Spatial distribution of brominated very short-lived substances in the eastern Pacific


Corresponding author: Y. Liu, Department of Oceanography, Texas A&M University, College Station, TX 77843, USA. (


[1] Seawater concentrations and distributions of brominated very short-lived substances (BrVSLS), including bromoform (CHBr3), dibromomethane (CH2Br2), bromodichloromethane (CHBrCl2), chlorodibromomethane (CHClBr2), were measured in the upper water column (5–750 m) in the eastern Pacific. Inorganic nutrient, pigment concentrations, and picoplankton cell counts were measured to determine biogeochemical factors that affect the production and distribution of these BrVSLS. Elevated concentrations of BrVSLS were observed in coastal and tropical seawater. Concentration maxima for CHBr3, CH2Br2, and CHClBr2 were observed below the mixed layer, near the subsurface chlorophyll a maxima, which suggest BrVSLS production may be related to photosynthetic biomass production. Our results also suggest that heterotrophic bacteria may also contribute to CH2Br2 and CHBrCl2 production in the water column. The maximum CHBrCl2 concentration was observed at a depth much deeper than the euphotic zone, which suggests sources other than photosynthetic biomass. Elevated CHBrCl2 concentrations in deeper waters were coincident with elevated CHCl3 concentrations, which may be an evidence for successive chlorine substitution of CHBr3 in deeper and older water masses.

1. Introduction

[2] Bromoform (CHBr3), dibromomethane (CH2Br2), bromodichloromethane (CHBrCl2), chlorodibromomethane (CHClBr2), along with bromochloromethane (CH2BrCl), are the most abundant brominated very short-lived substances (BrVSLS) in the atmosphere [Montzka et al., 2011]. These compounds are important sources of inorganic bromine (Bry) to the troposphere and lower stratosphere, in addition to the longer-lived compounds such as halons and methyl bromide. Thus, these BrVSLS play an important role in catalytic ozone destruction in the atmosphere. BrVSLS mainly originate from natural sources in seawater and enter the marine boundary layer via sea-to-air fluxes [Quack and Wallace, 2003]. Anthropogenic sources, such as seawater cooled power plants, also contribute to elevated concentrations of CHBr3, CHClBr2, and CHBrCl2 observed in coastal waters; however, these sources may only be significant on a local scale [Quack and Wallace, 2003].

[3] Elevated concentrations of BrVSLS in seawater and the adjacent atmosphere are thought to be affected by macroalgae and phytoplankton production [Carpenter and Liss, 2000; Goodwin et al., 1997; Karlsson et al., 2008; Manley et al., 1992; Moore et al., 1995, 1996; Quack et al., 2007; Sturges et al., 1992, 1993]. For example, elevated BrVSLS concentrations in surface seawater have been observed near regions, where surface chlorophyll a concentrations were also elevated [Carpenter et al., 2009; Schall et al., 1997]. Although phytoplankton production has been suggested as a possible source, poor correlation between chlorophyll a and BrVSLS concentrations observed in different regions have raised questions regarding the sources and mechanisms of formation and degradation of these compounds in seawater. Therefore, many uncertainties remain concerning environmental factors that influence production and distributions of BrVSLS and the role of phytoplankton in BrVSLS production.

[4] The enzyme bromoperoxidase is thought to be responsible for the production of some polybrominated compounds, such as CHBr3 and CH2Br2, in the presence of hydrogen peroxide (H2O2) and dissolved organic matter (DOM) as the substrate [Hill and Manley, 2009; Lin and Manley, 2012; Manley, 2002; Manley and Barbero, 2001; Moore et al., 1996; Theiler et al., 1978; Wever et al., 1993]. The presence of bromoperoxidase in macroalgae and microalgae appears to be class and species-specific, which may partly explain the observed weak correlations between BrVSLS and chlorophyll a concentrations. Several studies have used pigment biomarkers to identify specific taxa of phytoplankton that may be significant BrVSLS sources in seawater [Abrahamsson et al., 2004; Karlsson et al., 2008; Mattson et al., 2012; Quack et al., 2007; Raimund et al., 2011; Roy, 2010]. For example, Quack et al. [2007] showed significant correlations between CHBr3 seawater concentrations with the abundances of cyanobacteria (Synechococcus spp.) and diatoms in the Mauritanian upwelling region. Karlsson et al. [2008] also suggested that some species of cyanobacteria in the Baltic Sea were possible sources of CHBr3. Other phytoplankton groups, such as prymnesiophytes, green algae, and prochlorophytes, may have also contributed to the production of BrVSLS, as observed by Raimund et al. [2011] near the Iberian upwelling system. However, in certain regions, such as the Antarctic, elevated diatom abundances may, in part, have contributed to the observed BrVSLS concentrations in seawater; yet, no significant correlations were found between the BrVSLS and phytoplankton pigments [Abrahamsson et al., 2004; Mattson et al., 2012].

[5] In this study, depth profiles of CHBr3 (219 samples), CH2Br2 (236 samples), CHClBr2 (238 samples), and CHBrCl2 (238 samples) were measured during the Halocarbon Air-Sea Transect-Pacific (HalocAST-P) cruise. Phytoplankton pigment concentrations, picoplankton abundance, and inorganic nutrient concentrations were measured to determine relationships between phytoplankton abundance and BrVSLS concentrations over the large spatial range covered by this cruise.

2. Sampling and Analyses Methods

[6] The HalocAST-P cruise, conducted onboard the R/V Thomas G. Thomson, departed from Punta Arenas, Chile, on 31 March 2010 and arrived in Seattle, WA on 28 April 2010 (Figure 1). The cruise track was designed to observe oceanic and atmospheric latitudinal distributions of a suite of halocarbon compounds including CHBr3, CH2Br2, CHBrCl2, and CHClBr2. A total of 24 Conductivity, Temperature, and Depth (CTD)/rosette casts were conducted along the cruise track (one cast per day at local noon) and water was collected from 12 depths ranging approximately from 5 to 750 m. Here onwards, we refer the “surface” values to measurements made at ∼5 m. For stations with water depths less than 750 m, i.e., casts 1, 2, and 3, the water columns were sampled from the surface to just above the sediment. Data collected from casts 1 and 2, where water depths were less than 200 m, are considered as coastal data. These definitions are approximately equivalent to “coastal and coastally influenced waters” defined by Carpenter et al. [2009] and “shelf regime” defined by Quack et al. [2004]. Cast 3 with water depth less than 400 m, although not coastal by our definition, was located on the continental shelf. Seawater samples were collected using a CTD/rosette sampler with 24 Niskin bottles of 10 liters. The CTD/rosette package was also equipped with an oxygen sensor, a fluorescence sensor, and a transmissometer.

Figure 1.

Sampling station map for the Halocarbon Air-Sea Transect—Pacific cruise, superimposed on March to April 2010 monthly average Sea-viewing Wide Field-of-view Sensor chlorophyll a concentrations [Acker and Leptoukh, 2007].

2.1. Halocarbons

[7] Samples for halocarbon analysis were collected using 100 mL gas-tight ground glass syringes immediately after the rosette unit was on deck. The samples were immediately transferred to a purge-and-trap sample storage module. The purge-and-trap sample storage module is a compact thermoelectric refrigerator held at ∼6°C. Inside, sixteen 70 mL glass bulbs were connected to a 16-position loop selection valve (Valco Instrument Co.), which was connected to the purge-and-trap system. This configuration allowed seawater samples to be stored in a cold, dark, and gas-tight environment until analysis [Yvon-Lewis et al., 2004]. Each seawater sample was gently injected into the appropriate bulb through a 0.2 µm membrane filter to remove microorganisms. All the samples were analyzed within 12 h of collection.

[8] For gas analysis, each seawater sample in the glass bulb was sequentially transferred into a glass sparger with an ultrapure helium gas stream, and purged at 144 mL min−1 at 40°C. The whole sample line was then flushed with the same ultrapure helium gas stream until it was time for the next sample to ensure no residue from the previous sample was left in the line. The sample purge stream was dried with two inline Nafion driers (Perma Pure, NJ), and the analytes were preconcentrated on one cryotrap and focused on a second cryotrap that were both held at −75°C, and then desorbed at 200°C. The analytes were desorbed from the focusing trap and transferred by chromatography grade helium carrier gas into the gas chromatograph-mass spectrometer for quantification (Agilent GC 6890N and MS 5973N; DB-VRX 0.180 ID column with 10 m precolumn and 30 m main column). Detailed analytical method was described in Yvon-Lewis et al. [2004] and references therein.

[9] Moist whole air gas calibration standards used in the field were calibrated before and after the cruise against gas standards obtained from the National Oceanic and Atmospheric Administration Global Monitoring Division. The field calibration standard was run after every fourth injection. The ultrapure helium gas stream was run as a blank after every three samples to monitor flushing efficiency. The detection limit for CHBr3 and CH2Br2 were 1.00 × 10−3 pmol L−1. For CHClBr2 and CHBrCl2, the detection limits were 5.00 × 10−3 pmol L−1. The averaged analytical uncertainty was 7.00% for CHBr3, 6.00% for CH2Br2, and 9.00% for CHClBr2 and CHBrCl2. Purge efficiency was determined by restripping seawater samples three times, and the percentage in total concentration for the first stripping was 72.3% for CHBr3, 72.4% for CH2Br2, 77.0% for CHClBr2, and 93.7% for CHBrCl2. BrVSLS concentrations reported here were corrected for purge efficiencies.

2.2. Pigment, Picoplankton Counts, and Nutrients

[10] Pigment samples were collected from designated Niskin bottles at 5 m, the chlorophyll maximum depth at 750 m. For each cast, 5–10 L of seawater at each depth were filtered through a precombusted 0.7 µm nominal pore size (GF/F) filter until color was visible on the filter. The filters were frozen immediately at −80°C and kept frozen until analysis back in the laboratory. The phytoplankton pigment extraction procedure was followed as described in Wright et al. [1991]. Briefly, filters were extracted ultrasonically with acetone, centrifuged, and the supernatant blown to dryness under a nitrogen stream at room temperature in the dark. The concentrated residue was redissolved in 150 µL of acetone prior to the analysis by high-performance liquid chromatography (HPLC). Pigments were analyzed using a Waters HPLC coupled to an online 996 photodiode array detector, a fluorescence detector (Shimadzu RF535) (Ex: 440 nm; Em: 660 nm), on a reversed-phase Grace Adsorbosphere C18 column (5 µm, 250 mm × 4.6 mm internal diameter), using the gradient flow described by Chen et al. [2003]. A total of 18 pigment biomarkers were analyzed, which included total chlorophyll a (chlorophyll a + divinylchlorophyll a), chlorophyll b, c2 and c3, total carotene (α + β), peridinin, 19-butanoyloxyfucoxanthin, fucoxanthin, 19-hexanoyloxyfucoxanthin, prasinoxanthin, pheophorbide a, violaxanthin, diadinoxanthin, alloxanthin, diatoxanthin, lutein, pheophytin a, and zeaxanthin. Pigment standards were purchased from DHI Water and Environment, Denmark. The pigment standards were run individually and as a mixed standard to determine retention times, spectra, and the response factors.

[11] The pigment biomarker zeaxanthin can provide information about the presence of cyanobacteria [Jeffrey et al., 2011; Wright and Jeffrey, 2006]. However, this pigment is common in both Prochlorococcus and Synechococcus, two genera of cyanobacteria that often dominate the picoplankton. To gain specific information on the potential picoplankton contribution to the BrVSLS seawater concentrations, these two genera were distinguished and enumerated using flow cytometry. In addition, flow cytometry counts were used to provide abundance information on picoeukaryotes (<3 µm algae including chlorophytes, pelagophytes, and haptophytes) and heterotrophic bacteria. Picoplankton samples were collected from the Niskin bottles, 1 mL of the sample was filtered through a 35 µm mesh cell strainer and fixed with 20 µL 10% paraformaldehyde for 10 min (final concentration 0.2%) and then quickly frozen at −80°C and kept frozen until analysis. Enumeration of picoplankton and heterotrophic bacteria was performed using a Becton Dickinson FACSCalibur flow cytometer [Campbell, 2001]. For the nonpigmented heterotrophic bacteria, the cells were stained with SYBR Green (Molecular Probes) prior to introduction into the flow cytometer. Data were analyzed using CytoWIN [Vaulot, 1989].

[12] Inorganic nutrient samples, nitrate (NO3), nitrite (NO2), orthophosphate (HPO42−), and silicate (SiO44−), were sent to the Geochemical and Environmental Research Group at Texas A&M University for analysis. Inorganic nutrient concentrations were determined using the Astoria-Pacific autoanalyzer following the methodologies of Armstrong et al. [1967] and Bernhardt and Wilhelms [1967].

2.3. Data Handling and Statistical Methods

[13] All data were checked for quality. Exceptionally high or low concentrations measured due to occasional instrumental error, which were determined from CFC-11 concentration as an inert tracer in the same sample, were removed. Subsequently, the data were tested for normality to determine the most appropriate statistical method used for data interpretation. Spearman's rank correlation was chosen to assess the association between parameters, due to the nonnormal distributions of the data. Spearman's rank correlation analyses were evaluated at 95% confidence level. Significance of the correlation was determined using a two-tailed test. Significant differences in means between data groups were determined by one-way ANOVA test at 95% confidence level. All statistical analyses were performed using the OriginLab® version 9.0 statistic module.

3. Results and Discussion

3.1. Hydrography and Water Column Layers During the HalocAST-P Cruise

[14] Based on profiles of salinity, temperature, potential density (σθ), dissolved oxygen, and CFC-11 (Figures 2-4), there were four main hydrographic features observed during the HalocAST-P cruise: (1) freshwater discharged from the coasts, (2) the entrainment of surface waters into deeper water column, (3) the Antarctic intermediate water (AAIW) in the southern hemisphere, and (4) the north Pacific intermediate water (NPIW) in the northern hemisphere. There was apparent entrainment of surface waters into the deeper water column, which was probably due to seasonal fluctuation of thermocline depths and mixing. Based on the dissolved oxygen profile, surface water entrainment reached down to ∼300 m in the southern temperate waters and to ∼500 m in the northern temperate waters. Lower salinity and elevated CFC-11 concentrations were also observed in these regions. The AAIW is characterized as low salinity (∼34.2) [Pickard, 1990] located at 26.98 < σθ < 27.80 isopycnal [Schmidtko and Johnson, 2011] in the southern hemisphere, which was ∼500 m and below in the temperate waters during the HalocAST-P cruise (Figures 2 and 3a). Near the Chilean shelf, the top of AAIW can be as shallow as ∼300 m [Palma et al., 2005]. The AAIW is also characterized as enriched in NO3 and HPO42− and low in HSiO3 concentrations (Figure 5) [Sarmiento and Gruber, 2006], and enriched in oxygen and CFC-11 [Hartin et al., 2011] (Figure 4). The NPIW is characterized as low salinity located at 26.60 < σθ < 27.40 isopycnal [Talley, 1997] in the north Pacific, which was ∼300 m and below in the temperate waters during the HalocAST-P cruise (Figures 2 and 3b). This water mass is enriched in nutrients such as NO3, HPO42−, and HSiO3 [Sarmiento and Gruber, 2006] (Figure 5), and depleted in CFC-11, with an apparent age of a few decades in the eastern boundary of the Pacific Ocean [Watanabe et al., 1994] (Figure 4).

Figure 2.

Salinity profile for the HalocAST-P cruise.

Figure 3.

Temperature-salinity (TS) diagram. (a) Below ∼500 m of casts 4 (red), 5 (green), 6 (blue), and 7 (magenta) were the Antarctic intermediate water (AAIW; marked with black square). Casts 2 and 8 were plotted in gray lines as reference of water masses not in the AAIW. (b) Below approximately 300 m of casts 21 (red), 22 (green), 23 (blue), and 24 (magenta) were the North Pacific intermediate water (NPIW; marked with black square). Casts 22 and 25 (TS only, no BrVSLS data collected for Cast 25, conducted in Puget Sound, Seattle) were plotted in gray lines as reference of water masses not in the NPIW.

Figure 4.

(a) Dissolved oxygen profile and (b) CFC-11 profile during the HalocAST-P cruise.

Figure 5.

(a) Silicate (HSiO3), (b) orthophosphate (HPO42−), and (c) nitrate (NO3) profile during the HalocAST-P cruise.

[15] To better understand BrVSLS distributions in the water column, the water column was divided into three layers: mixed layer, below mixed layer within euphotic zone, and below euphotic zone. The bottom of the mixed layer was defined following Brainerd and Gregg [1995]. The bottom of the euphotic was defined as depths, where photosynthetic active radiation is 1% of surface value, as measured at ∼5 m [Kirk, 1994].

3.2. Spatial Distributions of CHBr3, CH2Br2, CHClBr2, and CHBrCl2

[16] In the mixed layer, CHBr3, CH2Br2, CHClBr2, and CHBrCl2 concentrations were in general higher in the coastal ocean (depth ≤ 200 m) and the tropics (∼20°S to 20°N; Table 1 and Figure 6). Such a general trend is consistent with that observed in other regions [Butler et al., 2007; Carpenter et al., 2007b; Class and Ballschmiter, 1988; Liu et al., 2011; Quack and Wallace, 2003]. CHBr3 and CH2Br2 concentrations were higher below the mixed layer within the euphotic zone than those observed in the mixed layer and below euphotic zone (Table 1 and Figures 7 and 8). The maximum seawater concentrations for CHBr3, CH2Br2, and CHClBr2 were observed near the depths of subsurface chlorophyll a maxima in subtropical waters (Table 1 and Figure 7), which suggests that the production of these compounds is likely related to photosynthetic biomass production and associated biochemical processes. Such features were consistent with the report by Quack et al. [2004], who also found that CHBr3 concentration maxima in the Atlantic tropical ocean were located below the mixed layer near the chlorophyll a maxima. The authors reported subsurface CHBr3 concentration maxima ranged from 14 to 60 pmol L−1. Subsurface CHBr3 concentration maxima in the Pacific tropical ocean (∼20°S to 20°N) were within the range reported by Quack et al. [2004] (Figure 7). Patches of elevated CHBr3, CH2Br2, and CHClBr2 below the bottom of tropical euphotic zone (∼20°S to 20°N) were likely due to entrainment of water from the euphotic zone, where high concentrations of BrVSLS were produced.

Table 1. Mean (Range; Number of Samples) BrVSLS Concentrations (pmol L−1) in Coastal and Open Ocean Water Columna
  1. a

    Bold text highlighted the minimum and maximum of the data.

Coastal ocean    
Mixed layer17.46 (13.80–20.67; 13)4.79 (3.36–6.51; 13)2.59 (2.00–3.11; 13)1.69 (1.28–2.10; 13)
Below mixed layer, within euphotic zone 2.19 (1.78–2.59; 2)0.29 (0.10–0.49; 2)0.15 (0.11–0.19; 2)
Below euphotic zone 1.79 (1.62–1.96; 2)0.49 (0.46–0.52; 2)0.31 (0.11–0.52; 2)
Open ocean    
Mixed layer2.47 (0.15–11.32; 65)1.61 (0.70–3.91; 69)0.51 (0.14–1.39; 69)0.24 (0.01−0.95; 69)
Below mixed layer, within euphotic zone6.32 (0.08–31.77; 35)6.00 (1.25–24.97; 37)1.11 (0.29–3.72; 37)0.40 (0.04–2.26; 37)
Below euphotic zone2.78 (0.01−19.44; 106)1.99 (0.04−17.23; 113)0.63 (0.05−2.76; 115)0.64 (0.03–4.23; 115)
Figure 6.

Latitudinal distributions of mean mixed layer CHBr3, CH2Br2, CHClBr2, and CHBrCl2 concentrations, error bars indicate ±1 standard deviation of the mixed layer concentrations.

Figure 7.

Depth profiles of (a) CHBr3, (b) CH2Br2, (c) CHClBr2, and (d) CHBrCl2. White line marks the bottom of mixed layer, black line marks the depths of chlorophyll a maxima, and red line marks the bottom of the euphotic zone.

Figure 8.

Box plots of data grouped based on geographical setting (open ocean versus coastal ocean), and data grouped based on water column layers (mixed layer, below mixed layer within euphotic zone, and below euphotic zone), for (a and b) CHBr3, (c and d) CH2Br2, (e and f) CHClBr2, and (g and h) CHBrCl2. Vertical line in the box indicates median of data, filled square indicates mean of data, box range indicates 25–75th percentile of data, whisker indicates 10–90th percentile of data, open box indicates 5–95th percentile of data. Stars indicate minimum and maximum of data. Note that the y axis is a log scale.

[18] CHClBr2 and CHBrCl2 concentrations were higher below the euphotic zone than in the mixed layer (Figures 7 and 8), which suggest there were sources of these two BrVSLS in deeper waters. In fact, the CHBrCl2 concentration range below the euphotic zone was even higher than near the chlorophyll a maxima. CHBrCl2 concentration was greatest below the euphotic zone in the northern temperate water column (Figure 7), which suggests source strengths of CHBrCl2 in deeper waters could be higher than in the euphotic zone in certain regions (Figure 8). Our results were consistent with the observations of Moore and Tokarczyk [1993] in the Northwest Atlantic Ocean, where CHClBr2 and CHBrCl2 concentrations were higher in deeper waters compared to surface waters. The authors suggested slow successive CHBr3 chlorine substitutions to CHClBr2, then CHBrCl2 [Class and Ballschmiter, 1988; Geen, 1992], may in part explained such features, as CHBr3 produced at the upper water columns mixed and/or diffused into deeper waters.

[19] Patches of elevated CHBr3, CHClBr2, and CHBrCl2 were observed below the euphotic zone, particularly in the AAIW and NPIW (Figure 7). Other studies have reported high concentrations of BrVSLS in surface Antarctic waters. For example, concentration ranges from 5.25 to 280 pmol L−1 and 2.99 to 30.0 pmol L−1 were observed in Antarctic surface seawater for CHBr3 and CH2Br2, respectively, depending on the location of observations [Abrahamsson et al., 2004; Butler et al., 2007; Carpenter et al., 2007a; Hughes et al., 2011; Mattson et al., 2012]. As suggested by these studies, the high CHBr3 and CH2Br2 concentrations may be attributed to sources such as phytoplankton blooms and sea ice. Hydrolysis half-lives of CHBr3, CH2Br2, CHClBr2, and CHBrCl2 at 20°C are 686, 183, 137, and 274 years, respectively [Vogel et al., 1987]. Chlorine substitution half-life of CHBr3 is ∼5–74 years depending on seawater temperature [Geen, 1992]. With such long half-lives of these BrVSLS in seawater, it is not surprising that low concentrations of these BrVSLS were retained in the AAIW, as the water mass subsided from the surface where prolific sources of these BrVSLS existed. While elevated CHClBr2 and CHBrCl2 concentrations in the AAIW may be partially attributed to successive chlorine substitution of CHBr3, it is difficult to assess the magnitude of the contributions of this process to the features observed in this study.

[20] Nonetheless, it would be reasonable to speculate that the CHBrCl2 maxima and elevated CHClBr2 located in the NPIW were, at least partially, the result of successive chlorine substitution of CHBr3. The apparent age of NPIW (few decades) in this location was old enough relative to chlorine substation half-life of CHBr3 to result in such a feature. In this water mass, CHBr3 concentrations were not elevated, which may potentially suggest conversions to CHClBr2 and then CHBrCl2. Although chloroform (CHCl3) is not discussed in detail in this study, it is worth noting that CHCl3 concentrations were also elevated in the NPIW (Figure 7e), probably due to further chlorine substitution of CHBrCl2. Such features of CHCl3 in NPIW further support the hypothesis of successive chlorine substitution of CHBr3.

3.3. Sources of CHBr3, CH2Br2, CHClBr2, and CHBrCl2 in the Euphotic Zone

3.3.1. Sources of CHBr3, CH2Br2, CHClBr2, and CHBrCl2 in the Coast

[21] Elevated CHBr3, CH2Br2, CHClBr2, and CHBrCl2 concentrations were observed in Chilean coastal waters (i.e., Casts 1 and 2; Figures 6-8). Distinctly lower salinity in this region implies freshwater input from the continent (Figure 2). The southern Chilean coast, where Casts 1 and 2 were collected, is characterized as a fjord coast, dominated by islands, channels, and fjords [Fariña et al., 2008]. The subtidal kelp Macrocystis pyryfera and the intertidal kelp Durvillea antarctica are the most representative macroalgae in this region [Fariña et al., 2008; Lüning, 1990; Ríos et al., 2007]. In this region, phytoplankton assemblages have heterogeneous spatial and temporal distribution patterns, with diatoms being the most abundant phytoplankton group year round [Alves-de-Souza et al., 2008]. The giant kelp Macrocystis pyryfera has been shown to produce considerable amounts of CHBr3 and CH2Br2 [Manley et al., 1992]. Numerous laboratory studies also showed production of CHBr3 and CH2Br2 from certain diatoms, which are likely associated with bromoperoxidase activity [Hill and Manley, 2009; Moore et al., 1995, 1996; Tokarczyk and Moore, 1994]. Therefore, it is not surprising that the maximum mixed layer concentrations of CHBr3, CH2Br2, CHClBr2, and CHBrCl2 were observed in this region (Table 1 and Figure 6), which may be explained either due to in situ biologically mediated production or transport from the channels.

[22] During the BLAST-I cruise conducted in 1994, high CHBr3 and CH2Br2 concentrations, up to 90.23 and 25.83 pmol L−1, respectively, were observed in the inland passage [Butler et al., 2007], which supports the role of channels on the Chilean coast as significant sources of BrVSLS to the atmosphere. The HalocAST-P cruise track followed the coast outside of the inland passage and mixed layer CHBr3 and CH2Br2 concentrations were much lower than those observed during the BLAST-I cruise (Table 1 and Figure 6). Such a substantial difference in CHBr3 and CH2Br2 concentrations may suggest significant spatial variability in their concentrations if we assume interannual differences in CHBr3 and CH2Br2 production was not significant. The fact that CHBr3 and CH2Br2 concentrations were higher inside the inland passage may suggest more prolific sources. Moreover, unlike outside the inland passage, the restricted water circulations in the inland passage may also allow CHBr3 and CH2Br2 to accumulate. Despite the fact that the CHBr3 and CH2Br2 mixed layer concentration maxima observed in this study were lower than other studies in macroalgal-dominated regions, where extremely high concentrations of BrVSLS were observed in surface waters above macroalgal beds, the concentration patterns observed in the Chilean coast were consistent with other studies [Carpenter and Liss, 2000; Fogelqvist and Krysell, 1991; Goodwin et al., 1997; Gschwend and MacFarlane, 1986; Jones and Carpenter, 2005; Laturnus, 1996, 2001; Laturnus et al., 1996; Liu et al., 2011; Moore and Tokarczyk, 1993; Yokouchi et al., 2005]. Lin and Manley [2012] found that CHBr3 and CH2Br2 production is a function of coastal DOM characteristics (i.e., size fraction and source). Therefore, terrestrial-derived DOM carried by the channels may also influence CHBr3 and CH2Br2 concentrations observed in this region.

[23] Flow cytometry cell counts of picoplankton, which have the same sampling resolution as the BrVSLS, allowed us to examine BrVSLS biological sources in further detail. In the Chilean coastal region, CH2Br2 was significantly correlated with Synechococcus spp. and picoeukaryotes (<3 µm algae including chlorophytes, pelagophytes, haptophytes), and CHClBr2 was significantly correlated with heterotrophic bacteria (Table 2). However, CHBr3 and CHBrCl2 did not show any association with any of the picoplankton groups. Relationship between BrVSLS with phytoplankton taxonomic group derived from pigment analysis was not considered in the coastal region due to small sample size (n = 5).

Table 2. Spearman's Rank Correlation Coefficient (ρ) of BrVSLS With Picoplanktona
 Heterotrophic BacteriaProchlorococcusSynechococcusPicoeukaryotes
  1. a

    p value and number of samples (n) are presented in the parentheses. Relationships between the BrVSLS with photosynthetic picoplanktons were not considered below the euphotic zone. Below the euphotic zone, only relationships between BrVSLS and heterotrophic bacteria were considered. “ns” indicates no significant correlation.

Coastal euphotic zone    
CH2Br2nsns0.92 (<<0.001; 13)0.90 (<<0.001; 13)
CHClBr20.59 (<<0.03; 13)nsnsns
Open ocean    
Mixed layer    
CHBr3nsns0.27 (0.03; 65)ns
CH2Br20.36 (<<0.003; 65)−0.25 (0.04; 67)0.46 (<<0.001; 68)ns
CHClBr20.24 (<<0.05; 65)ns0.27 (0.02; 68)ns
Below mixed layer, within euphotic zone
CHClBr2ns0.36 (0.03; 36)nsns
CHBrCl2nsnsns0.37 (0.03; 36)
Below euphotic zone    
CH2Br20.29 (0.003; 109)   

3.3.2. Sources of CHBr3, CH2Br2, CHClBr2, and CHBrCl2 in the Open Ocean

[25] In the surface open ocean, CH2Br2 was significantly correlated with 19′-hexanoyloxyfucoxanthin (ρ = 0.52, p = 0.03, n = 17), which is the pigment biomarker for prymnesiophytes [Jeffrey et al., 2011; Wright and Jeffrey, 2006]. In the open ocean chlorophyll a maximum depths, CHClBr2 was significantly correlated with violaxanthin (ρ = 0.54, p = 0.03, n = 19), which is the pigment biomarker for green algae [Jeffrey et al., 2011; Wright and Jeffrey, 2006]. CHBr3 and CHBrCl2 were not significantly correlated with any of the pigment biomarkers. While some studies have observed correlations between CHBr3 and pigment biomarkers [Quack et al., 2007; Raimund et al., 2011], the lack of correlation between CHBr3 and pigment biomarkers is not uncommon. Abrahamsson et al. [2004] and Mattson et al. [2012] consistently found that CHBr3 did not yield significant correlations with any of the pigment biomarkers in Antarctic seawater. Such a lack of correlation between CHBr3 and pigment biomarkers may be due to differences in their turnover times in the water column [Mattson et al., 2012], short-term (diurnal) variability in CHBr3 production rate [Ekdahl et al., 1998], species specificity to CHBr3 production (which cannot be resolved by using pigment biomarkers), and/or the interplay between different phytoplankton and bacterioplankton groups.

[26] Zeaxanthin is a pigment biomarker commonly used for assessing the presence of cyanobacteria [Jeffrey et al., 2011; Wright and Jeffrey, 2006]. However, unlike the flow cytometry data, information from zeaxanthin distributions cannot distinguish important cyanobacteria taxa such as Synechococcus and Prochlorococcus. Several studies have indicated cyanobacteria as a possible source of CHBr3 [Karlsson et al., 2008; Quack et al., 2007]. Therefore, data from flow cytometry were employed to examine the relationship between the BrVSLS and picoplankton in greater detail. In addition, heterotrophic bacteria counts were also obtained, which allowed us to examine the potential influence of heterotrophic bacteria in BrVSLS biogeochemistry. However, it should be noted that heterotrophic bacteria abundance does not necessarily correlate with bacterial activity. We found that the BrVSLS were correlated with various picoplankton groups at different layers of the water column (Table 2). For example, CHBr3, CH2Br2, and CHBrCl2 were significantly correlated with Synechococcus spp. in the open ocean mixed layer (Table 2). Despite the fact that CHBr3 was only weakly correlated with Synechococcus, more detailed phytoplankton assemblage analyses could reveal relationships between CHBr3 and specific groups of phytoplankton.

[27] CH2Br2 was significantly correlated with heterotrophic bacteria in the mixed layer and below the euphotic zone (Table 2), which may indicate bacterial production. While bacterial degradation of CH2Br2 has been observed [Goodwin et al., 1998; Hughes et al., 2013], to date, no information is available on bacterial production of CH2Br2. In the open ocean mixed layer, CHClBr2 was also correlated with heterotrophic bacteria. Moore [2003] indicated bacteria are unlikely to affect CHBr3 production, which supports our observations of a lack of correlation of CHBr3 with bacteria. Our results suggest that heterotrophic bacteria may also contribute to BrVSLS biogeochemistry in terms of their destruction and production.

[28] In the open ocean water column, mean CHBr3 and CH2Br2 concentrations were comparable in the mixed layer, below the mixed layer within the euphotic zone, and below the euphotic zone. In contrast, CHBr3 concentrations in the coastal ocean mixed layer were more than three times higher than CH2Br2. While these observations may suggest differences in their formation and degradation rates in the coastal ocean and open ocean, it is also possible that different DOM characteristics were influencing such trends. Lin and Manley [2012] found that CH2Br2 was not always formed when different DOM size fractions collected from various locations in California were exposed to bromoperoxidase. This finding may indicate that terrestrially derived DOM is potentially more reactive for CHBr3 formation, compared to DOM derived in the marine environment.

3.4. Correlations Between CHBr3 and CH2Br2, CHClBr2, and CHBrCl2

[29] Despite the differences in the degree of correlations, CHBr3 was significantly correlated with CH2Br2 and CHClBr2 throughout the water column for the entire HalocAST-P cruise (Table 3). Significant correlation between CHBr3 and CH2Br2 has been frequently interpreted as the two are derived from common sources. However, CHBr3 and CH2Br2 significantly correlated with different biological parameters during the HalocAST-P cruise. For example, heterotrophic bacteria abundance only significantly correlated with CH2Br2 but not with CHBr3. Such a pattern suggests different biogeochemical factors within the ecosystem can affect the concentrations of these compounds separately. This finding is consistent with observations by Quack et al. [2007] in the Mauritanian upwelling system. A recent laboratory study by Hughes et al. [2013] found that CHBr3 and CH2Br2 concentrations in seawater were controlled by different processes. The authors hypothesized that CH2Br2 is transformed from CHBr3 via biologically mediated processes. Such a hypothesis was supported by their experimental results and other lines of evidence previously observed in the laboratory and field [Hughes et al., 2009; Tokarczyk and Moore, 1994]. Therefore, the cooccurrence of CHBr3 and CH2Br2 is potentially due to processes occurring in a common type of ecosystem and controlled by different factors, rather than their being derived from a common biological source.

Table 3. Spearman's Rank Correlation Coefficient (ρ) of CHBr3 With CH2Br2, CHClBr2, and CHBrCl2a
  1. a

    p value and number of samples (n) are presented in the parentheses. “ns” indicates no significant correlation.

Mixed layer   
CHBr30.67 (<<0.001; 78)0.81 (<<0.001; 78)0.57 (<<0.001; 78)
Below mixed layer within the euphotic zone
CHBr30.56 (<<0.001; 35)0.66 (<<0.001; 35)0.30ns (0.08; 35)
Below euphotic zone
CHBr30.70 (<<0.001; 106)0.29 (0.003; 106)0.06ns (0.56; 106)

[31] CHBr3 was only significantly correlated with CHBrCl2 in the mixed layer (Table 3). It is interesting to consider CHBrCl2 below the euphotic zone, where successive chlorine substitution of CHBr3 was proposed. CHBrCl2 was significantly correlated with CHClBr2 (ρ = 0.55, p << 0.001, n =106) but not associated with CHBr3 below the euphotic zone, which may further support the idea of chlorine substitution, as CHBrCl2 is not directly linked to CHBr3 in such a chemical process.

4. Conclusions

[32] Our findings suggest BrVSLS production is in general related to photosynthetic biomass production, and heterotrophic bacteria may be sources for certain BrVSLS, for example, CH2Br2. However, the relationship between BrVSLS production and photosynthetic biomass production is not straightforward and likely involves complex factors, such as phytoplankton species composition, interactions between phytoplankton groups and/or bacterioplankton, and enzymatic reactions with specific DOM moieties. Our results also suggest CHBr3 and CH2Br2 may be derived from disparate sources and are controlled by different biogeochemical processes in certain regions. Finally, CHBrCl2 may have more significant sources in deeper waters than biochemical sources found in the euphotic zone, and successive chlorine substitution of CHBr3 may be one of them. To better understand the source of the BrVSLS and their feedbacks under climate change, fundamental studies concerning their formation and degradation mechanisms are required. A better understanding of these mechanisms is important as most of the anthropogenic ozone depleting substances (ODS) were controlled by the Montreal Protocols, and the naturally derived ODS will be more important in controlling ozone chemistry in the future atmosphere.


[33] We thank the crew of the UNOLS R/V Thomas G. Thompson for their support during the cruise. We thank the laboratory technicians at the Geochemical and Environmental Research Group (GERG) at Texas A&M University for analysis the inorganic nutrient samples. We also thank Lei Hu and Richard Smith for their help in collecting samples during the cruise. Finally, we sincerely thank the three anonymous reviewers, who provided valuable insights and constructive comments for improving the quality of this manuscript. This work was supported by National Science Foundation grant OCE 0927874 to S.A.Y-L.