Transport and fate of hexachlorocyclohexanes in the oceanic air and surface seawater

Transport and fate of hexachlorocyclohexanes in the oceanic air and surface seawater Z. Xie, B. P. Koch, A. Möller, R. Sturm, and R. Ebinghaus Helmholtz-Zentrum Geesthacht, Centre for Materials and Coastal Research GmbH, Institute of Coastal Research, Max-Planck Str. 1, 21502 Geesthacht, Germany Alfred Wegener Institute for Polar and Marine Research, Bremerhaven, Germany University of Applied Sciences, Bremerhaven, Germany


Introduction
Hexachlorocyclohexanes (HCHs) are ubiquitous organic pollutants derived from pesticide application.They are subject to long-range transport, persistent in the environment, and capable of accumulation in biota (Nizzetto et al., 2010).HCHs can enter the coast, marine and oceanic environment by a number of processes, once introduced they are subject to biogeochemical cycling, sinks, and bioaccumulation processes.Apart from river discharge and continental runoff, the atmospheric deposition is considered to be the primary and most rapid pathway for persistent organic pollutants to the coast and the marine environment as a result of their hydrophobic and semi-volatile nature (Lohmann et al., 2007).Besides, it has been discussed that re-emission of HCHs from indirect sources such as soils, sediments, vegetation, phytoplankton and "old" concentrations in the ocean, which may interfere the air-water exchange process and governs their circulation and transport in the marine environment (Dachs et al., 2002;Fenner et al., 2004;Jaward et al., 2004;Lohmann et al., 2006).

Sample
Date Latitude Longitude Volume Temp.α-HCH γ -HCH β-HCH  (Lakaschus et al., 2002;Lohmann et al., 2009).As a consequence of declining atmospheric concentrations, the air-water change of α-HCH was expected changing from net deposition to volatilization, and the oceans will subsequently turn into sources (Jantunen and Bidleman, 1995).
Along the Atlantic, in 1999/2000, air-sea exchange of αand γ -HCH reached an equilibrium in the North Atlantic, whereas the surface waters of the tropical and southern Atlantic were strongly undersaturated with γ -HCH (Lakaschus et al., 2002).Due to the variability of climate and different intrinsic physical-chemical properties of organo-chlorine pesticides, it is necessary to further investigate the state of HCHs in the oceanic environment.
In this study, we analyzed marine boundary layer air and surface water samples in the Atlantic transect and the Southern Ocean for HCHs.The objectives of this study are (1) to update the levels of HCHs in the atmosphere and the surface seawater, (2) to estimate the air-sea gas exchange directions and fluxes of HCHs, and (3) to evaluate the temporal and latitudinal variability of dissolved HCHs.

Sampling protocol
The sampling of air (Fig. 1a) and seawater (Fig. 1b) has been described in Xie et al. (2011).Briefly, seawater and air samples were collected onboard the R/V Polarstern in the Atlantic and Southern Ocean (51 • N-67 • S) in October to De-cember 2008.Sampling locations, dates and general sampling conditions were recorded aboard from PODAS (Polarstern Data System) and are summarized in Tables 1 and 2. Seawater samples were collected from the ship's intake system located in the keel (depth: 11 m) using a combination of PAD-2 resins (Polystyrene-DVB-copolymer resin, SERVA GmbH, Heidelberg, Germany) and glass fibre filters (GFF: pore size, 0.7 µm).Air samples were collected using GFF filters combined with a glass column packed with PUF/PAD-2 at the upper deck (Altitude: 20 m).Water and air samples were stored at 4 • C and −20 • C, respectively.

Chemicals
All solvents (methanol, acetone, dichloromethane and nhexane) were residue analysis grade and additionally distilled in a full glass unit prior to use.Analytical standards of HCHs and deuterated α-HCH (d6-HCH) were purchased from Dr. Ehrenstorfer GmbH (Augsburg, Germany), and 13 C-HCB was obtained from Cambridge Isotope Laboratories.

Extraction, clean-up and analysis
Extraction, clean-up and analysis of the samples were done based on our previously published method (Xie et al., 2011).Briefly, samples were spiked with the surrogate standard d6-HCH prior to extraction, then Soxhlet extracted and purified on 10 % water deactivated silica column.Analysis was done by a GC/

QA/QC
Breakthrough of the target analytes of the sampling methods has been checked on board R/V Polarstern (Lakaschus et al., 2002), and further proved during this cruise.Three field blanks were run for each sample type while blank showed very low values which were 22, 13, 10 pg in air sample and 12, 7, 43 pg in water sample for α-, γ -and β-HCH, respectively.Method detection limits (MDLs) were derived from mean blank values plus three times the standard deviation (σ ) (for compounds showing no blanks a peak area of 100 was adopted as background response).Atmospheric MDLs were 0.03 pg m −3 for α-HCH, 0.01 pg m −3 for γ -and β-HCH, and seawater MDLs were 0.02, 0.06 and 0.01 pg l −1 for α-, γand β-HCH.Recoveries of internal standard d6-HCH were 81 ± 23 % for water samples and 89 ± 35 % for air samples, respectively.

Air mass back trajectories
Air mass origins along the cruise segments of the individual air samples were calculated using NOAA's HYSPLIT model.Air mass back trajectories were calculated in 6 h steps tracing back the air masses for 7 day using the sampling height as arrival height (Fig. 2).

Results and discussion
Individual concentrations of HCHs in air and seawater are given in Tables 1 and 2. For aqueous samples, only dissolved concentrations were considered, as concentrations of HCHs are below the method detection limits in all filter samples.It is shown in Table 3 for Comparison of HCH concentrations measured in the present study with previous data in seawater and air of the oceans and Polar Regions.

Temporal and latitudinal trends of HCHs in surface seawater
Comparison of HCH in 1999/2000 with those obtained between 1987 and 1997 have been performed in Lakaschus et al. (2002), and exhibited a strong decline for α-HCH between 50 • N and 60 • S, and no clear trend for γ -HCH.To evaluate updated temporal variation in the Atlantic, the new data set from this study and the historical data were merged into The different trends for α-and γ -HCH suggest (i) the influence of international regulation on technical HCHs and lindane; and (ii) variable environmental behavior and fate for α-and γ -HCH.There was a rather high variability presented in the tropic region for both α-and γ -HCH.Unlike explanation by Lakaschus et al. (2002) for γ -HCH in 1999, the high precipitation rate of approximately 2000 mm yr −1 in the Intertropical Convergence Zone could cause significantly dilution rather than addition due to the intensive wet deposition.Another important factor is intensive biomass blooming in the tropical region, which has been observed during this cruise as well.The Equator tread winds bring massive Sahara dust containing nutrients and elements into the tropic ocean (Jullien et al., 2007;Cole et al., 2009;Pohl et al., 2011), which accelerates phytoplankton and zooplankton blooming in surface water of the Atlantic (Fernández et al., 2010;Guieu et al., 2010;Neogi et al., 2011;Taylor et al., 2011).The adsorption of α-and γ -HCH to biomass and par-  (1993), ANT-XV (1997), ANT-XVII (1999/2000) and ANT-XXV (2008, this work).
has been reported in a recent study for HCHs in mediterrancean seawater (Berrojalbiz et al., 2011).The concentrations of α-HCH in the Southern Ocean are quite variable, which can be addressed to the complex frontal system.It has been pointed out that elevated [α-HCH] diss was present between 40 • S and 50 • S from 1993 to 2000 (Lakaschus et al., 2002), this phenomenon was also found in this present study.The sampling data of W27, W29 and W31 showed a salinity decrease from 35.5 to 33.9 and 34.0, and temperature decreased from 22.3 • C down to 4.6 and −1.6 • C as well.This is caused by an influx of fresh melting sea ice and snow water from the Antarctic shelf and results in a transfer the "old" contamination back to the Southern Ocean (Dickhut et al., 2005).Moreover, the Southern African current may transport HCHs from Indian Ocean to the Atlantic and moves them northward by the path of thermohaline circulation.This input may significantly contribute to the elevated HCHs in the Southern Ocean.

α/γ -HCH and α/β-HCH ratios in surface water
Variations of α/γ -HCH and α/β-HCH ratios in space and time are highly influenced by the historical usage of technical HCH and lindane, and the environmental behaviors of different isomers.The most widely quoted composition of technical HCH is 60-70 % α/γ -HCH, 5-12 % α/γ -HCH and other isomers, resulting α/γ -HCH and α/β-HCH ratio about 4-7 (Iwata et al., 1993).As show in Fig. 4, obviously high α/γ -HCH values were present in 2008 in comparison to 1987-2000, and mostly above the range of 5-7 of the α/γ -HCH ratio in technical mixture (Iwata et al., 1993).In contrast to the present results, α/γ -HCH ratios in the NH were mostly less than 4.5, and reached 0.1-2 in 1991 and 1999, indicating intensive application of lindane after a ban for technical γ -HCH during long range transport (Oehme et al., 1996).The highest α/γ -HCH ratio 33 was found in the Southern Ocean; again indicates high persistence of α-HCH in remote region.
So far, α/β-HCH ratios were only insufficiently studied.In this study, α/β-HCH ratios varied from 0.5 to 52, highlighted different environmental behavior of α-and β-HCH isomers.Generally, β-HCH is one of the five stable isomer of technical HCH, and accounts for 5-14 % of the technical formation.Unlike α-and γ -HCH, β-HCH has a higher affinity to water than air, reflected in its lower Henry's law constant and higher water solubility.Low α/β-HCH ratios (<4) may suggest input sources of HCHs from adjacent landmasses.In the Southern Ocean as atmospheric transport and deposition is the major pathway, high α/β-HCH ratios were present in this region (31 for W27 and 17 for W31).

Air-water gas exchange
The direction (or equilibrium status) of the gas exchange was estimated based on the fugacity ratio f A /f W , and the exchange fluxes were calculated using the two-film model which have been applied in Xie et al. (2011) andLohmann et al. (2009).Briefly, the fugacity ratio was calculated using Eq. ( 1).
where f W and f A are the fugacities in water and air, C W and C A are the dissolved and gaseous concentrations in water and air (pg m −1 ), H is the Henry's Law constant (Pa m −3 mol −1 ) at the given water temperature and corrected by the salinity according to Schwarzenbach et al. (2003), R is the gas constant (8.31 Pa m −3 K −1 mol −1 ) and T A is the air temperature (K).The Henry's Law constant of HCH and its temperature dependence was taken from Sahsuvar et al. (2003) and Cetin and Odabasi (2005).Generally, a fugacity ratio f A /f W = 1 means a system at equilibrium, whereas f A /f W > 1 and f A /f W < 1 indicates deposition and volatilization, respectively (Eq. 1).Due to uncertainties of knowing air-water transfer coefficient, a significant deviation from equilibrium cannot be assessed within a factor of 3 around a fugacity ratio of 1 (Bruhn et al., 2003;Lohmann et al., 2009).The air-seawater gas exchange was calculated based on following Eq.( 2) (Liss and Slater, 1974;Bidleman and Mc-Connell, 1995;Schwarzenbach et al., 2003) where H is the dimensionless temperature and salinity corrected Henry's Law constant defined as H = H/RT (R = gas constant, T = Temperature).K OL (m h −1 ) is the overall airwater mass transfer coefficient compromising the resistances to mass transfer in both water (K W , m h −1 ) and air (K A , m h −1 ) and is defined by Schwarzenbach et al. (2003): D air is the diffusity in air, U 10 is the wind speed at 10 m height above sea level (m s −1 ), and Sc is the water phase Schmidt number which was taken from Schwarzenbach et al. (2003) for CO 2 .D air was calculated using the method described in Fuller et al. (1966) and Sc was calculated using the method described in Hayduk and Laudi (1974).The uncertainty of the flux can be estimated by propagation of the uncertainties in C W (23 %), C A (35 %), K OL (40 %) and H (20 %, Sahsuvar et al., 2003), which is 61 %. f A /f W of α-HCH ranged from 0.8 to 27 with most values >3, indicated air to water deposition dominating airseawater gas exchange directions (Fig. 5), which might also be caused by important loss terms in the water mass e.g.setting and degradation.Two values (0.8 for W1 and 1.9 for W3) were within 0.3 to 3, which showed a dynamic equilibrium reached near the western European coast (51 • N-45 • N).For the γ -isomer, f A /f W indicated net deposition in all samples (774 > f A /f W > 3.8, n = 17).Although β-HCH has relatively low levels in the atmosphere, because it's lower H value, f A /f W varied between equilibrium (volatilization) and net deposition.Lakaschus et al. (2002) found that in the North Atlantic air-water exchange status of α-HCH changed from net deposition in 1990 to equilibrium in 1999, and a new equilibrium was being established on a lower concentration level than 1990 (Lakaschus et al., 2002).Obviously, the results from this work showed that a new equilibrium has established for α-HCH on a lower level than 1999, while equilibrium status in the Atlantic and the Southern Ocean from 45 • N to 67 • S has been broken up and changed to net deposition again.This variability was also observed in the north Atlantic and the Arctic Ocean (Harner et al., 1999;Lohmann et al., 2009).Gas exchange directions of α-HCH between seawater and air reversed in the western Arctic from net deposition in the 1980s to net volatilization in the 1990s with declined primary emissions (Bidleman et al., 1995;Jantunen and Bidleman, 1995;Jantunen et al., 2008).While air-water exchange direction for γ -HCH have been dominated by net deposition in most studies (Harner et al., 1999;Jantunen et al., 2008;Lohmann et al., 2009), with volatilization occasionally reported in the western Arctic.
The air-water deposition fluxes were quite high for α-HCH with a median of 3800 pg m −2 day −1 through the cruise (Fig. 5), except W1 for a volatilization of 820 pg m −2 day −1 .Elevated net deposition occurred in the European and northwest African coast ranging from 3800 to 11 000 pg m −2 day −1 , and was just slightly lower than those measured in mid-Atlantic region in 2000/2001 (Gioia et al., 2005).In the SH, net deposition of α-HCH ranged from 570 to 4700 pg m −2 day −1 , which are ∼10 times lower than those in the NH.Relatively constant levels in atmosphere and decreasing water concentrations of α-HCH in recent   years may contribute to the changing direction of the airseawater gas exchange and high deposition fluxes.A recent study in the Canadian Arctic (2007)(2008) (Wong et al., 2011) showed net volatilization for α-HCH with a mean of 6800 ± 3200 pg m −2 day −1 , where the air concentrations were similar to this present work; but the water concentra-tions were 2-3 orders of magnitude higher than our results.Nevertheless, α-HCH undergoes the iterative process of deposition and adsorption onto soil and vegetation, reemission into the atmosphere and re-deposition because of reduction in primary emissions and climate change.
The γ -HCH was undergoing net deposition to surface waters with deposition fluxes ranging 400-5600 pg m −2 day −1 (mean: 1987 pg m −2 day −1 ) (Fig. 6), indicating continually loading into the Atlantic and the Southern Ocean since several decades.In comparison to the other two HCH species, β-HCH showed relatively low exchange fluxes (6-690 pg m −2 day −1 for net deposition and <12 pg m −2 day −1 for net volatilization).

Conclusions
α-, γ -, and β-HCH have been simultaneously measured in air and seawater of the Atlantic and the Southern Ocean in 2008.There was a marked difference between contaminant trends in atmosphere and surface waters highlighting the different time-scales affecting compounds in either compartment, reemission from the continental soil, vegetation and forest fire are considered sources for atmospheric HCHs.Climate change may significantly accelerate the releasing process and drive long-range transport from sources to deposition in the open oceans.In the tropic ocean higher biological productivities may increase metabolism, sorption and settling fluxes and lead to lower concentrations of HCHs in the surface water, and thus control the air-water exchange process.Consequently, further investigation is necessary to elucidate the long term trends and the biogeochemistry process of HCHs in the oceanic environment.

Fig. 1 .
Fig. 1.(a) Gaseous (pg m −3 ) and (b) dissolved (pg l −1 ) concentrations of α-, γ -and β-HCH in the Atlantic and the Southern Ocean.The bars are placed on the average position for each air and water sample. al

Fig. 2 .
Fig.2.96 h air mass back trajectories (6 h steps) and altitudinal profiles of the air mass parcels for the cruises ANT-XXV/1+2 (A1-A17).For samples longer than 72 h, only every second BT was plotted.The black line indicates the cruise leg.
Figure 3a and b for a close comparison.Similar latitudinal trends in the NH have been observed in all cruises from 1987 to 2008.Slightly increasing tendency from the Equator to the Southern Ocean also appeared in the SH.From 50 • N to 30 • S, concentrations of α-HCH in the present work were lower by factor of 10-50 than those measured in 1987-1997 and just slightly lower than those in 1999/2000.The concentrations of γ -HCH obviously decreased in comparison to those reported in 1987-2000, especially showed clearly declining trend from 2000-2008.

Fig. 5 .
Fig.5.Air-water fugacity ratio (f A /f W ) of α-, γ -and β-HCH in the Atlantic and Southern Ocean, a f A /f W within the range 0.3-3 means a system at equilibrium.

Table 2 .
Individual concentrations of HCHs in surface seawater (pg l −1 ) in the Atlantic and the Southern Ocean.

Table 3 .
Comparison of HCH concentrations measured in the present study with previous data in seawater and air of the oceans and Polar Regions.