Occurrence of volatile organic compounds (VOCs) in Liverpool Bay, Irish Sea

Available online at www.sciencedirect.com

Marine Pollution Bulletin 54 (2007) 1742–1753 www.elsevier.com/locate/marpolbul

Occurrence of volatile organic compounds (VOCs) in Liverpool Bay, Irish Sea C.M. Bravo-Linares a, S.M. Mudge b

a,*

, R.H. Loyola-Sepulveda

b

a School of Ocean Sciences, University of Wales-Bangor, Menai Bridge, Anglesey LL59 5AB, UK Universidad Cato´lica de la Santı´sima Concepcio´n, Facultad de Ciencias, Alonso de Ribera 2850, Concepcio´n, Chile

Abstract Surface seawater samples were collected in the Irish Sea and Liverpool Bay area from the R.V. Prince Madog during the period of 25–31 of March 2006. VOCs were purged with nitrogen, pre-concentrated on a SPME fibre and analysed immediately on a GC–MS. Target compounds quantified were halogenated (0.2–1400 ng L1), BTEXs and mono-aromatics (1.5–2900 ng L1), aliphatic hydrocarbons and others (0.6–15,800 ng L1). Day and night sampling was performed at a single station and suggested that factors such as sunlight and tide affect the presence of many of these compounds. Sample variability was high due to the variable weather conditions at the station. Poor correlations were found between marine phytopigments and selected VOCs. Principal component analysis (PCA) analysis showed that chlorinated compounds such as 1,2-dichloroethane, 1,1,1-trichloroethane, trichloroethene, tetrachloroethene and carbon tetrachloride, predominantly from anthropogenic sources, originated from the River Mersey. Other brominated and iodinated compounds quantified were more likely to be from biogenic sources including novel marine compounds such as 2-chloropropane, 1-bromoethane and 1-chlorobutane.  2007 Elsevier Ltd. All rights reserved. Keywords: Halocarbons; Seawater; Pigments; River mersey; Solid-phase microextraction (SPME); BTEXs

1. Introduction Halogen-containing VOCs are potential greenhouse gases and ozone depletors and since their use and production has been regulated by the Montreal and Kyoto Protocol (Buchmann et al., 2003), their natural production has received more attention. Together, this group of small molecular weight compounds may have dramatic effects on the future global climate. Anthropogenic sources for many of these compounds have been identified and effort is being directed towards reducing their release to the environment. Anthropogenic sources of these compounds are well known and they can arise from several industrial processes and human activities (McCulloch et al., 1999). However, natural marine sources may be important too: oceans

*

Corresponding author. Fax: +44 (0) 1248 382879. E-mail address: [email protected] (S.M. Mudge).

0025-326X/$ - see front matter  2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2007.07.013

may act as a sink or reservoir of these compounds (Chester, 1990) and the marine boundary layer is one of the most important places for gas exchange between water/atmosphere (Jones, 1980). Little is the known about the mechanisms inducing the production of these compounds by macro and microalgae. There are many hypotheses regarding their production such as for removal of metabolic wastes (Gagosian and Lee, 1981), as chemical communicants (Gagosian and Lee, 1981), for chemical protection (Laturnus, 1996), to rid the cell of harmful oxidants such as hydrogen peroxide (Manley, 2002), maybe as a form of oxidative stress relief (Mtolera et al., 1996), for anti-herbivore activity (Wolfe et al., 1997), anti-microbial properties (Fenical, 1981; Fenical, 1982; Neidleman and Geigert, 1987), to facilitate food gathering or as hormones (Gribble, 2000) and other reasons not yet clearly understood. It is proposed that the mechanisms of production are mediated by enzymes called haloperoxidases (van Pee and Unversucht, 2003). Biogenic halocarbons are broadly

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

in the area of Liverpool Bay or Irish Sea. Liverpool Bay and especially the Mersey River are areas of intense human activity (Fox et al., 2001): the port and rivers have played a major role in the city development and many industries are next to the river and gas extraction activities occur offshore. This bay is influenced by several rivers including the Conwy, Dee, Ribble and Mersey. Tide plays an important role in water masses in this area with ranges up to 10 m. The aim of this research was to quantify the distribution of halocarbons and other selected volatile organic compounds which may have both natural and anthropogenic sources in the area of Liverpool Bay and Irish Sea.

produced in the marine environment by macro and microalgae, making the ocean a significant natural source of halocarbons and other VOCs (Chuck et al., 2005). The incidence and distribution of halocarbons and other VOCs in marine water have been widely studied e.g. in South Asian and Western Pacific Oceans (Yokouchi et al., 1997), the Atlantic Ocean (Baker et al., 2000), the Northwest Atlantic and Pacific Oceans (Tokarczyk and Saltzman, 2001), the Arctic Ocean (Krysell, 1991), the North Sea (Huybrechts et al., 2005), estuaries (Christof et al., 2002), bays (Yamamoto et al., 2001) and different marine environments (Dewulf and VanLangenhove, 1997). The distribution and concentration of VOCs depended greatly on the location and, for biogenic sources, whether or not an algal bloom was present in ocean system or macroalgae in coastal zones. In relation to anthropogenic inputs, concentrations are dependent on the proximity of production to the sampling site. VOCs are of special interest as they may photo-dissociate producing radicals that can participate in atmospheric reactions and able to deplete tropospheric and stratospheric ozone (Barrie et al., 1988). They also may participate in cloud condensation nuclei formation and reduce the oxidation capacity of the atmosphere (Liss et al., 1997). However, no studies of the incidence and distribution of these compounds in seawater have been reported

2. Materials and methods 2.1. Seawater sampling and extraction procedure Surface seawater samples were collected across Liverpool Bay and River Mersey between 25 and 31 March 2006 at eighteen stations (Fig. 1). Water samples were collected from the RV Prince Madog using a polished stainless steel container and transferred very slowly to avoid sample disturbance to a 4.85 L amber glass bottle; 4.5 L of water were added leaving a head-space of approximately 350 mL. The seawater was purged with extra pure nitrogen (150 mL min1) for 1 h at ambient temperature with

54.2

54.0

LATITUDE ˚N

Irish Sea River Ribble

53.8

8

10 11

7

53.6

12

14

6

9

13 4 5 53.4

3

2

1 Anglesey

M3 M2

Liverpool M1

15 River Dee

River Mersey

River Conwy -4.0

-3.8

-3.6

1743

-3.4

-3.2

LONGITUDE ˚W Fig. 1. Map of the sampling locations.

-3.0

-2 .8

1744

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

continuous stirring to homogenize the sample and release the VOCs in the seawater (see Fig. 2). The nitrogen passing through the seawater was purified through an organic trap to minimize the possible volatile organic content in this gas. The seawater was spiked with an internal standard (4-bromofluorobenzene, Supelco) to a final concentration of 1776 pg L1 prior to purging and VOCs were trapped on a SPME fibre for an hour. The fibre was retracted and transferred to the GC–MS for analysis. All the analyses were conducted immediately after purging on board the ship. Sampling was performed with the ship’s prow towards the dominant wind and with the engine off to avoid sampling the gases coming from the ship and stirring from the ship’s propeller. 2.2. Solid-phase microextraction (SPME) procedure The purgeable VOCs contained in the samples were trapped on a SPME fibre. Solid-phase microextraction (SPME) was developed by Pawliszyn and co-workers (Chai et al., 1993) and combines sampling and concentration in one step. The method requires no solvent and provides good results for a wide range of analyte concentrations. This technique has been used to analyze different compounds in water (Eisert and Levsen, 1996), but the analysis of halocarbons or other VOCs in seawater has yet to be reported.

N2

2.3. Gas chromatographic system and conditions The VOCs adsorbed onto the fibre were analyzed on a Fisons MD800 Gas Chromatograph coupled to a mass spectrometry detector (GC–MS). A fused silica capillary column AT-502.2 (Alltech 60 m · 0.32 mm i.d. and film thickness 1.8 lm) was used for the separation of all halocarbons and other VOCs. The chromatographic conditions were: helium as a carrier gas (4.3 ml min1), at a pressure of 140 KPa; temperature programme: isothermal at 40 C held for 5 min, 4 C min1 to 60 C held for 3 min, 4 C min1 to 80 C and held for 2 min, 4 C min1 to 150 C and isothermal at 150 C for 10 min. The injector temperature was kept in 230 C throughout analysis. The mass spectrometer conditions were: electron impact ionisation (70 eV), source temperature at 200 C and interface temperature at 160 C.

b

a

N2

A Jade valve kit for SPME for Fisons series 8000 from Alltech was used as an injector device for the fibre’s syringe. A SPME fibre of 2 cm-50/30 lm DVB/carboxen/ PDMS for manual sampling (Supelco) was employed. Before sample extraction, the fiber was conditioned for 1 h at 230 C according to the manufacturer’s specifications to remove compounds adsorbed on the fibre and to stabilize the phase. Conditioning was carried out in a split/splitless port (split open) with helium (99.999% pure). The compounds adsorbed on the fibre were immediately desorbed into the injector for 5 min at 230 C with the split/ splitless valve closed for thermal desorption of the extracted compounds.

c

Fig. 2. Diagram showing the purging and concentration system employed for VOCs analysis in seawater. (a) Organic filter, (b) SPME fibre and (c) Amber bottle of 4.8 L (4.5 L of seawater and 0.35 L of head-space) with continuous stirring.

2.3.1. Identification and quantification The target compounds were previously identified and classified in seawater taken from the Menai Strait (BravoLinares and Mudge, submitted for publication). Single ion monitoring (SIM) was used to identify and quantify the compounds of interest. This methodology was at least 10 times more sensitive than scan mode. The target compounds were equally divided through two SIM programs and care was taken to have the same dwell time and interchannel delay among the fragments. Some of the targets compounds sought and their respective fragments were: halogenated compounds such as 1,1-dichloroethene (m/z 61), dichloromethane (m/z 49), chloroform (m/z 83), 1,1,1-trichloroethane (m/z 97), tetrachloroethene (m/z 166), carbon tetrachloride (m/z 117), bromoform (m/z 173), dibromomethane (m/z 93), iodoethane (m/z 156), diiodomethane (m/z 268), iodomethane (m/z 142), chloroiodomethane (m/z 176), bromodichloromethane (m/z 83) among others; some sulphur containing compounds such as dimethyl sulphide (m/z 62) and dimethyl disulphide (m/z 94) as well some mono-aromatics such as BTEXs in general (m/z 91 or 105), benzene (m/z 78) and aliphatics hydrocarbons (m/z 43 or 57). Quantification was performed using the internal standard method and external calibration using calibration

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

curves from a halocarbon standard (JMHW VOC Mix A, 16 analytes 1000 lg mL1, Supelco). 2.4. Marine phytopigments sampling and analysis Seawater sub-samples were also taken for photosynthetic pigment analysis by HPLC. Waters were filtered on board through a Whatman 47 mm GF/F filter using a vacuum pump. The filters were folded and wrapped using aluminium foil and kept frozen at 20 C for up to 10 days. From prior analysis, no discernible changes were detected when storing pigments samples and standards at this temperature for a month. The volumes filtered varied between 2 and 5 L, due to the high amount of suspended particles, especially near to river mouths. All the samples were extracted at the same time. The filters were unwrapped, and cut in thin slices and extracted in 5 mL of 90:10 acetone/water for 24 h in the freezer at 20 C. A 100 lL aliquot was analysed by HPLC using the joint global ocean flux study (JGOFS) method. The solvent program utilized for the analysis was reported by Jeffrey et al. (1997). Standards were used to determine the concentration of pigments present in the seawater (b,b-carotene, fucoxanthin, 19 0 -hexanoyloxyfucoxanthin, alloxanthin, chlorophyll b and phaeophytin a from DHI Water and Environment, Denmark and chlorophyll a from Sigma–Aldrich). The pigments were identified in the spectral range of 200–700 nm and the areas quantified at 436 nm. Identification was done using wavelengths reported by Jeffrey et al. (1997). 3. Results and discussion 3.1. Limit of detection (LOD) For blank purposes, filtered seawater (0.1 lm), sterilized with UV light and degassed for two days with nitrogen was used. The limit of detection was determined by multiple analyses of the blanks and determining their standard deviation (SD). The standard deviation was multiplied for three ensuring that the signal is different with the blank. These values were extrapolated in the calibration curve. The results were calculated using the follow relation LOD = 3 · SDblank/ Slope. Limits of detection varied from 25 to 49 pg L1 for all the VOCs employed for calibration purposes.

1745

gas transference, varied from 3.6 to 26.2 m s1. Atmospheric pressure was uniform (ca. 1000 mbar) and solar irradiation reached 868 W m2. Weather in general was changeable and this made it difficult to follow the original sampling grid. 3.3. Surface seawater sampling for VOCs Seventeen surface seawater samples were collected. Only one sample per station was sampled. Variograms suggest that classed postings are a better way of displaying the results than contour plots as there was no systematic gradient across the region of interest. 3.3.1. Halocarbons Twenty one different halogenated compounds were detected during the sampling period in the range of 0.2 to 1400 ng L1. The compounds were consistent with the use of chlorinated solvents in industrial processes and biological production. In addition, brominated and iodinated VOCs that are rarely reported as industrial chemicals and are more likely to be produced from biogenic sources were also identified. Some chlorinated compounds (see Fig. 3a–e) were clearly coincident with riverine inputs especially the River Mersey; these included 1,2-dichloroethane, 1,1,1-trichloroethane, trichloroethene, tetrachloroethene and carbon tetrachloride. Carbon tetrachloride was one of the compounds whose production was phased out under the Montreal Protocol (Buchmann et al., 2003). However, it is still found in the environment due to its long lifetime in seawater of approximately 30 years (Huhn et al., 2001) and its production from natural sources (Bravo-Linares and Mudge, submitted for publication). The mean measured concentration of carbon tetrachloride (2.3 ng L1) is similar to those reported by Schall et al. (1997) and Class et al. (1986) in the Atlantic and South Atlantic Ocean, respectively. The remaining halogenated compounds were heterogeneously distributed and are known to be produced from biological sources; these were brominated, iodinated or contained mixed halogens. Table 1 shows that halogenated compounds found in this cruise are commonly reported from biogenic sources have similar concentration to those reported in the literature. However, there are some compounds which were also detected which may be novel biogenic compounds in the marine environment such as 2-chloropropane, 1-bromoethane and 1-chlorobutane.

3.2. Conditions during sampling Across the sampling area, water depth varied from 8 to 47 m. Salinity varied according to the influences of the rivers, with stations far from rivers having higher values (32– 33) and stations close to or in the River Mersey had lower values (19–30). Surface water temperature was relatively constant ranging from 5.7 to 7.9 C. A wider fluctuation was observed with air temperature ranging from of 2.4 to 11.7 C. Wind speed, one of the main factors affecting the

3.3.2. BTEXs and mono-aromatic compounds At least 13 different mono-aromatic compounds were found in a range of 1.5–2900 ng L1. The sources of BTEX are diverse and, due to this, their distribution is not clear. Some originate from terrestrial sources (combustion of fuel oils, solvents, etc.) and also from oil and gas extraction facilities in the Irish Sea; the high maritime traffic of fossil fuel powered vessels in this particular area may also contribute. Selected distribution plots are shown in Fig. 3f–i.

1746

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

Results from the Menai Strait (Bravo-Linares and Mudge, submitted for publication), an isolated portion of the Irish Sea ranged between 0.038 and 4282 ng L1. Gschwend et al. (1982) and Mantoura et al. (1982) have reported a maximum concentration of 500 ng L1 in coastal waters ostensibly free from anthropogenic inputs. However, Marti et al. (2001) found aromatic hydrocarbons in surface seawater in the North-eastern Atlantic in the range 23–68 pg L1, values very low in comparison with those measured in the Irish Sea. The amounts found at each place depend intimately on local contamination and the uses of the water: the highest concentrations were found near the river mouths and close to the port activities. Lower concentrations were found in waters remote from the coast and towards the west (see Fig. 3).

a

1,2-dichloroethane

b

c

Trichloroethene

0.2 to 0.6

0.0 to 0.1

0.6 to 0.9

0.1 to 0.7

0.5 to 1.3

0.9 to 1.0

0.7 to 0.8

1.3 to 4.3

0.8 to 2.5

4.3 to 14.1

2.5 to 24.3

14.1 to 44.0

2.1 to 10.5

g

1,1,1-trichloroethene

0.0 to 0.5

1.0 to 2.1

d

3.3.3. Aliphatic hydrocarbons and others The distribution plots for some hydrocarbons detected are shown in Fig. 3j–m. Their concentrations ranged between 0.6 and 15,800 ng L1. Results in the relatively clean area of the Menai Strait had values much lower than those found in this study (Bravo-Linares and Mudge, submitted for publication) suggesting a greater impact of anthropogenic hydrocarbons in the Liverpool Bay area. In this study n-C5 correlated with other small hydrocarbons such as cyclohexane (R = 0.886) and hexane (R = 0.793) and these behaved differently to longer hydrocarbons such as n-C9, n-C10, n-C11 and branched aliphatic hydrocarbons. McDonald et al. (1988) suggested that hydrocarbons n-C6 to n-C10 are the dominant alkanes in areas affected by petroleum sources and the n-C5 has a

Tetrachloroethene

e

Carbon tetrachloride

f

Benzene

0.0 to 0.8

0.3 to 0.7

3.9 to 31.4

0.8 to 1.5

0.7 to 1.3

31.4 to 59.0

1.5 to 3.4

1.3 to 1.9

59.0 to 86.6

3.4 to15.1

1.9 to 4.1

86.6 to 114.1

15.1 to 45.5

4.1 to 11.4

114.1 to 141.7

Toluene

h

Ethylbenzene

i

m-xylene

3.6 to 37.3

2.5 to 8.1

4.3 to 70.0

37.3 to 71.0

8.1 to 16.9

70.0 to 135.6

71.0 to 104.7

16.9 to 18.5

104.7 to 138.3 138.3 to 172.1

135.6 to 201.3

18.5 to 33.8

201.3 to 266.9

33.8 to 119.7

266.9 to 332.6

Fig. 3. Distributions for some selected VOCs (ng L1) across Liverpool Bay and the River Mersey.

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

j

n-C5

k

n-C9

0.0 to 213.8

0.0 to 21.8

0.0 to 7.4

213.8 to 427.5

21.8 to 43.5

7.4 to 14.6

43.5 to 65.3

14.6 to 19.4

65.3 to 87.1

19.4 to 41.1

87.1 to 108.9

41.1 to 15770.0

427.5 to 641.3 641.3 to 855.0 855.0 to 1069.0

m

l

n-C8

n-C10

n

o

Isoprene

Dimethyl sulphide

0.0 to 15.4

3.9 to 31.4

1.7 to 2.6

15.4 to 28.9

31.4 to 59.0

2.6 to 4.0

28.9 to 50.3

59.0 to 86.6

4.0 to 4.4

50.3 to 71.9 71.9 to 7430.0

1747

86.6 to 114.1

4.4 to 6.2

114.1 to 141.7

6.2 to 18.8

Fig. 3 (continued)

Table 1 Concentrations of halogenated compounds commonly reported from natural sources in this study compared with others locations Compound

Range (ng L1)

CHBr3

<0.03–4.30 3–7

Mean (ng L1) 0.74 0.80

CH3CH2Br CH3I CH3CH2I CHBrCl2

CHBr2Cl

0.8–>6 <0.03–15 1–100 100.29–110,376 <0.03–1.42 0.049–20.73 <0.03–1.52 1.60–17.75 <0.03–183.06 0.14–49.29 <0.03–1.68 <0.03–15 <0.03–13.98 6.90–12,463 <0.03–15 0.1–2.2

0.26 0.43 30.51 0.16 0.10 2.90

0.12

presumably biogenic source. Thurman (1985) attributed the presence of n-C5 and n-C7 to marine phytoplankton and n-C8, n-C9, n-C10 and n-C11 were concentrated in areas with activities involving oil.

Place

Reference

Irish Sea Arctic Ocean South Atlantic North Atlantic Atlantic Ocean Eastern Arctic Ocean Coastal east Atlantic Irish Sea Coastal east Atlantic Irish Sea Coastal east Atlantic Irish Sea Coastal east Atlantic Irish Sea Atlantic Ocean South Atlantic Irish Sea Coastal east Atlantic Atlantic Ocean North Atlantic South Atlantic

This study Class et al. (1986) Class et al. (1986) Schall et al. (1997) Dyrssen and Fogelqvist (1981) Carpenter et al. (2000) This study Carpenter et al. (2000) This study Carpenter et al. (2000) This study Carpenter et al. (2000) This study Schall et al. (1997) Class et al. (1986) This study Carpenter et al. (2000) Schall et al. (1997) Class et al. (1986) Class et al. (1986)

Isoprene (Fig. 3n) was found in the range 0.62 to 14.24 ng L1, with a mean concentration of 3.49 ng L1. Similar quantities (0.66–3.45 ng L1) were detected by Milne et al. (1995) in the Gulf Stream off the Florida coast.

1748

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

Isoprene was reported as an indicator of phytoplankton presence by Yokouchi et al. (1999), however, no correlation was found between isoprene and chlorophyll a in this study (R = 0.063, P = 0.798). Gist and Lewis (2006) found alkenes such as ethene (0.47–26.62 ng L1) and propene (0.54–13.86 ng L1) in the North Sea. Their results demonstrated that their marine concentrations and the processes controlling their production are heterogeneous. 3.3.4. Sulphur containing compounds Dimethyl sulphide (DMS) was the only sulphur containing compound detected (Fig. 3o). Its concentration ranged between 1.7 and 19 ng L1 with a mean concentration of 5.6 ng L1. It is well known that this compound is produced by marine phytoplankton (Simo, 2001) and higher concentration might be expected during algal bloom periods. Turner et al. (1996) measured a concentration of 8.06 ng L1 during the winter time and 1550 ng L1 during the spring bloom in May in the North Sea. Other investigators have measured 124 ng L1 in Antarctic waters (Turner et al., 1995) and 68 ng L1 in the east Atlantic (Andreae et al., 2003). The concentrations of DMS are strongly influenced by the seasons. This work was conducted in early spring and suggests the early stages of a phytoplankton bloom. Higher concentration may be expected during May with presence of Phaeocystis and other microflagellates. Good correlations were found by Shenoy et al. (2006) between phytoplankton population and chlorophyll a/phaeopigments ratios and DMS production; no correlation (R = 0.013, P = 0.968) was found in this study between chlorophyll a and DMS concentrations. Results obtained by Bravo-Linares and Mudge, submitted for publication suggested a highly seasonal variation throughout the year with low values during the winter and high values during the algal bloom periods ranging from 0.078 to 71 ng L1. A review of DMS in seawater (Buckley and

Mudge, 2004) showed there are wide spatial differences in DMS concentrations and it might not practical to compare results from different places. 3.4. Surface seawater sampling for pigments During the period of sampling (end of March) chlorophyll a concentrations were low (1.50–8.91 lg L1) compared to the expected peak blooms periods a month later. Satellite images from remote sensing data analysis service (RSDAS) (Fig. 4) for the corresponding period (29th March–4th April 2006) were obtained. The presence of clouds during the exact dates of sampling meant no coverage was available but these images are within a few days. These images showed concentrations of chlorophyll a between 2 and 8 lg L1. However, is possible to see higher values in the coastal zone, although this may be due to the high suspended solid loadings coming from the rivers or resuspension in this area and not necessarily to a high primary production. Despite that, values of chlorophyll a were similar to the ones found in this investigation. Major pigments such as chlorophyll a (1.50– 8.91 lg L1), chlorophyll b (0.06–0.37 lg L1), b,b-carotene (0.02–0.10 lg L1) and fucoxanthin (0.05– 0.72 lg L1) were present in all the samples. Minor pigments such as chlorophyllide a (0.10–1.46 lgL1), 19 0 -butanoyloxyfucoxanthin (0.48–4.42 lg L1), 19 0 -hexanoyloxyfucothanthin (0.11–0.73 lg L1), cis-19 0 -hexanoyloxyfucothanthin (0.11–0.79 lg L1), diadinoxanthin (0.04–0.17 lg L1), diatoxanthin (0.01–0.22 lg L1) and chlorophyll allomer (0.10–0.82 lg L1) were also detected. The findings of Gowen et al. (2000) in Liverpool Bay and Irish Sea suggest that the main phytoplankton bloom occurs at the beginning of April and lasts to the middle of May according to the chlorophyll a values. Concentrations of chlorophyll a during the spring bloom in coastal

Fig. 4. Satellite image of (a) chlorophyll a (lg L1) and (b) suspended particulate matter (mW cm2 lm1 sr1) for the period 29th March–4th April 2006. Satellite data were acquired and processed by NERC Earth Observation Data Acquisition and Analysis Service (NEODAAS: www.neodaas.ac.uk).

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

ranged from 11 to 170 RSD% with the lowest values for major pigments and higher values for minor pigments.

and offshore waters of the western Irish Sea can reach 23 and 16 lg L1, respectively; values of up to 44 lg L1 in Liverpool Bay, dominated by Phaeocystis and other microflagellates (Gowen and Stewart, 2005), also occur. The small concentrations of some accessory pigments (chlorophyll b, fucoxanthin, diadinoxanthin, diatoxanthin, etc.) suggest that the phytoplankton were at an early stage of bloom development and may be principally diatoms.

3.6. Statistical analysis Principal component analysis (PCA) was employed as a tool to investigate the relationships among compounds. The concentrations of the VOCs analyzed were standardized using proportions and, since most of them were not normally distributed, a logarithmic transformation was performed.

3.5. Samples variability 3.5.1. VOCs variability In order to measure the variability of VOC concentrations, three samples were collected at station 2 on different days during the sampling period. The concentrations of VOCs in surface water were highly variable with a relative standard deviation (RSD%) of 380% although these differences were not unexpected as the VOCs fluxes are highly effected by physico-chemical variables such as wind speed (Liss et al., 1993) which varied considerably over this period. For instance, the highest concentrations of hydrocarbons and BTEXs were coincident with high solar irradiation and low wind speeds; the opposite was found for halogenated compounds. This suggests different sources, mechanisms of formation or different physical and chemical properties. The meteorological conditions also lead to different stripping mechanisms.

3.6.1. Halocarbon sources The loadings plot for the halocarbons alone is shown in Fig. 5. The results show that compounds such as carbon tetrachloride, tetrachloroethene, trichloroethene, 1,2dichloroethane and 1,1,1-trichloroethane clearly separate on PC1 from the other (natural–source) halocarbons. From the scores plot in Fig. 5 is possible to see a gradient of different compounds with sampling sites in the River Mersey characterized by the chlorinated solvents above; a group of samples characterized by bromoform, DMS and 2-chloropropane was seen in the near coast environment. A third offshore area had a diverse variety of halogenated compounds brominated and iodinated with a primarily biogenic source. 3.6.2. Correlation with pigments Chlorophyll a is usually used as an indicator for phytoplankton, one of the main producers of halocarbons in the open oceans (Tokarczyk and Moore, 1994). Although it is improbable that chlorophyll a is directly involved in the marine production of halocarbons, it has been found that it could be used as an indicator for the biogenic production of these compounds (Schall et al., 1997). The correlation

3.5.2. Pigment variability The same sampling procedure was followed to see the variability of pigments at the same station. The variability was not great and may be due to mixing of water masses with different properties under the influence of tides. Another factor is the influence of the river’s discharge near this station, diluting and mixing the water. The variability Tetrachloroethene Trichloroethene 1,2-dichloroethane Carbon tetrachloride

0.40

2-chloropropane 0.20

0.10

0.00

Bromoform Dimethyl sulphide

Predominantly anthropogenic

M3 M1 5 6 7

3 1

9 10 413 14

11

1-chlorobutane 1-bromoethane Iodoethane

12

8

-0.10

1-bromopentane Iodomethane

Scores on PC1 -0.10

cis-1,2-dichloroethene 2-chlorobutane Dichloromethane 1-Iiodobutane Chloroform 1-bromopropane 2-Iiodopropane Dibromochloromethane

M2

Scores on PC2

Loadings on PC2

Predominantly biogenic

1,1,1-trichloroethane

0.30

1749

0.00

0.10

0.20

0.30

Loadings on PC1 Fig. 5. PCA loadings plot for halogenated compounds showing a separation on the first major axis of those compounds from biogenic or anthropogenic sources. Scores plot are included inside the figure.

1750

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

Interestingly, the degradation pigments such as phaeophytin a and chlorophyllide a correlated with the chlorinated solvents predominantly produced by anthropogenic sources. This correlation may suggest that those chemical might be produced, in part, by senescent phytoplankton cells, zooplankton faecal pellets (grazing activity) or due to the presence of greater amounts of terrestrial detritus in the river area.

between chlorophyll a and iodocarbons is not high (Schall et al., 1997). However, in the surface water of the Southern North Sea, Liss et al. (1994) found a correlation between iodomethane and chlorophyll a concentrations although Moore et al. (1996) did not find a correlation between chlorophyll and methyl chloride. Connan et al. (1997) observed a good correlation between phytoplankton presence and chlorophyll a concentration and suggested that the occurrence of bromocarbons in coastal zones is associated with phytoplankton growth or simply their presence. The correlation between halocarbons might indicate the same sources and origins; good correlation has been found among bromocarbons compounds by Yokouchi et al. (1997) and Schall et al. (1997) proving the same biogenic origin and mechanism of formation. PCA was performed using the halogenated compounds and the marine phytopigments; those compounds from biogenic sources such as bromoform, dimethylsulphide, etc. correlated with the major pigments such as chlorophyll a, fucoxanthin, etc.

a

20

Water Temp (°C) Air Temp (°C) Wind Speed (m/s)

15

b

200

Solar Irradiation (W/m2)* Tide (m)

3.6.3. Day and night sampling Samples for VOCs and pigments analysis were collected over a day/night cycle at station 15 (see Fig. 1). Meteorological as well physico-chemical parameters were recorded at the time that the samples were taken (Fig. 6a). The sampling interval was 3–5 h. Fig. 6 shows how some VOCs behaved over this period. Pentane behaved differently to the C8 to C11 hydrocarbons (Fig. 6b); its concentration decreased considerably from day to night. C8 and C9 maintained their concentration Pentane (C5) Octane (C8)*

Nonane (C9)* Decane (C10)

Undecane (C11)

2000

150

150

1500

5

ng.l-1

100

100

ng.l-1

10

1000 50

50

500 0

0 12:00

15:00

18:00

21:00

00:00

0

05:00

0 12:00

Sampling Time

15:00

18:00

21:00

00:00

05:00

Sampling Time

Benzene Toluene*

140

m-Xylene p-Xylene

d 400 60

120

40

3 ng.l-1

200 60

4

40

ng.l-1

80

Dimethyl sulphide Carbon tetrachloride Iiodoethane

50

300

100 ng.l-1

1-bromopropane 1-Iodobutano Chloroform Dichloromethane

2

30 20

100

ng.l-1 Thousands

c

1

20

10

0

0 12:00

15:00

18:00

21:00

00:00

0

05:00

0 12:00

15:00

Sampling Time

18:00

21:00

00:00

05:00

Sampling Time

e

Chlorophillade a 19'-butanoyloxyfucoxanthin

5

Fucoxanthin Chlorophyll a

10 9 8 7

3

6 2

ug.l-1

ug.l-1

4

5 4

1

3 0

2 12:00

15:00

18:00

21:00

00:00

05:00

Sampling Time

Fig. 6. Tendencies during the day and night sampling. (a) Physico-chemical parameters, solar irradiation (W m2) right axis and seawater, air temperature (C) and wind speed (m s1) left axis. (b–d) Selected VOCs in ng L1 (iodoethane, chloroform, dichloromethane, C8 to C11 and toluene are plotted against the right axis). (e) Some marine photopigments (lg L1), chlorophyll a and 19 0 -butanoeloxyfucoxanthin plotted against right axis.

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

the VOCs were normally distributed, no further transformations were necessary. PCA with VOCs showed a day/ night pattern of distribution on PC1. Iodinated, brominated, light hydrocarbons and some chlorinated compounds, previously identified as predominantly from biogenic sources, correlated well with solar irradiation. This suggests that some of them might be produced as a product of photoproduction or photodegradation as well by photosynthesis. Bromoform correlated with tide. At high tide the macroalgae near to the coast are covered with water and compounds, especially bromoform, are released into the water (Carpenter and Liss, 2000; Bravo-Linares and Mudge, submitted for publication). Chlorophyll a and other pigments did not have a clear relationship with the VOCs measured.

during the sampling; C10 and C11 correlated well, with a lowering of their concentration during the day and relatively constant through the night. The differences between the hydrocarbons might be due to different sources or due to the lighter hydrocarbons being influenced by sunlight or produced during photosynthesis. BTEXs are more likely to be produced by anthropogenic sources. However, toluene behaved differently to other BTEXs (Fig. 6c) and this might be due to its biological production in anoxic environments as well anthropogenic sources. It has been reported by Mrowiec et al. (2005) that toluene can be produced during the acid phase of anaerobic degradation in sludge; similar processes may happen with anoxic sediments (Bravo-Linares and Mudge, 2007) or micro-anaerobic sites in the water column. The concentrations of some halogenated compounds decreased dramatically after dark (Fig. 6d). Similar results were found by Ekdahl et al. (1998) in a diurnal study and concluded that it was due to losses to the atmosphere. Dimethylsulphide (Fig. 6d) did not vary significantly during the day cycle and it seems to not be affected by day and night transitions. Compounds which were identified as chlorinated solvents such as carbon tetrachloride did not vary considerably throughout the sampling. Marine phytopigments did not correlate with the production of VOCs. Fig. 6e shows that they are inversely correlated with tide cycle, probably owing to water masses movement with different characteristics. Factor such as water and air temperature, wind speed, solar irradiation and tide height are shown in Fig. 6a. In order to see if these difference in concentration through the day are due to physical or biological factors, PCA was performed (Fig. 7). The VOC concentrations were standardized using proportions (to avoid the concentration effect) and the values of physico-chemical parameters and pigments were used without modifications. Since

Day Loadings on PC2

4. Conclusions Liverpool Bay, especially the area influenced by the River Mersey have significant concentrations of halogenated compounds such as 1,2-dichloroethane, 1,1,1-trichloroethane, trichloroethene, tetrachloroethene and carbon tetrachloride which are widely used as a industrial solvents. Biogenic sources of VOCs, with emphasis on the halogenated ones, may also be an important source in this area. These results indicated that three halogenated compounds might be novel biogenic VOCs in the marine environment; these are 2-chloropropane, 1-bromoethane and 1-chlorobutane. Sampling one location under different environmental conditions may dramatically affect the presence of VOCs due to such factors as wind speed, solar irradiation and tide. Day and night sampling showed that for this area halogenated as well other VOCs typically from natural sources decreased noticeably after sunset; this can be

Wind Speed

0.2

1-Iiodobutano 1-chlorobutanecis -1.2-dichloroethene

Fucoxanthin

2-chloropropane HeptanalOctane Decane

Night

β-carotene Chlorophyll a

0.1 1-bromopropane

1751

Water T

Dichloromethane Diiodomethane 0.0 Dibromochloromethane 2-Iodopropane Solar Irradiation PentaneCiclohexane iodomethane Dibromomethane 1-bromoethane -0.1 Air T 2-chlorobutane Chloroform Iodoethane 1-bromopentane

-0.2

Tide

α-Pinene Isoprene Alkyne unknown Carbon tetrachloride Nonane 1,2-dichloroethane Salinity Dimethyl sulphideBenzene Hexanal Tetrachloroethenem-Xylene p-Xylene Undecane Limonene Octanal

Bromoform Toluene cis-1,3-dichloropropene

Trichloroethene

-0.10

0.00

0.10

Loadings on PC1

Fig. 7. PCA analysis for day/night sampling showing how physico-chemical, meteorological and some pigments (underlined and italics) correlate with the distribution and production of some selected VOCs.

1752

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753

explained due to lower production or transference of these compounds to the atmosphere during the night. Marine pigments may indicate the presence of phytoplankton but do not have a direct relationship with VOCs presence in this survey. SPME is a good technique for shipborne analysis and enables quantification of wide range of VOCs in seawater at low concentrations. Further seasonal sampling is needed to understand the annual processes of production and distribution of volatile organic compounds in the Irish Sea. Acknowledgements The authors would like to thank to the R.V. Prince Madog crew members, specially the Captain Steve Duckworth for his important cooperation and advice through the cruise and all the enthusiastic scientists who helped in the sample preparation and extraction during the cruise. We would like to thank Peter Miller (PML) for the satellite images. Claudio Bravo-Linares would like to acknowledge the support of his Scholarship funded by the European Community through the ALbAn program code E03D00781CL. References Andreae, M.O., Andreae, T.W., Meyerdierks, D., Thiel, C., 2003. Marine sulfur cycling and the atmospheric aerosol over the springtime North Atlantic. Chemosphere 52 (8), 1321–1343. Baker, A.R., Turner, S.M., Broadgate, W.J., Thompson, A., McFiggans, G.B., Vesperini, O., Nightingale, P.D., Liss, P.S., Jickells, T.D., 2000. Distribution and sea-air fluxes of biogenic trace gases in the eastern Atlantic Ocean. Glob. Biogeochem. Cycle 14 (3), 871–886. Barrie, L.A., Bottenheim, J.W., Schnell, R.C., Crutzen, P.J., Rasmussen, R.A., 1988. Ozone destruction and photochemical-reactions at polar sunrise in the lower Arctic atmosphere. Nature 334 (6178), 138–141. Bravo-Linares, C.M., Mudge, S.M., submitted for publication. Temporal variation of volatile organic compounds (VOCs) in seawater. Mar. Chem. Bravo-Linares, C.M., Mudge, S.M., 2007. Analysis of volatile organic compounds (VOCs) in sediments using in situ SPME sampling. J. Environ. Monit. 9 (5), 411–418. Buchmann, B., Stemmler, K., Reimann, S., 2003. Regional emissions of anthropogenic halocarbons derived from continuous measurements of ambient air in Switzerland. Chimia 57 (9), 522–528. Buckley, F.S.E., Mudge, S.M., 2004. Dimethylsulphide and oceanatmosphere interactions. Chem. Ecol. 20 (2), 73–95. Carpenter, L.J., Liss, P.S., 2000. On temperate sources of bromoform and other reactive organic bromine gases. J. Geophys. Res.-Atmos. 105 (D16), 20539–20547. Carpenter, L.J., Malin, G., Liss, P.S., Kupper, F.C., 2000. Novel biogenic iodine-containing trihalomethanes and other short-lived halocarbons in the coastal East Atlantic. Glob. Biogeochem. Cycle 14 (4), 1191– 1204. Chai, M., Arthur, C.L., Pawliszyn, J., Belardi, R.P., Pratt, K.F., 1993. Determination of volatile chlorinated hydrocarbons in air and water with solid-phase microextraction. Analyst 118 (12), 1501–1505. Chester, R., 1990. Marine Geochemistry, pp. 233–269, Unwin Hyman Ltd., London. Christof, O., Seifert, R., Michaelis, W., 2002. Volatile halogenated organic compounds in European estuaries. Biogeochemistry 59 (1–2), 143–160. Chuck, A.L., Turner, S.M., Liss, P.S., 2005. Oceanic distributions and air– sea fluxes of biogenic halocarbons in the open ocean. J. Geophys. Res. Oceans 110(C10).

Class, T., Kohnle, R., Ballschmiter, K., 1986. Chemistry of organic traces in air 7. Bromochloromethanes and bromochloromethanes in air over the Atlantic Ocean. Chemosphere 15 (4), 429–436. Connan, O., LeCorre, P., Marty, Y., 1997. Formation of dibromomethane and bromodichloromethane by phytoplankton in the coastal waters of the Western English Channel. Comptes Rendus Acad. Sci. Ser II-A 324 (1), 25–31. Dewulf, J., VanLangenhove, H., 1997. Chlorinated C-1- and C-2hydrocarbons and monocyclic aromatic hydrocarbons in marine waters: an overview on fate processes, sampling, analysis and measurements. Water Res. 31 (8), 1825–1838. Dyrssen, D., Fogelqvist, E., 1981. Bromoform concentrations of the Arctic Ocean in the Svalbard Area. Oceanol. Acta 4 (3), 313– 317. Eisert, R., Levsen, K., 1996. Solid-phase microextraction coupled to gas chromatography: a new method for the analysis of organics in water. J. Chromatogr. A 733 (1–2), 143–157. Ekdahl, A., Pedersen, M., Abrahamsson, K., 1998. A study of the diurnal variation of biogenic volatile halocarbons. Mar. Chem. 63 (1– 2), 1–8. Fenical, W., 1981. Marine organic chemistry, evolution, composition, interactions and chemistry of organic mater in seawater. In: Duursma, E.K., Dawson, R. (Eds.), pp. 375–393, Elsevier, New York. Fenical, W., 1982. Natural-products chemistry in the marine environment. Science 215 (4535), 923–928. Fox, W.M., Connor, L., Copplestone, D., Johnson, M.S., Leah, R.T., 2001. The organochlorine contamination history of the Mersey estuary, UK, revealed by analysis of sediment cores from salt marshes. Mar. Environ. Res. 51 (3), 213–227. Gagosian, R.B., Lee, C., 1981. Marine organic chemistry, evolution, composition, interactions and chemistry of organic mater in seawater. In: Duursma, E.K., Dawson, R. (Eds.), pp. 91–107, Elsevier, New York. Gist, N., Lewis, A.C., 2006. Seasonal variations of dissolved alkenes in coastal waters. Mar. Chem. 100 (1–2), 1–10. Gowen, R.J., Mills, D.K., Trimmer, M., Nedwell, D.B., 2000. Production and its fate in two coastal regions of the Irish Sea: the influence of anthropogenic nutrients. Mar. Ecol.-Prog. Ser. 208, 51–64. Gowen, R.J., Stewart, B.M., 2005. The Irish Sea: nutrient status and phytoplankton. J. Sea Res. 54 (1), 36–50. Gribble, G.J., 2000. The natural production of organobromine compounds. Environ. Sci. Pollut. Res. 7 (1), 37–49. Gschwend, P.M., Zafiriou, O.C., Mantoura, R.F.C., Schwarzenbach, R.P., Gagosian, R.B., 1982. Volatile organic compounds at a coastal site 1 Seasonal – variations. Environ. Sci. Technol. 16 (1), 31–38. Huhn, O., Roether, W., Beining, P., Rose, H., 2001. Validity limits of carbon tetrachloride as an ocean tracer. Deep-Sea Res. Part IOceanogr. Res. Pap. 48 (9), 2025–2049. Huybrechts, T., Dewulf, J., Van Langenhove, H., 2005. Priority volatile organic compounds in surface waters of the southern North Sea. Environ. Pollut. 133 (2), 255–264. Jeffrey, S.W., Mantoura, R.F.C., Wright, S.W., 1997. Phytoplankton pigments in oceanography: guidelines to modern methods, Unesco, Paris. Jones, E.P., 1980. Air–Sea interaction (instrument and methods). In: Dobson, F., Hasse, L., Davis, R. (Eds.), Plenum Press, New York, 433–445. Krysell, M., 1991. Bromoform in the Nansen basin in the arctic-ocean. Mar. Chem. 33 (1–2), 187–197. Laturnus, F., 1996. Volatile halocarbons released from Arctic macroalgae. Mar. Chem. 55 (3–4), 359–366. Liss, P.S., Hatton, A.D., Malin, G., Nightingale, P.D., Turner, S.M., 1997. Marine sulphur emissions. Philos. Trans. R. Soc. Lond. Ser. BBiol. Sci. 352 (1350), 159–168. Liss, P.S., Watson, A.J., Liddicoat, M.I., Malin, G., Nightingale, P.D., Turner, S.M., Upstillgoddard, R.C., 1993. Trace gases and air–sea exchanges. Philos. Trans. R. Soc. Lond. Ser. A-Math. Phys. Eng. Sci. 343 (1669), 531–541.

C.M. Bravo-Linares et al. / Marine Pollution Bulletin 54 (2007) 1742–1753 Liss, P.S., Watson, A.J., Liddicoat, M.I., Malin, G., Nightingale, P.D., Turner, S.M., Upstill-Goddard, R.C., 1994. Understanding the North Sea system. In: Charnock, H., Dyer, K.R., Huthnance, J.M., Liss, P.S., Simpson, J.H., Tett, P.B. (Eds.), Chapman and Hall for The Royal Society, London, 153–163. Manley, S.L., 2002. Phytogenesis of halomethanes: a product of selection or a metabolic accident? Biogeochemistry 60 (2), 163–180. Mantoura, R.F.C., Gschwend, P.M., Zafiriou, O.C., Clarke, K.R., 1982. Volatile organic compounds at a coastal site 2 Short-term variations. Environ. Sci. Technol. 16 (1), 38–45. Marti, S., Bayona, J.M., Albaiges, J., 2001. A potential source of organic pollutants into the North-eastern Atlantic: the outflow of the Mediterranean deep lying waters through the Gibraltar Strait. Environ. Sci. Technol. 35 (13), 2682–2689. McCulloch, A., Aucott, M.L., Graedel, T.E., Kleiman, G., Midgley, P.M., Li, Y.F., 1999. Industrial emissions of trichloroethene, tetrachloroethene, and dichloromethane: reactive chlorine emissions inventory. J. Geophys. Res.-Atmos. 104 (D7), 8417–8427. McDonald, T.J., Kennicutt, M.C., Brooks, J.M., 1988. Volatile organic compounds at a coastal Gulf of Mexico site. Chemosphere 17 (1), 123– 136. Milne, P.J., Riemer, D.D., Zika, R.G., Brand, L.E., 1995. Measurement of vertical-distribution of isoprene in surface seawater, its chemical fate, and its emission from several phytoplankton monocultures. Mar. Chem. 48 (3–4), 237–244. Moore, R.M., Groszko, W., Niven, S.J., 1996. Ocean-atmosphere exchange of methyl chloride: results from NW Atlantic and Pacific Ocean studies. J. Geophys. Res. Oceans 101 (C12), 28529–28538. Mrowiec, B., Suschka, J., Keener, T.C., 2005. Formation and biodegradation of toluene in the anaerobic sludge digestion process. Water Environ. Res. 77 (3), 274–278. Mtolera, M.S.P., Collen, J., Pedersen, M., Ekdahl, A., Abrahamsson, K., Semesi, A.K., 1996. Stress-induced production of volatile halogenated organic compounds in Eucheuma denticulatum (Rhodophyta) caused by elevated pH and high light intensities. Eur. J. Phycol. 31 (1), 89–95. Neidleman, S.L., Geigert, J., 1987. Biological halogenation – roles in nature, potential in industry. Endeavour 11 (1), 5–15. Schall, C., Heumann, K.G., Kirst, G.O., 1997. Biogenic volatile organoiodine and organobromine hydrocarbons in the Atlantic Ocean from 42N to 72S. Fresenius J. Anal. Chem. 359 (3), 298–305.

1753

Shenoy, D.M., Paul, J.T., Gauns, M., Ramaiah, N., Kumar, M.D., 2006. Spatial variations of DMS, DMSP and phytoplankton in the Bay of Bengal during the summer monsoon 2001. Mar. Environ. Res. 62 (2), 83–97. Simo, R., 2001. Production of atmospheric sulfur by oceanic plankton: biogeochemical, ecological and evolutionary links. Trends Ecol. Evol. 16 (6), 287–294. Thurman, E.M., 1985. Organic geochemistry of natural waters. Kluwer Academic Publishers, Boston. Tokarczyk, R., Moore, R.M., 1994. Production of volatile organohalogens by phytoplankton cultures. Geophys. Res. Lett. 21 (4), 285–288. Tokarczyk, R., Saltzman, E.S., 2001. Methyl bromide loss rates in surface waters of the North Atlantic Ocean, Caribbean Sea, and eastern Pacific Ocean (8–45N). J. Geophys. Res.-Atmos. 106 (D9), 9843–9851. Turner, S.M., Malin, G., Nightingale, P.D., Liss, P.S., 1996. Seasonal variation of dimethyl sulphide in the North Sea and an assessment of fluxes to the atmosphere. Mar. Chem. 54 (3–4), 245–262. Turner, S.M., Nightingale, P.D., Broadgate, W., Liss, P.S., 1995. The distribution of dimethyl sulfide and dimethylsulphoniopropionate in Antarctic waters and sea-ice. Deep-Sea Res. Part Ii-Topical Studies Oceanogr. 42 (4–5), 1059–1080. van Pee, K.H., Unversucht, S., 2003. Biological dehalogenation and halogenation reactions. Chemosphere 52 (2), 299–312. Wolfe, G.V., Steinke, M., Kirst, G.O., 1997. Grazing-activated chemical defence in a unicellular marine alga. Nature 387 (6636), 894–897. Yamamoto, H., Yokouchi, Y., Otsuki, A., Itoh, H., 2001. Depth profiles of volatile halogenated hydrocarbons in seawater in the Bay of Bengal. Chemosphere 45 (3), 371–377. Yokouchi, Y., Li, H.J., Machida, T., Aoki, S., Akimoto, H., 1999. Isoprene in the marine boundary layer (Southeast Asian Sea, eastern Indian Ocean, and Southern Ocean): Comparison with dimethyl sulfide and bromoform. J. Geophys. Res.-Atmos. 104 (D7), 8067– 8076. Yokouchi, Y., Mukai, H., Yamamoto, H., Otsuki, A., Saitoh, C., Nojiri, Y., 1997. Distribution of methyl iodide, ethyl iodide, bromoform, and dibromomethane over the ocean (east and southeast Asian seas and the western Pacific). J. Geophys. Res.-Atmos. 102 (D7), 8805–8809.