Biogeosciences, 10, 1379–1390, 2013www.biogeosciences.net/10/1379/2013/doi:10.5194/bg-10-1379-2013© Author(s) 2013. CC Attribution 3.0 License.

EGU Journal Logos (RGB)

Advances in Geosciences

Open A

ccess

Natural Hazards and Earth System

Sciences

Open A

ccess

Annales Geophysicae

Open A

ccess

Nonlinear Processes in Geophysics

Open A

ccess

Atmospheric Chemistry

and Physics

Open A

ccess

Atmospheric Chemistry

and Physics

Open A

ccess

Discussions

Atmospheric Measurement

Techniques

Open A

ccess

Atmospheric Measurement

Techniques

Open A

ccess

Discussions

Biogeosciences

Open A

ccess

Open A

ccess

BiogeosciencesDiscussions

Climate of the Past

Open A

ccess

Open A

ccess

Climate of the Past

Discussions

Earth System Dynamics

Open A

ccess

Open A

ccess

Earth System Dynamics

Discussions

GeoscientificInstrumentation

Methods andData Systems

Open A

ccess

GeoscientificInstrumentation

Methods andData Systems

Open A

ccess

Discussions

GeoscientificModel Development

Open A

ccess

Open A

ccess

GeoscientificModel Development

Discussions

Hydrology and Earth System

Sciences

Open A

ccess

Hydrology and Earth System

Sciences

Open A

ccess

Discussions

Ocean Science

Open A

ccess

Open A

ccess

Ocean ScienceDiscussions

Solid Earth

Open A

ccess

Open A

ccess

Solid EarthDiscussions

The Cryosphere

Open A

ccess

Open A

ccess

The CryosphereDiscussions

Natural Hazards and Earth System

SciencesO

pen Access

Discussions

Technical Note: A simple method for air–sea gas exchangemeasurements in mesocosms and its application in carbon budgeting

J. Czerny, K. G. Schulz, A. Ludwig, and U. Riebesell

GEOMAR – Helmholtz Centre for Ocean Research Kiel, 24105 Kiel, Germany

Correspondence to:J. Czerny (jczerny@geomar.de)

Received: 31 July 2012 – Published in Biogeosciences Discuss.: 3 September 2012Revised: 27 January 2013 – Accepted: 11 February 2013 – Published: 1 March 2013

Abstract. Mesocosms as large experimental units providethe opportunity to perform elemental mass balance calcula-tions, e.g. to derive net biological turnover rates. However,the system is in most cases not closed at the water surfaceand gases exchange with the atmosphere. Previous attemptsto budget carbon pools in mesocosms relied on educatedguesses concerning the exchange of CO2 with the atmo-sphere. Here, we present a simple method for precise deter-mination of air–sea gas exchange in mesocosms using N2Oas a deliberate tracer. Beside the application for carbon bud-geting, transfer velocities can be used to calculate exchangerates of any gas of known concentration, e.g. to calculateaquatic production rates of climate relevant trace gases. Us-ing an arctic KOSMOS (K iel Off ShoreMesocosms for fu-tureOceanSimulation) experiment as an exemplary dataset,it is shown that the presented method improves accuracy ofcarbon budget estimates substantially. Methodology of ma-nipulation, measurement, data processing and conversion toCO2 fluxes are explained. A theoretical discussion of pre-requisites for precise gas exchange measurements providesa guideline for the applicability of the method under variousexperimental conditions.

1 Introduction

Pelagic mesocosms represent large volume (mostly betweenone and fifty m3) experimental enclosures used to gatherdata on natural plankton communities (Petersen et al., 2003).Generally open towards the atmosphere, mesocosms, how-ever allowing for air–sea gas exchange, make it difficultto calculate production or consumption of CO2 and othervolatile compounds inside an experimental unit. Climate rel-

evant trace gases and other volatile carbon compounds pro-duced in marine environments are increasingly investigatedfor their potential climate feedbacks and have been mea-sured in previous mesocosm experiments (Sinha et al., 2007;Archer et al., 2012; Hopkins et al., 2012). Observed concen-trations in a mesocosm are a product of water-column reac-tions and losses or gains from the atmosphere. Precise knowl-edge of air–sea gas exchange rates can be used to calculatenet production rates of these compounds in the water col-umn, which can be compared between various experiments.Aquatic production rates in concert with data on biologicalcommunity composition and physiological state would helpto understand observed open-ocean distributions.

Not only in the context of global change, biological CO2fixation and consequent carbon export by sinking particles isof special interest to biogeochemical experimentalists. Mostmesocosm studies currently focus on investigating ecolog-ical interactions applying standard oceanographic methodson subsamples of the enclosed water. In principal, mesocosmexperiments also provide the opportunity to compare biogeo-chemical element fluxes such as air–sea gas exchange andexport to water-column production. With production rates,as usually measured in side experiments (e.g. O2 produc-tion or 14C incorporation), uncertainties arise from sampletransfer into bottle incubations and from extrapolating backfrom incubation conditions to temperature and light gradi-ents present in mesocosms. In situ primary production mea-surements using the whole enclosure as experimental vesselhave to be elaborated, in order to produce estimates com-parable to total mesocosm fluxes like sedimentation of or-ganic matter. Calculating carbon fluxes from water-columnpools of inorganic and organic carbon quantitatively relatedto air–sea fluxes and export rates could largely improve the

Published by Copernicus Publications on behalf of the European Geosciences Union.

1380 J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms

understanding of the system (Czerny et al., 2012a). To di-rectly estimate cumulative net community production (NCP),changes in total dissolved inorganic carbon (CT) have to becorrected for CO2 air–sea gas exchange and eventually forcalcification and evaporation. In previous mesocosm experi-ments in a Norwegian fjord (Delille et al., 2005) and indoors(Wohlers et al., 2009; Taucher et al., 2012), net commu-nity production (NCP), calculated on the basis of measuredchanges in CT, were presented. To calculate air–sea gas ex-change, Delille et al. (2005) used a parameterisation for winddependent boundary layer thickness achieved from experi-mental data compiled by Smith (1985). Wind speed, the cru-cial input parameter, was set to zero, because the mesocosmswere closed to the atmosphere and moored in a sheltered sur-rounding. Whereas most parameterisations result in zero gasexchange at zero wind speed (Wanninkhof, 1992), labora-tory derived wind dependent parameterisations by Smith etal. (1985) resulted in positive exchange at zero wind speed.Under calm conditions, gas exchange is low, but not zero; it isgoverned by other energy inputs than wind, e.g. thermal con-vection due to evaporation and temperature changes (Liss,1973; Wanninkhof et al., 2009). Although direct wind forc-ing might be negligible in most mesocosms, the general as-sumption that overall energy input is comparable to the con-ditions in the experimental tanks used by Smith et al. (1985),however, is not justifiable. Surface turbulence in many meso-cosm experiments is unlikely to be very low. Active mixingsystems, wave movement of the surrounding water, thermalmixing or the deployment of sampling gear might create tur-bulence within the enclosures, comparable to quite windyconditions. Taucher et al. (2012), for example, found windspeeds of more than 6 m s−1 to be necessary for balancingthe carbon budget in a Kiel indoor mesocosm experiment,applying the Smith et al. (1985) calculation. Parameterisa-tions for wind speed dependent gas exchange over the oceanare obviously not suitable for calculating mesocosm air–seagas exchange. Other than open-ocean gas exchange measure-ments, direct measurement of exchange velocities in an en-closed water volume can be relatively easily done.

Here, we present a simple method for direct measurementsof air–sea gas exchange rates in mesocosm experiments us-ing N2O as a tracer. N2O is a perfect choice as a gas tracer inthis application, due to its well known atmospheric concen-tration, relatively simple detection and structural similarityto CO2. Although N2O is not an inert gas, conditions favour-ing its biological production are unlikely to occur in pelagicmesocosms. Possible bias by biological activity is assessd byparallel measurement of natural variations in N2O and willbe discussed later in the manuscript. The conversion of mea-sured N2O exchange rates to those of CO2 and other gasesis explained. We are providing a detailed description of themethod and calculations including a discussion of prerequi-sites to achieve high quality data.

The measurement protocol and results are explained usinga KOSMOS (Kiel Off Shore Mesocosm for Ocean Simula-

tion, Fig. 1) experiment on ocean acidification in the Arcticas a model. Applicability of the method in the Kiel indoormesocosm facility is further explained and discussed.

2 Methods

2.1 Setup of the Svalbard 2010 ocean acidificationexperiment

Nine 15 m deep KOSMOS mesocosms, each with a diameterof 2m were moored end of May 2010 in the Kongsfjorden,Svalbard. Seven different CO2 treatment concentrations wereachieved through addition of CO2 saturated seawater. Whilethe ambient (∼ 180 µatmpCO2) control treatment was repli-cated twice, the seven enriched mesocosms followed a gradi-ent up to∼ 1420 µatmpCO2. Development of the enclosednatural plankton community was followed for 30 days afterCO2 manipulation, including addition of inorganic nutrientson day 13. For more details see Riebesell et al. (2012) andSchulz et al. (2013).

2.2 N2O addition

One litre of saturated N2O solution (N2O medical, Air Liq-uide, purity> 98 %) was prepared via bubbling of seawa-ter for two days in a narrow measurement cylinder coveredwith Parafilm®. Amounts of the solution to be added tothe mesocosm were calculated using solubility constants byWeiss and Price (1980) with respect to in situ salinities (S)and temperatures (T ). The targeted concentrations of N2Oshould be adapted to the setup in order to achieve meso-cosm to air fluxes, which can be measured at good preci-sion over reasonable time scales. Here, seawater tracer con-centrations were chosen in accordance to the highest certifi-cated reference material for N2O analyses available in ourlab (∼ 55 nmol kg−1).

Assuming a background concentration of 13 nmol kg−1,40 nmol kg−1 of medical grade N2O was added. Based onexperience, a surplus of approximately 20 % was added tothe mesocosms to account for losses unavoidable during han-dling of the solution.

Addition of the solution to the mesocosms (about 1–2 mL m−3) can be calculated according to the formulation:

Vad =Vw · ad

KT S · p, (1)

whereVad is the volume of N2O stock solution added (L),Vw the volume of the mesocosm (L), ad the desired addi-tion (mol L−1), andKT S is the solubility constant by Weissand Price (1980) forS and T of the N2O stock solution(mol L−1 atm−1) prepared at a pressure,p (atm), of one at-mosphere.

A syringe with a large inlet diameter was used to transferthe stock solution carefully. Filling of the syringe was done

Biogeosciences, 10, 1379–1390, 2013 www.biogeosciences.net/10/1379/2013/

J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms 1381

23

Figures 792

793

Fig. 1 794

Drawing of a KOSMOS mesocosm in the configuration used for the Svalbard experiment. 795

Fig. 1.Drawing of a KOSMOS mesocosm in the configuration usedfor the Svalbard experiment.

slowly as vacuum increases undesired outgassing of N2O.The stock solution was first diluted with filtered seawaterin 25 L carboys, which were filled almost to the rim. Thecontent of the carboys was homogeneously distributed to themesocosms by using the pumped injection device “Spider”(Riebesell et al., 2012).

2.3 Sampling

Three of the nine mesocosms were sampled every second dayusing integrating water samplers (IWS, Hydrobios). Equalamounts of sample were sucked into the sampling bottle ateach depth between 0 and 12 m, electronically controlled viahydrostatic pressure sensors. These integrated water samplesrepresent inventories of the 15 m deep water column. Tripli-cate samples were drawn directly from the sampler. The wa-ter was filled bubble free into 50 mL headspace vials via ahose reaching to the bottom of the vial. The vial volume wasallowed to overflow about four times before closing. Vialswere closed with butyl rubber plugs (N 20, Machery andNagel), crimp sealed and stored at room temperature afteraddition of 50 µL of saturated mercury chloride solution.

2.4 Measurement procedures

Measurement of aquatic N2O concentrations was performedvia gas chromatography (GC) with electron capture detec-tion (Hewlett Packard 5890 II), using a headspace staticequilibration procedure as described by Walter et al. (2006,precision∼ ±1.8 %). The GC was equipped with a 6′/1/8′′

stainless steel column packed with a 5A molecular sieve(W. R. Grace & CO) and operated at a constant oven tem-perature of 190◦C. A 95/5 argon-methane mixture (5.0, AirLiquide) was used as carrier gas. 10 mL of helium (5.0, AirLiquide) headspace was added to the sample vials and laterinjected into the sample loop of the GC after equilibrationwas achieved by manual shaking and storage of the vials forat least 10 h at a temperature of 21◦C. Certified gas mix-tures of N2O in artificial air (Deuste Steininger GmbH) withmixing ratios of 87.2± 0.2, 318± 0.2 and 1002± 0.2 ppb aswell as 1 : 1 dilutions with helium were used to construct cal-ibration curves with a minimum of three data points closeto sample concentrations. Headspace to water phase ratios inthe vials was determined gravimetrically.

Total dissolved inorganic carbon (CT) was determined viacoulometric titration using a SOMMA system and total al-kalinity (TAlk) via potentiometric titration (Dickson, 1981)(standard error of both methods∼ ±1 µmol kg−1). CO2 con-centrations, partial pressures and pH (total scale) were cal-culated from CT and TAlk measurements with the programCO2SYS by Lewis and Wallace (1995). For more details oncarbonate chemistry see Bellerby et al. (2012).

Determination of salinity and temperature in the meso-cosms was performed with a data logger-equipped hand heldmultisensory CTD 60M (Sea and Sun Technology). Volume

www.biogeosciences.net/10/1379/2013/ Biogeosciences, 10, 1379–1390, 2013

1382 J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms

of the mesocosms was determined with the same instrumentusing sodium chloride additions of∼ 0.2 g kg−1 as a tracer(Czerny et al., 2012b).

Wind velocity and direction measured at 10 m height on-shore, about one mile from the mooring site, were providedby the staff of the AWI-PEV Station in Ny Alesund.

Atmospheric measurements of N2O and CO2 were mea-sured on close by Zeppelin Mountain (∼ 4.5 km from the ex-perimental side) and provided to us by the NOAA CarbonCycles Gases Group in Boulder, CO, USA and ITM in Stock-holm University, Sweden, for N2O and CO2, respectively.

3 Results and discussion

Concentrations of N2O added on day 4 decreased in the en-riched mesocosms from initially measured∼ 50 nmol kg−1

on day 6 to∼ 30 nmol kg−1 on day 28 (Fig. 2). Concen-trations measured in the fjord close to the mesocosms wereslightly oversaturated compared to atmospheric equilibriumvalues, calculated for in situ seawaterT , S and atmosphericmixing ratios measured close by on Zeppelin Mountain. De-spite variable wind conditions, the concentration decrease in-side the mesocosms could be fitted (R2

= 0.96) using a stan-dard diffusion relationship:

CN2O = 60.556· e−0.0241·d (2)

where the concentration of N20 (CN2O) is described as anexponential function of the sampling day (d).

3.1 Calculation of CO2 fluxes from changes in N2Oconcentrations

Daily N2O fluxes were calculated from the fitted N2O con-centration decrease over time and converted to volumet-ric units. Changes in the N2O inventory, derived using thedetermined volume of the mesocosms (method describedin Czerny et al., 2012b) were used to calculate fluxes inµmol cm−2 h−1 (FN2O) across the water surface according to

FN2O =Iw1 − Iw2

A · 1t(3)

whereIw1 is the fitted bulk water N2O inventory in µmol permesocosm ont1 andIw2 on t2 with 1t as the time intervalbetweent1 and t2 in h, while A is the nominal surface areaof the mesocosm in cm2. A N2O transfer velocity (kN2O) incm h−1 is then calculated by dividingFN2O by the concentra-tion gradient according to Eq. (4):

kN2O =FN2O

(CN2O w − CN2O aw)(4)

where CN2O w is the fitted bulk water N2O concentration(µmol cm−3) at the point in time andCN2O aw the calculated(Weiss and Price, 1980) equilibrium concentration of N2O

24

0

10

20

30

40

50

60

70

0 5 10 15 20 25 30

Day of the experiment

N2O

(n

mo

l L-1

)

796

797

Fig. 2 798

N2O concentrations during the experiment. Diamonds ( ) represent the measured 799

concentration inside the 3 examined mesocosms; blue squares ( ) are fitted daily 800

concentrations according to equation 2, circles ( ) are background N2O concentrations 801

measured in the surrounding fjord, triangles ( ) are calculated equilibrium concentrations 802

from atmospheric measurements at in situ T and S. Shaded areas indicate periods when waves 803

(up to H1/3 = 0.8m) occurred at the mooring site. 804

805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822

Fig. 2. N2O concentrations during the experiment. Diamonds (�)represent the measured concentration inside the three examinedmesocosms; blue squares (�) are fitted daily concentrations ac-cording to equation 2, circles (◦) are background N2O concentra-tions measured in the surrounding fjord, triangles (4) are calcu-lated equilibrium concentrations from atmospheric measurementsat in situT andS. Shaded areas indicate periods when waves (up toH1/3 = 0.8 m) occurred at the mooring site.

with the atmosphere at prevailing bulk waterT andS. Usingbulk water concentrations to derive a surface diffusion gra-dient perfect mixing of the mesocosm appears to be an ab-solute requirement. Referring from N2O to inert gasses withair–sea gas exchange being the only exchange process, it isirrelevant whether exchange is limited by mixing processesclose to the air–sea interface or within the water column.Yet, if a permanent stratification is formed inside the meso-cosm, the decrease of N2O bulk water concentration cannotbe used to calculate mesocosm–atmosphere CO2 exchange.Processes modulating the concentrations of biologically ac-tive compounds such as CO2 are usually variable along thelight gradient. Therefore, due to shallow primary production,considerable differences in the surface gradient of CO2 mightemerge compared to N2O surface gradients that are governedby diapycnal mixing. For stratified mesocosms, gas exchangecalculations require the integration of information about ver-tical distribution of tracer and gases of interest. Here, N2Oand CO2 inventories have to be determined by integrated wa-ter samples independently from surface gradients determinedfrom discrete surface water samples. Regardless of the spe-cific sampling strategy applied, it is imperative to use thesame protocol for the tracer as for the gases of interest.

Due to convection caused by slight temperature changesin the surrounding water (Fig. 5) and an evaporation inducedsalinity increase (Schulz et al., 2013), the mesocosms in theSvalbard KOSMOS study could be considered to be homo-geneous on time scales relevant for air–sea gas exchange.N2O as well as CO2 surface concentrations are therefore ad-equately represented by bulk water measurements.

kN2O can be translated into a transfer velocity for any othergas using its Schmidt numbers to correct for gas specific

Biogeosciences, 10, 1379–1390, 2013 www.biogeosciences.net/10/1379/2013/

J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms 1383

properties as shown for the transfer coefficient of CO2 (kCO2)

in Eq. (5):

kCO2 =kN2O(

ScCO2ScN2O

)0.5. (5)

The Schmidt number for N2O (ScN2O) published by Rhee(2000), and the Schmidt number for CO2 (ScCO2) derivedfrom diffusion coefficients published by Jahne et al. (1987)were used. Using this pair of Schmidt numbers,kCO2 is gen-erally less than 1 % smaller thankN2O, similarly to when bothSchmidt numbers are derived from coefficients published byJahne et al. (1987). If both coefficients are taken from Wilkeand Chang (1955) the difference is∼ 10 % and can becomelarger than 25 % ifScCO2 from Wilke and Chang (1955) ispaired with (ScN2O) derived from coefficients published byBroecker and Peng (1974). More recent wind and wave tankexperiments have shown that a conversion ofkN2O to kCO2

is not necessary as gas exchange of the two gases is indis-tinguishable under conditions were chemical enhancementof CO2 exchange is not relevant (Degreif, 2006). Fluxes forCO2 (FCO2) can then be calculated by multiplication ofkCO2

with the diffusion gradient between bulk water CO2 concen-trations (CCO2 w) and calculated equilibrium concentrationswith the atmosphere (CCO2 aw) as

FCO2 = kCO2 · (CCO2 w − CCO2 aw). (6)

Daily CO2 fluxes were calculated for the nine mesocosmswithin the Svalbard ocean acidification experiment (Fig. 3).In the first days after CO2 addition was completed (day4), maximum efflux of ∼ 2 µmol CO2 per kg seawaterand day could be observed in the highest CO2 treatment(∼ 1400 µatm) at a CO2 gradient of∼ 980 µatm. In the fol-lowing two weeks, the CO2 gradient was diminished by out-gassing CO2 in concert with biological uptake, so that fluxeson day 27 were considerably lower (gradient∼ 450 µatm).Decrease of fluxes as a result of decreasing CO2 gradientswas less pronounced in the more moderately oversaturatedmesocosms due to a higher buffer capacity of the carbonatesystem. About 0.5 µmol kg−1 d−1 CO2 gassed into the waterfrom the atmosphere in the non-manipulated control treat-ments (∼ 175 µatm). Here, biological uptake was roughlybalanced by influx so that the gradient remained rather con-stant over time.

3.2 Chemical enhancement of CO2 air–sea gasexchange

Another correction has to be applied to derive accurate CO2fluxes in calm environments like the KOSMOS mesocosms.As CO2 reacts with water, unlike N2O, CO2 gas exchangemight be chemically enhanced due to buffering of diffu-sive concentration change by equilibration reactions withinthe boundary layer. Other than inert gases, CO2 diffuses

25

Outgas.

Ingas.

-1

-0.5

0

0.5

1

1.5

2

2.5

3

-2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Day of the experiment

CO

2 flu

x (µ

mo

l kg-1

d-1

)

Outgas.

Ingas.

-1

-0.5

0

0.5

1

1.5

2

2.5

3

-2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Day of the experiment

CO

2 flu

x (µ

mo

l kg-1

d-1

)

823 Fig. 3 824

Daily CO2 fluxes for all CO2 treatments over time. High CO2 treatments are shown in red, 825

medium in gray and low CO2 are blue. Estimates include chemical enhancement according to 826

Hover and Berkshire (1969). 827

828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851

Fig. 3. Daily CO2 fluxes for all CO2 treatments over time. HighCO2 treatments are shown in red, medium in gray and low CO2 areblue. Estimates include chemical enhancement according to Hoverand Berkshire (1969).

are not necessarily exchanged through the boundary layer,but can also be formed from bicarbonate close to the inter-face. This applies only at low wind speeds and not whenmixing is considerably faster because CO2 hydration kinet-ics are slow. Thus, chemical enhancement is thought to berather insignificant under turbulent conditions relevant foropen-ocean CO2 exchange (e.g. whenk > 5 cm h−1), but ap-plies to the conditions found inside the mesocosms (k ∼ 1.8–2.5 cm h−1) (Wanninkhof and Knox, 1996). Moreover, thestate of the carbonate system determines the extent of chem-ical enhancement, being negligible at pH< 6 and substantialat pH> 8. In the Svalbard ocean acidification experiment, thetreatment pHtot (total scale) ranged from 7.5 to 8.3 (Bellerbyet al., 2012), therefore chemical enhancement created a pHeffect on carbon flows that must be considered. To correctfor this, theoretical parameterisations by Hoover and Berk-shire (1969) were chosen, as currently no empirical parame-terisations exist sufficiently describing the process in naturalseawater (Wanninkhof and Knox, 1996). The enhancementfactor α, the ratio between chemical enhanced flux and notenhanced flux can be calculated using Eq. (7):

α =τ

[(τ − 1) + tanh(Qz)/(Qz)]. (7)

Here, dimensionlessτ = 1+[H+

]2(K∗

1K∗

2 + K∗

1

[H+

])−1,with K∗

1 and K∗

2 being the first and second stoichiometricequilibrium constants for carbonic acid and

[H+

]the pro-

ton concentration.Q =(rτD−1

)0.5in cm−1, whereD is the

diffusion coefficient for CO2 by Jahne et al. (1987) andrdescribes the hydration of CO2 either directly or via true car-bonic acid.r in the unit s−1 can be calculated using Eq. (8):

r = KCO2 + KOH−K∗w

[H+

]−1(8)

with KCO2 being the CO2 hydration rate constant (s−1),KOH− is the CO2 hydroxylation rate constant (L mol−1 s−1)

www.biogeosciences.net/10/1379/2013/ Biogeosciences, 10, 1379–1390, 2013

1384 J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms

from Johnson (1982) andK∗w is the equilibrium constant for

water. The boundary layer thicknessz (cm) can be calculatedfrom determined transfer velocity (z = DkCO2

−1). All con-stants used here can be found in Zeebe and Wolf-Gladrow(2001). Using the Hoover and Berkshire (1969) model, in-put conditions similar to our experimental conditions in Sval-bard (T = 5◦C, S = 35, z = 0.002 cm, pHtot = 8.2) result inenhancement of about 8 % (α = 1.082). For the same condi-tions, but at a temperature of 25◦C, CO2 gas exchange wouldbe enhanced by about 48 % (α = 1.479).

Chemical enhancement factors using more complex mod-els published by Quinn and Otto (1971), Emerson (1975),Smith (1985), and Keller (1994) give very similar resultsto the Hoover and Berkshire (1969) model (Wanninkhofand Knox, 1996). Experimental data from tank experimentsreproduce calculated chemical enhancement relatively well(i.e. Hoover and Berkshire, 1969; Liss, 1973; Wanninkhofand Knox, 1996; Degreif, 2006). The simple pH dependentfit derived from enhancement experiments in natural Balticseawater published by Kuss and Schneider (2004) is not rec-ommended for use, as influences ofT , S andz are not con-sidered.

The relevance of chemical enhancement for open-oceanCO2 exchange is controversial as the calculation ofk fromwind speed over the ocean itself still bears considerable un-certainty. Ask in our experiments is measured directly, com-parability to experimental results is quite straight forward.

Due to low temperatures during the Svalbard experiment,chemical enhancement of∼ 3 to 7 % is very low (Fig. 7).The influence of about three degree warming during the ex-periment in June 2010 is overall larger than the calculateddifference arising from pH treatments (Fig. 4). Wrong pH-dependent chemical enhancement could produce artificialtreatment effects in the carbon budget estimates especiallyin warm water ocean acidification studies. NCP estimateswithin this experiment by Silyakova et al. (2012) and Czernyet al. (2012a) at arctic temperatures are relatively unaffectedby enhancement of this magnitude and possible uncertaintiestherein.

Evidence for a strong increase in chemical enhancementdue to enzymatic catalysis by free carbonic anhydrases assuggested by Berger and Libby (1969) was not found in laterexperiments (Goldman and Dennett, 1983; Williams, 1983),but it might be interesting to reconsider this question in futuremesocosm experiments.

The lack of empirical data coverage on chemical enhance-ment parameterisations in seawater poses the major quantita-tive uncertainty for NCP estimates based on CO2 air–sea gasexchange using the presented method. Especially in setupswere temperatures are high, the proportion of CO2 exchangerelying on theoretical considerations is high compared to thedirectly measured flux.

26

Effectof pH

Effectof T

0

1

2

3

4

5

6

7

8

9

7.4 7.6 7.8 8.0 8.2 8.4 8.6

pH

Che

mic

al e

nhan

cem

ent (%

) (

%)

(%

) (

%)

Effectof pH

Effectof T

0

1

2

3

4

5

6

7

8

9

7.4 7.6 7.8 8.0 8.2 8.4 8.6

pH

Che

mic

al e

nhan

cem

ent (%

) (

%)

(%

) (

%)

852 853 Fig. 4 854

Chemical enhancement of CO2 compared to N2O according to Hover and Berkshire (1969) 855

calculated for measured k, S, T and pHtot during the Svalbard experiment. High CO2 856

treatments are red, medium are gray and low CO2 are blue. The effect of pHtot on chemical 857

enhancement is indicated by the black arrow, while the effect of the ~3°C temperature 858

increase during the experiment is indicated by the red arrow. 859

860

861

862

863

864

865

866

867

868

869

870

871

872

Fig. 4.Chemical enhancement of CO2 compared to N2O accordingto Hover and Berkshire (1969) calculated for measuredk, S, T andpHtot during the Svalbard experiment. High CO2 treatments arered, medium are gray and low CO2 are blue. The effect of pHtoton chemical enhancement is indicated by the black arrow, while theeffect of the∼ 3◦C temperature increase during the experiment isindicated by the red arrow.

3.3 The choice of N2O as a gas exchange tracer and itsbiological stability

The N2O molecule strongly resembles CO2 in most physicalproperties; it has the same mass, nearly the same solubilityand diffusivity. Other than CO2, equilibrium reaction of N2Owith water lies strongly on the side of free N2O so that air–sea gas exchange can be approximated as for inert gases. Inlaboratory experiments N2O is a perfect tracer for CO2 gasexchange. Because of the similarity of both gases a conver-sion of measuredkN2O to kCO2 is small and so are potentialuncertainties. Wall effects relevant in laboratory experimentssuch as permeability or adhesion to plastic walls can be as-sumed to be comparable between similar molecules. In openwaters, background concentrations of N2O are slightly vari-able. Therefore,3He and SF6 were used as gas exchange trac-ers in open-ocean applications as they are highly inert, theirnatural background concentration and detection limit is verylow so that measurement is possible also after considerabledilution. Despite many practical advantages of N2O in theapplication in mesocosms its prominent role as a biologicallyproduced climate relevant trace gas is putting the inertness ofN2O into question.

The natural source of oceanic background N2O concen-trations is biological production. N2O is produced predom-inantly as a side product of nitrification, when ammonia isincompletely oxidised in the course of deep remineralisa-tion at low oxygen concentrations. Yet, most parts of theocean are near equilibrium with the atmosphere (mean globalsaturation 103 %) (see Bange et al., 1996 and referencesherein), whereas significant N2O oversaturation is predom-inantly found in tropical regions rather than in cold and tem-perate waters (Walter et al., 2006). Detectable nitrificationin the euphotic zone was hypothesised to also be a source

Biogeosciences, 10, 1379–1390, 2013 www.biogeosciences.net/10/1379/2013/

J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms 1385

27

0

1

2

3

4

5

6

-5 0 5 10 15 20 25 30

Day of the experiment

T (

°C)

873

Fig. 5 874

Mean water temperatures between 0-12 m measured during the Svalbard experiment. The 875

three examined mesocosms are shown as black symbols, while open circles represent 876

measurements from the surrounding fjord. 877

878

879

880

881

882

883

884

885

886

887

888

889

890

Fig. 5. Mean water temperatures between 0–12 m measured duringthe Svalbard experiment. The three examined mesocosms are shownas black symbols, while open circles represent measurements fromthe surrounding fjord.

of N2O (Dore and Karl, 1996; Santoro et al., 2010), but thiswas not yet directly observed. Physiological results (Goreauet al., 1980; Loescher et al., 2012) suggest possible N2O pro-duction by nitrification in fully oxygenated waters to be verylow. However, even N2O production at relatively high sur-face layer nitrification rates, as found in upwelling regions(Rees et al., 2011), are orders of magnitudes too low to sig-nificantly bias the large fluxes caused by deliberate N2O ad-dition. The only known pathway of biological N2O uptakeas a reactive nitrogen species is by denitrifiers at anoxic con-ditions (<∼ 10 µmol O2 kg−1) (Zumft, 1997; Zamora et al.,2012). Conditions favouring this process are unlikely to formin mesocosms and would result in a loss of the tracer (as N2)

but not into utilisation of N2O as a nitrogen source. Side-effects of N2O addition on biological activity in mesocosmexperiments can therefore not be expected. Remineralisationof detritus at the bottom of the mesocosm could possibly bea source of N2O. Conditions allowing for extensive reminer-alisation of accumulated organics inside pelagic mesocosmsshould thus be avoided. It is further strongly recommended tomeasure background natural N2O concentrations preferablyinside non-enriched experimental units, because N2O is notconsidered as an inert gas.

3.4 Lateral gas exchange by diffusion throughmesocosm wall material

Gas fluxes through the mesocosm walls can be calculated iftemperature dependent permeability coefficients of the foilmaterial and gases of interest are known.

Fluxes through the wall (Fwall) in mol d−1 can be derivedusing the equation:

Fwall =Dt · 1p · A · t

zm. (9)

HereDt is the permeability coefficient at a given temperaturein mol · µm atm−1 m−2 d−1, 1p is the partial pressure differ-ence between inside and outside in atm,A is the submerged

surface of the mesocosm walls in m2, t is the duration in daysandzm is the thickness of the material in µm. Turbulence ofthe media in and outside the mesocosm is not relevant to thisdiffusion gradient as permeation of the materials is generallyorders of magnitude slower than removal and advection ofgases in the media.

Estimates of lateral gas exchange for the Svalbard experi-ment were calculated based on permeability coefficients pub-lished for Desmopan® 385, (Bayer). Desmopan® 385 is theraw material of our bag foil (Walopur®, Epurex Films). Di-rect measurements for Walopur® are not available for CO2and N2O. Permeability for the specific temperatures was ex-trapolated.

For the experiments in the KOSMOS mesocosms, the frac-tion of the measured gas flux due to permeability of thebag material was maximal on the order of 1–2 %. Fluxesare low because of the relatively thick foil (0.5 to 1 mm) atcomparatively low temperatures. In the perspective of CO2gas exchange estimates for carbon mass balance it is gen-erally not interesting whether CO2 exchanges through thewalls or via the water surface. However, differences be-tween N2O and CO2 in the material specific permeabilityof the bags have the potential to cause systematic errors ifexchange is largely through the foil and not with the atmo-sphere. Such differences seemed at first unlikely because ofthe general similarity of N2O and CO2 in diffusivity and sol-ubility, but permeability specifications for Desmopan® 385suggest a considerably higher permeability for N2O (27.1and 51.7 mol· µm atm−1 m−2 d−1 for CO2 and N2O at 25◦C,respectively) (Bayer MaterialScience, TPU TechCentre). Forthe Svalbard experiment bias through lateral gas fluxes werenot corrected, as the overall magnitude of these fluxes wasnegligible. The data basis in terms of permeability measure-ments would not have allowed for an exact correction ofsuch bias. A set of permeability measurements at a relevanttemperature range would improve gas exchange estimatesespecially at temperatures above 10◦C. If thin foil is usedfor mesocosms, a material with good gas barrier propertiesshould be chosen and exact permeabilities should be knownfor the gases of interest.

WhenkN2O is translated into transfer velocities of poorlywater soluble gases, dissolution and adhesion of those gasesin and on the plastic material could cause a lateral sink ofthese substances in addition to the permeability issue.

3.5 Sensibilities of the results towards uncertainties inmeasured variables

Sensitivities of the overall resulting CO2 fluxes to uncer-tainties in the determination of the most important measuredvariables for the presented method were estimated (Table 1).Water temperatures are used on numerous occasions for thecalculation of gas exchange rates, e.g. for the calculation ofCO2 from CT and TAlk, for Schmidt numbers, for solubility

www.biogeosciences.net/10/1379/2013/ Biogeosciences, 10, 1379–1390, 2013

1386 J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms

Table 1. Sensitivities of the overall resulting CO2 fluxes to uncer-tainties in the determination of the most important measured vari-ables of the presented method. The effect of systematic N2O un-derestimation was tested using an alternative fit including only up-per end values from the Svalbard dataset. The influence of errors inCO2 gradient determination is denoted for an intermediate gradientof 400 µatm.

Parameter Uncertainty Uncertaintyin parameter in CO2fluxes

Sea surface temperature±1◦C ±3 %Mesocosm volume ±1 % ±1 %Mesocosm surface ±1 % ±1 %Systematic error in Outliers due 0.26± 0.29 %N2O measurement to losses

during samplingAir–sea CO2 gradient ±2 µmol in CT ±5 %

and TAlk

and chemical enhancement. Errors inkCO2, in response touncertainties in temperature, are mainly caused by the tem-perature dependence of N2O solubility. Errors in resultingCO2 fluxes are largely balanced by errors in CO2 solubilitycalculated using identical temperatures. The remaining sen-sitivity of 3 % for 1◦C appears to be relatively low in respectof usually very precise temperature measurements using cur-rent technology. It has to be kept in mind that gas exchangeis a continuous process; therefore, measurement frequencyshould be adequate for sufficient description of relevant tem-perature changes during the entire duration of the experimen-tal duration. At water temperatures above 10◦C, chemicalenhancement corrections become important so that precisetemperature records gain additional relevance. Uncertaintiesin mesocosm volume or surface area translate directly into er-rors in calculated CO2 fluxes (one to one %) whenkCO2 mea-sured in one mesocosm is applied to calculate gas exchangein a parallel mesocosm with different volume or surface area.A random uncertainty in N2O determination of 1.8 % as de-noted by Walter et al. (2006) would be averaged out dueto the large number of fitted measurements. However, a hy-pothetical systematic underestimation of N2O by includingoutliers possibly caused by N2O losses from oversaturatedsamples would influencekCO2. The effect of systematic N2Ounderestimation was tested using an alternative fit includingonly upper end values. Resulting CO2 fluxes differ by 0.7 %in the beginning of the experiment when N2O concentrationswere high and roughly equal fluxes calculated including allN2O measurements at the end of the experiment. Maximumuncertainties in CO2 fluxes of 5 % due to errors in the deter-mination of air–sea CO2 gradients are calculated on the basisof a maximum uncertainty of±2 µmol kg−1 in CT as wellas TAlk and an intermediate CO2 gradient of∼ 400 µatm.As absolute errors in CO2 determination have to be seen inrelation to the air–sea gradient, percentile errors are small

when gradients are large and vice versa. The effects of ran-dom errors in CO2 determination on uncertainties in cumu-lative CO2 mass flux over time are averaged out over time.A cumulative error of the applied constants cannot be givenbut it has to be highlighted that this is an additional source ofuncertainty. Parameterisations of Schmidt numbers, solubil-ity and rate constants, as well as diffusion coefficients citedin the text, were chosen to the best of our knowledge.

3.6 Processes driving gas exchange in mesocosms

The concentration of N2O (CN2O) decreased quite steadilyover the whole experimental period (Fig. 2). This indicatesthat N2O fluxes were controlled by the diffusion gradient tothe atmosphere. Variable external forcing by wind or wavesas commonly observed in natural environments was of minorimportance. Wind measurements at Bellevaja station at 10 mabove sea level (U10) reported velocities of up to 5 m s−1 dur-ing the experiment (Fig. 6). The water surface of the meso-cosms, however, is sheltered from direct wind sheer by thetwo meter high plastic walls of the bag (Fig. 1; Riebesell etal., 2012).

Fetch, the distance wind could act on the water surface,was dependent on wind direction ( maximum of∼ 10 nmin the Kongsfjorden). Waves that were able to propagatethrough the mesocosms (significant wave height (H1/3) upto ∼ 0.8 m) were only observed on the mooring site on threedays when stronger winds were blowing along the fjord fromsoutheast, the most exposed wind direction. Enhanced gasexchange during the days with waves could not be resolvedby our measurements. However, CO2 gas exchange insidethe mesocosms was measured to be constantly about threetimes higher than calculated flux at zero wind as performedby Delille et al. (2005) (Fig. 7, stagnant film thickness cal-culated according to Smith, 1985, chemical enhancement ac-cording to Hoover and Berkshire, 1969).

Applying a quadratic wind dependent function (Wan-ninkhof, 1992) at constant wind speed of 3.15 m s−1, result-ing fluxes are very close to our empirical estimate over thewhole period. MeasuredU10 wind speeds at the experimen-tal site were generally lower than this (mean 2.1 m s−1), andaccordingly calculated mean air–sea gas exchange was alsolower outside in the fjord than inside the mesocosms. Com-pared to relevant open-ocean gas transfer, estimated meso-cosm CO2 transfer velocities between∼ 1.9 to 2.5 cm h−1 inthe Svalbard experiment are low. They are within a gray zoneof baseline gas exchange where buoyancy fluxes and chem-ical enhancement contribute largely to gas exchange so thatpurely wind depended parameterisations are not applicable(Wanninkhof et al., 2009). Additional factors can be arguedto be driving gas exchange in mesocosms compared to openwaters. Rinsing of the plastic walls when waves are propagat-ing through the setup presumably leads to enhanced air–seasurface renewal compared to open water. Slight temperaturechanges in the surrounding water mass were immediately

Biogeosciences, 10, 1379–1390, 2013 www.biogeosciences.net/10/1379/2013/

J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms 1387

28

0

1

2

3

4

5

6

Win

d sp

eed

(m s-1

)

0

50

100

150

200

250

300

350

5 7 9 11 13 15 17 19 21 23 25 27 29

Day of the experiment

Win

d d

irect

ion

(°)

N

N

E

W

S

891

Fig. 6 892

Wind speed and direction during the course of the experiment, measured at 10 m height at 893

Bayelva station, Ny Alesund. Waves (up to H1/3 = 0.8m) were observed at the mooring site 894

during time intervals indicated by shaded areas, when relatively strong wind was blowing 895

along the fjord from south east. 896

897

898

899

900

901

902

903

904

905

906

Fig. 6. Wind speed and direction during the course of the exper-iment, measured at 10 m height at Bayelva station, Ny Alesund.Waves (up toH1/3 = 0.8 m) were observed at the mooring site dur-ing time intervals indicated by shaded areas, when relatively strongwind was blowing along the fjord from southeast.

heating or cooling the bags (Fig. 5), this probably causedconsiderably enhanced buoyancy fluxes that kept the exper-imental units relatively homogenous throughout the experi-ment. Last but not least, the extensive daily sampling withwater samplers and probes contributed to gas exchange byactive perturbation of the mesocosm surface.

3.7 Mesocosm proportions

Transfer velocities (k) in other mesocosm setups deployed inmore sheltered surroundings, standing on land or inside cli-mate controlled rooms might be lower or higher, dependingon methodology used for sampling, temperature control, ac-tive mixing and gas specific permeability of the mesocosmmaterial. Even more important than these influences onk,is the ratio between the mesocosm volume and its surfacearea (A/V ), when exchange rates are normalised to units ofwater (kg−1 or L−1). In an exemplary 15 m deep KOSMOSmesocosm (Fig. 1), holding∼ 45 m3 of water, CO2 gas ex-change over 3.14 m2 (A/V = 0.07) surface area is causingrelatively moderate changes in aquatic concentrations despitelarge diffusion gradients (Fig. 3). Taking the example of theKiel indoor mesocosm (Fig. 8a) of about 1.4 m3 at 2 m2 sur-face (A/V = 1.4), concentration change in response to thesame gas exchange flux is 20 times faster. Additionally, air–sea gas exchange velocities are accelerated by continuous ac-tive mixing, necessary to keep plankton organisms in suspen-sion (Fig. 8a). While after 20 days∼ 50 % of the N2O addedwas still present during the Svalbard study (Fig. 1), the sametracer concentration was virtually gone after five days in theshallow indoor mesocosm (Fig. 8b) in the uncovered config-uration. Here, inorganic carbon uptake by phytoplankton canbe rapidly compensated by ingassing of CO2 from the atmo-sphere. Ocean acidification experiments in setups withA/V

29

-1

-0.5

0

0.5

1

1.5

2

2.5

5 10 15 20 25 30

Day of the experiment

CO

2 flu

x (µ

mol

* k

g-1 d

-1)

Outgas.

Ingas.

-1

-0.5

0

0.5

1

1.5

2

2.5

5 10 15 20 25 30

Day of the experiment

CO

2 flu

x (µ

mol

* k

g-1 d

-1)

Outgas.

Ingas.

907

Fig. 7 908

Comparison of different approaches to estimate CO2 air-sea gas exchange. Daily ingassing 909

rates in the low CO2 control treatment (average ~180µatm) are shown in blue, while 910

outgassing from the highest CO2 treatment (average ~1085µatm) is red. Filled symbols are 911

estimates from the N2O tracer approach, circles for chemical enhanced flux ( / ) and 912

triangles ( / ) for non chemical enhanced flux. Open squares ( / ) are an estimate using a 913

quadratic wind dependent function according to Wanninkhof (1992) at a constant wind speed 914

of 3.15 m s-1, chosen to match the N2O results. Open circles ( / ) is a chemically enhanced 915

zero wind speed output, according to Smith (1985), while open triangles ( / ) are the same 916

estimate without chemical enhancement. 917

918

919

920

921

922

923

924

Fig. 7. Comparison of different approaches to estimate CO2 air–sea gas exchange. Daily ingassing rates in the low CO2 controltreatment (average∼ 180 µatm) are shown in blue, while outgassingfrom the highest CO2 treatment (average∼ 1085 µatm) is red. Filledsymbols are estimates from the N2O tracer approach, circles (•)for chemical enhanced flux and triangles (N) for non chemical en-hanced flux. Open squares (�) are an estimate using a quadraticwind dependent function according to Wanninkhof (1992) at a con-stant wind speed of 3.15 m s−1, chosen to match the N2O results.Open circles (◦) are a chemically enhanced zero wind speed output,according to Smith (1985), while open triangles (4) are the sameestimate without chemical enhancement.

similar to the Kiel indoor mesocosm would lose their treat-ment CO2 within a few days. To maintain treatment levels insuch shallow experiments, continuous measurement and con-trol technology can be used (see e.g. Widdicombe and Need-ham, 2007) . Resulting controlled treatment levels are ben-eficial when physiological questions are investigated. How-ever, CO2 drawdown does not occur, and therefore DIC con-centration change cannot be used to calculate NCP. Anotheroption is to artificially decrease the surface area by cover-ing the mesocosm with a low permeability transparent film.For comparison 50 nmol kg−1 N2O was added to two Kielindoor mesocosms, one in an uncovered configuration andone covered with a transparent floating foil to reduce surfacein contact with the atmosphere (Fig. 8a). Both mesocosmswere stirred at the same speed; samples were drawn usinga tube. The thin polyurethane foil mounted on a light frameand floating on the surface, efficiently minimised air–sea gasexchange (Fig. 8b). If covers are used, reducing the surfacearea to a minimum, it has to be considered that the remainingopen surface should be equally large. As the working prin-ciple of this approach is to minimise surface area, it can beassumed to be very sensible to the size of the remaining inter-face (leaks). Therefore, air–sea gas exchange should be mea-sured in all experimental units to check for reproducibility ofresults.

www.biogeosciences.net/10/1379/2013/ Biogeosciences, 10, 1379–1390, 2013

1388 J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms

30

0

10

20

30

40

50

60

0 2 4 6 8 10 12 14 16

Day of the experiment

N2O

(nm

ol k

g-1)

a b

0

10

20

30

40

50

60

0 2 4 6 8 10 12 14 16

Day of the experiment

N2O

(nm

ol k

g-1)

a b

925

Fig. 8 926

a) Schematic drawing of the Kiel indoor mesocosm. With stirrer, floating lid and sampling 927

hose, below: vertical view on floating foil lid with enforcement frame. 928

b) Comparison of N2O tracer outgassing in a Kiel indoor mesocosm, between a simple 929

uncovered setup (black) and a setup using a floating foil lid reducing the water surface area 930

available for air-sea gas exchange (blue). 931

932

933

934

935

936

937

938

939

940

941

942

943

944

945

Fig. 8. (a) Schematic drawing of the Kiel indoor mesocosm. Withstirrer, floating lid and sampling hose, below: vertical view on float-ing foil lid with enforcement frame.(b) Comparison of N2O traceroutgassing in a Kiel indoor mesocosm, between a simple uncoveredsetup (black) and a setup using a floating foil lid reducing the watersurface area available for air–sea gas exchange (blue).

4 Conclusion and outlook

Direct measurement of N2O air–sea gas exchange can beused to estimate accurate CO2 fluxes in various mesocosmsetups, whereas common wind dependent parameterisationsfor air–sea gas exchange cannot be adapted to mesocosmconditions and their application is therefore prone to system-atic errors. Within the mesocosm low energy physical sur-rounding, N2O gas exchange measurements allow for directestimation of CO2 fluxes with uncertainties far below open-ocean dual tracer measurements due to a known, constantwater volume. Measured transfer velocities are within therange of zero wind open-ocean baseline velocities assumedto be dominated by chemical enhancement and buoyancyfluxes (Wanninkhof et al., 2009). The influence of sea surfacemicrolayers of surface active organic molecules is discussedto be responsible for large discrepancies in gas exchange be-tween productive coastal waters and open-ocean conditions(Frew, 1997; Kock et al., 2012). The decrease in open-oceank by 20–50 %, due to surfactants smoothening wind effectson the water surface (Tsai and Liu, 2003), can be expectedto be of minor relevance to mesocosm gas exchange wherewind stress is not the dominating energy input. The effectof these surfactants, possibly produced in high amounts dur-ing phytoplankton blooms in mesocosms, is difficult to in-clude in theoretical calculations, but is inherently includedin our direct measurements. Future mesocosm experimentscombining the close observation of biological, chemical andphysical processes might offer the opportunity to bring morelight into origin and composition of organic surface micro-layers.

The application of gas exchange measurements for calcu-lation of NCP inside the mesocosm delivered satisfying re-sults. Of the four community production estimates publishedfor the Svalbard 2010 experiment, NCP calculated fromchanges in dissolved inorganic carbon corrected for air–seagas exchange (Czerny et al., 2012a; Silyakova et al., 2012)

seems to be quantitatively most plausible. Although over-all quantity compares relatively well with results from oxy-gen and in situ13C- primary production estimates (Tanakaet al., 2013; de Kluijver et al., 2013), comparability to14Cincubation data presented by Engel et al. (2012) is weak.Much higher14C fixation rates can be plausibly explainedby the shallow (∼ 1 m) incubation depth, while oxygen incu-bations obviously experienced more intermediate light andtemperature conditions at∼ 4 m depth.13C tracer incorpora-tion measurements inside the mesocosm deliver results rep-resentative for the entire mesocosm but only for a period be-fore organic matter approached saturation with the tracer (deKluijver et al., 2013). Rates measured in side experimentsare more useful to compare community production betweenthe treatments rather than giving quantitative cumulative es-timates for in situ carbon uptake (see discussion in Engel etal., 2012). As incubations were performed only at one depth,it is impossible to integrate these data over time and depth inrespect of variable light and temperature gradients. Summingup incubation results to achieve cumulative estimates couldlead to further error propagation, whereas NCP calculatedfrom in situ inorganic carbon measurements is per se cumu-lative and propagation of single measurement errors cannotoccur.

Further development of the N2O tracer concept is focussedon using it not only to determine air–sea gas exchange instratified mesocosms, but also to estimate diapycnal mixingbetween surface layer and deep water inside mesocosms. Forthis purpose, N2O gradients developing over time will be cor-related to high resolution profiles of oxygen, pH and salinity,measured with CTD sensors. Especially in temperate turbidwaters, mesocosm NCP is mostly not restricted by mesocosmlength but by light penetration. The photoautotrophic surfacelayer communicates to some extent with a more nutrient richdeep layer where heterotrophic processes dominate. Budget-ing these more naturally structured mesocosms is not only aninteresting challenge, but will also introduce new ecologicalaspects connected to upward and downward elemental fluxesinto the biogeochemical interpretation of the mesocosm sys-tem.

Acknowledgements.This work is a contribution to the “EuropeanProject on Ocean Acidification” (EPOCA) which received fundingfrom the European Community’s Seventh Framework Programme(FP7/2007-2013) under grant agreement no. 211384. Financialsupport was provided through Transnational Access funds by theEuropean Union Seventh Framework Program (FP7/2007-2013)under grant agreement no. 22822 MESOAQUA and by the FederalMinistry of Education and Research (BMBF, FKZ 03F0608)through the BIOACID (Biological Impacts of Ocean ACIDifica-tion) project. We acknowledge Annette Kock and Hermann Bangefor support in measuring N2O as well as Allanah Paul andScarlett Sett for providing measurements from the Kiel indoormesocosms. We thank Thomas Conway from the NOAA CarbonCycles Gases Group in Boulder, US, and Johan Strom from ITM,

Biogeosciences, 10, 1379–1390, 2013 www.biogeosciences.net/10/1379/2013/

J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms 1389

Stockholm University, Sweden, for providing us with atmosphericN2O and CO2 measurements, respectively, from close by ZeppelinMountain. We gratefully acknowledge the logistical support ofGreenpeace International for its assistance with the transport ofthe mesocosm facility from Kiel to Ny-Alesund and back to Kiel.We also thank the captains and crews of M/VESPERANZAofGreenpeace and R/VViking Explorerof the University Centre inSvalbard (UNIS) for assistance during mesocosm transport andduring deployment and recovery in Kongsfjorden. We thank thestaff of the French-German Arctic Research Base at Ny-Alesund,in particular Marcus Schumacher, for on-site logistical support.

The service charges for this open access publicationhave been covered by a Research Centre of theHelmholtz Association.

Edited by: J. Middelburg

References

Archer, S. D., Kimmance, S. A., Stephens, J. A., Hopkins, F. E.,Bellerby, R. G. J., Schulz, K. G., Piontek, J., and Engel, A.:Contrasting responses of DMS and DMSP to ocean acidifica-tion in Arctic waters, Biogeosciences Discuss., 9, 12803–12843,doi:10.5194/bgd-9-12803-2012, 2012.

Bange, H. W., Rapsomanikis, S., and Andreae, M. O.: Nitrous ox-ide in coastal waters, Global Biogeochem. Cy., 10, 197–207,doi:10.1029/95gb03834, 1996.

Bellerby, R. G. J., Silyakova, A., Nondal, G., Slagstad, D., Czerny,J., de Lange, T., and Ludwig, A.: Marine carbonate system evo-lution during the EPOCA Arctic pelagic ecosystem experimentin the context of simulated Arctic ocean acidification, Biogeo-sciences Discuss., 9, 15541–15565,doi:10.5194/bgd-9-15541-2012, 2012.

Berger, R. and Libby, W. F.: Equilibration of Atmospheric CarbonDioxide with Sea Water: Possible Enzymatic Control of the Rate,Science, 164, 1395–1397,doi:10.1126/science.164.3886.1395,1969.

Broecker, W. S. and Peng, T.-H.: Gas exchange rates between airand sea, Tellus, 26, 21–35, 1974.

Czerny, J., Schulz, K. G., Boxhammer, T., Bellerby, R. G. J.,Budenbender, J., Engel, A., Krug, S. A., Ludwig, A., Nachtigall,K., Nondal, G., Niehoff, B., Siljakova, A., and Riebesell, U.: El-ement budgets in an Arctic mesocosm CO2 perturbation study,Biogeosciences Discuss., 9, 11885–11924,doi:10.5194/bgd-9-11885-2012, 2012a.

Czerny, J., Schulz, K. G., Krug, S. A., Ludwig, A., and Riebesell,U.: Technical Note: On the determination of enclosed water vol-ume in large flexible-wall mesocosms, Biogeosciences Discuss.,9, 13019-13030,doi:10.5194/bgd-9-13019-2012, 2012b.

de Kluijver, A., Soetaert, K., Czerny, J., Schulz, K. G., Boxham-mer, T., Riebesell, U., and Middelburg, J. J.: A13C labellingstudy on carbon fluxes in Arctic plankton communities under el-evated CO2 levels, Biogeosciences, in press, 2013.

Degreif, K. A.: Untersuchungen zum Gasaustausch: Entwicklungund Applikation eines zeitlich aufgelosten Massenbilanzver-fahrens, Ph.D., Naturwissenschaftlich-Mathematische Gesamt-fakultat, Ruprecht-Karls-Universitat, Heidelberg, 183 pp., 2006.

Delille, B., Harlay, J., Zondervan, I., Jacquet, S., Chou, L., Wollast,R., Bellerby, R. G. J., Frankignoulle, M., Borges, A. V., Riebe-sell, U., and Gattuso, J.-P.: Response of primary production andcalcification to changes ofpCO2 during experimental bloomsof the coccolithophorid Emiliania huxleyi, Global Biogeochem.Cy., 19, GB2023,doi:10.1029/2004gb002318, 2005.

Dickson, A. G.: An exact definition of total alkalinity and a proce-dure for the estimation of alkalinity and total inorganic carbonfrom titration data., Deep-Sea Res., 28, 609–623, 1981.

Dore, J. E. and Karl, D. M.: Nitrification in the euphotic zone as asource for nitrite, nitrate, and nitrous oxide at Station ALOHA,Limnol. Oceanogr., 41, 1619–1628, 1996.

Emerson, S.: Chemically enhanced CO2 gas exchange in a eu-trophic lake: a general model, Limnol. Oceanogr., 20, 743–753,1975.

Engel, A., Borchard, C., Piontek, J., Schulz, K., Riebesell, U., andBellerby, R.: CO2 increases14C-primary production in an Arcticplankton community, Biogeosciences Discuss., 9, 10285–10330,doi:10.5194/bgd-9-10285-2012, 2012.

Frew, N. M.: The role of organic films in air-sea gas exchange, in:The Sea Surface And Global Change, edited by: Liss, P. S. andDuce, R. A., Cambridge University Press, Cambridge, 121–171,1997.

Goldman, J. C. and Dennett, M. R.: Carbon Dioxide Exchange Be-tween Air and Seawater: No Evidence for Rate Catalysis, Sci-ence, 220, 199–201,doi:10.1126/science.220.4593.199, 1983.

Goreau, T. J., Kaplan, W. A., Wofsy, S. C., McElroy, M. B., Val-ois, F. W., and Watson, S. W.: Production of NO2- and N2O byNitrifying Bacteria at Reduced Concentrations of Oxygen, Appl.Environ. Microb., 40, 526–532, 1980.

Hoover, T. E. and Berkshire, D. C.: Effects of Hydration on CarbonDioxide Exchange across an Air-Water Interface, J. Geophys.Res., 74, 456–464,doi:10.1029/JB074i002p00456, 1969.

Hopkins, F. E., Kimmance, S. A., Stephens, J. A., Bellerby, R. G.J., Brussaard, C. P. D., Czerny, J., Schulz, K. G., and Archer, S.D.: Response of halocarbons to ocean acidification in the Arc-tic, Biogeosciences Discuss., 9, 8199–8239,doi:10.5194/bgd-9-8199-2012, 2012.

Jahne, B., Heinz, G., and Dietrich, W.: Measurement of the Dif-fusion Coefficients of Sparingly Soluble Gases in Water, J. Geo-phys. Res., 92, 10767–10776, 10.1029/JC092iC10p10767, 1987.

Johnson, K. S.: Carbon dioxide hydration and dehydration kineticsin seawater, Limonol. Oceanogr., 27, 849–855, 1982.

Keller, K.: Chemical enhancement of CO2 transfer across the Airsea interface, M.S., Mass. Inst. Technol., Boston, 120 pp., 1994.

Kock, A., Schafstall, J., Dengler, M., Brandt, P., and Bange, H.W.: Sea-to-air and diapycnal nitrous oxide fluxes in the east-ern tropical North Atlantic Ocean, Biogeosciences, 9, 957–964,doi:10.5194/bg-9-957-2012, 2012.

Kuss, J., and Schneider, B.: Chemical enhancement of the CO2 gasexchange at a smooth seawater surface, Mar. Chem., 91, 165–174, 2004.

Lewis, E. L. and Wallace, D. W. R.: Basic programs for the CO2system in seawater, Brookhaven National Laboratory InformalReport, BNL# 61827, 1995.

Liss, P. S.: Processes of gas exchange across an air-water interface,Deep Sea Res., 20, 221–238, 1973.

Loscher, C. R., Kock, A., Konneke, M., LaRoche, J., Bange, H.W., and Schmitz, R. A.: Production of oceanic nitrous oxide

www.biogeosciences.net/10/1379/2013/ Biogeosciences, 10, 1379–1390, 2013

1390 J. Czerny et al.: A simple method for air–sea gas exchange measurements in mesocosms

by ammonia-oxidizing archaea, Biogeosciences, 9, 2419–2429,doi:10.5194/bg-9-2419-2012, 2012.

Petersen, J. E., Kemp, W. M., Bartleson, R., Boynton, W. R., Chen,C.-C., Cornwell, J. C., Gardner, R. H., Hinkle, D. C., Houde,E. D., Malone, T. C., Mowitt, W. P., Murray, L., Sanford, L. P.,Stevenson, J. C., Sundberg, K. L., and Suttles, S. E.: Multiscaleexperiments in coastal ecology: Improving realism and advanc-ing theory, BioScience, 53, 1181–1197, 2003.

Quinn, J. A. and Otto, N. C.: Carbon Dioxide Exchange at theAir-Sea Interface: Flux Augmentation by Chemical Reaction, J.Geophys. Res., 76, 1539–1549,doi:10.1029/JC076i006p01539,1971.

Rees, A. P., Brown, I. J., Clark, D. R., and Torres, R.: The La-grangian progression of nitrous oxide within filaments formedin the Mauritanian upwelling, Geophys. Res. Lett., 38, L21606,doi:10.1029/2011gl049322, 2011.

Rhee, T. S.: The process of air water gas exchange and its appli-cation, Ph.D. thesis, College Station, A&M University, Texas,2000.

Riebesell, U., Czerny, J., von Brockel, K., Boxhammer, T.,Budenbender, J., Deckelnick, M., Fischer, M., Hoffmann, D.,Krug, S. A., Lentz, U., Ludwig, A., Muche, R., and Schulz, K.G.: Technical Note: A mobile sea-going mesocosm system – newopportunities for ocean change research, Biogeosciences Dis-cuss., 9, 12985–13017,doi:10.5194/bgd-9-12985-2012, 2012.

Santoro, A. E., Casciotti, K. L., and Francis, C. A.: Activity, abun-dance and diversity of nitrifying archaea and bacteria in thecentral California Current, Environ. Microbiol., 12, 1989–2006,doi:10.1111/j.1462-2920.2010.02205.x, 2010.

Schulz, K. G., Bellerby, R. G. J., Brussaard, C. P. D., Budenbender,J., Czerny, J., Engel, A., Fischer, M., Koch-Klavsen, S., Krug,S. A., Lischka, S., Ludwig, A., Meyerhofer, M., Nondal, G.,Silyakova, A., Stuhr, A., and Riebesell, U.: Temporal biomassdynamics of an Arctic plankton bloom in response to increasinglevels of atmospheric carbon dioxide, Biogeosciences, 10, 161–180,doi:10.5194/bg-10-161-2013, 2013.

Silyakova, A., Bellerby, R. G. J., Czerny, J., Schulz, K. G., Non-dal, G., Tanaka, T., Engel, A., De Lange, T., and Riebesell, U.:Net community production and stoichiometry of nutrient con-sumption in a pelagic ecosystem of a northern high latitude fjord:mesocosm CO2 perturbation study, Biogeosciences Discuss., 9,11705–11737,doi:10.5194/bgd-9-11705-2012, 2012.

Sinha, V., Williams, J., Meyerhofer, M., Riebesell, U., Paulino, A.I., and Larsen, A.: Air-sea fluxes of methanol, acetone, acetalde-hyde, isoprene and DMS from a Norwegian fjord following aphytoplankton bloom in a mesocosm experiment, Atmos. Chem.Phys., 7, 739–755,doi:10.5194/acp-7-739-2007, 2007.

Smith, S. V.: Physical, chemical and biological characteristics ofCO2 gas flux across the air-water interface, Plant Cell Environ.,8, 387–398,doi:10.1111/j.1365-3040.1985.tb01674.x, 1985.

Tanaka, T., Alliouane, S., Bellerby, R. G. B., Czerny, J., de Klui-jver, A., Riebesell, U., Schulz, K. G., Silyakova, A., and Gattuso,J.-P.: Effect of increasedpCO2 on the planktonic metabolic bal-ance during a mesocosm experiment in an Arctic fjord, Biogeo-sciences, 10, 315–325,doi:10.5194/bg-10-315-2013, 2013.

Taucher, J., Schulz, K. G., Dittmar, T., Sommer, U., Oschlies,A., and Riebesell, U.: Enhanced carbon overconsumption in re-sponse to increasing temperatures during a mesocosm exper-iment, Biogeosciences, 9, 3531–3545,doi:10.5194/bg-9-3531-2012, 2012.

Tsai, W.-T. and Liu, K.-K.: An assessment of the effect of sea sur-face surfactant on global atmosphere-ocean CO2 flux, J. Geo-phys. Res., 108, 24.21–24.16, 2003.

Walter, S., Bange, H. W., Breitenbach, U., and Wallace, D. W. R.:Nitrous oxide in the North Atlantic Ocean, Biogeosciences, 3,607–619,doi:10.5194/bg-3-607-2006, 2006.

Wanninkhof, R.: Relationship between wind speed and gas ex-change over the ocean, J. Geophys. Res., 97, 7373–7382,doi:10.1029/92jc00188, 1992.

Wanninkhof, R. and Knox, M.: Chemical enhancement of CO2exchange in natural waters, Limonol. Oceanogr., 41, 689–697,1996.

Wanninkhof, R., Asher, W. E., Ho, D. T., Sweeney, C., andMcGillis, W. R.: Advances in quantifying air-sea gas exchangeand environmental forcing, Annu. Rev. Mar. Sci., 1, 213–244,2009

Weiss, R. F. and Price, B. A.: Nitrous oxide solubility in water andseawater, Mar. Chem., 8, 347–359, 1980.

Widdicombe, S. and Needham, H. R.: Impact of CO2-induced sea-water acidification on the burrowing activity of Nereis virensand sediment nutrient flux, Mar. Ecol.-Prog. Ser., 341, 111–122,2007.

Wilke, C. R. and Chang, P.: Correlation of diffusion co-efficients in dilute solutions, AICHE J., 1, 264–270,doi:10.1002/aic.690010222, 1955.

Williams, G. R.: The Rate of Hydration of Carbon Dioxide in Nat-ural Waters, Environ. Biogeochem., 35, 281–289, 1983.

Wohlers, J., Engel, A., Zollner, E., Breithaupt, P., Jurgens, K.,Hoppe, H.-G., Sommer, U., and Riebesell, U.: Changes in bio-genic carbon flow in response to sea surface warming, P. Natl.Acad. Sci., 106, 7067–7072,doi:10.1073/pnas.0812743106,2009.

Zamora, L. M., Oschlies, A., Bange, H. W., Huebert, K. B.,Craig, J. D., Kock, A., and Loscher, C. R.: Nitrous ox-ide dynamics in low oxygen regions of the Pacific: insightsfrom the MEMENTO database, Biogeosciences, 9, 5007–5022,doi:10.5194/bg-9-5007-2012, 2012.

Zeebe, R. E. and Wolf-Gladrow, D.: CO2 in seawater: equilib-rium, kinetics, isotopes, Elsevier Oceanography Series, editedby: Halpen, D., Elsevier, Amsterdam, 2001.

Zumft, W. G.: Cell biology and molecular basis of denitrification,Microbiol. Mol. Biol. Rev., 61, 533–616, 1997.

Biogeosciences, 10, 1379–1390, 2013 www.biogeosciences.net/10/1379/2013/