Extreme Variations of pC02 and pH in a Macrophyte Meadow of … · In situ pC02, pH (total scale), ... [11-13]) and (2) reefs formed by corals or calcifying algae because of the direct

OPEN 3 ACCESS Freely available online © PLOSI o -

Extreme Variations of pC02 and pH in a Macrophyte Meadow of the Baltic Sea in Summer: Evidence of the Effect of Photosynthesis and Local UpwellingVincent Saderne1'2*, Peer Fietzek3, Peter Maria Jozef Herman21 Benthic Ecology group, GEOMAR: Helmholtz Center for Ocean Research in Kiel, Kiel, Schleswig-Holstein, Germany, 2 Spatial Ecology, NIOZ: Royal Netherlands Institute for Sea Research, Yerseke, Zeeland, The Netherlands, 3 CONTROS Systems & Solutions GmbH, Kiel, Schleswig-Holstein, Germany

AbstractThe impact o f ocean acidification on benthic habitats is a major preoccupation o f the scientific community. However, the natural variability o f pC02 and pH in those habitats remains understudied, especially in temperate areas. In this study we investigated temporal variations o f the carbonate system in nearshore macrophyte meadows o f the western Baltic Sea. These are key benthic ecosystems, providing spawning and nursery areas as well as food to numerous commercially im portant species. In situ pC02, pH (total scale), salinity and PAR irradiance were measured w ith a continuous recording sensor package dropped in a shallow macrophyte meadow (Eckernförde bay, western Baltic Sea) during three different weeks in July (pC02 and PAR only), August and September 2011.The mean (± SD) pC02 in July was 383±117 patm. The mean (± SD) pC02 and pHtot in August were 239±20 patm and 8.22±0.1, respectively. The mean (± SD) pC02 and pHtot in September were 1082±711 patm and 7.83±0.40, respectively. Daily variations o f pC02 due to photosynthesis and respiration (difference between daily maximum and m inimum) were o f the same order o f magnitude: 281 ±88 patm, 219±89 patm and 1488±574 patm in July, August and September respectively. The observed variations o f pC02 were explained through a statistical model considering w ind direction and speed together w ith PAR irradiance. At a tim e scale o f days to weeks, local upwelling o f elevated pC02 water masses w ith offshore winds drives the variation. W ithin days, primary production is responsible. The results demonstrate the high variability o f the carbonate system in nearshore macrophyte meadows depending on meteorology and biological activities. We h igh light the need to incorporate these variations in future pC02 scenarios and experimental designs for nearshore habitats.

C i t a t i o n : Saderne V, Fietzek P, Herman PMJ (2013) Extreme Variations of pC02 and pH in a Macrophyte Meadow of the Baltic Sea in Summer: Evidence of the Effect of Photosynthesis and Local Upwelling. PLoS ONE 8(4): e62689. doi:10.1371/journal.pone.0062689

E d i to r : Wei-Chun Chin, University of California, Merced, United States of America

R e c e iv e d October 9, 2012; A c c e p t e d March 25, 2013; P u b l i s h e d April 23, 2013

C o p y r i g h t : © 2013 Saderne et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

F u n d i n g : This work was funded by the European community, Marie Curie ITN CALMARO (PITN-GA-2008-215157). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

C o m p e t i n g I n t e r e s t s : One of the authors is employed by a commercial company (CONTROS SYSTEMS AND SOLOUTIONS GmbH). This does not alter theauthors' adherence to all the PLOS ONE policies on sharing data and materials.

* E-mail: [email protected]

IntroductionH u m an activities since the 19th century led to an increase o f

atm ospheric p C 0 2 from 280 to 392 pa tm [1] and the trend is rising. Som e scenarios expect an elevation o f atm ospheric p C 0 2 up to 1000 patm during the 21th century, peaking around 1400 patm in the year 2300 [2,3]. As oceans equilibrate with the atm osphere, dissolution o f C 0 2 in w ater induces a decrease in pH . G lobal change has already led to an average seawater “ acidification” o f 0. 1 p H units in the w orld ocean [4]. Acidification is enhancing the corrosiveness o f seawater to calcite and aragonite, the two isoforms o f calcium carbonates com posing the shells and skeletons o f m arine organisms. Corrosiveness is expressed by the saturation states f í arat! an d H caic. A saturation state below 1 indicates a tendency towards dissolution o f the crystal. A ragonitic structures (e.g. scleratinian corals, nacre o f bivalve shells) are m ore soluble th an calcific (e.g. outer shell o f oyster) [4-5].

O nce dissolved, C 0 2 becom es pa rt o f the carbonate system, alm ost entirely com posed o f b icarbonate ( H C 0 3_ ) an d carbonate ions ( C 0 3~ ). U nder actual atm ospheric concentrations o f C 0 2,

C 0 2(aq) in the surface ocean represents less th an 1% of the dissolved inorganic carbon (DIC) while H C 0 3_ represents ~ 9 0 % and C 0 3~ ~ 1 0 % . T h e carbonate system in open oceanicenvironm ents is well know n and m ost o f the data form ing the basis for the predictive m odels (see G L O D A P database, [5,6]) are derived from the open ocean. T hose oceanic actual and future p C 0 2/p H values are the ones referred to w hen designing ocean acidification studies [7]. How ever, the biogeochem istry o f nearshore ecosystems features m ore variations an d differs widely th an offshore conditions [8,9], Shallow nearshore and estilarm e areas are the hab ita t o f num erous benthic calcifiers. As highlighted by Andersson and M ackenzie (2012) [10], investigations on the effects o f ocean acidification on calcifiers are neglecting this na tu ra l variability.

In nearshore habitats, the few available investigations were conducted on: (1) estuaries, salt m arshes, m angroves an d mudflats w here transfers o f carbon from land are occurring (e.g. [11-13]) and (2) reefs form ed by corals or calcifying algae because o f the direct effect o f calcification on the carbonate system (e.g. [14,15]). T h e carbonate chem istry o f nearshore habitats dom inated by m acrophytes (kelp forests and seagrass a n d /o r seaweed meadows)

PLOS ONE I www.plosone.org 1 April 2013 I Volume 8 | Issue 4 | e62689

mailto:[email protected]://www.plosone.org

C arbonate Chemistry in M acrophyte M eadows

is alm ost unknow n, even though they represent ~ 5 % o f the prim ary production in the ocean and % o f the vegetal biomass [16,17]. In these habitats, the carbonate system is driven by the photosynthetic physiology o f the m acrophytes, taking up carbon during the day and releasing it by respiration during the night [18,19]. Since C 0 2 is a m inor com ponent o f D IC an d its diffusion ra te in seawater is 10,000 times lower than in air [20], m ost m acrophytes are relying on H C 0 3- for photosynthesis [21-24], M acrophyte m eadows are highly productive and n e t autotrophic (carbon sink) [25,26] exporting their excess p roduction to the neighboring ecosystems under the form o f litter o r dissolved organic carbon [27]. M icrobial degradation o f exported organic carbon leads to elevated p C 0 2 an d hypoxia o f deeper offshore waters [28]. This phenom enon is particularly im portan t in eutroph ied ecosystems like the Baltic Sea [29]. In the inner bays o f the w estern Baltic, local upwelling o f hypercapnie w ater masses is regularly observed, increasing the surface p C 0 2 up to 2500 patm . This typically happens a t the end o f sum m er when offshore winds are upwelling deep waters after long periods o f stratification [30-32].

T h e aim of this study, conducted in the w estern Baltic Sea over three weeks o f sum m er 2011, was to m easure (1) the d a y /n ig h t variations o f p C 0 2 due to photosynthesis and respiration o f a m acrophyte bed, (2) the evolution o f the baseline p C 0 2 over sum m er and (3) the effect o f local upwellings on the carbonate system and its diel variations. T h e variations o f D IC , m easured in the m acrophyte m eadow , were m odeled with a statistical m odel considering w ind direction and speed together with photosynthet- ically active radiation (PAR).

Materials and Methods

2.1 Study sitePhysico-chem ical param eters o f seawater were recorded in a

m acrophyte m eadow (3 m depth) in Eckernförde Bay (western Baltic Sea, G erm any, 54°27 ' N, 9°54 ' E, see Fig. 1), during 3 weeks o f sum m er 2011: July: 2 9 .0 6 -0 8 .0 7 , August: 29 .07-05.08 and Septem ber: 09 .09-16.09. In Ju ly , only p C 0 2 and PA R irradiance were recorded by two independent sensors. For the August an d Septem ber deploym ents, a m ultim eter recording pH , salinity and tem perature was added to the C 0 2 and PA R sensor. W ind speed (m s ’) an d direction (rad) with a resolution o f 10 m in were provided by the m eteorological station o f A schau in Eckernförde Bay (54°27'40 N, 9°55 '30 E) belonging to the division o f m arine technology and research o f the G erm an Navy. T h e nearshore h ab ita t o f Eckernförde is a m ixed bo ttom type dom inated by m acrophyte vegetation. T h e m acrophyte vegetation covers approxim ately 75% o f the surface. D om inan t species are the brow n algae Fucus serratus (40-60% o f the m acrophytes), growing on stones and rocks an d the seagrass postera marina ( 0 .9 9 . T h e resulting equation was used to correct the sam pled tem perature. T he sam pled voltage an d corrected tem perature were converted to p H tot by m aking use o f the initial T R IS buffer calibration o f the electrode an d by using the equations given in the S O P 6a. In the lab, work conducted on Certified Reference M aterial (CRM) (Andrew Dickson, Scripps Institution o f O ceanography) together with 35 psu buffers dem onstra ted an accuracy o f 0.003 to 0.005 p H units and a precision be tte r than 0.001 p H unit. H ow ever, this accuracy was no t reached in the field and therefore the p H series were no t used to derive the carbonate system b u t instead the alkalinity from discrete samples (see 2.3 for m ethod and 4.4 for discussion on the method).

Salinity was m easured w ithin 0.01 psu by a M ettler Toledo Inlab 738 conductivity probe after calibration a t 25°C w ith KC1 0.1 m ol L -1 (Fischer Scientific, USA). T h e PA R irradiance (400 to


http://www.plosone.org


E9°50‘ 52' 54' 56'

N54°29'

28'

leteo station AschauE ckern fö rde

Sampling site27'

Baltic ^ » _Sea %

Schnellm arkwood

Figure 1. Study site . Map of the inner Eckernförde Bay indicating th e locations of the survey site and th e m eteorological sta tion, urban area (brown) and forests area (green). doi:10.1371/journal.pone.0062689.g001

700 nm) was recorded every 5 m in using a L I-192 q uan tum sensor connected to a L I-1400 da ta logger (Li-Cor Biosciences, LTSA). All series were extended to one m inute interval series by linear interpolation.

2.3 Calculation o f the carbonate systemSam ples for alkalinity were taken a t the beginning, m iddle and

end o f the m easurem ent periods o f August (the 29.07, 02.08 and 05.08, no replication) and Septem ber (the 09.09, 13.03 and 16.09, one replicate). Alkalinity was m easured w ith an accuracy of ± 5 pinoi kg 1 using an open cell potentiom etric titra to r as described in the S O P 3b of [38] (T itrando 888, M ethrom ,Switzerland). A regression was calculated betw een the total alkalinity an d salinity o f the samples from August and Septem ber pooled together. T h e regression was highly significant (F-statistic: pCO.001, R 2 = 0 .8 3 , n = 9): Alkalinity = 9 05 .17 * Salinity +59.83, with bo th slope and intercept significant at pCO.OOl and p < 0 .0 1 (t-test, n = 9). T herefore, the alkalinity for the entire period was estim ated from the salinity series. T h e D IC , f íarat! and fecale w ere derived from calculated alkalinity and the m easured pC O o according to the equations described in [39] w ith the R package Seacarb [40] using first an d second carbonate system dissociation constants for estilarm e systems from [41] and the dissociations constants o f H F and H S 0 4 - o f [42,43].

2.4 Statistical modelingT h e calculated D IC series o f August and Septem ber was

explained th rough a statistical m odel. It considered that (1) the weekly trend of the D IC series is caused by the upwelling o f DICI- rich bo ttom water, (2) the diel variation o f the D IC series is caused by p rim ary p roduction an d respiration of the m eadow. T h e weekly trends o f D IC (Cw) were separated from the diel oscillations (Cid) by a central running average with a tim e fram e o f 24 h. LTpwelling was assum ed to be driven by w ind. Cw was fit by a function o f

PLOS ONE I w ww.plosone.org 3 April 2013 | Volume 8 | Issue 4 | e62689

w ind speed w eighted by w ind direction (WWt).

W W j = w ind speed \ms ~ 1 ] • eos ( w ind direction [rod] + W up ),

W ith W up being a p a ram ete r betw een 0 and 2tt, corresponding to the w ind direction for w hich w ind-induced upwelling is m axim al. As exam ples, if W up = 0 or 2tt, m axim al weights are given to no rth ern winds, if W up = Jt m axim al weights are given to Southern winds. A “left-sided” runn ing integration was perform ed on the weighted w ind tim e series over a period kw:

Cm y = /i,,, + ctw Jrr= hv W W {x)dx,

W here Cwt is the 24 h running-average series o f D IC a t tim e t, W W , is the weighted w ind speed a t tim e t, p w an d are regression param eters and kw is determ ined as the integration period yielding the best fit. Param eters in the regression were chosen to m inim ize residual variance. In the regression betw een tim e series, form al hypothesis tests, F-statistics and regression coefficients were no t considered as they w ould be biased due to auto-correlation.

T h e w ithin-day variations o f D IC , Cid were m odeled by exponentially weighted runn ing integration o f the PA R series over a period o f 12 hours with an exponential decay rate X (m in-1 ):

f X = t

Cd, = fid + ad P A R {x)-e ( - X(,- l))dxx = t - 1 2 h



In a final step, the irradiance an d w ind sub-models were sum m ed to obtain the final m odel. T he standard deviation o f the residuals betw een m odel an d D IC observation were considered for the param eterization o f W up, kw and X.

Results

3.1 MeasuresT h e three weeks revealed im portan t d a y /n ig h t oscillations o f

p H an d p C 0 2. T h e p C 0 2 increased during the night and decreased during the day reaching m inim a and m axim a at 18:00 and 6:00 respectively. T h e p H inversely m irrored these p C 0 2 variations.

In Ju ly , only p C 0 2 and light was recorded. T h e m ean p C 0 2 of the week was 390 pa tm (m ean ± SD), corresponding to the atm ospheric C 0 2. How ever, diel oscillations o f 281 ± 8 8 patm (m ean ± SD) were recorded, w ith a phase shift o f ~ 6 h betw een light and p C 0 2 cycles. This shift, com m on to all three m easurem ent periods is illustrated for Ju ly in Fig. 2.

In August, seawater was undersatu ra ted for C 0 2 com pared to the atm osphere with a week m ean o f 239 patm only. How ever, the diel variations rem ained com parable to those o f Ju ly with a m ean o f 2 1 9 ± 2 4 patm (± SD) (Fig. 3, top left panel). T h e m ean ± SD daily m axim al an d m inim al p C 0 2 values were 3 7 4 ± 6 7 and 155 ± 3 2 , respectively. T h e m axim al am plitude was 416 patm . T he p H variations m irrored fluctuations in p C 0 2 with a weekly m ean o f 8.21, m ean diel am plitude o f 0 .3 4 ± 0 .1 5 (± SD) and m ean daily m inim um and m axim um o f 8 .0 3 ± 0 .0 7 and 8 .3 7 ± 0 .0 8 (± SD). T h e average salinity an d tem perature during the week were 1 4 .5± 0 .3 psu and 19 .1±0 .7°C , respectively (Fig. 3, lower left panel).

T h e recordings o f Septem ber exhibit a strong difference betw een the beginning (09.09 to 13.09) and the end o f the week (13.09 to 16.09). T h e m ean ± SD salinity and tem perature during the first three days were 17.8± 0.6 psu and 15 .9±0 .6"C while during the three last days they increased and decreased to 20 .7 ± 0 .5 psu an d 12 .4±0 .4"C respectively (Fig. 3, lower right panel). This sudden decrease in seawater density is revealing an upwelling occurring in the m iddle o f the week. C onsequently, we observed a large discrepancy of the p C 0 2/p H betw een those two periods w ith m eans o f 426 p a tm /8 .1 4 an d 1593 p a tm /7 .4 6 respectively (Fig. 2, top right panel). In that second pa rt o f the week, oscillations o f p C 0 2 becam e extrem e. T h e m axim al daily am plitudes recorded were 2184 pa tm p C 0 2 an d 1.15 p H units.

800

n 600 -

* 400 ■ 0U 2 0 0 -

1500

ID 1 0 0 0

3. %%

D a te (U TC)

Figure 2. Com parison betw een Light and p C 02 series in July.Dark bands: period of darkness. Dashed lines: estim ated cen ter o f thedaylight distribution.doi:10.1371/journal.pone.0062689.g002

T h e night peaks (m ean ± SD) o f p C 0 2 were 239 7 ± 4 2 5 patm , with drops o f p H to (m ean ± SD) 7 .26± 0 .07 . D uring daytim e, m inim um p C 0 2 levels w ere still m uch h igher th an atm ospheric concentrations with m inim a o f (means ± SD) 681 ± 211 pa tm and corresponding m axim a of p H of 7.77 ±0 .18 .

3.2 Carbonate systemIn August, m eans ± SD o f the week an d d a y /n ig h t variations o f

D IC were 1609± 37 pm ol kg-1 and 131 ± 4 5 pm ol kg-1 respectively (Fig. 4 central left panel). T h e seawater was always supersaturated with respect to bo th aragonite an d calcite w ith diel m eans ± SD of 2 .2 ± 0.2 and 3 .7 ± 0 .3 an d m eans ± SD d ay /n ig h t variations o f 1 .4 ± 0 .6 and 2 .4 ± 0 .9 respectively (see Fig. 4, top left panels).

In Septem ber the daily m ean ± SD D IC concentrations preceding an d during the upwelling event were 1829±45 and 215 8 ± 27 pm ol kg *, respectively (Fig. 4 central right panel). T he diel variations o f D IC were h igher before than during the upwelling, w ith m ean ± SD of 2 5 9 ± 10 pm ol kg 1 against 20 5 ± 9 1 pm ol kg *, respectively.

Before the upwelling, the seawater was alm ost always supersatu ra ted for aragonite and calcite (see Fig. 4 top right panel, 09.09 to 13.09). H ow ever a t the hours o f m axim al p C 0 2 (6:00), aragonite saturation d ropped below 1: m ean ß arat!: 0 .8 ± 0 .3 (± SD). Oppositely, seawater reached m axim a for ß arat! and ß caic a round 18:00. T h e resulting diel oscillations were o f 4 .3 ± 0 .3 and 2 .6 ± 0 .2 for calcite an d aragonite, respectively.

D uring the upwelling event the seaw ater was constantly undersaturated with m ean ± SD ß caic and ß araa o f 1 ± 0 .5 and 0 .6 ± 0 .3 , exception m ade of the hours a round 18 h, were m odest m axim a (m ean ± SD) of 1 .9 ± 0 .6 and 1.1 ± 0 .4 were reached (see Fig. 4 top right panel, 13.09 to 16.09).

3.3 ModelT h e D IC calculated from p C 0 2 and alkalinity and the D IC

predicted by the statistical m odel for August and Septem ber are presented in Fig. 4 central panels. T h e differences betw een bo th are p resented in Fig. 4. lower panels (residuals). In August, the m odel predicted a diel D IC m ean (± SD) o f 1610 ± 16 pinoi kg-1 and diel am plitudes o f D IC o f 8 6 ± 1 4 pinoi kg 1 (m ean ± SD). T h e differences betw een m odeled and calculated D IC is 1 pinoi kg-1 for the diel average and 44 pinoi kg-1 for the diel am plitude.

In Septem ber, the m odel predicted a diel D IC m ean (± SD) of 1852 ± 83 pm ol kg-1 before the upwelling an d 2143 ± 4 7 pinoi kg 1 during the upwelling. T h e differences w ith the calculated D IC are 23 pinoi kg 1 an d 15 pinoi kg *, respectively. Before and during upwelling, the diel am plitudes o f D IC predicted by the m o d e la re 141 ±61 pinoi kg 1 and 15 0 ± 107 pinoi kg 1 (m ean ± SD), respectively. T h e difference from the calculated D IC is 117 pinoi kg 1 and 55 pinoi kg ’.

T h e set o f param eters producing the best fitting m odels are W up = 3 tt/2 , corresponding to westerly w ind, a period o f integration kw o f 55 h an d X o f 0.0025 m in _1an d 0.001 m in -1 for August and Septem ber respectively. T h e standard deviation o f the residuals betw een m odeled an d calculated D IC are 25 pm ol kg 1 and 49 pm ol kg- 1 , corresponding to percentages o f unexplained variation o f 1.6% and 2.5% for August an d Septem ber respectively (Fig. 3 lower panels). Fig. 5 presents, for Septem ber only, the evolution of the m odel residuals as function o f the w ind direction (Wup) and the integration period (kw). Easterly winds as m axim al weights produce the worst fitting m odels while westerly winds the best fitting ones, southerly an d northerly winds are in term ediate. T h e param eter X had less influence on the outcom e o f the m odel (Fig. 6), increasing its accuracy by a m axim um o f 2 to




9.0 -,

8.5 ■

5 8.0

« 7.5 Io- 7.0 J

0u

E 3 E ai1S

3000 'E' 2500 - % 2000 - 3 1500 O* 1000 £ 500-

0 Ji ^ 2 3 1

20 - g 21 -18 - 3 19 -16 - i 17 -14 - £ 15 -12 - ™ 1 3 -10-1 11 J

T i m e (U TC)

Figure 3. In situ m easurem ents for A ugust and Septem ber. pC 02 (cyan), pH (red) Salinity (black) and tem pera tu re (dark blue) recorded in A ugust (left) and Sep tem ber (right). The recording frequency was one m inute for pC 02 and 45 m inutes for th e 3 o ther param eters. Symbols in the plots mark the recording events; linkage betw een the m easurem ents every 45 min was achieved by linear interpolation. Secondary vertical gridlines unit: 6 h.doi:10.1371/journal.pone.0062689.g003

4 pm ol kg- 1. Despite its poor effect on the m odel, the presence of this param eter is justified by the consideration o f the decrease o f the day length betw een August and Septem ber (the period of integration o f the irradiance is fixed to 12 h in both).

Discussion

In our study, we showed the im portance o f the diel variations o f p C 0 2 an d p H due to photosynthesis and the im portance of the variations o f the carbonate chem istry baseline over sum m er. W e dem onstra ted the interactive effects o f upwelling an d algal m etabolism on the carbonate chem istry w ith a simple statistical m odel.

4.1 Inter-weekly dynamicsT h e p C 0 2 of o u r m easurem ent site exhibited very different

weekly trends over the sum m er. In August, we observed an im portan t and stable p C 0 2 undersaturation . This reflects the conditions o f the whole Baltic a t this period, generated by the succession o f blooms o f phytoplankton an d cyanobacteria [44,45]. In Septem ber this stable condition is in terrup ted by the strengthening o f the westerly winds, leading to the upwelling o f the w ater masses isolated until then below the pycnocline. T he chem istry o f the deep waters o f Eckernförde bay are m onitored since 1957 and p C 0 2 o f about 2500 patm , linked to heterotrophic processes, are yearly observed in sum m er (Boknis Eck tim e series:

765 -

3 4 -3 - 2 - 1 - 0 J

2400 -,

b 2150 - XÖ 1900 -

3 1650 -

1400 -

Ijl

0

g - 1 0 0 -

® -200 -

A A A A A A A A

Figure 4. Calculated and m odeled carbonate chem istry. Top panels: Saturation sta tes for calcite (light brown) and aragonite (dark brown) for A ugust (left) and Sep tem ber (right). Dashed line: Saturation threshold. Central panels: Observed (cyan) and m odeled (red) DIC for August (left) and Sep tem ber (right). The residuals betw een m odel and observations are p resen ted in bo ttom left and right for August and Sep tem ber respectively. Secondary vertical gridlines unit: 6 h. doi:10.1371/journal.pone.0062689.g004




North South West North2t t

W up (R a d ia n )

Figure 5. Sensitivity o f th e m odel to the integration period le, and wind direction Wup. Contour plot o f the standard deviation of the residuals betw een m odel and observation in pmol kg-1 DIC (color palette) for the Sep tem ber series as a function of Wup, the azim uth of reference used as maximal w eight, and th e kw, th e period o f the running integration of the w eighted wind series. doi:10.1371/journal.pone.0062689.g005

[46]). O u r study shows that the nearshore b io ta in shallow w ater is exposed to these high concentrations during upwelling events. T he deep-w ater p C 0 2 is p resum ed to increase in the future by the conjunction of h igher atm ospheric p C 0 2 and reduced salinity and alkalinity o f the Baltic Sea [32]. M elzner et al. (2012) [29] are expecting this deep w ater body to reach p C 0 2 of 4000 to 5000 patm an d the hypercapnie events o f Septem ber to increase in m agnitude. Besides, the frequency and duration o f those upwelling events could also increase as a reinforcem ent o f the westerly w inds in the Baltic region is expected w ith global w arm ing[47].

4.2 Daily oscillationsDiel oscillations, related to photosynthesis and respiration, are

superim posed to the week scale dynamics. In norm al sum m er conditions (without upwelling), the m ean (± SD) am plitudes o f the diel variations w ere 2 4 3 ± 9 5 pa tm C 0 2 (July, August) and 0 .3 4 ± 0 .1 5 p H units (August only). Such diel variations have already been observed in m acrophyte stands worldwide: in seagrass beds o f A ustralia [16], M ed iterranean [48] an d Z anzibar [19], in tidal rocky-shores o f the northeastern Pacific [49] an d in algal m eadows o f the D anish islands [18], A m ong nearshore ecosystems, the highest diel variations are occurring in m acrophyte dom inated ecosystems, upwelling areas and estuaries [50], O u r study site cum ulates those three characteristics: brackish, weakly buffered an d eutrophied ecosystem dom inated by m acrophytes and subm itted to upwelling. W e recorded diel oscillations o f 1604.9±795.7 pa tm (m ean ± SD) during the upwelling event o f Septem ber. T o our knowledge only near-shore m angroves exhibit w ider diel variations with 3500 patm recorded in the B erm uda [10].

H ow ever, the im portance o f these variations is relative, as the fraction o f D IC present as C 0 2(aq) in high p C 0 2/lo w p H seawater is h igher th an in low p C 0 2/h ig h p H seawater (see B jerrum plot in e.g. [39]). T hus, in Septem ber, the diel variations o f C 0 2 are

80

70

60

50O)

0 0.002 0.004 0.006 0.008 0.01A (m irr1)

Figure 6. Sensitivity o f th e m odel to 4. Evolution of the standard deviation o f th e residuals betw een m odel and observation in pmol kg-1 DIC in August (red) and Sep tem ber (black) as a function of X (m in-1 ), the instantaneous rate o f increase used for th e exponentially w eighted running integration of the irradiance series. doi:10.1371/journal.pone.0062689.g006

stronger during the upwelling th an before bu t we observe the opposite for D IC (mean ± SD D IC : 258.6 ± 10.5 pm ol kg-1 against 205.1 ± 9 0 .7 pm ol kg '), possibly revealing a reduction o f photosynthesis during the upwelling. This could be explained by an osm otic stress engendered by a rap id increase o f salinity [51,52] ra th e r th an by the elevated p C 0 2. Indeed, any increase in D IC and p C 0 2 is p resum ed to be beneficial to m arine m acrophytes [53], even if this have never been form ally p roven [54],

4.3 ModelW e were able to explain the D IC variations o f August and

Septem ber to an accuracy o f± 5 0 pm ol kg- w ith a simple statistical m odel (three param eters only), based on w ind an d PAR. T h e w eighing procedure o f the w ind speed, although simple in its geom etric approach, appeared to w ork very well as an explanation o f the m ulti-day trends.

O u r m odel designates westerly winds as responsible for upwelling. This result is slightly different from previous observations and simulations in o ther regions o f the Baltic Sea identifying southwesterly winds as being the m ost effective [30], This small discrepancy is certainly due to the east-west orientation o f E ckernförde bay com pared to the general north-east/south-w est orientation o f the Baltic Sea. T h e use o f exponentially w eighted integration o f irradiance as a proxy for p rim ary production was less effective than the similar approach used for the upwelling. Indeed, while the daily D IC m eans are quite accurately predicted (< 2 0 pinoi kg '), the diel am plitude is systematically underestim ated by 40 to 70 pinoi kg- 1 D IC . This variation, unexplained by the m odel, could be due to heterotrophic respiration, air-sea exchange o f C 0 2 [17] or diel variation o f alkalinity as discussed in 4.1. How ever, despite the simplistic na ture o f our m odel, it is a first step in the understand ing and prediction, in a context o f global change, o f the carbonate system dynam ics in the Baltic nearshore areas.

4.4 Measurement reliabilityT h e accuracy o f the post-processed p C 0 2 da ta is expected to be

< ± 1 0 patm for values w ithin the calibrated m easuring range of 100-1000 patm . An additional erro r can be expected w hen the p C 0 2 exceeds the m easuring range as in this case the calibration polynom ial o f the instrum ent is extrapolated. W e expect this erro r to be ± 1 5 0 pa tm a t m axim um for the highest p C 0 2 recorded in Septem ber at a round 2600 patm .




Despite our efforts, we did no t achieve in the field the p H accuracy necessary to use it as inpu t param eter for the derivation o f the carbonate system, as we are able to in laboratory. W e estim ate the erro r in p H of the o rder o f 0.01 due to uncorrectable drift during the m easurem ent periods. In the conditions o f tem peratu re and salinity o f the w estern Baltic, such inaccuracy in p H produces anom alies in derived alkalinity o f the o rder o f 10— 100 pm ol kg 1 a t p H / p C 0 2 inferior to 7 .8 /8 0 0 patm an d up to 1000 a t h igher p H /lo w e r p C 0 2. For future field studies, Durafets sensors [55] o r spectrophotom etric sensors [56] represent p ro m ising alternatives to glass electrodes, bo th capable o f reaching accuracies o f 0.001 to 0.0001 p H units. W e had to perform alkalinity titrations an d rely on a salinity to alkalinity relationship to achieve the calculations. T his m ethod is relevant [57,58] and we estim ate the erro r on the derived D IC to be < 1 5 pm ol kg-1 D IC . How ever, it ignores any changes o f alkalinity a t constant salinity. This phenom enon is very im portan t in coral and shellfish reefs due to the uptake or release o f C a2+ by calcification or dissolution [59,60]. How ever, in m acrophyte ecosystems, we expect very m arginal diel variations o f alkalinity due to photosynthesis and respiration [61].

4.5 Effect on faunaT h e daily oscillations o f p C 0 2 generated by photosynthesis

could be o f prim e im portance for calcifiers, creating a t daytim e periods o f high saturation states favourable to C a C 0 3 precip itation. Such coupling betw een photosynthesis an d calcification has already been observed in a H aw aiian coral reef by D rupp et al. (2011) [59] an d Sham berger et al. (2011) [60] w here calcification is m axim al a t m idday w hen the p C 0 2 is m inim al due to planktonic photosynthesis. In th a t reef, the intensity o f the photosynthesis is m odulated by w ind driven inputs o f nutrients from the flume o f a neighboring estuary.

In general, studies conducted on w estern Baltic populations o f anim als, calcifying or not, tend to dem onstrate their tolerance to acidic conditions [62]. Also, the Baltic population o f the mussels Mytilus edulis experiences reduced grow th and dissolution o f the shell only w hen Í2arag < 0 .1 5 corresponding to a p C 0 2 of ~ 4000 patm [31]. Despite this w eakening of their shells, the p redation by sea stars Asterias rubens an d crabs Carcinus maenas

References1. N O A A — ES R L (2011) Available: h ttp ://w w w .e sr l.n o a a .g o v /g m d /c c g g /tr e n d s .2. S olom on S, Q in D , M anning M , C h en Z, M arquis M , et al. (2007) IP C C , 2007:

C lim ate C h ange 2007: T h e Physical Science Basis. Contribution o f W orking G roup I to the Fourth A ssessm ent R eport o f the Intergovernm ental Panel on C lim ate C hange, Cam bridge: C am bridge University Press. 996 p.

3. R idgw ell A , Zondervan I, H argreaves J , Bijma J , L enton T (2007) A ssessing the potential long-term increase o f oceanic fossil fuel C O 2 uptake due to C O 2 — calcification feedback. Biogeosciences 4: 481—492.

4. O rr J C , Fabry VJ, A u m ont O , Bopp F , D on ey SC, et al. (2005) A nthropogenic ocean acidification over the twenty-first century and its im pact on calcifying organism s. Nature 437: 681—686.

5. Feely RA , Sabine G F, F ee K , Berelson W , K leypas JA , et al. (2004) Im pact o f anthropogenic C O 2 on the C a C 0 3 system in the oceans. Science 305: 362—363.

6 . K ey R M , K ozyr A , Sabine G F, F ee K , W anninkhof R , et al. (2004) A global o cea n carb on clim atology: R esu lts from G lobal D ata A nalysis P roject (G FO D A P). G lob B iogeochem Gy 18: 1—23.

7. R iebesell U , Fabry VJ, H ansson F (2010) G uide to best practices for ocean acidification research and data reporting. Fuxem bourg: Publication office o f the European U n ion . 260 p.

8 . Borges A V , D elille B, Frankignoulle M (2005) Budgeting sinks and sources o f C O 2 in the coastal ocean: D iversity o f ecosystem s counts. G eophys R es Fett 32: 1—6 .

9. C h en C T A , Borges A V (2009) R econ cilin g opposing view s on carbon cycling in the coastal ocean: C ontinental shelves as sinks and near-shore ecosystem s as sources o f atm ospheric C O 2 . D eep Sea R es Pt II 56: 57 8 —590.

10. A ndersson AJ, M ackenzie F T (2012) R evisiting four scientific debates in ocean acidification research. B iogeosciences 9: 893—905.

m aintained a t similar p C 0 2 are reduced by 56% and 41% respectively [63]. Besides, the grow th of the Baltic brackish barnacle Amphibalanus improvisus, com petitor o f Mytilus for space, rem ains unaffected a t bo th larval an d adult stage a t p C 0 2 > 3 0 0 0 pa tm [64], M acrophyte m eadows are also transient habitats, sheltering early life stages o f num erous anim al species. Those m ight exhibit m ore tolerance to ocean acidification as well. As exam ple, the spawns o f Baltic herrings Clupea harengus, deposited on m acrophyte beds [65], are no t affected in their em bryonic developm ent by high p C 0 2/lo w p H [66]. T h e volatility o f the carbonate system and the extrem e acidic events we observed could have exerted a selective pressure on Baltic populations, explaining the resistance o f the local fauna to acidification stress in laboratory. Nevertheless, all the studies quoted previously were conducted under stable elevated p C 0 2. N one coupled elevated p C 0 2 baseline and the diel variations we observed.

Conclusion

O u r study represents one o f the first attem pts o f high resolution continuous m easurem ent o f the carbonate system in the highly variable environm ent th a t is the Baltic Sea’s nearshore. T h e three weeks showed quite different results related to the dynam ics o f the whole Baltic carbonate system, to the m eteorological condition and to very local processes o f photosynthesis and respiration. This study highlights the im portance o f the natural variations o f p C 0 2 and p H and emphasizes the consideration o f these in ocean acidification studies on nearshore organisms.

AcknowledgmentsWe thank M. Fischer, C. Hiebenthal, C. Howe, B. H uang X uan, C. Lieberum, K. Maczassek and C. Pansch, Y. Sawall and the crew of the F. B. Polarfuchs for their field assistance.

Author ContributionsM anuscript improvement: PF PMJFi. Conceived and designed the experiments: VS. Performed the experiments: VS. Analyzed the data: VS PF PM JH. Contributed reagents/m aterials/analysis tools: VS PF PMJH. W rote the paper: VS.

11. Brasse S, R eim er A , Seifert R , M ichaelis W (1999) T h e influence o f intertidal m udflats on the dissolved inorganic carbon and total alkalinity distribution in the G erm an Bight, southeastern N orth Sea. J Sea R es 42: 93—103.

12. Gattuso J P , Frankignoulle M , W ollast R (1998) C arbon and carbonate m etabolism in coastal aquatic ecosystem s. A n nu R ev E col Syst 29: 4 0 5 —434.

13. G azeau F, Borges A V , Barrón G, D uarte G M , Iversen N , et al. (2005) N e t ecosystem m etabolism in a m icro-tidal estuary {Randers Fjord, Denmark): evaluation o f m ethods. M ar E col P rog Ser 301: 23—41.

14. R ees SA, O pdyke B N , W ilson PA, H enstock TJ (2006) Significance o f Halimeda bioherm s to the global carbonate budget based on a geological sedim ent budget for the N orthern G reat Barrier R eef, Australia. Coral R eefs 26: 177—188.

15. K leypas JA , A n thon y K R N N , Gattuso J P (2011) C oral reefs m odify their seawater carbon chem istry — case study from a barrier reef {M oorea, French Polynesia). G lob C h ange B iol 17: 3667—3678.

16. Sm ith S (1981) M arine m acrophytes as a global carbon sink. Science 211: 838— 840.

17. Gattuso J P , Payri G E, P ichón M , D elsalle B, Frankignoulle M (1997) Primary production , calcification and air-sea C O 2 fluxes o f a m acroalgal-dom inated coral-reef com m unity {M oorea, French Polynesia). J Phycol 33: 729—738.

18. M id delboe AL, H ansen IJ {2007) H igh p H in shallow-water m acroalgal habitats. M ar Ecol Prog Ser 338: 10 7 -1 1 7 .

19. Sem esi IS, Beer S, Björk M {2009) Seagrass photosynthesis controls rates o f calcification and photosynthesis o f calcareous m acroalgae in a tropical seagrass m eadow . M ar E col Prog Ser 382: 4 1 -4 8 .

20. Fobban C S , H arrison IJ {1994) Seaw eed Ecology and Physiology. Cam bridge: C am bridge University Press. 366 p.

21. Beer S, R eh n b erg J {1997) T h e acquisition o f inorganic carbon by the seagrass ¡Tosiera marina. A quat Bot 56: 277—283.


http://www.esrl.noaa.gov/gmd/ccgg/trendshttp://www.plosone.org


22. Badger M R (2003) T h e roles o f carbonic anhydrases in photosynthetic C O 2 concentrating m echanism s. Photosynth R es 77: 83—94.

23. G iordano M , Beardall J , R aven J A (2005) C O 2 concentrating m echanism s in algae: m echanism s, environm ental m odulation, and evolution. A nnu R ev Plant Biol 56: 9 9 -1 3 1 .

24. H epburn C D , Pritchard D W , Cornw all C E , M cL eod RJ, Beardall J , et al. {2011) Diversity o f carbon use strategies in a kelp forest com m unity: im plications for a high C 0 2 ocean. G lob C h ange B iol 17: 2 4 8 8 -2 4 9 7 .

25. Barrón C , M arbà N , D uarte C M , Pedersen M F, Lindblad C , et al. (2003) H igh O rganic C arbon Export Precludes Eutrophication R esponses in Experim ental R ocky Shore Com m unities. Ecosystem s 6 : 144—153.

26. D uarte C M , M arbà N , G acia E, Fourqurean JW , Beggins J , et al. (2010) Seagrass com m unity metabolism : A ssessing the carbon sink capacity o f seagrass m eadow s. G lob B iogeochem C y 24: 1—8.

27. Barrón C , D uarte C M (2009) D issolved organic m atter release in a Posidonia oceanica m eadow . M ar E col Prog Ser 374: 75—84.

28. G ra y J S , W u R S , O r Y Y (2002) Effects o f hypoxia and organic enrichm ent on the coastal m arine environm ent. M ar Ecol Prog Ser 238: 249—279.

29. M elzner F, T h om sen J , K oeve W , O schlies A , G utowska M A , et al. (2012) Future ocean acidification w ill be am plified by hypoxia in coastal habitats. M ar B iol. A v a ila b le:h ttp :/ /w w w .s p r in g e r l in k .c o m /in d e x /1 0 .1 0 0 7 /s 0 0 2 2 7 -0 1 2 - 1954-1. A ccessed 5 J u n e 2012.

30. M yrberg K , Andrejev O (2003) M ain upw elling regions in the Baltic Sea — a statistical analysis based on three-dim ensional m odelling. Boreal Environ R es 8 : 9 7 -1 1 2 .

31. T h om sen J , Gutowska M A , Saphörster J , H einem an n A , Trüb en bach K , et al. (2010) Calcifying invertebrates succeed in a naturally C 0 2-rich coastal habitat but are threatened b y high levels o f future acidification. Biogeosciences 7: 3 879— 3891.

32. G raw e U , Burchard H (2011) G lobal C hange and Balde Coastal Zones. In: Schernewski G, H ofstede J , N eu m a n n T , editors. G lobal C hange and Balde C oastal Zones. Dordrecht: Springer N etherlands, V ol. 1. 3—22.

33. K arez R (2008) K artierung m ariner Pflanzenbestände im Flachwasser der O stseeküste — Schw erpunkt Fucus und 2jostera\ Außenküste der schleswigholsteinischen O stsee und Schlei. Landesam t für Landwirtschaft, U m w elt und ländliche R äum e. 310 p.

34. K rey J , K oske P, Szekielda K (1965) Produktionsbiologische und hydrographische Untersuchungen in der Eckernförder Bucht. K ieler M eeresforschungen 21: 1 3 5 -1 4 3 .

35. B ange H , Bergm ann K , H ansen H , K ock A, K oppe R , et al. (2010) D issolved m ethane during hypoxic events at the Boknis Eck T im e Series Stadon {Eckernförde Bay, SW Balde Sea). B iogeosciences 7: 1 2 7 9 -1 2 8 4 .

36. Fietzek P, K örtzinger A {2010) O ptim izadon o f a m em brane-based N D IR sensor for d isso lved carb on d iox id e. In: H arrison D E , Stam m er D , editors. O ceanO b s'09: Sustained O cean O bservations and Inform ation for Society Conference, venice: ESA publication. 1—f.

37. Fietzek P, K ram er S, Esser D {2011) D ep loym en t o f the H y d r o C ™ { C 0 2 / C H 4) on stationary and m obile platform s — M erging the trends in the field o f platform and sensor developm ent. O cean s 11 M T SIE E E Conference. 1—9.

38. D ickson A G , Sabine CL, C h ristianJR {2007) G uide to Best Practices for O cean C 0 2 M easurem ents. PIC E S spec. D ickson A G , Sabine CL, Christian J R , editors. 191 p.

39. Z eebe R E , W olf-G ladrow D A {2001) C 0 2 in seawater: equilibrium , kinetics, isotopes. Dordrecht: Elsevier. 346 p.

40. Lavigne H , Gattuso J P {2010) seacarb: seawater carbonate chemistry with R. R package version 2 .3 . 3. Softw are. Available: h t tp :/ /s c h o la r .g o o g le .c o m / scholar? hi — en&btnG — Search&q — intide:seacarb:+seawater-l-carbonate-l- chemistry +with+R.+R+package+version+2.3:#=0.

41. M illero IJ {2010) C arbonate constants for estuarine waters. M ar Freshwater Res 61: 1 3 9 -1 4 2 .

42. Perez FF, Fraga F {1987) T h e p H m easurem ents in seawater on the N B S scale. M ar C h em 21: 3 1 5 -3 2 7 .

43. D ickson A G {1990) Standard potential o f the reaction: AgCl{s) + '/zH 2 {g) - Ag{s) + HCl{aq), and the standard acidity constant o f the ion H S O 4 in synthetic sea water from 273.15 to 318 .15 K . C h em T h erm odyn 22: 11 3 -1 2 7 .

44. T h om as H , Schneider B {1999) T h e seasonal cycle o f carbon d ioxide in Baltic Sea surface waters. J M ar Syst 22: 53—67.

45. Schneider B {2011) T h e C 0 2 system o f the Baltic Sea: B iogeochem ical control and im pact o f anthropogenic C 0 2. In: Schernewski G , H ofstede J , N eu m an n T , editors. G lobal C h an ge and Baltic C oasta l Z on es. D ordrecht: Springer N etherlands, V ol. 1. 33—49.

46. B ange H , H ansen H , M alien F, Laß K , K arstensen J , et al. {2011) Boknis Eck tim e series station {SW Baltic Sea): M easurem ents from 1957 to 2010. L O IC Z inprint: 16—22.

47. N eu m an n T {2011) Clim ate C h ange Im pacts on the Baltic Sea. In: Schernewski G, H ofstede J , N eu m a n n T , editors. G lobal C h ange and Baltic Coastal Zones. Dordrecht: Springer N etherlands, V ol. 1. 23—32.

48. Invers O , R om ero J , Pérez M {1997) Effects o f p H on seagrass photosynthesis: a laboratory and field assessm ent. A quat Bot 59: 185—194.

49. W ootton J T , Pfister C A , Forester J D {2008) D yn am ic patterns and ecological im pacts o f declining ocean p H in a high-resolution multi-year dataset. P N a d A cad Sei U S A 105: 1 8 8 4 8 -1 8 8 5 3 .

50. H ofm ann G E, Sm ith J E , J oh n son K S , Send U , Levin LA, et al. {2011) H igh- frequency dynam ics o f o cean pH : a m ulti-ecosystem com parison. PloS on e 6 : e 2 8 9 8 3 . A v a ila b le : h t tp : / / w w w .p u b m e d c e n tr a l .n ih .g o v /a r tic le r e n d e r . fcg i?artid = 3 2 4 2 7 7 3 & to o l = p m cen trez& ren d ertyp e = abstract. A ccessed 6 M arch 2012.

51. O h n o M {1976) Som e observations on the influence o f salinity on photosynthetic activity and chloride ion loss in several seaweeds. Internationale R evu e der gesam ten H ydrobiologie und H ydrographie 61: 66 5 —672.

52. Bäck S, Collins J C , Russel G {1992) C om parative ecophysiology o f Baltic and A dantic Fucus vesiculosus. M ar E col Prog Ser 84: 71—82.

53. Beer S, K o ch E {1996) Photosynthesis o f m arine m acroalgae and seagrasses in globally changing C 0 2 environm ents. M ar Ecol Prog Ser 141: 199—204.

54. Israel A , H op h y M {2002) G rowth, photosynthetic properties and Rubisco activities and am ounts o f m arine m acroalgae grown under current and elevated seawater C 0 2 concentrations. G lob C hange Biol 8 : 8 3 1 —840.

55. M artz T R , C on nery J G , Joh n son K S {2010) T esting the H oneyw ell D urafet for seawater p H applications. L im nol O ceanogr -M eth 8 : 172—184.

56. A ß m ann S, Frank C , K örtzinger A {2011) Spectrophotom etric high-precision seawater p H determ ination for use in underw ay m easuring systems. O cea n Sei 7: 5 9 7 -6 0 7 .

57. Beldowski J , Löffler A , Schneider B, Joensuu L {2010) D istribution and biogeoch em ical control o f total C 0 2 and total alkalinity in the Baltic Sea. J M arSyst 81: 2 5 2 -2 5 9 .

58. H jalm arson S, W essiander K , A nderson LG , O m sted t A , Perttila M , et al. {2008) Distribution, long-term developm ent and m ass balance calculation o f total alkalinity in the Baltic Sea. C on t Shelf R es 28: 5 9 3 -6 0 1 .

59. D rupp P, D e Carlo E H , M ackenzie F T , Bienfang P, Sabine C L {2011) N utrient inputs, phytoplankton response, and C 0 2 variations in a sem i-enclosed subtropical em baym ent, K an eoh e bay, H aw aii. A quat G eochem 17: 4 7 3 —4-98.

60. Sham berger K EF, Feely R A , Sabine CL, Atkinson MJ, D eC arlo E H , et al. {2011) Calcification and organic production on a H aw aiian coral reef. M ar C h em 127: 6 4 -7 5 .

61. C hisholm J , G attuso J P {1991) V alidation o f the alkalinity anom aly technique for investigating calcification and photosynthesis in coral reef com m unities. L im nol O ceanogr 36: 1 2 3 2 -1 2 3 9 .

62. T h om sen J , M elzner F {2010) M oderate seawater acidification does n ot elicit long-term m etabolic depression in the b lue m ussel M ytilus edulis. M ar Biol 157: 2 6 6 7 -2 6 7 6 .

63. A ppelhans Y, T h om sen J , Pansch C , M elzner F, W ahl M {2012) Sour times: seawater acidification effects on growth, feed ing behaviour and acid—base status o f Asterias rubens and Carcinus maenas. M ar Ecol Prog Ser 459: 85—98.

64. Pansch C , Nasrolahi A , A ppelhans Y S, W ahl M {2012) Im pacts o f ocean w arm in g and acid ification on the larval d eve lop m en t o f the barnacle Amphibalanus improvisus. J Exp M ar B iol E col 42 0 —4-21: 48—55.

65. A n eer G {1989) H erring {Clupea harengus L.) spaw ning and spaw ning ground characteristics in the Baltic Sea. Fish R es 8 : 169—195.

6 6 . Franke A, C lem m esen C {2011) Effect o f ocean acidification on early life stages o f A dantic herring {Clupea harengus L.). B iogeosciences 8 : 3697—3707.

http://www.springerlink.com/index/10.1007/s00227-012-http://scholar.google.com/http://www.pubmedcentral.nih.gov/articlerenderhttp://www.plosone.org

Documents

Extreme Variations of pC02 and pH in a Macrophyte Meadow of … · In situ pC02, pH (total scale), ... [11-13]) and (2) reefs formed by corals or calcifying algae because of the direct