Hydroxy fatty acids in marine aerosols as microbial tracers: 4-year study on β- and ω-hydroxy fatty acids from remote Chichijima Island in the western North Pacific

Accepted Manuscript Hydroxy fatty acids in marine aerosols as microbial tracers: 4-year study on β- and ωhydroxy fatty acids from remote Chichijima Island in the western North Pacific Poonam Tyagi, Yutaka Ishimura, Kimitaka Kawamura PII:

S1352-2310(15)30110-2

DOI:

10.1016/j.atmosenv.2015.05.038

Reference:

AEA 13845

To appear in:

Atmospheric Environment

Received Date: 10 December 2014 Revised Date:

17 May 2015

Accepted Date: 18 May 2015

Please cite this article as: Tyagi, P., Ishimura, Y., Kawamura, K., Hydroxy fatty acids in marine aerosols as microbial tracers: 4-year study on β- and ω-hydroxy fatty acids from remote Chichijima Island in the western North Pacific, Atmospheric Environment (2015), doi: 10.1016/j.atmosenv.2015.05.038. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

ACCEPTED MANUSCRIPT

Hydroxy fatty acids in marine aerosols as microbial tracers: 4-year study on β- and ω-

2

hydroxy fatty acids from remote Chichijima Island in the western North Pacific

SC

RI PT

1

1

M AN U

Poonam Tyagi1,2,Yutaka Ishimura1,2, Kimitaka Kawamura2,* Graduate School of Environmental Sciences, Hokkaido University, Japan Institute of Low Temperature Science, Hokkaido University, Japan

*

AC C

EP

TE D

2

Corresponding author:

E-mail: [email protected]

Running Title: Hydroxy fatty acids in marine aerosols 3

2

ACCEPTED MANUSCRIPT 4

Abstract To better understand the long-range atmospheric transport of microbial aerosols from

6

Asia to the western North Pacific, marine aerosols were collected from Chichijima Island

7

(27°04'N; 142°13'E) on a biweekly basis during 1990–1993. These samples were investigated

8

for β– and ω–hydroxy fatty acids (FAs) as terrestrial biomarkers of Gram-negative bacteria

9

(GNB) and higher plants, respectively. The average concentrations of β–hydroxy (C8–C31) and

10

ω–hydroxy (C11–C28) FAs show pronounced seasonal variability with maxima in spring

11

(300±70 pg m-3) and winter (650±330 pg m-3), respectively. Airmass back trajectories clearly

12

indicate the continental outflow from Asia during winter to spring, whereas maritime

13

airmasses dominate in summer to autumn over Chichijima. It is noteworthy that atmospheric

14

abundances of β–hydroxy FAs and, thus, the estimated mass concentration of GNB have not

15

been significantly varied between polluted (continental) and pristine (oceanic) airmasses

16

during the study period. However, the relative source strength observed from cluster analysis

17

of β–hydroxy FAs in the polluted continental airmassess vary significantly among seasons

18

(winter: 98%, spring: 63%, summer; 11%, autumn: 26%). In addition, there were

19

distinguishable differences between polluted continental and pristine maritime airmasses with

20

regard to C-number predominance. The even C-number predominance of β– and ω–hydroxy

21

FAs (~80 and 98% of total mass concentration, respectively) in marine aerosols could be due

22

to their significant contribution from GNB, terrestrial plants and soil microorganisms. These

23

results have implications towards assessing the atmospheric transport of bacterial and plant

24

lipids in the continental outflow over the open ocean.

25

Keywords: β− and ω−hydroxy fatty acids, terrestrial biomarkers, marine aerosols, North

26

Pacific, East-Asian outflow, Gram-negative bacteria.

27

AC C

EP

TE D

M AN U

SC

RI PT

5

3

ACCEPTED MANUSCRIPT 28

1. Introduction A wide range of hydroxy fatty acids (FAs) are found in lipid fractions isolated from a

30

variety of organisms (Downing, 1961) including bacteria (Wilkinson, 1988), algae (Blokker

31

et al., 1998) and higher plants (Pollard et al., 2008; Molina et al., 2006). Several studies have

32

been carried out to assess the abundances of hydroxy FAs in environmental samples such as

33

lacustrine and marine sediments (Eglinton et al., 1968; Cardoso and Eglinton, 1983;

34

Kawamura and Ishiwatari, 1984 a,b; Kawamura and Ishiwatari, 1982) and lake waters

35

(Kawamura et al., 1987) as well as marine aerosols (Kawamura, 1995). Despite having their

36

wide spread occurrence, unfortunately these compounds have received little attention from

37

atmospheric chemists. In general, hydroxy FAs can be categorized into different classes on

38

the basis of the number and position of the hydroxyl (OH) group. In the literature, the

39

position of OH group in hydroxy FAs reported so far, are mainly limited to α–, β–, C–9, (ω–

40

l)–, and ω–positions. Among these, aliphatic α−, β– and ω– monohydroxy FAs have been

41

used as tracers of soil microorganisms (Kawamura, 1995), bacteria (Lee et al., 2004) and

42

terrestrial plants (Kawamura et al., 2003) in ambient aerosols, respectively. Likewise,

43

hydroxy FAs have been used as geochemical tracers to evaluate microalgal (Gelin et al.,

44

1997), bacterial and terrestrial plant contributions to sediments (Cordoso and Eglinton, 1983).

45

Long chain β–hydroxy FAs (C10–C18) are important structural constituents of

46

lipopolysaccharides (LPS) and lipid A, especially in the cell wall of Gram-negative bacteria

47

(GNB). These β–hydroxy FAs are essentially associated with bacterial endotoxin activity

48

(Wilkinson, 1988). Studies have been performed for the quantification and characterization of

49

endotoxins and GNB in ambient aerosols by using β–hydroxy FAs as biomarkers (Lee et al.,

50

2004; Hines et al., 2003; Laitinen et al., 2001). Similarly, long chain aliphatic β–hydroxy

51

FAs have also been reported as intermediates in the β−oxidation of monocarboxylic acids by

AC C

EP

TE D

M AN U

SC

RI PT

29

4

ACCEPTED MANUSCRIPT 52

microorganisms in the old lacustrine sediments from the English Lake district (Eglinton et al.,

53

1968). It has been suggested that ω–hydroxy FAs (C11–C28) are the pivotal structural

55

components of cell wall of plants (Pollard et al., 2008; Molina et al., 2006) and green algae

56

(Blokker et al., 1998). A possible bacterial source was also identified by Skerratt et al. (1992)

57

who found C26, C28 and C30 (ω–1)−hydroxy FAs in methane utilizing bacteria. Recent

58

scientific work has suggested that some microalgae can be a potential source of mid-chain

59

hydorxy FAs (C30–C34) in marine environments (Gelin et al., 1997). Thus, these compounds

60

can also be used as biomarkers for soil microorganisms and higher plant metabolites in

61

ambient aerosols. Rather few studies exist in the literature that focuses on tracing the

62

pathways of atmospheric hydroxy FAs in remote marine aerosols (e.g., Kawamura, 1995).

63

However such studies are, in particular, useful in understanding the changes in atmospheric

64

circulation over the geological past and to assess the long-range atmospheric transport of

65

microorganisms along with the continentally derived particulate matter.

TE D

M AN U

SC

RI PT

54

Several studies have documented the impact of Asian outflow to the western North

67

Pacific during winter and spring in delivering the airborne particulates and trace gases.

68

However, to the best of our knowledge, no single study exists in the literature for assessing

69

the long-range atmospheric transport of soil microbes in the continental outflow from Asia to

70

the open ocean waters of the North Pacific. In this regard, we aim to trace the presence of soil

71

microbes in aerosols collected from the Asian outflow in the western North Pacific using

72

certain biomarkers. The present study was carried out over remote marine island, Chichijima,

73

located in the western North Pacific to observe the changes in the long-range atmospheric

74

transport of bacterial tracers to the ocean.

AC C

EP

66

75

In this study, we made an attempt to use hydroxy FAs as a proxy to understand the

76

linkage between the terrestrial components and those in the oceans and pelagic sediments. In

5

ACCEPTED MANUSCRIPT

a previous study, Kawamura et al. (2003) have documented the chemical compositions of

78

marine aerosols collected from a remote island, Chichijima during 1990–1993. We have

79

investigated these aerosol samples for the atmospheric abundances of hydroxy FAs in order

80

to assess the long-range transport of terrestrial microbes.

81 82

2. Material and methods

83

2.1. Aerosol collection and prevailing meteorology

RI PT

77

Aerosol samples were collected on a biweekly basis from April 1990 to November

85

1993 at the Ogasawara downrange station of the Japan Aerospace Exploration Agency

86

(JAXA, elevation: 254 m a.s.l.) in Chichijima (27°04'N; 142°13'E) in the western North

87

Pacific (Figure 1). For this study, bulk aerosol (TSP) samples (N=69) were collected using a

88

high volume air sampler and quatz filter with a nominal flow rate of 1.0 m3 min-1 (Kawamura

89

et al., 2003). In general, the collection period for each aerosol sample varied in between 4

90

and 6 days. The air sampler was setup on the top (5 m above the ground) of the base of the

91

parabola antenna for the satellite tracking. During the study period, the relative humidity

92

changed from 60–90% (av. ~78%). In general, the sampling site is characterized by

93

prevailing westerlies in winter and spring, whereas trade winds dominate in summer and

94

autumn seasons.

M AN U

TE D

EP

AC C

95

SC

84

All samples used in this study were collected on a pre-combusted (500°C, 3 hours)

96

quartz fiber filter (PALLFLEX 2500QAT-UP, 20 x 25 cm2). After collection, quartz filters

97

were placed into a precleaned glass jar equipped with Teflon liner and screw cap, transported

98

to the laboratory and stored at – 20°C until the analysis of chemical constituents. Prior to the

99

chemical analyses, samples were kept in the desicator for near about overnight to remove the

100

moisture content. Owing to technical difficulties arose in the high volume air sampler,

101

aerosol samples were not collected in winter and autumn seasons of year 1990–91 and 1993–

6

ACCEPTED MANUSCRIPT 102

94. Overall, we have categorized the each year samples into respective seasons, winter

103

(December–February),

104

(September–November).

105

2.2. Sample preparation for the quantification of hydroxy FAs

spring

(March–May),

summer

(June–August)

and

autumn

The analytical protocol used for assessing the atmospheric abundances of hydroxy

107

FAs is adapted from Kawamura et al. (2003). Briefly, an aliquot of filter was extracted with

108

methanolic KOH under reflux (~2 hrs) and the neutral fraction was seperated by extraction

109

with n-hexane/methylene chloride (10:1) mixture. Then, the acid fraction was isolated with

110

methylene chloride (i.e., CH2Cl2) after the acidification of the remaining solution with 6M

111

HCl. The isolated acids were converted to methyl esters using 14% BF3 in methanol

112

(SUPELCO Analytical). Soon after, methyl ester derivatives of FAs (FAMEs) were further

113

seperated into three subfractions using silica gel column chromatography. The third fraction

114

(A-3) containing hydroxy FAs was then derivatized with N,O-bis-(trimethylsilyl)

115

trifluoroacetamide (BSTFA) (SUPELCO Analytical) to derive the OH groups to trimethyl

116

silyl (TMS) ethers. After reaction, 50 µL of hexane containing 1.43 ng µL-1 of internal

117

standard (C13 n–alkane) was added to dilute the derivatives prior to GC/MS injection. The

118

extraction of the samples and subsequent separation into ester fractions were completed in

119

1998 and the A-3 fractions were stored at − 20 °C in a glass vial with a Teflon-lined screw

120

cap.

SC

M AN U

TE D

EP

AC C

121

RI PT

106

For the quantification of hydroxy FAs, GC/MS system (Hewelett-Packard, model

122

6890; GC coupled to Hewelett-Packard model 5973 mass-selective detector, MSD) was used.

123

The GC was equipped with a split/splitless injector and a DB-5MS fused silica capillary

124

column. The GC oven temperature was programmed from 50 °C (2 min) to 305 °C (15 min)

125

at a rate of 5 °C min-1. The mass spectrometer was operated in an electron impact mode at 70

126

eV and scanned from 50 to 650 dalton. After the analysis, the data were acquired and

7

ACCEPTED MANUSCRIPT

processed with the Chemstation software. Structural identification of hydroxy FAs was

128

performed using the TMS derivatives of authentic nC12, nC14, nC15 and nC16 β–hydroxy FAs

129

and nC16, nC20 and nC22 ω–hydroxy FAs, which were also used as external standards for the

130

comparison of retention times. The recoveries of authentic hydroxy FAs standards were

131

confirmed to be better than 80%. Field blanks were analyzed as the real samples for quality

132

assurance. Target compounds were not found in the blanks.

134

3. Results and discussion

135

3.1. Air mass back trajectory analysis

SC

133

RI PT

127

In order to assess the sources that contribute to aerosol composition over Chichijima,

137

7–day isentropic air mass backward trajectories (AMBTs) were computed for the sampling

138

days using HYSPLIT (Hybrid Single Particle Lagrangian Integrated Trajectory Model,

139

version 4; Draxler and Rolph, 2014; Rolpf, 2014) and NCEP-reanalysis data from NOAA air

140

resources laboratory for meteorological parameters. Figure 2 depicts the AMBT clusters for

141

the sampling days over Chichijima during 1990−1993. From this figure, it is implicit that the

142

generalized pattern of back trajectories show a seasonal shift in the wind regimes with the

143

impact of continental outflow over the sampling site during winter and spring in comparison

144

with the dominance of maritime airmasses that prevail in summer and autumn. However,

145

occasional shift in wind regimes is noteworthy for the airmasses that are also coming from

146

continents due to reverse shifting in air circulation during summer and autumn (see

147

supplementary information).

AC C

EP

TE D

M AN U

136

148

Previous studies suggested that Chichijima Island can be a suitable site to investigate

149

the mineral dust transport from East-Asia along with the polluted air masses from China

150

(Chen et al., 2013; Yamamoto et al., 2011; Mochida et al., 2003). It is well known that East

151

asian dust is very common during late winter and spring (Wang et al., 2000; Lin et al., 2001;

8

ACCEPTED MANUSCRIPT 152

Chun et al., 2001; Murayama et al., 2001) and originate in arid and semi-arid regions in

153

northern China, Mongolia, and Central Asia under high surface wind conditions (Duce et al.,

154

1980). In addition to dust transport, previous studies reported on terrestrial lipid compounds

156

(Kawamura et al., 2003), dicarboxylic acids (Mochida et al., 2003), levoglucosan and

157

saccharides (Chen et al., 2013) in marine aerosols collected from Chichijima. However, these

158

studies have focussed mainly on changes in the atmospheric abundances and molecular

159

compositions with respect to the seasons (Gagosian and Peltzer, 1986). No studies have been

160

performed to understand the seasonal patterns of bacterial transport by continental outflows

161

to the North Pacific region. Bacteria in marine ecosystem are significantly and positively

162

correlated with phytoplankton biomass (Linley et al., 1983; Bird and Kalff, 1984; Cho and

163

Azam, 1990; Li et al., 2004). It has been suggested that bacteria mainly contribute to

164

insoluble organic fraction of ambient aerosols and their concentrations found to be lower in

165

marine atmospheric boundary layer (MABL) than those over continents (Prospero et al.,

166

2005; Harrison et al., 2005; Griffin et al., 2006). Subsequent studies have shown that

167

concentrations of culturable bacteria in marine air are closely related to dust concentrations,

168

suggesting their sources to be dust plumes in non-biologically active regions (Prospero et al.,

169

2005; Griffin et al., 2006). So far, there is no study reporting the total bacterial concentration

170

in remote marine air. In the present study, we attempted to assess the bacterial mass over the

171

remote island (Chichijima) on the basis of their hydroxy FAs distribution.

173

SC

M AN U

TE D

EP

AC C

172

RI PT

155

3.2. Concentrations of hydroxy fatty acids

174

In this study, the measured β–hydroxy FAs (C8−C31) were classified into three groups

175

(i.e., C8–C9, C10–C18 and C19–C31) and ω–hydroxy FAs (C11−C28) into two groups (i.e., C11–

176

C14 and C16–C28) based on the chain length of the components. In addition, we also present a

9

ACCEPTED MANUSCRIPT

comprehensive discussion on the contributing source regions of these species in the long-

178

range atmospheric transport from East-Asia. A brief statistical summary regarding range,

179

mean and standard error of the concentrations of β–hydroxy (C8–C31) and ω–hydroxy (C11–

180

C28) FAs are presented in Table 1. From this table, it is noteworthy that all the even carbon

181

numbered β–hydroxy FAs show high concentrations than odd carbon ones and follow the

182

hierarchy (C12 > C10 > C8 > C16 > C18 > C14). Among the odd carbon numbered β–hydroxy

183

FAs, high concentrations (29±2 pg m-3) were found only for C9. However, the concentrations

184

of measured β–hydroxy FAs vary over a wide range (0–557 pg m-3). To the best of our

185

knowledge, rather poor studies exist in the literature on the quantification of β–hydroxy FAs

186

in remote marine aerosols. Although previous studies have focused on atmospheric mass

187

concentrations of hydroxy FAs, no single study is carried out over the open ocean in order to

188

assess the long-range transport of soil derived bacterial species.

M AN U

SC

RI PT

177

Table 1 also reports the statistical summary for ω–hydroxy FAs (C11–C28). The

190

dominance of even carbon numbered ω–hydroxy FAs (C16 > C24 > C22 > C14 > C12) is also

191

clearly apparent from the concentration data during the study peirod. Among the odd carbon

192

numbered ω–hydroxy FAs, C13 was predominant (10.4±5.6 pg m-3). In this study, the sum of

193

mass concentrations of measured ω–hydroxy FAs varied from 0.03 to 2500 pg m-3. As

194

mentioned earlier, not many studies exist in the literature that documents the atmospheric

195

mass concentration of ω–hydroxy FAs over the open oceans. However, Kawamura (1995)

196

reported the presence of ω–hydroxy FAs (C12 –C30) in one marine aerosol sample (collected

197

during August 1-3, 1989) whose concentration was ~900 pg m-3. It is interesting to note that

198

atmospheric concentrations of ω–hydroxy FAs (C11–C28) in summer from this study (av.

199

450±250 pg m-3) are somewhat consistent with that reported in Kawamura (1995). In our

200

remote marine aerosol samples from the western North Pacific, ω–hydroxy FAs (av. C11–C28;

AC C

EP

TE D

189

10

ACCEPTED MANUSCRIPT 201

480±120 pg m-3) are more abundant than β–hydroxy FAs (C8–C31) (230±200 pg m-3), being

202

consistent with the previous study by Kawamura (1995).

203

3.3. Relative abundances of hydroxy fatty acids Figure 3 shows the relative abundances of β– and ω–hydroxy FAs with respect to C

205

chain number in the marine aerosols collected over the Chichijima during 1990–1993. The

206

molecular distributions of β–hydroxy FAs are characterized by even C predominance (80%)

207

with a maxima at C12 (19%) followed by C10 (17%), C16 (14%) and both C18, C14 (9%)

208

(Figure 3). Among the odd carbon numbered β–hydroxy FAs, C9 (9.7%) is the only dominant

209

species. Saraf et al. (1997) suggested that the classification of β–hydroxy FAs on the basis of

210

C number could provide crucial information regarding the Gram-negative bacterial species,

211

owing to the fact that the contributions of individual β–hydroxy FAs of their LPS vary among

212

different species. Thereafter, subsequent studies have documented that C10–C18 β–hydroxy

213

FAs can be used as bacterial biomarkers in aerosols (Lee et al., 2004; Hines et al., 2003;

214

Wilkinson, 1988). Based on the preliminary analyses, the relative abundances of Σβ–hydroxy

215

FAs (C10 to C18) in their total mass concentration (ΣC8–C31 β–hydroxy FAs) are found to be

216

high (~69%) in marine aerosols sampled over the western North Pacific. This observation

217

clearly emphasizes the applicability of β–hydroxy FAs as a biomarker for assessing their

218

atmospheric abundances of GNB in marine aerosols.

SC

M AN U

TE D

EP

AC C

219

RI PT

204

We found that the relative abundance of β–hydroxy C8–C9 FAs to be ~16.7%, which

220

can be explained due to their contribution from photochemical and bacterial oxidation of

221

higher molecular weight FAs during long-range atmospheric transport (Kawamura, 1995).

222

However, the relative abundance of β–hydroxy C19-C31 FAs in their total mass concentration

223

found to be minimum (~7%) in marine aerosols during the study period (1990–1993). The

224

relatively low atmospheric abundances over the sampling site may be due to the fact that

11

ACCEPTED MANUSCRIPT 225

these higher molecular weight β–hydroxy FAs might have undergone intense photochemical

226

oxidation during the long-range transport of terrestrial plant metabolites. Analogous to β−hydroxy FAs, molecular distributions of ω−hydroxy FAs are

228

characterized by the even carbon predominance with their relative abundance of sum

229

concentrations is found to be as ~98% of total ω−hydroxy FAs. However, the relative

230

abundance of measured ω−hydroxy FAs (C11–C28) follows the order with characteristic

231

dominance of C16 (56%) and to a lesser extent followed by C14 (14.5%), C24 (10.5%), C22

232

(7%) and C12 (6%). In a previous study, Kawamura (1995) has also documented the even

233

carbon number predominance of ω–hydroxy FAs (C12−C30) with a peak of C16 or C22 in

234

marine aerosols. It has been suggested by several studies that a strong even-to-odd carbon

235

preference is indicative of biogenic sources from lipid residues of microflora (bacteria, algae,

236

fungi, etc.) and epicuticular waxes of vascular plants (Simoneit, 1989; Rogge et al., 1993). It

237

is implicit from Figure 3 that ω–hydroxy FAs with odd C numbers are insignificant and were

238

not detectable in most of the samples except for C13, whose relative abundance is on average

239

~1%. However, relative abundances of ω–hydroxy C11−C14 and C16−C28 FAs in their total

240

mass concentration are around 22 and 78%, respectively.

EP

TE D

M AN U

SC

RI PT

227

Relatively high concentrations of ω–hydroxy C16–C28 FAs in marine aerosols is due to

242

their omnipresence in cutin and suberin of terrestrial plants (Pollard et al., 2008; Molina et al.,

243

2006; Graca et al., 2006). In contrast, the low-molecular weight ω–hydroxy FAs (C11–C14)

244

are probably produced by ω–oxidation of corresponding monocarboxylic acids present in soil

245

particles. These ω–hydroxy FAs are probably associated with soil dust particles and have also

246

been reported to be present in terrestrial higher plants (Eglinton et al., 1968) and in lacustrine

247

and marine sediments (Cardoso and Eglinton, 1983; Kawamura and Ishiwatari, 1984a,b).

248

Overall, we found higher abundances of ω–hydroxy FAs than β–hydroxy FAs in marine

249

aerosol samples during the study period. This could be due to the reason why microbes prefer

AC C

241

12

ACCEPTED MANUSCRIPT 250

the pathway of ω–oxidation of fatty acids and other lipids present in soils (Kawamura, 1995).

251

It is very likely that these hydroxy FAs are transported to the open ocean during long-range

252

atmospheric transport along with soil particles during the study period.

253

3.4. Seasonal variability of hydroxy fatty acids The seasonal variability of average atmospheric relative abundances among the

255

three groups of β–hydroxy FAs (C8–C9, C10–C18 and C19–C31) for all the four years (1990–

256

1993) is shown in Figure 4. The relative abundance of sum of concentrations of lower

257

molecular weight (LMW) β–hydroxy (C10–C18) FAs dominate (~63–82%) the total mass

258

(ΣC8–C31 β–hydroxy FAs) in all seasons, followed by the sum of atmospheric abundances of

259

C8 and C9 (7–36%). However, rather insignificant differences were observed for ΣC10–C18

260

during all seasons. It has been suggested that LMW β–hydroxy FAs (C10-C18) could be used

261

as a proxy for assessing the atmospheric abundances of endotoxin and the GNB (Lee et al.,

262

2004; Hines et al., 2003). However, Wilkinson (1988) suggested that the outer membrane of

263

GNB comprised of LPS which contributes primarily for the atmospheric abundances of even

264

carbon numbered β–hydroxy FAs (C10-C18). Therefore, we have estimated the relative

265

abundances of even carbon numbered β–hydroxy FAs from C10 to C18 and their temporal

266

trend during the sampling days, as shown in Figure 5.

EP

TE D

M AN U

SC

RI PT

254

It is noteworthy that relatively high atmospheric concentrations of even carbon

268

numbered β–hydroxy FAs (C10-C18) were observed in winter and spring during the study

269

period. This observation perhaps indicates their contribution from soil-derived bacteria in the

270

continental outflow from East Asia. Although differences persist for Σ(C8–C9) β–hydroxy

271

FAs during sampling days, the seasonal variability is not clear and therefore could not be

272

interpreted to assess their sources. It has been hypothesized that C8 and C9 β–hydroxy FAs

273

could originate from the photochemical oxidation of unsaturated fatty acids such as oleic acid.

274

Analogous to ΣC10–C18, the relative abundances of ΣC19–C31 (range:1–12 %) exhibit little

AC C

267

13

ACCEPTED MANUSCRIPT

variability among all the seasons during the study period. To further check the seasonal

276

variability in β–hydroxy FAs (C8–C31) over the four years, we combined the average

277

concentrations of all the seasons over the study period. We observed maximum average

278

concentrations in spring (297±66.4 pg m-3) followed by summer (222±49.6 pg m-3), winter

279

(220±40 pg m-3) and autumn (160±27.8 pg m-3). We found the similarity in atmospheric

280

abundances of β–hydroxy FAs among four seasons which could be explained by AMBTs

281

(see supplementary information). During winter and spring, airmasses were transported from

282

continents whereas during summer and autumn, most airmasses were originated in ocean but

283

sometimes they came from continents due to the reverse shifting in atmospheric circulation.

284

Both continental and oceanic airmasses contribute equally to the relative abundances of β–

285

hydroxy FAs, which clearly indicate that both polluted (continents) and pristine (ocean)

286

ecosystems are having microorganisms. It has been suggested that Gram-negative bacteria

287

are abundant in seawater and their cell membrane is comprised of long-chain fatty acids with

288

OH groups positioned at β− and ω− (Giovannoni and Rappe, 2000). Likewise, the analysis of

289

lipid biomarkers in the ultra-filtered seawater samples from the equatorial Pacific Ocean and

290

North Sea revealed the presence of C12−C20 fatty acids and C10−C18 β−hydroxy fatty acids

291

(FAs), which are the constituents of marine bacterial membrane (Wakeham et al., 2003).

292

Another study by Mochida et al. (2002) had documented an increase in atmospheric

293

abundances of saturated fatty acids (C14−C20) during a summer cruise conducted in the North

294

Pacific, for which the surface waters are characterized by high primary productivity. Since

295

bacterial cells contain β− and ω−hydroxy FAs, they can be co-emitted with sea spray from

296

the productive ocean surfaces to the air like fatty acids. Taken together, we argue that

297

apparent equal concentrations of β− and ω−hydroxy FAs (i.e., comparison with winter) in the

298

pristine oceanic air masses during summer can be ascribed to their marine source (e.g.,

299

marine bacteria) via sea-to-air emissions.

AC C

EP

TE D

M AN U

SC

RI PT

275

14

ACCEPTED MANUSCRIPT

Figure 6 depicts the bar graph, representing the seasonally averaged (mean±S.E.)

301

relative abundances of ω–hydroxy FAs (C11–C14 and C16–C28) in marine aerosols collected

302

over the western North Pacific. As mentioned earlier, the measured ω–hydroxy FAs (C11 to

303

C28) have been further divided into two categories, C11–C14 and C16–C28. It has been

304

suggested that latter group is primarily originated from terrestrial plant waxes (Molina et al.,

305

2006; Graga et al., 2006). The relative abundances of ω–hydroxy C16–C28 FAs are somewhat

306

higher in spring than summer during 1990–1993. The observed high relative abundance of

307

ΣC16–C28 ω–hydroxy FAs in spring season can be explained based on their AMBTs (see

308

Figure 2), which exhibit long-range atmospheric transport from East Asia. We, therefore,

309

attribute high relative abundances of ΣC16–C28 ω–hydroxy FAs in spring to enhanced

310

contribution from terrestrial higher plant or soil organic matter. In contrast, the relative

311

abundances of ΣC11–C14 ω–hydroxy FAs show similar distribution in summer and spring,

312

perhaps indicating the contribution of these species from biogenic fatty acids from the ocean

313

as well as from the continental surface.

TE D

M AN U

SC

RI PT

300

From Table S3, it is obvious that no significant differences in the relative abundances

315

were observed for the three classes of β−hydroxy FAs (C8−C9, C10−C18, and C19−C31) among

316

all seasons. In contrast, the one-way ANOVA analysis has revealed significant differences

317

with regard to the relative abundances of the two classes of ω−hydroxy FAs (C11−C14 and

318

C16−C28). Therefore, we have conducted post-hoc tests to further ascertain which seasons

319

show similarity for the observed relative abundances of ω−hydroxy FAs and which are not

320

(Table S4). In case of C11−C14 ω−hydroxy FAs, relative abundances in winter are not

321

significantly different than those in spring and summer, whereas higher than those

322

documented in autumn. Likewise, relative abundances of C11−C14 ω−hydroxy FAs in spring

323

season are consistent with those reported for autumn as both are transition periods. As

AC C

EP

314

15

ACCEPTED MANUSCRIPT

mentioned earlier, the strength of the East-Asian outflow decreases from winter to summer

325

over the sampling site. However, in both spring and autumn, we found a weak continental

326

influence from the East-Asian outflow, explaining the similar relative abundances of C11−C14

327

ω−hydroxy FAs observed between both seasons. On the other hand, relative abundances of

328

C11−C14 ω−hydroxy FAs in summer are consistent with those reported in winter, but

329

significantly different than those documented for spring and autumn. In spite of different

330

origin of AMBTs for winter (i.e., continental origin) and summer (i.e., marine origin), almost

331

near equal relative abundances of C11−C14 ω−hydroxy FAs suggest their ubiquitous

332

occurrence in terrestrial vegetation, soil microbes as well as marine algae and bacteria

333

(Wakeham et al., 2003).

M AN U

SC

RI PT

324

In case of C16−C28 ω−hydroxy FAs, relative abundances in winter are significantly

335

lower than those in spring and autumn, whereas they are consistent with those reported for

336

summer. Likewise, relative abundances in spring are significantly higher than those obtained

337

in winter and summer, while they overlap with those occurred in autumn. From these

338

observations, it is implicit that the transition periods (spring and autumn) have higher

339

atmospheric abundances of C16−C28 ω−hydroxy FAs than winter and summer. The East-

340

Asian outflow is associated with the convective dust storm events during spring season

341

(March-May) (Matsumoto et al., 2003). It has been suggested that soil microbes and can be a

342

source of ω−hydroxy FAs (Zelles, 1999 and references therein). Therefore, we believe that

343

observed higher relative abundances of C16−C28 ω−hydroxy FAs in spring are due to

344

enhanced contribution of soil microbes associated with dust events. However, it is not clear at

345

this moment for the observed higher abundances for autumn.

AC C

EP

TE D

334

346

Simoneit (1989) noted that fatty acids < C20 are derived primarily from microbial

347

sources whereas > C20 are associated largely with plant wax. As there is not much variability

348

in atmospheric relative abundances of ω–hydroxy FAs among seasons with respect to years,

16

ACCEPTED MANUSCRIPT

we computed the combined seasonal average concentrations of C11–C28 ω–hydroxy FAs over

350

the four years (1990−1993). We observed clear seasonal trend with winter maxima (648±332

351

pg m-3) followed by spring (513±252 pg m-3), summer (449±236 pg m-3) and autumn

352

(298±123 pg m-3) for the years of 1990 to 1993. This observation can be strengthen by

353

AMBTs which show the continental outflow from continents was dominant in winter and

354

spring and maritime airmasses were in summer and autumn. This point clearly signifies that

355

these ω–hydroxy FAs can be used as biomarkers of terrestrial plants or soil organic matter

356

transported to the open ocean.

357

3.5. Relative contribution of continental vs. marine soruces to β−hydroxy FAs β−

SC

RI PT

349

Although having similar atmospheric abundances of measured β−hydroxy FAs

359

between polluted continental airmasses (in winter/spring) and pristine maritime airmasses (in

360

summer/autumn), considerable differences are noteworthy in them with regard to carbon

361

number predominance. It is noteworthy that the polluted air masses sampled in winter show a

362

temporal distribution characterized by the predominance of low molecular weight β−hydroxy

363

FAs with the following order: C10 > C12 > C14 > C16 (Figure 7). However, a slight change in

364

the molecular distribution is observed for spring samples (C16 > C14 > C18 > C12 > C10; Figure

365

7), which is significantly different during winter and summer (C10 > C16 > C18 > C20 > C14;

366

Figure 7). Since autumn season is a transition period between the two wind regimes (as

367

discussed in AMBTs section), the temporal abundance distribution of β−hydroxy FAs is as

368

follows, C10 ≈ C16 > C18 ≈ C14 > C8 (Figure 7).

TE D

EP

AC C

369

M AN U

358

Several studies have documented the Asian dust transport to the western North Pacific

370

during spring season (Duce et al., 1980; Uematsu et al., 1983). Therefore, we attribute the

371

observed temporal distribution of β−hydroxy FAs during spring season over Chichijima to

372

their contribution from soil bacteria and higher plant metabolites. In contrast, the change in

373

temporal variability of mass concentration of β−hydroxy FAs in summer is also associated

17

ACCEPTED MANUSCRIPT 374

with a shift in AMBTs from polluted continental airmasses in spring to pristine oceanic

375

airmasses. Therefore, the photochemical oxidation of biogenic unsaturated fatty acids in

376

the

377

predominance of C10 followed by C16, C14 and C18 and C20 β−hydroxy FAs. A supportive

378

evidence for this argument can be obtained in Mochida et al. (2003), whose study

379

demonstrated that the oceanic airmasses in summer over the North Pacific are characterized

380

by increased atmospheric concentrations of C14−C19 fatty acids. Their study also reported that

381

the atmospheric abundances of C14−C19 fatty acids show strong seasonal dependence with

382

maximum concentration observed in spring and summer.

SC

RI PT

marine atmospheric boundary layer (MABL) during summer can explain the

Overall, the assessed atmospheric abundances of β−hydroxy FAs exhibit no

384

significant differences between winter and summer, which are influenced by continental and

385

marine airmasses, respectively. As mentioned earlier, previous studies have suggested that

386

these β−hydroxy FAs are the constituents of LPS of GNB (Wilkinson, 1988). Despite having

387

similar atmospheric concentrations of β−hydroxy FAs, the differences in wind regimes can

388

provide the relative importance of continental and marine GNB over the sampling site during

389

all seasons. Here we applied empirical formula and cluster analysis approach (see section 3.1)

390

to evaluate the percentage contribution from continental and marine airmasses to the

391

observed concentrations of hydroxy FAs.

TE D

EP

AC C

392

M AN U

383

Since our sampling site is a remote island in the western North Pacific, it is likely that

393

oceanic airmasses can also contribute to the chemical composition during winter and spring.

394

Therefore, we define a factor to distinguish continental versus marine contribution of

395

atmospheric abundances of β−hydroxy FAs and thereby GNB in various seasons, based on

396

the computed AMBTs. The factor f is defined as the number of trajectories passing over the

397

continent to the total trajectories computed for the sampling days, which will provide

398

continental contribution.

18

ACCEPTED MANUSCRIPT 399

fcontinental = [(AMBTs)continental /[(AMBTs)continental + (AMBTs)marine] x 100

400

fmarine (%) = [1 – fcontinental] x 100 where (AMBTs)continental and (AMBTs)marine refer to air mass backward trajectories passing

402

over the continent and ocean, respectively. The estimated fraction of continental sources in

403

contributing to the atmospheric β−hydroxy FAs over the Chichijima during winter and spring

404

correspond to 92% and 70%, respectively. However, due to the shift in wind regimes from

405

westerlies, influence of maritime airmasses is expected to maximize during summer and

406

become somewhat higher in autumn season. Therefore, the fractional contribution of

407

continental sources to atmospheric β−hydroxy FAs correspond to be 9% for summer and

408

18% for autumn. Although these empirical relation provided us the information regarding

409

relative source contribution (i.e., continental vs. oceanic) in various seasons, it is important to

410

assess their geographical variability within the source region. Therefore, we have performed

411

cluster analysis during sampling days to decipher the potential source regions that contribute

412

to atmospheric abundances of β−hydroxy FAs in airmasses originating from the continent

413

and ocean.

414

3.6. Potential source regions of bacterial and plant biomarkers

EP

TE D

M AN U

SC

RI PT

401

In this study, a HYSPLIT model (Version 4.0) was installed on a PC and run using its

416

Graphical User Interface (GUI). The application of HYSPLIT requires gridded

417

meteorological data at regular time intervals. The operational system (NCEP/NCAR) was

418

used to produce these data. Based on the results of HYSPLIT model, the AMBTs were

419

assigned to distinct clusters according to their moving speed and direction using k that means

420

clustering algorithm (Hartigan and Wong, 1979). In this study, various pre-selected number

421

of clusters (e.g., 3, 4, 5, and 6) were tested, and we found that 4 clusters can best represent

AC C

415

19

ACCEPTED MANUSCRIPT 422

the meteorological characteristics of the transport pathways over Chichijima Island. As a

423

result, this number was selected to expect number of AMBT clusters. Since the AMBTs show a similarity with respect to seasons, we have combined the

425

average mass concentrations of β− and ω−hydroxy FAs for the seasonal data from all years

426

and compared them with the output contribution of source regions from the cluster analysis.

427

The results of cluster analysis for the combined seasonal data during sampling days are

428

shown in Figure 8. In winter, the trajectory clusters show a long-range atmospheric transport

429

from Europe and Russia through westerly winds before arriving to Chichijima Island. These

430

airmasses would, thus, be responsible for the significant contribution of hydroxy FAs (this

431

study) along with other lipid-class compounds from the Asian continent to the sampling site.

M AN U

SC

RI PT

424

Among the mean trajectory path of four clusters shown in Figure 8, maximum

433

concentrations of hydroxy FAs (i.e., both β− and ω−hydroxy FAs) were observed in winter

434

during which most of the airmasses were originated from Europe and Russia and passed over

435

Mongolia and Korea. In contrast, the airmasses in summer are mainly transported from the

436

oceans by easterly trade winds. However, few trajectories show transport from Indonesian

437

Islands to the sampling site during summer. Despite having differences in the AMBT cluster,

438

the measured atmospheric abundances of β−hydroxy FAs (C10-C18) (i.e., proxy for GNB) do

439

not vary between winter and summer.

EP

AC C

440

TE D

432

The concentrations of β−hydroxy FAs (C8-C31) and β−hydroxy FAs (C10−C18), GNB

441

tracers, do not vary in winter (sum of average concentrations in each cluster (1, 2, 3 and 4);

442

C8–C31: 210 pg m-3, C10–C18: 190 pg m-3) and summer (C8–C31: 210 pg m-3 and C10–C18: 180

443

pg m-3) (Table 2). This observation supports the fact that these GNB are ubiquitous in the

444

environment whether the airmasses are transported from continents or originated within the

445

MABL of open oceans. Surprisingly, atmospheric abundances of ω−hydroxy FAs (C11-C28)

446

in winter (sum of average concentrations in each cluster (1, 2, 3 and 4), 640 pg m-3) are

20

ACCEPTED MANUSCRIPT

higher than those in summer (440 pg m-3). This could be due to the reason why these

448

compounds are primarily originated from terrestrial plants and soil microorganisms

449

(Kawamura, 1995; Eglinton et al., 1968) and were transported from the Asian continent to

450

Chichijima Island via westerlies. As mentioned earlier, spring and autumn seasons are the

451

transition periods from winter to summer and summer to winter, respectively. In this regard,

452

we observe high contribution of hydroxy FAs from continents (61−64%) in spring and from

453

oceans (~75%) in autumn. During spring and autumn, high atmospheric abundances of

454

hydroxy FAs are noteworthy for the trajectory clusters that are passing over Korea and Japan.

455

However, lower concentrations of hydroxy FAs were observed for clusters showing transport

456

from Russia. Another interesting feature of the data is that the trajectory clusters obtained

457

from HYSPLIT model have similar pattern in both spring and autumn, suggesting the similar

458

source regions of hydroxy FAs in both seasons.

M AN U

SC

RI PT

447

460

4. Conclusions

TE D

459

To the best of our knowledge, the results obtained in this study is the first of its

462

kind in applying a biomarker approach to determine the long-term contributions and transport

463

of hydroxy FAs of Gram-negative bacteria and terrestrial plant origin to remote marine

464

aerosols. However, our study is in continuation to that documented for marine aerosols (in

465

only one sample) by Kawamura (1995), who reemphasizes the significance of these

466

compounds as biomarkers for bacterial transport through aeolian pathway. In the present

467

study, we have used sixty nine marine aerosol samples to determine β– and ω–hydroxy FAs

468

collected in Chichijima over the period of four years (1990–1993). We observed substantial

469

seasonal differences in the carbon number predominance of β– and ω–hydroxy FAs. These

470

biomarkers are primarily associated with terrestrial plants and soil microorganisms; but our

471

study shows that continents and oceans equally contribute to the microbial biomarkers in

AC C

EP

461

21

ACCEPTED MANUSCRIPT

marine aerosols. The observed similarity in relative abundances of β−hydroxy FAs in marine

473

aerosols indicates the presence of GNB in both airmasses transported from continents and

474

originated in the western North Pacific. Thus, our study characterize the source regions of β–

475

and ω–hydroxy FAs. We also show that ω–hydroxy FAs are mainly contributed by terrestrial

476

plants and soil microbes whereas β−hydroxy FAs are ubiquitous in envrionment owing to

477

their bacterial sources in both polluted continental and pristine oceanic airmasses. We also

478

show that carbon number of β−hydroxy FAs is also variable among different seasons, which

479

can be potentially used as specific tracer for the particular bacterial species. Thus, our study

480

demonstrates the usage of β– and ω–hydroxy FAs in marine aerosol samples as a potential

481

proxy to trace the Gram-negative bacterial species and terrestrial plant metabolites and also

482

their sources (continental vs. marine) regions.

M AN U

SC

RI PT

472

483 484

Acknowledgement

The authors thank S. Ogishima for his help in the aerosol sampling and the Japan

486

Aerospace Exploration Agency (JAXA) for the courtesy to use the facility in the Satellite

487

Tracking Center at Chichijima. Authors are also thankful to Y. Ishimura for his help in the

488

sample extraction, separation and derivatization. This study is in part supported by Japan

489

Society for the Promotion of Science (Grant-in-aid No. 24221001). PT is thankful to

490

Japanese Ministry of Education, Science and Culture (MEXT) for the Monbukagakusho

491

Scholarship.

EP

AC C

492

TE D

485

22

ACCEPTED MANUSCRIPT References

494

Bird, D.F., Kalff, J., 1984. Empirical relationships between bacterial abundance and

495

chlorophyll concentration in fresh and marine waters. Canadian Journal of Fisheries

496

and Aquatic Sciences 41, 1015-1023.Blokker P., Schouten S., Van den Ende H., De

497

Leeuw J.W., Sinninghe Damsté J.S., 1998. Cell wall specific ω-hydroxy fatty acids in

498

some freshwater green microalgae. Phytochemistry 49, 691-695.

500

Cardoso, J.N., Eglinton, G., 1983. The use of hydroxyacids as geochemical indicators. Geochimica et Cosmochimica Acta 47, 723-730.

SC

499

RI PT

493

Chen, J., Kawamura, K., Cong-Qiang Liu, Fu, P., 2013. Long-term observations of

502

saccharides in remote marine aerosols from the western North Pacific: A comparison

503

between 1990-1993 and 2006-2009 Periods. Atmospheric Environment 67, 448-458.

504 505

M AN U

501

Cho, B., Azam, F., 1990. Biogeochemical significance of bacterial biomass in the ocean’s euphotic zone. Marine Ecological Program Series 63, 253-259. Chun, Y.S., Boo, K.O., Kim, K., Park, S.U., Lee, M., 2001. Transport, and physical

507

characteristics of Asian dust in Korea. Journal of Geophysical Research 106, 18461-

508

18469.

511

EP

510

Downing, D.T., 1961. Naturally occurring aliphatic hydroxyacids. Reviews of Pure and Applied Chemistry 11, 196-211.

AC C

509

TE D

506

Draxler, R.R., Rolph, G.D., 2014. HYSPLIT (HYbrid Single-Particle Lagrangian Integrated

512

Trajectory)

Model

access

via

NOAA

ARL

READY

Website

513

(http://ready.arl.noaa.gov/HYSPLIT.php). NOAA Air Resources Laboratory, Silver

514

Spring, MD.

515

Duce, R.A., Unni, C.K., Ray, B.J., Prospero, J.M., Merrill, J.T., 1980. Long-range

516

atmospheric transport of soil dust from Asia to the tropical North Pacific: Temporal

517

variability. Science 209, 1522-1524.

23

ACCEPTED MANUSCRIPT 518

Eglinton, G., Hunneman, D.H., Douraghi-Zadeh, K., 1968. Gas chromatographic-mass

519

spectrometric studies of long chain hydroxy acids-II. The hydroxy acids and fatty

520

acids of a 5000-years old lacustrine sediment. Tetrahedron 24, 5929-5941.

522

Gagosian, R.B., Peltzer, E.T., 1986. The importance of atmospheric input of terrestrial organic material to deep sea sediments. Organic Geochemistry 10, 661- 669.

RI PT

521

Gelin, F., Volkman, J.K., de Leeuw, J.W., Sinninghe Damste Â, J. S., 1997. Mid-chain

524

hydroxy long-chain fatty acids in microalgae from the genus Nannochloropsis.

525

Phytochemistry 45, 641- 646.

SC

523

Giovannoni, S., Rappe, M., 2000. Evolution, diversity and molecular ecology of marine

527

prokaryotes. In: Kirchman, D.L. (Eds.), Microbial Ecology of the Oceans. Wiley-Liss,

528

New York, 47-84.

529 530

M AN U

526

Graça, J., Santos, S., 2006. Glycerol-derived ester oligomers from cork suberin. Chem Phys Lipids 144 (1), 96-107.

Griffin, D., Westphal, D., Gray, M., 2006. Airborne microorganisms in the African desert

532

dust corridor over the mid-Atlantic ridge, Ocean Drilling Program, Leg 209.

533

Aerobiologia 22, 211-226.

TE D

531

Harrison, R., Jones, A., Biggins, P., Pomeroy, N., Cox, C., Kidd, S., Hobman, J., Brown, N.,

535

Beswick, A., 2005. Climate factors influencing bacterial count in background air

536

samples. International Journal of Biometeorology 49, 167-178.

538

AC C

537

EP

534

Hartigan, J.A., Wong, M.A., 1979. A k-means clustering algorithm. Journal of Applied Statistics 28, 100-108.

539

Hines, C.J., Waters, M.A., Larsson, L., Petersen, M.R., Saraf, A., Milton, D.K., 2003.

540

Characterization of endotoxin and 3-hydroxy fatty acid levels in air and settled dust

541

from commercial aircraft cabins. Indoor air 13(2), 166-173.

24

ACCEPTED MANUSCRIPT 542

Kawamura K, Ishiwatari R., 1984a. Fatty acid geochemistry of a 200 m sediment core from

543

Lake Biwa, Japan. Early disgenesis and paleoenvironmental information. Geochimica

544

et Cosmochimica Acta 48, 251-266. Kawamura, K., 1995. Land-derived lipid class compounds in the deep-sea sediments and

546

marine aerosols from North Pacific, in: H. Sakai and Y. Nozaki (Eds.), Biogeochemical

547

Processes and Ocean Flux in the Western Pacific, Terra Scientific Publishing Co.,

548

Tokyo, pp. 31-51.

RI PT

545

Kawamura, K., Ishimura, Y., Yamazaki, K., 2003. Four year’s observations of terrestrial lipid

550

class compounds in marine aerosols from the western North Pacific. Global

551

Biogeochemical Cycles 17 (1), 1003-1019.

554 555

M AN U

553

Kawamura, K., Ishiwatari, R., 1982. Tightly bound β−hydroxy acids in a recent sediment. Nature (London) 297, 144-145.

Kawamura, K., Ishiwatari, R., 1984b. Tightly bound aliphatic acids in Lake Biwa sediments: their origin and stability. Organic Geochemistry 7, 121-126.

TE D

552

SC

549

Kawamura, K., Ishiwatari, R., Ogura, K., 1987. Early diagenesis of organic matter in the

557

water column and sediments: microbial degradation and resynthesis of lipids in Lake

558

Haruna. Organic Geochemistry 11, 251-264.

EP

556

Laitinen, S., Kangas, J., Husman, K., Susitaival, P., 2001. Evaluation of exposure to airborne

560

bacterial endotoxins and peptidoglycans in selected work environments. Annals of

561

Agricultural and Environmental Medicine 8, 213–219.

AC C

559

562

Lee, A.K.Y., Chan, C.K., Fang, M., Lau, A.P.S., 2004. The 3-hydroxy fatty acids as

563

biomarkers for quantification and characterization of endotoxins and Gram-negative

564

bacteria in atmospheric aerosols in Hong Kong. Atmospheric Environment 56, 124-135.

565

Li, W., Head, E., Glen Harrison, W., 2004. Macro ecological limits of heterotrophic bacterial

566

abundance in the ocean. Deep-Sea Research Part I 51, 1529-1540.

25

ACCEPTED MANUSCRIPT 567 568

Lin, T.H., 2001. Long-range transport of yellow sand to Taiwan in spring 2000: Observed evidence and simulation. Atmospheric Environment 35, 5873-5882. Linley, E., Newell, R., Lucas, M., 1983. Quantitative relationships between phytoplankton,

570

bacteria and heterotrophic microflagellates in shelf waters. Marine Ecological Program

571

Series 12, 77-89.

RI PT

569

Matsumoto, K., Uematsu, M., Hayano, T., Yoshioka, K., Tanimoto, H., Iida, T., 2003.

573

Simultaneous measurements of particulate elemental carbon on the ground observation

574

network over the western North Pacific during the ACE-Asia campaign. Journal of

575

Geophysical Research: Atmospheres 108, 8635.

SC

572

Mochida, M., Kawamura, K., Umemoto, N., Kobayashi, M., Matsunaga, S., Ho-Jin Lim,

577

Turpin, B. J., Bates, T. S., Simoneit, B. R. T., 2003. Spatial distributions of oxygenated

578

organic compounds (dicarboxylic acids, fatty acids, and levoglucosan) in marine

579

aerosols over the western pacific and off the coast of east asia: continental outflow of

580

organic aerosols during the ace-asia campaign. Journal of Geophysical Research 108

581

(D23), 8638.

TE D

M AN U

576

Mochida, M., Kitamori, Y., Kawamura, K., Nojiri, Y., Suzuki, K., 2002. Fatty acids in the

583

marine atmosphere: Factors governing their concentrations and evaluation of organic

584

films on sea‐salt particles. Journal of Geophysical Research: Atmospheres 107, 1-10.

586 587 588 589 590

AC C

585

EP

582

Molina, I., Bonaventure, G., Ohlrogge, J., Pollard, M., 2006. The lipid polyester composition of Arabidopsis thaliana and Brassica napus seeds. Phytochemistry 67(23), 2597-2610. Murayama, T., 2001. Ground-based network observation of Asian dust eveint of April 1998 in east Asia. Journal of Geophysical Research 106, 18345-18359. Pollard, M., Beisson, F., Li, Y.H., Ohlrogge, J.B., 2008. Building lipid barriers: biosynthesis of cutin and suberin. Trends in Plant Science 13, 236-246.

26

ACCEPTED MANUSCRIPT 591 592

Prospero, J., Blades, E., Mathison, G., Naidu, R., 2005. Inter hemispheric transport of viable fungi and bacteria from Africa to the Caribbean with soil dust. Aerobiologia 21, 1-19. Rogge, W.F., Mazurek, M.A., Hildemann, L.M., Cass, G.R., Simoneit, B.R.T., 1993.

594

Quantification of urban organic aerosols at a molecular level: Identification, abundance

595

and seasonal variation. Atmospheric Environment 27, 1309-1330.

RI PT

593

Rolph, G.D., 2014. Real-time Environmental Applications and Display sYstem (READY)

597

Website (http://ready.arl.noaa.gov). NOAA Air Resources Laboratory, Silver Spring,

598

MD.

SC

596

Saraf, A., Larsson, L., Burge, H., Milton, D., 1997. Quantification of ergosterol and 3-

600

hydroxy fatty acids in settled house dust by gas chromatography–mass spectrometry:

601

comparison with fungal culture and determination of endotoxin by a limulus amebocyte

602

lysate assay. Applied and Environmental Microbiology 63, 2554-2559.

M AN U

599

Simoneit, B.R.T., 1989. Organic matter of the troposphere, V: Application of molecular

604

marker analysis to biogenic emissions into the troposphere for source reconciliations,

605

Journal of Atmospheric Chemistry, 8, 251-275.

TE D

603

Skerratt, J.H., Nichols, P.D., Bowman, J.P., Sly, L.I., 1992. Occurrence and significance of

607

long-chain (ω−1)− hydroxy fatty acids in methane-utilizing bacteria. Organic

608

Geochemistry 18, 189-194.

AC C

EP

606

609

Uematsu, M., Duce, R.A., Prospero, J.M., Chen, L., Merrill, J.T., McDonald, R.L., 1983.

610

Transport of mineral aerosol from Asia Over the North Pacific Ocean. Journal of

611

Geophysical Research: Oceans 88, 5343–5352.

612

Wakeham, S.G., Pease, T.K., Benner, R., 2003. Hydroxy fatty acids in marine dissolved

613

organic matter as indicators of bacterial membrane material. Organic Geochemistry 34,

614

857-868.

27

ACCEPTED MANUSCRIPT 615

Wang, Z., Ueda, H., Huang, M., 2000. A deflation module for use in modeling long-range

616

transport of yellow sand over East Asia. Journal of Geophysical Research 105, 26947-

617

26960.

619

Wilkinson, S.G., 1988. Gram-negative bacteria in: Ratledge, C., Wilkinson, S.G. (Eds.), Microbial Lipids, vol. 1. Academic Press, New York, 199-488.

RI PT

618

Yamamoto, S., Kawamura, K., Seki, O., 2011. Long-range atmospheric transport of

621

terrestrial biomarkers by the Asian winter monsoon: Evidence from fresh snow from

622

Sapporo, northern Japan. Atmospheric Environment 45, 3553-3560.

SC

620

Zelles, L., 1999. Fatty acid patterns of phospholipids and lipopolysaccharides in the

624

characterisation of microbial communities in soil: a review. Biology and Fertility of

625

Soils 29, 111-129.

AC C

EP

TE D

M AN U

623

28

ACCEPTED MANUSCRIPT

Table 1. Statistical summary (Min - Max, average and standard deviation, and median) of mass concentrations of β− and ω−hydroxy fatty acids (pg m-3) in marine aerosols collected

ω─Hydroxy FAs

Min - Max

Av. ± S.D.

Median

C8

7.41 - 81.4

36.6 ± 4.1

31.4

C9

5.8 - 63.7

29.1 ± 2.3

26

C10

6.2 - 308

44.2 ± 5.8

30

C11

3.0 - 29.6

12.1 ± 0.8

C12

2.95 - 557

46 ± 12

C13

0.8 - 22.1

6.6 ± 0.7

C14

3.0 - 98.5

21 ± 2

C15

0.07 - 19.4

6.1 ± 0.8

C16

4.4 - 140

C17

0.5 - 50.9

C18

2.4 - 94.5

C19

0.3 - 16.6

C20

Median

0.00 - 30.4

7 ± 1.4

4.2

16.3

0.00 - 575

62 ± 18

1.1

5.8

0.00 - 331

10.4 ± 5.85

0.71

17

0.72 - 1168

74 ± 25

6.7

0.34 - 2418

174 ± 45

23.8

0.00 - 93

11.4 ± 3

1.76

0.00 - 139

14 ± 3

3

0.00 - 843

82 ± 27

9

0.00 - 2497

89 ± 40.8

8

0.00 - 199

22.8 ± 9.5

1.6

0.00 - 62.6

6.7 ± 1.8

0

TE D

5.3 18

5.6 ± 1.7

3.4

22 ± 3

11.5

5.4 ± 1.3

4.1

1.03 - 56.6

11.3 ± 1.6

56.6

C21

1.23 - 18.5

5.2 ± 1.6

3.3

C22

0.3 - 40.3

7.4 ± 1.6

4.7

C23

0.9 - 15.4

4.7 ± 1.7

2.7

C24

0.53 - 23.8

6.3 ± 1.6

3.9

C25

1.55 - 5.7

4.2 ± 1.3

5.3

C26

2.6 - 8.1

5.1 ± 1.0

5.5

C27

1.62 - 4.6

3.1 ± 1.5

3.1

C28

3.3 - 6.3

5.1 ± 0.9

5.5

C30

0.00 - 2.9

2.9

2.9

EP

Av. ± S.D.

11.7

32.1 ± 4

AC C

Min - Max

M AN U

number

SC

β─Hydroxy FAs

Carbon

RI PT

from Chichijima Island during 1990-1993.

29

ACCEPTED MANUSCRIPT 0.00 - 1.5

1.5

1.5

AC C

EP

TE D

M AN U

SC

RI PT

C31

30

ACCEPTED MANUSCRIPT

Table 2. Concentrations (average and range, pg m-3) of (a) β−hydroxy FAs (C8−C31 ), (b) Gramnegative bacterial (GNB) tracers: β−hydroxy FAs (C10−C18), and (c) ω−hydroxy FAs (C10-C18) in

No. of clusters

Winter

Spring

RI PT

different seasons according to the cluster mean trajectories.

Summer

Autumn

Av.

Range

Av.

Range

Av.

Range

Av.

Range

Cluster 1

100

30-230

60

0-160

70

0-270

50

10-120

Cluster 2

90

30-210

130

10-360

70

0-290

70

20-150

Cluster 3

20

0-40

60

0-160

20

0-90

30

10-70

Cluster 4

0

0-10

M AN U

(b) GNB tracers: β−Hydroxy FAs (C10−C18) β−

SC

(a) β−Hydroxy β− FAs (C8−C31)

60

0-160

50

0-200

10

0-30

90

20-210

40

0-140

60

0-220

40

10-100

Cluster 2

80

90-190

100

10-330

60

0-230

50

10-120

Cluster 3

20

0-40

40

0-140

20

0-70

20

10-50

Cluster 4

TE D

Cluster 1

0

0-10

40

0-140

40

0-160

10

10-20

310

0-1810

100

0-960

140

0-1540

100

0-460

270

0-1610

220

0-2210

150

0-1630

120

0-570

50

0-310

100

0-960

50

0-530

50

0-250

10

0-80

100

0-960

100

0-1100

20

0-110

Cluster 1 Cluster 2

AC C

Cluster 3

EP

(c) ω−Hydroxy FAs (C11−C28) ω−

Cluster 4

31

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

Figure 1. Map with the sampling site (Chichijima Island) used for the collection of aerosols from the western North Pacific.

32

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 2. Typical 7−days isentropic air mass back trajectories (AMBTs) during the four seasons (winter, spring, summer, autumn) of (a) 1990, (b) 1991-92, (c) 1992-93 and (d) 1993 starting from 9th April 1990.

33

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 3. The relative abundances of β− and ω−hydroxy (OH) fatty acids with respect to

AC C

EP

North Pacific.

TE D

carbon numbers in marine aerosol samples collected over Chichijima Island in the western

34

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 4. Bar graphs showing the seasonally averaged (mean ± S.E.) relative abundances of β-hydroxy fatty (C8−C9, C10–C18, C19–C31) during the years of (a) 1990, (b) 1991−92, (c) 1992−93, and (d) 1993 in the marine aerosols collected from Chichijima Island, the western North Pacific.

35

SC

RI PT

ACCEPTED MANUSCRIPT

M AN U

Figure 5. Temporal variability of relative abundances of even carbon number β−hydroxy FAs

AC C

EP

TE D

(C10−C18) in marine aerosols collected over Chichijima Island in the western North Pacific.

36

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 6. Bar graphs, showing the seasonally averaged (mean ± S.E.) relative abundances of ω-hydroxy fatty acids (C11−C14 and C16–C28) during (a) 1990, (b) 199−92, (c) 1992−93 and (d) 1993 in marine aerosols collected from Chichijima Island in the western North Pacific.

37

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 7. Temporal variability of atmospheric abundances of β−hydroxy fatty acids with regard to chain length (C−number) among different seasons.

38

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

39

ACCEPTED MANUSCRIPT

Figure 8. Mean horizontal plots of AMBTs trajectories for 7 days at 500 m above ground level were drawn with the NOAA HYSPLIT model. The trajectories in each season were grouped into four clusters as suggested by the model. The numbers 1 to 4 indicate the name of mean clusters. The numbers in parenthesis indicate the percentage of total air mass

M AN U

SC

RI PT

transport from the origins. The star indicates the location of the sampling site.

TE D

Table S3. One-way ANOVA analyses for the relative abundances of β− and ω−hydroxy FAs measured to ascertain the significant differences in various seasons.

EP

β−hydroxy FAs

Type

F(3, 50) = 2.1

AC C

C8-C9

One-way ANOVA P-value

Post-hoc test Inference

> 0.05

Not Significant

N/A

C10-C18

F(3, 60) = 0.9

> 0.05

Not Significant

N/A

C19-C31

F(3, 49) = 0.9

> 0.05

Not Significant

N/A

ω−hydroxy FAs

C11-C24

F(3, 56) = 4.7

< 0.05

Significant

A

C16-C28

F(3, 61) = 3.3

< 0.05

Significant

A

Note: N/A: not applicable; A: Applicable.

40

ACCEPTED MANUSCRIPT

Table S4. A Post-hoc test conducted for estimating least significant differences (LSD) in the

M AN U

SC

RI PT

relative abundances of ω−hydroxy FAs among four seasons.

Post-hoc Test: LSD for ω−hydroxy FAs Type 1

Spring

3

Summer

4

Fall

< 0.053

< 0.052

< 0.051

> 0.053

> 0.054

< 0.054

> 0.052

< 0.054

> 0.051

> 0.051

< 0.053

< 0.052

< 0.053

< 0.052

< 0.051

> 0.053

> 0.054

< 0.054

> 0.052

< 0.051

> 0.051

< 0.053

EP

C16-C28

> 0.052

2

TE D

C11-C14

Winter

AC C

< 0.054

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Highlights

EP

TE D

M AN U

SC

RI PT

A biomarker based approach for bacterial transport through aeolian pathway Seasonal, temporal & molecular distributions of β− & ω−hydroxy fatty acids Ubiquitous occurrence of hydroxy fatty acids in marine aerosols Continental & marine sources equally contribute to these hydroxy fatty acids

AC C

1. 2. 3. 4.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Spring

Winter

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Autumn

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Summer

ACCEPTED MANUSCRIPT Table S3. One-way ANOVA analyses for the relative abundances of and hydroxy FAs measured to ascertain the significant differences in various seasons. hydroxy FAs One-way ANOVA P-value

Type

F(3, 50) = 2.1

> 0.05

Not Significant

N/A

C10-C18

F(3, 60) = 0.9

> 0.05

Not Significant

N/A

C19-C31

F(3, 49) = 0.9

> 0.05

Not Significant

hydroxy FAs C11-C24

< 0.05

F(3, 56) = 4.7

Significant

N/A A A

SC

< 0.05 C16-C28 F(3, 61) = 3.3 Note: N/A: not applicable; A: Applicable.

Significant

RI PT

C8-C9

Post-hoc test

Inference

M AN U

Table S4. A Post-hoc test conducted for estimating least significant differences (LSD) in the relative abundances of hydroxy FAs among four seasons.

Post-hoc Test: LSD for hydroxy FAs

Type C11-C14

2

> 0.052

< 0.053

< 0.052

< 0.051

> 0.053

> 0.054

< 0.054

> 0.052

< 0.054

> 0.051

> 0.051

< 0.053

< 0.052

< 0.053

< 0.052

< 0.051

> 0.053

> 0.054

< 0.054

> 0.052

< 0.054

< 0.051

> 0.051

< 0.053

AC C

EP

C16-C28

Winter

Spring

TE D

1

3

Summer

4

Fall