Biogeosciences Dissolved methane during hypoxic events at the Boknis Eck time series station ( Eckernförde Bay , SW Baltic Sea )

Dissolved CH4 was measured in the water column at the Boknis Eck (BE) time series station in the Eckernf ö de Bay (SW Baltic Sea) on a monthly basis from June 2006 to November 2008. The water column at BE was always supersaturated with CH 4 and, therefore, CH 4 was released to the atmosphere throughout the sampling period: the mean CH4 surface (1 m) saturation at BE was 554 ±317%. A pulse of enhanced CH4 emissions occurs when the CH 4 accumulation in the hypoxic bottom layer during summer is terminated in late summer/autumn. We did not detect a straightforward relationship between periods of enhanced CH 4 in the bottom layer and hypoxic events at BE: the sedimentary release of CH4 seemed to be mainly triggered by sedimenting organic material from phytoplankton blooms. We conclude that future CH4 emissions from BE will be determined by the intensity of phytoplankton blooms, which in turn will be influenced by eutrophication. However, hypoxic events seem to have only a modulating effect on the enhancement of sedimentary methanogenesis and the subsequent release of CH 4 to the water column.


Introduction
The worldwide increasing number of coastal areas with extremely (persistent or temporary) depleted dissolved oxygen (O 2 ) concentrations [O 2 ] has received considerable scientific and public attention during the last years (Diaz and Rosenberg, 2008;Rabalais et al., 2009;UNEP, 2004).So-called coastal hypoxic events or hypoxia occur when [O 2 ] falls below ∼90 µmol L −1 (2 mL L −1 ) (Diaz and Rosenberg, 2008).Anoxic conditions are defined by [O 2 ] = 0 µmol L −1 and the occurrence of dissolved hydrogen sulphide (H 2 S).The increasing number of coastal hypoxia is attributed mainly to increasing nutrient inputs (i.e.eutrophication) by anthropogenic activities (UNEP, 2006).
The duration and actual [O 2 ] of hypoxia and anoxia have profound influences on both ecosystems and biogeochemical cycles in coastal zones (Diaz and Rosenberg, 2008;Malakoff, 1998).This is especially alerting in view of the fact that the microbial formation (methanogenesis) of the climate-relevant trace gas methane (CH 4 ) is a strictly Introduction

Conclusions References
Tables Figures

Back Close
Full Screen / Esc Printer-friendly Version Interactive Discussion anaerobic process which is favoured under hypoxic and anoxic conditions in the water column as well as in the sediments (Cicerone and Oremland, 1988).Therefore, on the one hand, it was speculated that an expansion of eutrophication-caused hypoxia might lead to an increase in the accumulation of dissolved CH 4 and thus lead to increasing coastal CH 4 emissions to the atmosphere (Bange, 2006).On the other hand, Naqvi et al. (2009) concluded that an expansion of hypoxic areas not necessarily would lead to an increase in CH 4 emissions.
We measured CH 4 in the water column at the Boknis Eck coastal time series station in the Eckernf örde Bay (SW Baltic Sea) on a monthly basis from June 2006 to November 2008.Because of seasonally occurring hypoxic events, Boknis Eck provides an ideal natural laboratory to investigate the effects of coastal hypoxia on the distribution of dissolved CH 4 .

Study site description
The time series station Boknis Eck (BE) is located at the entrance of the Eckernf örde Bay (54 • 31 N, 10 • 02 E; Fig. 1) and has a water depth of about 28 m.The hydrographic setting of BE is dominated by the inflow of North Sea water via the Kattegat and the Great Belt.Seasonal stratification is associated with the development of a pycnocline at about 15 m water depth from mid of March until mid of September.During this period, vertical mixing is restricted and bacterial decomposition of organic material in the deep layer causes pronounced hypoxia and sporadically occurring anoxia during late summer (Hansen et al., 1999).(aq).After replacing 10 mL of the water sample with helium, the samples were allowed to equilibrate for at least 2 hours.A 9 mL subsample of the headspace was used to flush a 2 mL sample loop.The volume of the sample loop was then injected onto a packed stainless steel column filled with molsieve 5A for isothermal gas chromatographic separation.Detection of CH 4 took place with a flame ionization detector.Further details of the method are described in Kock (2007).Two-point calibration curves were measured daily before and after the measurements of the samples.For the calibration we used two standard gas mixtures of 1.779±0.002and 2.543±0.002ppm CH 4 in synthetic air.The gravimetrically prepared gas standards (made by DEUSTE Steininger GmbH, M ühlhausen, Germany) were calibrated against the actual NOAA standard scale (Dlugokencky et al., 2005) in the laboratories of the Max Planck Institute for Biogeochemistry in Jena, Germany.
[CH 4 ] in nmol L −1 was calculated with where x is the dry mole fraction of CH 4 in the headspace, P is the ambient pressure (set to 10 5 Pa), V hs and V wp are the volumes of the headspace and the water phase, respectively.R stands for the gas constant (8.31451 m 3 Pa K −1 mol −1 ), T is the temperature during equilibration and β is the Bunsen solubility of CH 4 (Wiesenburg and Guinasso Jr., 1979).CH 4 saturations (Sat in %) are expressed as where [CH 4 ] eq is the equilibrium concentration calculated with the in-situ temperature and salinity at BE and an atmospheric CH 4 dry mole fraction of 1.85 ppm, which was considered to be a representative mean for the Baltic Sea region from 2006 to 2008 (Rigby et al., 2008).for the minimal and maximal CH 4 concentrations of the triplicate samples.The factor 1.91 is derived from the statistical method by David (1951).Triplicates with a standard deviation of >10% were omitted.The mean analytical error of [CH 4 ] was ±4.7%.Concentrations of dissolved O 2 and chlorophyll a were determined using standard methods (Grasshoff et al., 1999).Water temperature and salinity were taken from the routine CTD measurements at BE. Water transparency was measured with a Secchi disk (∼30 cm in diameter, white).

Results and Discussion
Three hypoxic events with [O 2 ] <90 µmol L −1 have been observed at BE from 2006 to 2008 (Fig. 2 The water column at BE was always supersaturated with CH 4 (Fig. 3) and, therefore, CH 4 was released to the atmosphere throughout the sampling period (for an estimate of the sea-to-air flux density see below).Our findings are in agreement with the measurements of Bussmann and Suess It is well known that the sediments of the Eckernf örde Bay are sites of high CH 4 accumulation resulting from in situ methanogenesis which, in turn, is fuelled by the high amount of organic material in the sediments (Abegg and Anderson, 1997;Treude et al., 2005;Whiticar, 2002).(The CH 4 in the Eckernf örde Bay sediments does not come from the underlying Schwedeneck petroleum reservoir (Whiticar, 2002).)Despite the fact that at BE anaerobic CH 4 oxidation in the sediments can effectively prevent the CH 4 release to the water column (Treude et al., 2005), some of the CH 4 obviously is released via ebullition (Jackson et al., 1998) or methanogenesis in the upper sediments (Treude et al., 2005).In the Eckernf örde Bay CH 4 can be also released to the water column from pockmark structures during episodes of groundwater efflux (Bussmann and Suess, 1998), however, the majority of the pockmarks are found in the southern part of Eckernf örde Bay (Whiticar, 2002) and, therefore, a CH 4 contribution via groundwater is negligible at BE.Moreover, we did not find any signs of groundwater release in the salinity data of BE.In view of the discussion above we conclude that CH 4 in the water column of BE is dominated by its release from the sediments.
From the data in Fig. 2 it is obvious that periods with enhanced [CH 4 ] in the bottom layers are not correlated with the hypoxic events.This was confirmed by statistical analysis which did not revealed any statistical significant correlation between [O 2 ] and [CH 4 ].In order to find out whether the observed variability of [CH 4 ] may be caused by the seasonality of sedimentary methanogenesis, we focus the following discussion on the bottom layer (25 m).We detected a bimodal seasonal CH 4 distribution in 25 m with Introduction

Conclusions References
Tables Figures

Back Close
Full Screen / Esc

Printer-friendly Version
Interactive Discussion two pronounced maxima in spring and fall (Fig. 4).The seasonal [CH 4 ] maxima in the bottom layer seem to be linked to both chlorophyll in 25 m and Secchi depth with a time lag of one month (Fig. 5).The correlation of [CH 4 ] with chlorophyll was significant at a significance level (α) of 0.05 (r = 0.67, n = 10), whereas the correlation with Secchi depth was significant at α>0.05 (r = 0.56, n = 10).We conclude that the major drivers of the seasonality of [CH 4 ] in the water column are the phytoplankton blooms which deliver organic material to the sediments leading in turn to enhanced sedimentary CH 4 formation (Treude et al., 2005) with a time lag of a month.Because the second (i.e. the fall) maximum of CH 4 in the bottom layer is more pronounced and it is concurrent with the hypoxic event, it might be reasonable to suggest that hypoxia in conjunction with the occurrence of sedimenting organic material emphasise the sedimentary CH 4 release at BE: During the hypoxic event in Fall, sedimentary methanogenesis at BE is significantly enhanced as has been recently shown by Treude et al. (2005).Therefore, the hypoxic events at BE modulate the sedimentary CH 4 release but they do not trigger enhanced CH 4 release to the water column.
The CH 4 release to the atmosphere was calculated with where F ase is the flux density in µmol m −2 d −1 , k w is the gas transfer coefficient (see below) and Sc is the Schmidt number which was calculated as the ratio of the kinematic viscosity of seawater (Siedler and Peters, 1986) and the diffusion of CH 4 in seawater (J ähne et al., 1987).For k w we used 0.65×10 −5 m s −1 as a lower limit and  , 1998).Therefore, we conclude that the CH 4 release from the sediments is sufficient to balance the CH 4 release to atmosphere.Additional source or sink terms such as advection or aerobic CH 4 oxidation in the water column seem to be negligible at BE.

Summary
Sedimentary release of CH 4 determined the CH 4 concentrations in the water column at the Boknis Eck time series station during the sampling period from June 2006 to November 2008.A rough flux estimate indicates that the CH 4 release to the atmosphere is balanced by the CH 4 flux from the sediments and no further sink or source processes are required to explain the CH 4 water column distribution.We did not detect a straightforward relationship between periods of enhanced CH 4 in the bottom layers and hypoxic events.Indeed we found a seasonality of [CH 4 ] in the bottom layer with two maxima in March and September.The sedimentary release of CH 4 seemed to be mainly triggered by sedimenting organic material from phytoplankton blooms.Therefore, we conclude that future CH 4 emissions from BE will be determined by both the occurrence and intensity of phytoplankton blooms, which in turn will be influenced by the future trend in nutrient inputs (i.e.eutrophication).Hypoxic events have only a modulating effect in the way that they enhance methanogenesis in the sediments.
dissolved CH 4 [CH 4 ] were measured with a static equilibration method: 25 mL glass vials were filled bubble-free and immediately poisoned with HgCl 2 Triplicate water samples for the determination of [CH 4 ] were taken from six standard depths (1, 5, 10, 15, 20 and 25 m) with a 5 L Niskin bottle.The standard deviation was approximated with ([CH 4 ] max -[CH 4 ] min )/1.91,where [CH 4 ] min and [CH 4 ] max stand 11466 ).The most pronounced event was observed in 2007 when [O 2 ] in the deep layer already started to drop below the threshold in May.The hypoxia in 2007 lasted until November and was most intensive in September when the waters became anoxic in the bottom layer (25 m).The hypoxic events in 2006 and 2008 were much shorter: They lasted from August to November in 2006 and August to September 2008.The lowest [O 2 ] observed during the 2006 and 2008 hypoxia were 6.0 and 6.1 µmol L −1 , respectively.The pronounced hypoxic event in 2007 was resulting from an intensive phytoplankton bloom which is clearly visible in both the enhanced chlorophyll a concentrations from November 2006 until March 2007 (Fig. 2) and a reduced water transparency (Secchi depth) of only ∼3 m and ∼4 m in February and March 2007, respectively, which were the lowest Secchi depths measured during the sampling period from June 2006 to November 2008.Because BE is not influenced by any river plumes, the reduced transparency was caused only by living phytoplankton as well as dead organisms and other decaying organic material.The intensive bloom led to enhanced sedimentation of organic material which, in turn, resulted in enhanced O 2 consumption during remineralisation and the development of the pronounced hypoxic event in 2007.The time series of CH 4 is shown in Fig. 2. CH 4 concentrations were in the range from saturation in 5 m in February 2008) to 235±17.4 nmol L −1 (= 8340% saturation in 20 m in October 2007).

(
1998) at two stations in the southern part of Eckernf örde Bay during February 1993 and May 1994.They found [CH 4 ] in the surface layer (1 m) in the range from 2.1 to 2815 nmol L −1 .
an upper limit as recommended by Raymond and Cole (2001) for calculations of flux densities in estuaries.[CH 4 ] was computed as the median of all measurements in 1 m depth (14.4 nmol L −1 ) and [CH 4 ] eq was calculated as the median of the equilibrium concentrations in 1 m (3.1 nmol L −1 ) (see Fig. 3).The resulting mean CH 4 flux densities for the sampling period 2006-2008 were 6.3 and 14.7 µmol m −2 d −1 .This is in reasonable agreement with a rough estimate of the CH 4 ebullition flux from the sediment to the overlying water column in the range of 3-20 µmol m −2 d −1 (Jack-

Fig. 3 :Fig. 3 .Fig. 4 : 5 Fig. 4 .
Fig. 3: Depth distribution of CH 4 at Boknis Eck (all data, open circles).The bold line indicates the mean CH 4 profile based on the median values calculated for each depth.The dashed line indicates the corresponding CH 4 equilibrium concentrations.

Fig. 5 .
Fig. 5. Mean monthly CH 4 vs. chlorophyll (upper panel) and CH 4 vs.Secchi depth (lower panel).The mean monthly CH 4 and chlorophyll data were calculated for 25 m.Please note that the CH 4 data were shifted by one month (see Fig. 4) to account for the apparent time lag of one month.

Dissolved methane during hypoxic events at the Boknis Eck Time Series Station (Eckernf örde Bay, SW Baltic Sea) H. W. Bange
1 , K. Bergmann 1 , H. P. Hansen 1 , A. Kock 1 , R. Koppe 2 , F. Malien 1 , and C. Ostrau 1 Farías et al., 2009)y presented here (and time series of dissolved CH 4 in other coastal regions worldwide, for example off Chile;Farías et al., 2009)seem to be a valuable tool to monitor CH 4 emissions from coastal areas in order to decipher the potential effects of eutrophication and climate change for the long-term trends of coastal CH 4 emissions to the atmosphere.Introduction