Terry F.
Bidleman
*a,
Agneta
Andersson
bc,
Sonia
Brugel
bc,
Lars
Ericson
b,
Peter
Haglund
a,
Darya
Kupryianchyk
a,
Danny C. P.
Lau
b,
Per
Liljelind
a,
Lisa
Lundin
a,
Anders
Tysklind
d and
Mats
Tysklind
a
aDepartment of Chemistry, Umeå University (UmU), SE-901 87 Umeå, Sweden. E-mail: terry.bidleman@umu.se
bDepartment of Ecology & Environmental Science, UmU, SE-901 87 Umeå, Sweden
cUmeå Marine Sciences Centre, SE-905 71 Hörnefors, Sweden
dKosterhavet National Park, Länsstyrelsen I Västra Götaland, SE-452 05 Sydkoster, Sweden
First published on 10th April 2019
Marine macroalgae are used worldwide for human consumption, animal feed, cosmetics and agriculture. In addition to beneficial nutrients, macroalgae contain halogenated natural products (HNPs), some of which have toxic properties similar to those of well-known anthropogenic contaminants. Sixteen species of red, green and brown macroalgae were collected in 2017–2018 from coastal waters of the northern Baltic Sea, Sweden Atlantic and Norway Atlantic, and analyzed for bromoanisoles (BAs) and methoxylated bromodiphenyl ethers (MeO-BDEs). Target compounds were quantified by gas chromatography-low resolution mass spectrometry (GC-LRMS), with qualitative confirmation in selected species by GC-high resolution mass spectrometry (GC-HRMS). Quantified compounds were 2,4-diBA, 2,4,6-triBA, 2′-MeO-BDE68, 6-MeO-BDE47, and two tribromo-MeO-BDEs and one tetrabromo-MeO-BDE with unknown bromine substituent positions. Semiquantitative results for pentabromo-MeO-BDEs were also obtained for a few species by GC-HRMS. Three extraction methods were compared; soaking in methanol, soaking in methanol–dichloromethane, and blending with mixed solvents. Extraction yields of BAs did not differ significantly (p > 0.05) with the three methods and the two soaking methods gave equivalent yields of MeO-BDEs. Extraction efficiencies of MeO-BDEs were significantly lower using the blend method (p < 0.05). For reasons of simplicity and efficiency, the soaking methods are preferred. Concentrations varied by orders of magnitude among species: ∑2BAs 57 to 57700 and ∑5MeO-BDEs < 10 to 476 pg g−1 wet weight (ww). Macroalgae standing out with ∑2BAs >1000 pg g−1 ww were Ascophyllum nodosum, Ceramium tenuicorne, Ceramium virgatum, Fucus radicans, Fucus serratus, Fucus vesiculosus, Saccharina latissima, Laminaria digitata, and Acrosiphonia/Spongomorpha sp. Species A. nodosum, C. tenuicorne, Chara virgata, F. radicans and F. vesiculosus (Sweden Atlantic only) had ∑5MeO-BDEs >100 pg g−1 ww. Profiles of individual compounds showed distinct differences among species and locations.
Environmental significanceMarine macroalgae (“seaweeds”) are used worldwide for human consumption and animal feed. In addition to beneficial nutrients, macroalgae contain brominated phenolic compounds, some of which have toxic properties similar to those of well-known anthropogenic contaminants. Knowledge of the bromophenolic content of macroalgae is needed to understand environmental pathways, including human exposure through consumption as well as bioaccumulation by grazers and transfer through the aquatic food web. Here we report bromoanisoles and methoxylated bromodiphenyl ethers in 16 species of macroalgae from the northern Baltic and Atlantic coasts of Sweden and Norway. This is the largest survey of these compounds in macroalgae from the Nordic region and the first for the Atlantic coasts. |
Many halogenated natural products (HNPs) are found in marine macroalgae. Bromophenolic compounds are a prominent subset of HNPs, comprising bromophenols (BPs) and their transformation products bromoanisoles (BAs), hydroxylated and methoxylated bromodiphenyl ethers (OH-BDEs, MeO-BDEs) and polybrominated dibenzo-p-dioxins (PBDDs).6–12 BPs add characteristic and desirable flavors to seafood,13,14 and due to their high food quality macroalgae have been used to supplement the feed of aquacultured fish15,16 and farm animals.4 Several brominated polyphenols have antioxidant, antimicrobial, anticancer, antidiabetic and antithrombotic activities.4,17 Bromophenolic compounds enter the human diet through macroalgae consumption and from compounds bioaccumulated in seafood, and profiles of MeO-BDE congeners in human serum reflect dietary exposure.18 Toxic properties associated with bromophenolic compounds include disruption of hormone synthesis or activity (OH-BDEs and MeO-BDEs)19 and oxidative phosphorylation (OH-BDEs),20 and binding to the aryl hydrocarbon (Ah) receptor (PBDDs).21,22
Bromophenolic compounds have been identified in several macroalgae species from the Baltic Sea: Dictyosiphon foeniculaceus, Ceramium tenuicorne, Polysiphonia fucoides and Pilayella littoralis, as well as in cyanobacteria Nodularia spumigena and Aphanizomenon flos-aquae,8–11 but quantitative data have only been reported for C. tenuicorne,11,23–25D. foeniculaceus8 and N. spumigena.8 Brominated polyphenols were isolated from the red alga Vertebrata lanosa collected on the coast of Norway.26 Several simple BPs were identified in macroalgae species from the families Ceramiaceae, Delesseriaceae, Bonnemaisoniaceae, Rhodophyllaceae, Corallinaceae and Rhodomelaceae, collected on the Swedish west coast.27 One of these compounds, lanosol (2,3-dibromo-4,5-dihydroxybenzyl alcohol) was also identified in seawater. To our knowledge, this is the only report of bromophenolic compounds in macroalgae from the west coast of Sweden.
Macroalgae and cyanobacteria are important sources of bromophenolic compounds to the Baltic ecosystem and they are transferred through the food web from macroalgae to invertebrate grazers to fish.24 Several studies have reported bromophenolic compounds in Baltic mussels, which filter-feed on cyanobacteria8,10,21,28–30 and transferred from mussels to seaducks.28 Macroalgae from the Nordic coastal region are being promoted as human food, with published recipes utilizing local species.31 Consideration should be given to the bromophenolic compounds and other HNPs present in macroalgae used for human consumption, to place exposure in perspective with anthropogenic compounds.
In 2017–2018, we collected >30 species of macroalgae from the Bothnian Sea (northern Baltic), Sweden Atlantic coast (Skagerrak) and Norway Atlantic coast. Objectives are to establish analytical methods and determine the variability of bromophenolic compounds among species and locations, as a prelude to estimating the role of macroalgae in supplying bromophenolic compounds to Nordic estuaries. Here we report concentrations of some neutral bromophenolic compounds (BAs and MeO-BDEs) in an initial survey of 16 species. A major task was to compare methods of extraction for these compounds, since different methods have been applied in previous studies. Compounds with free phenolic groups (BPs, OH-BDEs) were not included at this stage because of the additional steps required for their isolation and derivatization.
Abbreviation | Group | Speciesc | Latitude N | Longitude E | Collection date | pg g−1 ww | |
---|---|---|---|---|---|---|---|
∑2BAs | ∑5MeO-BDEs | ||||||
a ∑2,4-DiBA and 2,4,6-TriBA. b ∑2′-MeO-BDE68, 6-MeO-BDE47, one tetrabromo-MeO-BDE and two tribromo-MeO-BDEs with unknown bromine substituent positions. c Nomenclature follows Algae Base (www.algaebase.org). d Stonewort. | |||||||
Bothnian Sea | |||||||
Cet | Red alga | Ceramium tenuicorne | 60.769 | 17.349 | 2017-08-24 | 3360 | 199 |
Chv | Green algad | Chara virgata | 63.414 | 19.491 | 2017-08-26 | 57 | 103 |
Clg | Green alga | Cladophora glomerata | 63.461 | 19.805 | 2017-07-08 | 591 | 56 |
Dif | Brown alga | Dictyosiphon foeniculaceus | 63.462 | 19.803 | 2017-07-08 | 324 | 61 |
Fur | Brown alga | Fucus radicans | 60.806 | 17.356 | 2017-08-24 | 6690 | 476 |
Stt | Brown alga | Stictyosiphon tortilis | 60.791 | 17.381 | 2017-08-24 | 976 | 56 |
Uli | Green alga | Ulva intestinalis | 63.461 | 19.805 | 2017-07-08 | 726 | 45 |
Skagerrak | |||||||
Asn | Brown alga | Ascophyllum nodosum | 58.868 | 11.059 | 2017-10-06 | 41000 | 396 |
Cev | Red alga | Ceramium virgatum | 58.868 | 11.059 | 2017-10-06 | 1180 | 30 |
Ful | Red alga | Furcellaria lumbricalis | 58.869 | 11.143 | 2017-10-06 | 854 | <10 |
Fus | Brown alga | Fucus serratus | 58.868 | 11.059 | 2017-10-06 | 1160 | 85 |
Fuv | Brown alga | Fucus vesiculosus | 58.868 | 11.059 | 2017-10-06 | 5570 | 253 |
Rhc | Red alga | Rhodomela confervoides | 58.869 | 11.143 | 2017-10-06 | 437 | <10 |
Sal | Brown alga | Saccharina latissima | 58.868 | 11.059 | 2017-10-06 | 1120 | <10 |
Coastal Norway | |||||||
Ac/Sp | Green alga | Acrosiphonia/Spongomorpha sp. | 65.200 | 11.933 | 2018-05-13 | 1420 | <10 |
Asn | Brown alga | Ascophyllum nodosum | 65.200 | 11.933 | 2018-05-13 | 57700 | 34 |
Fuv | Brown alga | Fucus vesiculosus | 65.200 | 11.933 | 2018-05-13 | 3970 | 52 |
Lad | Brown alga | Laminaria digitata | 65.200 | 11.933 | 2018-05-13 | 12400 | <10 |
Fig. 1 Locations for sampling macroalgae in the Bothnian Sea (BS), Skagerrak (SK) and coastal Norway (NO). See Table 1 for coordinates and sampling dates. |
SOAK 1.7 Macroalgae pieces were placed in a glass jar with a polytetrafluoroethylene-lined cap and recovery surrogates were added: 15 ng 2,4,6-tribromoanisole-d5 (2,4,6-triBA-d5) and two PBDE congeners not found in commercial mixtures: 6.0 ng 3,3′,5-tribromodiphenyl ether (BDE-35) and 24 ng 2,3′,4,4′,6-pentabromodiphenyl ether (BDE-119) (Table 2). The macroalgae were covered with 15 mL methanol (MeOH) and allowed to stand in the refrigerator at 4 °C for one week.
2,4,6-TriBA-d5 | BDE-35 | BDE-119 | ||
---|---|---|---|---|
a 15 ng 2,4,6-triBA-d5, 6.0 ng BDE-35 and 24 ng BDE-119. b SOAK 1: methanol; SOAK 2: methanol–dichloromethane; BLEND: mixed solvents. c One-way ANOVA, post hoc Tukey HSD, p < 0.05. | ||||
SOAK 1 | ||||
Mean | 75.9 | 89.7 | 76.5 | |
SD | 10.4 | 19.3 | 16.6 | |
N | 18 | 17 | 17 | |
SOAK 2 | ||||
Mean | 85.1 | 103 | 98.4 | |
SD | 10.7 | 13.3 | 14.0 | |
N | 28 | 28 | 28 | |
BLEND | ||||
Mean | 81.4 | 109 | 103 | |
SD | 9.4 | 10.3 | 8.9 | |
N | 17 | 17 | 17 | |
Significance | ||||
SOAK 1 | SOAK 2 | Y | Y | Y |
SOAK 1 | BLEND | N | Y | Y |
SOAK 2 | BLEND | N | N | N |
SOAK 2. As for SOAK 1, but using 10 mL MeOH + 5 mL dichloromethane (DCM).
Extracts from SOAK 1 or SOAK 2 were diluted with 60 mL deionized water (DIW), 3 mL saturated potassium chloride was added, and the mixture was partitioned three times with 20 mL DCM followed by 20 mL hexane. Combined organic layers were filtered through glass wool and concentrated by rotary evaporation and nitrogen blowdown into 5 mL hexane.
BLEND.32,33 Macroalgae pieces were placed in a thick-walled glass jar and the above surrogates were added. A stainless steel stick blender (Ultra Turrax T25, Janke & Kunkel GmbH, IKA Labortechnik, Staufen, Germany) was used to homogenize the macroalgae with 25 mL acetone + 10 mL hexane followed by two portions of 20 mL hexane + 3 mL diethyl ether, blending for one minute each time. The combined extracts were diluted with 60 mL DIW, 3 mL saturated potassium chloride was added and the mixture was partitioned. The organic layer was removed and the aqueous phase was extracted twice more with 20 mL hexane + 3 mL diethyl ether. Combined organic layers were filtered and concentrated as in the SOAK methods.
Of the 18 specimens, three were extracted in replicate (3–4) with each method, five were extracted once with each method and ten were extracted only with SOAK 2 (Tables 3, 4 and S1†).
Sample | Methodc | 2,4-DiBA | 2,4,6-TriBA | triU1b | triU2b | 2′-68b | 6-47b | tetraU3b |
---|---|---|---|---|---|---|---|---|
a No entry indicates compound < LOD in more than one replicate. See Table S1. b Tri-U1, tri-U2 and tetraU3 = MeO-BDEs with unknown tribromo- or tetrabromo-substituent positions. 2′-28 = 2′-MeO-BDE68, 6-47 = 6-MeO-BDE47. c SOAK 1: methanol, SOAK 2: methanol–dichloromethane, BLEND: mixed solvents. d One-way ANOVA, post hoc Tukey HSD test. | ||||||||
Cladophora glomerata | ||||||||
Mean (n = 3–4) | SOAK 1 | 118 | 519 | 16.2 | 43.1 | |||
s.d. | 10.6 | 27.9 | 0.6 | 10.1 | ||||
RSD % | 9.0 | 5.4 | 3.5 | 23.3 | ||||
Mean (n = 3–4) | SOAK 2 | 114 | 524 | 15.6 | 36.5 | |||
s.d. | 14.2 | 79.5 | 2.6 | 5.9 | ||||
RSD % | 12.4 | 15.2 | 16.5 | 16.2 | ||||
Mean (n = 3–4) | BLEND | 90.1 | 403 | 11.6 | 34.3 | |||
s.d. | 30.8 | 36.5 | 2.3 | 16.3 | ||||
RSD % | 34.2 | 9.0 | 19.5 | 47.6 | ||||
Significance (p < 0.05)d | ||||||||
SOAK 1 | SOAK 2 | N | N | N | N | |||
SOAK 1 | BLEND | N | Y | Y | N | |||
SOAK 2 | BLEND | N | Y | N | N | |||
Fucus radicans | ||||||||
Mean (n = 3–4) | SOAK 1 | 829 | 6700 | 59.4 | 153 | 58.7 | 170 | 52.6 |
s.d. | 88.7 | 828 | 10.1 | 32.0 | 3.1 | 15.3 | 20.9 | |
RSD % | 10.7 | 12.3 | 17.0 | 20.9 | 5.3 | 9.0 | 39.7 | |
Mean (n = 3–4) | SOAK 2 | 711 | 6400 | 49.7 | 131 | 66.8 | 160 | 50.2 |
s.d. | 95.3 | 928 | 7.1 | 46.7 | 11.8 | 23.4 | 16.2 | |
RSD % | 13.4 | 14.5 | 14.2 | 35.6 | 17.6 | 14.6 | 32.2 | |
Mean (n = 3–4) | BLEND | 611 | 4830 | 25.0 | 89.6 | 40.7 | 89.8 | 32.5 |
s.d. | 37.3 | 940 | 3.4 | 22.2 | 3.6 | 11.9 | 13.3 | |
RSD % | 6.1 | 19.5 | 13.5 | 24.8 | 8.8 | 13.3 | 40.9 | |
Significance (p < 0.05)d | ||||||||
SOAK 1 | SOAK 2 | N | N | N | N | N | N | N |
SOAK 1 | BLEND | Y | Y | Y | N | Y | Y | N |
SOAK 2 | BLEND | N | N | Y | N | Y | Y | N |
Fucus vesiculosus | ||||||||
Mean (n = 4) | SOAK 1 | 967 | 5520 | 58.2 | 107 | 37.1 | 85.7 | |
s.d. | 178 | 1720 | 12.7 | 21.0 | 6.8 | 32.1 | ||
RSD % | 18.4 | 31.1 | 21.7 | 19.6 | 18.5 | 37.4 | ||
Mean (n = 4) | SOAK 2 | 928 | 5260 | 58.3 | 84.7 | 25.3 | 49.3 | |
s.d. | 287 | 1920 | 20.9 | 25.6 | 5.2 | 9.8 | ||
RSD % | 30.9 | 36.5 | 35.8 | 30.2 | 20.7 | 19.9 | ||
Mean (n = 4) | BLEND | 593 | 3440 | 26.1 | 46.8 | 14.6 | 24.8 | |
s.d. | 123 | 923 | 6.3 | 10.0 | 3.3 | 7.2 | ||
RSD % | 20.8 | 26.9 | 24.2 | 21.4 | 22.9 | 29.2 | ||
Significance (p < 0.05)d | ||||||||
SOAK 1 | SOAK 2 | N | N | N | N | Y | N | |
SOAK 1 | BLEND | N | N | Y | Y | Y | Y | |
SOAK 2 | BLEND | N | N | Y | N | Y | N |
Sample | Methodc | 2,4-DiBA | 2,4,6-TriBA | triU1 | triU2 | 2′-68 | 6-47 | tetraU3 |
---|---|---|---|---|---|---|---|---|
a tri-U1, tri-U2 and tetraU3 = MeO-BDEs with unknown tribromo- or tetrabromo-substituent positions. 2′-28 = 2′-MeO-BDE68, 6-47 = 6-MeO-BDE47. b Concentration data in Tables 3 and S2. No entry indicates compound < LOD. c SOAK 1: methanol, SOAK 2: methanol–dichloromethane, BLEND: mixed solvents. d Ratios calculated from mean concentrations in Table 3. e Ratio significantly different from 1.00, t-test at p < 0.05. | ||||||||
Ascophyllum nodosum | ||||||||
SOAK 1/SOAK 2 | 1.56 | 1.32 | 0.92 | 0.92 | 1.32 | |||
BLEND/SOAK 2 | 0.91 | 0.91 | 0.47 | 0.55 | 0.76 | |||
Cladophora glomerata | ||||||||
SOAK 1/SOAK 2 | 1.03 | 0.99 | 1.04 | 1.18 | ||||
BLEND/SOAK 2 | 0.79 | 0.77 | 0.67 | 0.94 | ||||
Fucus radicans | ||||||||
SOAK 1/SOAK 2 | 1.16 | 1.05 | 1.20 | 1.16 | 0.88 | 1.06 | 1.05 | |
BLEND/SOAK 2 | 0.86 | 0.75 | 0.50 | 0.68 | 0.61 | 0.56 | 0.65 | |
Fucus serratus | ||||||||
SOAK 1/SOAK 2 | 0.49 | 0.61 | 1.04 | 1.32 | ||||
BLEND/SOAK 2 | 0.89 | 1.30 | 0.83 | 0.60 | ||||
Fucus vesiculosus | ||||||||
SOAK 1/SOAK 2 | 1.04 | 1.05 | 1.00 | 1.26 | 1.46 | 1.74 | ||
BLEND/SOAK 2 | 0.64 | 0.65 | 0.45 | 0.55 | 0.57 | 0.45 | ||
Rhodomela confervoides | ||||||||
SOAK 1/SOAK 2 | 1.34 | 0.92 | ||||||
BLEND/SOAK 2 | 0.86 | 0.93 | ||||||
Saccharina latissima | ||||||||
SOAK 1/SOAK 2 | 0.73 | 0.58 | ||||||
BLEND/SOAK 2 | 1.02 | 0.78 | ||||||
Ulva intestinalis | ||||||||
SOAK 1/SOAK 2 | 1.14 | 0.89 | 0.56 | |||||
BLEND/SOAK 2 | 1.11 | 1.23 | 0.55 | |||||
Summary | ||||||||
Mean BAs | SD BAs | Significancee | Mean MeO-BDEs | SD MeO-BDEs | Significancee | Mean BAs | SD BAs | |
SOAK 1/SOAK 2 | 0.99 | 0.29 | N | 1.12 | 0.26 | N | 0.99 | 0.29 |
BLEND/SOAK 2 | 0.90 | 0.19 | N | 0.61 | 0.13 | Y | 0.90 | 0.19 |
Internal standard (59 ng 2,2′,6,6′-tetrachlorobiphenyl, CB-54) was added to the macroalgae extracts, which were then vortexed for 1 min with 2 mL 99% sulfuric acid and placed in the refrigerator overnight to allow the phases to separate. Fucus serratus required a second sulfuric acid treatment. The upper hexane layer was transferred with hexane rinsings to a Pasteur pipet containing 0.5 g Florisil topped with 0.5 cm granular anhydrous sodium sulfate and the column was eluted with 12 mL 3:2 (v/v) DCM–hexane. The eluate was concentrated into 100 to 200 μL isooctane. Blanks were run by carrying solvents through the extraction and cleanup procedures.
Four blanks were run by carrying solvents through the extraction and cleanup procedures. No target peaks were found, and blank quantities were estimated by integrating baseline noise in the vicinity of expected peaks. Blanks (mean ± SD) by compound class were: BAs 6.3 ± 2.6 pg, tribromo-MeO-BDEs 5.7 ± 3.4 pg, tetrabromo-MeO-BDEs 22 ± 9 pg. Some BAs and MeO-BDEs were absent from all macroalgae; 2,6-diBA and 2′-MeO-BDE28, and other MeO-BDEs were absent in some of the species. “Blanks” were also assessed using a larger data set from algae extractions and quantifying baseline signal in the absence of a noticeable target peak. The overall limits of detection, LOD = (mean blank + 3 × SD)/sample weight, from both types of blanks were: BAs 15 pg g−1, tribromo-MeO-BDEs 10 pg g−1 and tetrabromo-MeO-BDEs 17 pg g−1, based on a 2.5 g (ww) sample.
Percent recoveries of 2,4,6-triBA-d5, BDE-35 and BDE-119 surrogate spikes varied among the three extraction methods, with means ranging from 76% to 109% and relative standard deviations (% RSD) ranging from 8.7% to 21.7% (Table 2). While mean recoveries of all surrogates using each method exceeded 75%, higher yields were generally obtained by SOAK 2 (MeOH–DCM) and BLEND methods than by SOAK 1 (MeOH). Differences in mean recoveries between SOAK 2 and SOAK 1 were significant for all three surrogates (single factor ANOVA, post hoc Tukey HSD, p < 0.05), between SOAK 1 and BLEND for the two BDEs and not significant between SOAK 2 and BLEND for any surrogate (p > 0.05). Concentrations of BAs and MeO-BDEs in macroalgae were corrected on a per-sample basis for surrogate yields (2,4,6-triBA-d5 recovery for BAs, average of BDE-35 and BDE-119 recoveries for MeO-BDEs).
Five other macroalgae species, U. intestinalis from the Bothnian Sea, and A. nodosum, F. serratus, R. confervoides and S. latissima from Skagerrak, were extracted once with each method. Here the extraction methods were compared by expressing yields relative to SOAK 2; i.e. SOAK 1/SOAK 2 and BLEND/SOAK 2. Results for these five macroalgae plus the three species used in methods replication (Table 3) are summarized in Table 4. Averaged over all eight species and compounds, SOAK 1/SOAK 2 yielded ratios of BAs (0.99 ± 0.29) and MeO-BDEs (1.12 ± 0.26) that were not significantly different from 1.00 (t-test, p > 0.05). BLEND/SOAK 2 ratios for BAs were slightly, but not significantly, lower (0.90 ± 0.19, p > 0.05), while BLEND/SOAK 2 ratios for MeO-BDEs were significantly lower (0.61 ± 0.13, p < 0.001). Thus, the extraction efficiency of BAs from these test species did not differ significantly by the three methods. Yields of MeO-BDEs were not significantly different by SOAK 1 and SOAK 2. The extraction efficiency for MeO-BDEs was significantly lower using BLEND, confirming results from replicate extractions of three species. For reasons of simplicity and efficiency, the soaking methods are preferred.
Six macroalgae species (A. nodosum, C. glomerata, F. radicans, F. serratus, F. vesiculosus and R. confervoides) that had been extracted by SOAK 1 or SOAK 2 received second week-long extractions with fresh solvents. Second extraction percentages were 6.6 ± 3.5% of first-extraction yields for BAs and 21 ± 19% for MeO-BDEs. The MeO-BDE percentages are likely inflated because second extraction quantities were often near or below the LOD (LOD/2 was entered).
Fig. 3 Percent contribution of BAs (top) tribromo- and tetrabromo-MeO-BDEs (bottom) to ∑2BAs and ∑5MeO-BDEs in Nordic macroalgae. See Table 1 for species abbreviations. No bar indicates all compounds < LOD. |
The bromophenolic content varied over several orders of magnitude; Σ2BAs 57 to 57700 pg g−1 ww, and Σ5MeO-BDEs < 10 to 476 pg g−1 ww (Table 1). Macroalgae standing out with ∑2BAs >1000 pg g−1 ww were A. nodosum, C. tenuicorne, C. virgatum, F. radicans, F. serratus, F. vesiculosus, S. latissima, L. digitata and Acrosiphonia/Spongomorpha sp. The species A. nodosum, C. tenuicorne, C. virgata, F. radicans and F. vesiculosus (Skagerrak only) had ∑5MeO-BDEs >100 pg g−1 ww. The Bothnian Sea and Skagerrak species tended to have higher levels of ∑5MeO-BDEs, while the ∑5MeO-BDEs was low in all four species examined from coastal Norway. Considering all species and locations, there was no significant relationship between concentrations or logarithm transformed concentrations of ∑2BAs and ∑5MeO-BDEs (p > 0.05). The proportion of 2,4,6-triBA, expressed by the fraction 2,4,6-triBA/(2,4,6-triBA + 2,4-diBA), was significantly higher (p < 0.05) in Bothnian Sea macroalgae (0.883 ± 0.109), than in the combined set of Atlantic coast species (Skagerrak and Norway, 0.747 ± 0.154) (t-test of means, unequal variance).
Nine species were checked for pentabromo-MeO-BDEs by HRMS (Table S3†). Due to poor performance of the CB54 internal standard in GC-HRMS runs, semiquantitative results were estimated without use of an internal standard and assuming a 200 μL sample volume. Only C. tenuicorne and F. radicans from the Bothnian Sea, and A. nodosum and F. vesiculosus from Skagerrak contained one or more of 6-MeO-BDE85, 6-MeO-BDE90 and 6-MeO-BDE99 at concentrations ranging from 1.3 to 18 pg g−1 ww.
Group | Genera | Location | 2,4-DiBA | 2,4,6-TriBA | 2′-MeO-BDE68 | 6-MeO-BDE47 | Reference |
---|---|---|---|---|---|---|---|
a Including stoneworts. b Ranges in a single collection period. c Ranges from June to August. d Peak concentrations, July. e East coast, near Sydney. f Stonewort. | |||||||
Brown algae | 4 | Nordic | <15–18500 | 217–48000 | 18–280 | <17–165 | This study |
Red algae | 3 | Nordic | 71–275 | 344–3290 | <17–80 | <17–56 | This study |
Green algae | 3a | Nordic | <15–507 | 57–910 | <17–47 | <17–40 | This study |
Brown algae | 3 | Philippines | <20–700 | 300–229000 | 100–27600 | 7 | |
Red algae | 7 | Philippines | <20–1300 | 100–78000 | 50–29000 | 7 | |
Green algae | 5a | Philippines | 100–2200 | <20–5900 | <20–1700 | 7 | |
Group | Species | ||||||
Brown alga | Dictyosiphon foeniculaceus | Bothnian Sea | 106 | 217 | 20 | <17 | This study |
Brown alga | Dictyosiphon foeniculaceus | Baltic Proper | 33 | 230 | 120 | 150 | 8 |
Red alga | Ceramium tenuicorne | Bothnian Sea | 71 | 3290 | 80 | 56 | This study |
Red alga | Ceramium tenuicorne | Baltic Properb | 75–100 | 160–210 | 11 | ||
Red alga | Ceramium tenuicorne | Baltic Properc | 4–70 | 47–390 | 3–29 | 5–71 | 24 |
Red alga | Ceramium tenuicorne | Baltic Properd | 85 | 40 | 23 | ||
Red alga | Ceramium virgatum | Skagerrak | 210 | 970 | 30 | <17 | This study |
Green alga | Chara virgata | Bothnian Sea | <15 | 57 | 47 | 25 | This study |
Green alga | Chara sp.f | Philippines | 300 | <20 | 100 | 7 | |
Green alga | Cladophora glomerata | Bothnian Sea | 107 | 484 | <17 | 40 | This study |
Green alga | Cladophora sp. | Philippines | 1300 | <20 | <20 | 7 | |
Red alga | Polysiphonia sphaerocarpa | Australiae | 100–300 | 200–700 | 36 |
Wide ranges of BAs and MeO-BDEs have been reported in 15 genera of red, green and brown macroalgae from Philippine waters, all collected in the same month: 2,4,6-triBA < 20 to 2200 pg g−1 ww, 2′-MeO-BDE68 < 20 to 229000 pg g−1 ww and 6-MeO-BDE47 < 20 to 27600 pg g−1 ww.7 Ranges of 2,4-diBA and 2,4,6-triBA in the red alga Polysiphonia sphaerocarpa from the littoral zone near Sydney, Australia were 100 to 300 and 200 to 700 pg g−1 ww.36 Concentrations of BAs in our set of macroalgae are generally higher than those found in other investigations (Table 5). While BAs were highest in Fucus spp. and A. nodosum, which have not been analyzed by others, we also found higher levels in C. tenuicorne than previously reported. On the other hand, our levels of BAs in the brown alga D. foeniculaceus were comparable to those found by Löfstrand et al.,8 while their concentrations of tetrabromo-MeO-BDEs were higher than ours. The semiquantitative estimate of 17 pg g−1 ww 6-MeO-BDE85 in C. tenuicorne (Table S3†) compares favorably with 35 pg g−1 ww reported in the same species.11
In addition to variation among macroalgae species,7 concentrations and proportions of bromophenolic compounds fluctuate with the season.6,11,23,24 The collection months of Nordic macroalgae specimens was July–August in the Bothnian Sea, October in Skagerrak and May in coastal Norway (Table 1), which may account for observed differences. Concentrations of 2,4-diBA and 2,4,6-triBA in C. tenuicorne, collected from the Baltic Proper between June and August, ranged from 4 to 70 and 47 to 390 pg g−1 ww, respectively, while ranges of 2′-MeO-BDE68 and 6-MeO-BDE47 were 3 to 29 and 5 to 71 pg g−1 ww (Table 5).24 Each of these bromophenolic compounds peaked in August. Ratios of compounds in C. tenuicorne varied over the spring-summer; 2,4,6-triBA/2,4-diBA from 5.5 to 17 and 6-MeO-BDE47/2′-MeO-BDE28 from 1.1 to 2.7.24 Ratios of precursor 2,4,6-triBP/2,4-diBP concentrations also showed monthly fluctuations.23 In comparison, 2,4,6-triBA/2,4-diBA was 46, while 6-MeO-BDE47/2′-MeO-BDE28 was 0.70 in our specimen of C. tenuicorne, collected in August. Stresses from salinity variations, light intensity and grazing by predators also cause changes in concentrations of 2,4,6-triBP.23 These factors hinder comparisons among species, locations and sampling times.
Both BAs and MeO-BDEs have been determined in macroalgae from Nordic ecosystems and the Philippines, with strikingly different results (Table 5). Even considering the spread of concentrations within each compound class and among species, the general trend is BAs > MeO-BDEs in Nordic ecosystems versus MeO-BDEs > BAs in the Philippines. Reasons are not apparent, and it would be interesting to conduct more comparisons between tropical and cold ecosystems.
Methods of extraction varied in the other studies. SOAK 1 as used here was employed to extract macroalgae from the Philippines.7 Variations of the BLEND method were used for Baltic specimens.8,11,24 BPs in macroalgae from Hong Kong6 and Australia12,36 were isolated by combined steam distillation-solvent extraction. Free bromophenols (BPs) were also determined in the studies listed in Table 5, with levels similar or higher than those of BAs.
The more polar BPs and OH-BDEs were not considered in this initial survey. These are also abundant in macroalgae.6–12,36 BPs and OH-BDEs are partially ionized at seawater pH,37,38 but are nevertheless subject to bioaccumulation.24,28,39 Diverse toxic effects of 2,4,6-triBP40 and OH-BDEs18–20 have been reported, and beneficial antimicrobial activity has been attributed to 6-MeO-BDE47.41 Many studies have demonstrated interconversion of BPs/BAs and OH-BDEs/MeO-BDEs through O-methylation/demethylation reactions,18,19,40 and it has been suggested that wildlife acquire OH-BDEs through O-demethylation of accumulated MeO-BDEs.42 Thus, it is important to include both free phenolic and O-methylated forms in subsequent surveys, along with additional HNPs that have been reported in marine macroalgae; e.g. PBDDs,8,9,11,21 BDEs substituted with multiple OH- or MeO-groups, hydroxylated and methoxylated bromobiphenyls7 and brominated methylpyrroles.43 The SOAK 1 (methanol) method has been used to extract compounds with free phenolic groups from macroalgae.7 Non-target screening by tandem GC-time of flight mass spectrometry (GCxGC-ToF-MS) is very effective for identifying HNPs.44 GCxGC-ToF-MS45–47 and advanced data processing techniques based on GC-MS48 and LC-MS49,50 methods have identified hundreds to thousands of natural and anthropogenic halogenated compounds in marine mammals and sediments but so far these have not been applied to identification of HNPs in macroalgae.
Concentrations of BAs and MeO-BDEs in macroalgae were highly variable, spanning orders of magnitude, and compound profiles differed according to the species. The concentrations and compound profiles observed here represent only single sampling events, and it is likely that a more extensive survey would reveal seasonal and spatial variability. Other considerations for future research are expanding the list of target compounds to include the more polar BPs and OH-BDEs, search for other HNPs by nontarget screening, and examining the link between macroalgae and bioaccumulation of bromophenolic compounds and other HNPs in Nordic coastal waters.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9em00042a |
This journal is © The Royal Society of Chemistry 2019 |