latter, resuspendedanoxic sediment attenuates light and consumes O2 in the watercolumn leading to hypoxia (or even anoxia).14 This is especiallycritical for seagrasses during night-time, where photosyntheticO2 supply in the leaves is absent. To investigate such effects, wemonitored how defined changes in the O2 content of the watercolumn affected the rhizosphere O2 microenvironment of theseagrass (Figure 4). In 5 of the selected ROIs, O2 was rapidlydepleted in the seagrass rhizosphere under dark anoxicconditions in the overlaying water. After re-establishing fullatmospheric saturation in the water column, O2 diffusion fromthe water column into above-ground tissue resulted inincreased below-ground O2 levels in the rhizosphere thatapproached steady state (Figure 4; ROI 2, 3, 5, and 6). This wasfurther confirmed by measuring line profiles of the O2concentration across some smaller roots (Figure 4F). Notably,ROI 1 in Figure 4 showed no increase in O2 concentrationindicating that the prophyllums were not supplied with O2 viathe aerenchyma. The presented experimental setup thusprovides important information about the oxidation capabilityof the below-ground tissue of seagrasses under variousenvironmental scenarios. Spatially explicit investigations areimportant, especially seen in light of recent studies linkingseagrass die-backs with hypoxic water-column conditions,leading to internal anoxia, and thereby making the plantsmore susceptible to sulfide intrusion.11−13

In Figure 4, ROI 4 was chosen to follow the O2 dynamics inthe artificial sediment in close proximity to the artificialsediment surface. This area showed efficient O2 diffusion intothe uppermost few mm’s within the time frame of theexperiments. The sulfide in the top layer of the artificialsediment thus gets depleted over time, when O2 is present inthe overlaying water and such depletion has to be considered asthis limits the long-term applicability of the experimental setup.Possible ways to avoid such oxygenation over longer incubationtimes are presented elsewhere.29 The O2 concentrationdifference images comparing high light and darkness (underwater column anoxia) showed pronounced changes around thephotosynthetic prophyllum (Figure 4B), while O2 leakage waspredominantly observed at the root tips and the nodiums when

diffusion from the water column was the only supplymechanism (Figure 4C).While planar optrodes only generate luminescence images

when the focal plane of the camera matches with the plane ofthe optrode, the method presented here enables imaging atdifferent focal planes. As the entire artificial sediment is stainedwith optical sensor nanoparticles it is in principle possible toimage at different focal planes. In the Supporting InformationO2 distribution images at six different focal plains within theartificial sediment are shown (Figure S2). Only in one of thepictures is the rhizome in focus; the others present focal planesin front and in the back of the rhizome. The further away fromthe actual roots the picture is taken the lower the O2concentration. This explains why close contact to the roots isrequired when working with planar optrodes. When the sensorfilm is further away from the roots, only a blurry image can beobserved.In conclusion, the use of O2 sensitive nanoparticles in

artificial transparent sediment represents a powerful new tool toanalyze the microenvironment around the below-groundbiomass of seagrasses and to quantify O2 dynamics at highspatiotemporal resolution in the whole rhizosphere uponenvironmental changes. The possibility of mapping the wholebelow-ground O2 distribution and dynamics in the seagrassrhizosphere enables identification of particular hot spots withdifferent O2 supply mechanisms. In contrast to microsensormeasurements, this method can thus generate data for multipleparts of the plant simultaneously significantly acceleratingstudies of the chemical dynamics in the rhizosphere of aquaticplants. In future studies, O2 mapping can be easilysupplemented with detailed spot measurements with micro-sensors for O2 and other chemical parameters such as pH andH2S,

11,29 where the selection of specific measuring sites can beguided by O2 sensitive nanoparticle maps. The application ofthis new experimental approach is not limited to seagrasses andcan easily be adapted to studies of other waterlogged plants andenvironments. For example, studies of O2 dynamics in rice withthe new approach presented here could easily be combinedwith other techniques for monitoring metal concentrationsaround the roots.30

■ ASSOCIATED CONTENT

*S Supporting InformationVisualization of the image analysis and O2 concentration imagesobtained in different focal planes. A detailed video documenta-tion of the O2 sensitive sediment preparation and mounting ofseagrass in the experimental chamber. This material is availablefree of charge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATION

Corresponding Author*E-mail: mkuhl@bio.ku.dk.

Author Contributions#K.K. and K.E.B. contributed equally.K.K., K.E.B., and M.K. designed the research. K.K., K.E.B.,

M.K., and S.L.J. conducted experiments. K.K., K.E.B., and M.K.analyzed the data. K.K. wrote the manuscript with editorial helpfrom K.E.B. and M.K. All authors have given approval to thefinal version of the manuscript.

NotesThe authors declare no competing financial interest.

Environmental Science & Technology Article

DOI: 10.1021/es505734bEnviron. Sci. Technol. XXXX, XXX, XXX−XXX

F

■ ACKNOWLEDGMENTS

We thank Dr. Daniel A. Nielsen for help with the experimentalsetup and for fruitful discussions, Josua Hosch for help withvideo editing, and Sabrina Kapus for providing the TOCpicture. This study was funded by research grants from theAustralian Research Council (M.K.), the Danish Council forIndependent Research | Natural Sciences (M.K.), the VillumFoundation (M.K., K.K.), the Augustinus Foundation (K.E.B.),and the PA Fiskers Fund (K.E.B.).

■ REFERENCES(1) Orth, R. J.; Carruthers, T. J. B.; Dennison, W. C.; Duarte, C. M.;Fourqurean, J. W.; Heck, K. L., Jr.; Hughes, A. R.; Kendrick, G. A.;Kenworthy, W. J.; Olyarnik, S.; et al. A Global Crisis for SeagrassEcosystems. BioScience 2006, 56, 987−996.(2) Larkum, A. W. D.; Orth, R. J.; Duarte, C. M. Seagrasses: biology,ecology and conservation; Springer: Dordrecht, Netherlands, 2006.(3) Waycott, M.; Duarte, C. M.; Carruthers, T. J. B.; Orth, R. J.;Dennison, W. C.; Olyarnik, S.; Calladine, A.; Fourqurean, J. W.; Heck,K. L.; Hughes, A. R.; et al. Accelerating loss of seagrasses across theglobe threatens coastal ecosystems. Proc. Natl. Acad. Sci. U. S. A. 2009,106, 12377−12381.(4) Holmer, M.; Bondgaard, E. J. Photosynthetic and growthresponse of eelgrass to low oxygen and high sulfide concentrationsduring hypoxic events. Aquat. Bot. 2001, 70, 29−38.(5) Holmer, M.; Frederiksen, M. S.; Møllegaard, H. Sulfuraccumulation in eelgrass (Zostera marina) and effect of sulfur oneelgrass growth. Aquat. Bot. 2005, 81, 367−379.(6) Perez-Perez, M. E.; Lemaire, S. D.; Crespo, J. L. Reactive oxygenspecies and autophagy in plants and algae. Plant Physiol. 2012, 160,156−164.(7) Lamers, L. P. M.; Govers, L. L.; Janssen, I. C. J. M.; Geurts, J. J.M.; Van der Welle, M. E. W.; Van Katwijk, M. M.; Van der Heide, T.;Roelofs, J. G. M.; Sumolders, A. J. P. Sulfide as a soil phytotoxin-areview. Front. Plant Sci. 2013, 4, 268.(8) Borum, J.; Sand-Jensen, K.; Binzer, T.; Pedersen, O.; Greve, T.M. Oxygen Movement in Seagrasses. In Seagrasses: Biology, Ecology andConservation; Springer: Dordrecht, Netherlands, 2006; pp 255−270.(9) Jensen, S.; Kuhl, M.; Glud, R.; Jørgensen, L.; Prieme, A. Oxicmicrozones and radial oxygen loss from roots of Zostera marina. Mar.Ecol.: Prog. Ser. 2005, 293, 49−58.(10) Connell, E. L.; Colmer, T. D.; Walker, D. I. Radial oxygen lossfrom intact roots of Halophila ovalis as a function of distance behindthe root tip and shoot illumination. Aquat. Bot. 1999, 63, 219−228.(11) Brodersen, K. E.; Nielsen, D. A.; Ralph, P. J.; Kuhl, M. Oxicmicroshield and local pH enhancement protects Zostera muelleri fromsediment derived hydrogen sulphide. New Phytol. 2015, 205, 1264−1276.(12) Borum, J.; Pedersen, O.; Greve, T. M.; Frankovich, T. A.;Zieman, J. C.; Fourqurean, J. W.; Madden, C. J. The potential role ofplant oxygen and sulphide dynamics in die-off events of the tropicalseagrass Thalassia testudinum. J. Ecol. 2005, 93, 148−158.(13) Pedersen, O.; Binzer, T.; Borum, J. Sulphide intrusion ineelgrass (Zostera marina L.). Plant, Cell Environ. 2004, 27, 595−602.(14) Erftemeijer, P. L. A.; Lewis, R. R. R. Environmental impacts ofdredging on seagrasses: a review. Mar. Pollut. Bull. 2006, 52, 1553−1572.(15) Pedersen, O.; Borum, J.; Duarte, C. M. Oxygen dynamics in therhizosphere of Cymodocea rotundata. Mar. Ecol.: Prog. Ser. 1998, 169,283−288.(16) Schaferling, M. The art of fluorescence imaging with chemicalsensors. Angew. Chem., Int. Ed. Engl. 2012, 51, 3532−3554.(17) Glud, R. N.; Tengberg, A.; Kuhl, M.; Hall, P.; Klimant, I. An insitu instrument for planar O 2 optode measurements at benthicinterfaces. Limnol. Oceanogr. 2001, 46, 2073−2080.

(18) Kuhl, M.; Polerecky, L. Functional and structural imaging ofphototrophic microbial communities and symbioses. Aquat. Microb.Ecol. 2008, 53, 99−118.(19) Polerecky, L.; Volkenborn, N.; Stief, P. High TemporalResolution Oxygen Imaging in Bioirrigated Sediments. Environ. Sci.Technol. 2006, 40, 5763−5769.(20) Frederiksen, M. S.; Glud, R. N. Oxygen dynamics in therhizosphere of Zostera marina: A two-dimensional planar optode study.Limnol. Oceanogr. 2006, 51, 1072−1083.(21) Jovanovic, Z.; Pedersen, M.; Larsen, M.; Kristensen, E.; Glud, R.Rhizosphere O2 dynamics in young Zostera marina and Ruppiamaritima. Mar. Ecol.: Prog. Ser. 2015, 518, 95−105.(22) Meysman, F. J. R.; Galaktionov, O. S.; Glud, R. N.; Middelburg,J. J. Oxygen penetration around burrows and roots in aquaticsediments. J. Mar. Res. 2010, 68, 309−336.(23) Borisov, S. M.; Mayr, T.; Mistlberger, G.; Waich, K.; Koren, K.;Chojnacki, P.; Klimant, I. Precipitation as a simple and versatilemethod for preparation of optical nanochemosensors. Talanta 2009,79, 1322−1330.(24) Mistlberger, G.; Koren, K.; Scheucher, E.; Aigner, D.; Borisov, S.M.; Zankel, A.; Polt, P.; Klimant, I. Multifunctional Magnetic OpticalSensor Particles with Tunable Sizes for Monitoring MetabolicParameters and as a Basis for Nanotherapeutics. Adv. Funct. Mater.2010, 20, 1842−1851.(25) Larsen, M.; Borisov, S. M.; Grunwald, B.; Klimant, I.; Glud, R.N. A simple and inexpensive high resolution color ratiometric planaroptode imaging approach: application to oxygen and pH sensing.Limnol. Oceanogr.: Methods 2011, 9, 348−360.(26) Ralph, P. J.; Short, F. T. Impact of the wasting disease pathogen,Labyrinthula zosterae, on the photobiology of eelgrass Zostera marina.Mar. Ecol.: Prog. Ser. 2002, 226, 265−271.(27) Meier, R. J.; Fischer, L. H.; Wolfbeis, O. S.; Schaferling, M.Referenced luminescent sensing and imaging with digital colorcameras: A comparative study. Sens. Actuators, B 2013, 177, 500−506.(28) Macinnis-Ng, C. M. .; Ralph, P. J. Short-term response andrecovery of Zostera capricorni photosynthesis after herbicide exposure.Aquat. Bot. 2003, 76, 1−15.(29) Brodersen, K. E.; Nielsen, D. A.; Ralph, P. J.; Kuhl, M. A splitflow chamber with artificial sediment to examine the below-groundmicroenvironment of aquatic macrophytes. Mar. Biol. 2014, 161,2921−2930.(30) Williams, P. N.; Santner, J.; Larsen, M.; Lehto, N. J.; Oburger,E.; Wenzel, W.; Glud, R. N.; Davison, W.; Zhang, H. Localized FluxMaxima of Arsenic, Lead, and Iron around Root Apices in FloodedLowland Rice. Environ. Sci. Technol. 2014, 48, 8498−8506.

Environmental Science & Technology Article

DOI: 10.1021/es505734bEnviron. Sci. Technol. XXXX, XXX, XXX−XXX

G