Subscriber access provided by READING UNIV
Article
Boron dissolved and dust atmospheric inputs to a forest ecosystem (Northeastern France) Philippe Roux, Marie Pierre Turpault, Gil Kirchen, Paul-Olivier Redon, and Damien LEMARCHAND Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b03226 • Publication Date (Web): 17 Nov 2017 Downloaded from http://pubs.acs.org on November 20, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 28
Environmental Science & Technology
1 2 3 4 5 6
Boron Dissolved and Particulate Atmospheric Inputs to a Forest Ecosystem (Northeastern France)
7
8
9 10 11 12 13
PHILIPPE ROUX1.2 ; MARIE-PIERRE TURPAULT1 ; GIL KIRCHEN1 ; PAUL-OLIVIER REDON3
14
AND DAMIEN LEMARCHAND*
2
15
16 17 18
*corresponding author :
[email protected] 19 20
1
21
2
22
3
BEF-INRA, Centre de Nancy, 54280 Champenoux, France LHyGeS/CNRS, Université de Strasbourg, France Andra- Centre de Meuse Haute Marne 55290 Bure, France
1 Environment ACS Paragon Plus
Environmental Science & Technology
23
Abstract
24
Boron concentrations and isotopic compositions of atmospheric dust and dissolved depositions
25
were monitored over a two-years period (2012-2013) in the forest ecosystem of Montiers
26
(Northeastern France). This time series allows the determination of the boron atmospheric inputs to
27
this forest ecosystem and contributes to refine our understanding of the sources and processes that
28
control the boron atmospheric cycle. Mean annual dust and dissolved boron atmospheric depositions
29
are comparable in size (13 g.ha-1.yr-1 and 16 g.ha-1.yr-1, respectively), which however show significant
30
intra- and inter-annual variations. Boron isotopes in dust differ from dissolved inputs, with an annual
31
mean value of +1 ‰ and +18 ‰ for, respectively.
32
The notable high boron contents (190-390 µg.g-1) of the dust samples are interpreted as resulting
33
from localized spreading of boron-rich fertilizers, thus indicating a significant local impact of regional
34
agricultural activities. Boron isotopes in dissolved depositions show a clear seasonal trend. The
35
absence of correlation with marine cyclic solutes contradicts a control of atmospheric boron by
36
dissolution of seasalts. Instead, the boron data from this study are consistent with a Rayleigh-like
37
evolution of the atmospheric gaseous boron reservoir with possible but limited anthropogenic
38
and/or biogenic contributions.
39 40
Keywords: boron, isotopes, atmosphere, dust deposition, dissolved deposition
41
2 Environment ACS Paragon Plus
Page 2 of 28
Page 3 of 28
42
Environmental Science & Technology
Introduction
43
Boron is a micronutrient essential for plant growth1. Deficiency symptoms are observed in
44
more than 80 countries where 30% of the arable soils are diagnosed with too low boron
45
concentration to sustain perennial agriculture2,3. On the contrary, desertification, irrigation,
46
fertilizers, waste waters and mining cause boron local accumulations in soils with subsequent
47
dramatic effects on their fertility4 but also at wider scale on the quality of groundwaters5,6 and
48
possibly, but in a less documented way, on the atmosphere7. Finally, the range between boron
49
deficiency and toxicity levels is very restricted8, making boron the second global micronutrient
50
challenge after zinc9 and giving a critical interest to the understanding of the boron biogeochemical
51
cycle to develop sustainable agriculture practices and water management.
52
Boron is a trace constituent in the atmosphere that exists in both particulate and gaseous
53
forms with the latter representing about 90-95% of the total10. Published work on boron in
54
rainwaters have shown concentrations ranging over more than two orders of magnitude, from 0.2 to
55
385 µg.L-1, which contribute to about 20% of the dissolved boron in rivers11-13. Boron isotopic
56
compositions in rainwaters also show large variations, from -13‰ to +48‰, spanning approximately
57
60% of the total range observed in natural materials
58
the atmosphere considered seawater as the main source of boron
59
continental contribution evaluated at 0.2%
60
determined the global atmospheric boron dissolved deposition at about 2.75 Tg.yr-1. The major part
61
of it (2.54 Tg.yr-1) only impact coastal and marine sites leaving a much smaller global boron dissolved
62
deposition over lands (0.22 Tg.yr-1). These authors also expect a boron deposition over land as dry
63
falls equal to the dissolved ones and inferred a global boron particulate deposition from mobilization
64
of soil dust of 0.054 Tg.yr-1. The main conclusion of these authors is certainly that the global boron
65
cycle is increasingly impacted by anthropogenic activities, which now induce boron fluxes that about
7,12,14-22
. Early works on the sources of boron in 10,19,23
23
with a negligible 13
. In a recent review Schlesinger and Vengosh
3 Environment ACS Paragon Plus
Environmental Science & Technology
Page 4 of 28
66
equal the natural ones thus questioning the role of seawater as a net source of boron deposition
67
over lands.
68
It is now admitted that boron atmospheric wet deposits have three main sources, 1) a marine
69
component characterized by high 11B close to the water value (∼+40‰) composed of dissolved
70
marine aerosols and/or of evaporated/condensed gaseous boron; 2) a biogenic component with 11B
71
close to 0‰ or slightly negative that may, at least partly, derive from biomass burning, pollen or soil
72
organic compounds and 3) anthropogenic components deriving from leaching of aerosols including
73
coal combustion particles and from fertilizer spreading characterized by more scattered 11B values
74
but high SO4/B and NO3/B chemical ratios, which values depends on the fertilizer spread
75
7,14,15,17,21,24,25
76
combustion and fertilizers to rainwaters are determined from the concentration ratios of dissolved
77
sodium/boron, sulfate/boron and nitrate/boron, respectively. The contribution of each of these
78
sources is highly dependent on local conditions at the sampling site. Seawater is a major contributor
79
to dissolved boron in coastal rainwaters, which contribution rapidly decreases with the distance from
80
the coast or change of the wind direction7,14. This so-called marine contribution is inferred from the
81
B/Na chemical ratio of the collected rainwaters, assuming that seasalts are the only source of Na to
82
the rainwaters, which actually only accounts for the dissolution of marine aerosols. However, The
83
high B/Cl or B/Na chemical ratios (sometimes expressed in the literature using a boron enrichment
84
factor relative to seawater, EF) and the absence of correlation between boron isotopes in rainwaters
85
and the calculated contribution of marine seasalts (e.g. from 20 to 90% in Brest, France ), indicate
86
that dissolved B isotope in rainwaters cannot be explained solely by dissolution marine seasalts.
87
Alternatively, it has been proposed that evaporation of boric acid from seawater and later
88
condensation or interaction with particles in the atmosphere may control the dissolved boron budget
89
of rainwaters14,16,17 with a strong temperature dependence of the phase partition of boron in the
90
atmosphere23,26. The vapor-liquid fractionation of boron isotopes has been determined
. In all these studies, the respective contributions of marine sea-salts, coal
7
4 Environment ACS Paragon Plus
Page 5 of 28
Environmental Science & Technology
91
experimentally14 or empirically14,16,17, which therefore may explain most of the boron data in
92
rainwater following a Rayleigh-like distillation process. A comprehensive dataset including hydrogen
93
isotopes in rainwaters from various locations in the worlds also support such a gaseous atmospheric
94
transport of boron over long distance14,16,17. Compiling all these studies, it comes that seawater is
95
therefore considered as a major source of boron to continental rainwaters, through the combined
96
contribution of marine aerosols and water/vapor exchange processes with possible boron recharge
97
over lands, which may include biogenic and anthropogenic components.
98
Boron in atmospheric dust has received much less interest than rainwaters and has only been studies
99
through leachates of urban aerosols15 or water soluble aerosols25. In both studies, boron isotopes
100
have been used to trace the sources of particle emissions in two sites (Paris city center and Japan Sea
101
coast, respectively) chosen for their expected elevated anthropogenic contribution. Boron is here of
102
particular interest because it is abundant in coal and is characterized by relatively low δ11B values .
103
To our knowledge, no study has focuses on boron dust depositions and boron sources in rural areas.
27
104
With only few studies published so far, and none conducted in parallel on dissolved and dust
105
depositions, the sources and sinks of boron in the atmosphere are still a matter of
106
debate7,10,13,14,16,17,19,23. In this study, we present a 2-years (2012-2013) monitoring of boron
107
concentrations and isotopic composition in dissolved and dust depositions on a forest site in
108
Montiers (rural area located in Northeastern France). The dissolved and dust boron atmospheric
109
fluxes are then determined and discussed in terms of possible sources and processes affecting the
110
boron atmospheric transport.
111
Materials and methods
112
•
Site
113
The Montiers sampling site is an even-aged beech stand of approximately 50 years located in
114
Lorraine, Northeastern France (48°31'55"N / 5°16'8"E) managed by ANDRA-OPE and INRA-BEF. In this
5 Environment ACS Paragon Plus
Environmental Science & Technology
Page 6 of 28
115
region, 48% of the surface is dedicated to crop production (DRAAF, annual report 2014). The use of
116
boron fertilizer is widespread in this area, especially onto oleaginous products (colza and sunflower)
117
and beets, which corresponds to about 20% of the total cultivated surface. The Montiers area is very
118
rural with a low population density ( 70% of the total mineral phases). The rest of the particles (≈ 10-30%) consists in
209
interlayered clay minerals, micas, plagioclase, K-feldspar, Fe-oxides and carbonates. Sulfates and
210
gibbsite can be found occasionally35. From their chemical and mineralogical compositions, it has been
211
determined that dust result mainly from the erosion of regional temperate soil with only rare
212
episodes of marked Saharan contribution35. Between 2012 and 2013, dust deposition rate displays a
Mineralogy, organic content and deposition rate of dust.
9 Environment ACS Paragon Plus
Environmental Science & Technology
213
large variability ranging from 2.4 mg.m-2.d-1 to 23.3 mg.m-2.d-1. The highest dust deposition rates
214
were observed during the summer months and the lowest deposition rates during the winter
215
months. Despite seasonal changes of the dust deposition rates, the mineralogical composition of the
216
deposited dust has been observed to be about constant, with random occurrence of accessory
217
minerals however.
218
Major elements in dissolved deposition
219
Na and Cl concentrations in Aw range from 0.22 to 1.24 mg.L-1 and from 0.42 to 2.50 mg.L-1,
220
respectively (Table S1). The strong correlation between Na and Cl concentrations in our samples with
221
a molar Na/Cl = 0.86 (r² = 0.8) is consistent with the generally admitted marine origin of atmospheric
222
Na and Cl inputs37,38. NO3 concentrations vary between 1.02 mg.L-1 and 5.27 mg.L-1 with an outlier of
223
11.94 mg.L-1 measured in April 2013. Ca and SO4 concentrations vary between 0.18 mg.L-1 and 3.76
224
mg.L-1 while Mg concentrations range from 0.04 mg.L-1 in September 2013 to 0.21 mg.L-1 in August
225
2012 (Table S1). Boron
226
•
227
Dust deposition - Boron concentrations in Ap range from 150 mg.kg-1 in August 2013 to 502
228
mg.kg-1 in December 2013, with 67% of the data comprised between 300 and 400 mg kg-1 (Table 1
229
and Fig 1). These concentrations are very high compared to the mean continental crust (≈10 mg.kg-1)
230
but also compared to common soil minerals (≈10-100 mg.kg-1
231
between -4.7‰ during the winter months 2012 and +11.4‰ in December 2013 (Table 1). Over the
232
two-years sampling period, the δ11B values tend to show a bimodal distribution of boron isotope with
233
higher values (δ11B>0‰) between May and July and then between October and December. Boron
234
concentrations in Ap are statistically higher in 2012 than 2013. The annual oscillation of δ11B values in
235
2013 Ap samples is similar but of higher magnitude than in 2012.
16,31
10 Environment ACS Paragon Plus
). The δ11B values of Ap vary
Page 10 of 28
Page 11 of 28
Environmental Science & Technology
236
Dissolved deposition - Boron concentrations in Aw range from 1.1 µg.L-1 to 3.6 µg.L-1 with a mean
237
value of 2.2±1.6 µg.L-1 (±1SD, n=24, Table 1, Fig 1) and no clear seasonality is observed. A relative
238
high boron concentration is measured in July 2013 (9.1 µg L-1). The δ11B data in Aw show seasonal
239
variations over the two years ranging from 8.6‰ in March 2013 to 36.6‰ in July and August 2012.
240
December 2012 is characterized by a relative high δ11B value compared to other winter months
241
(+27‰). Overall, the ranges of boron concentrations and isotopic compositions in dissolved
242
deposition are consistent with previously published works7,12,14-17,21,22,24. Both years display similar
243
δ11B trends in Aw characterized by higher δ11B values in the summer months and lower values in
244
autumn and winter but we observe that the δ11B values in 2013 are significantly lower than those in
245
2012, especially during the summer months. We also observe that the period of highest δ11B value is
246
delayed by about one month in 2013 (max. in August-September) compared to 2012 (max. in July-
247
August).
248
Overall, dust and dissolved boron geochemical signals appear disconnected in our study. The
249
absence of correlation between boron data (concentrations and isotope) in Ap and Aw samples
250
together with the time lag between the observed seasonal oscillations of boron isotopes in dusts and
251
dissolved samples (Fig. 1) suggest that no significant boron exchange occurred between the solid and
252
liquid phases after collection. Additionally, the high boron concentrations measured in dust samples
253
contrast with the rather low ones in the recovered supernatants. It has however to be noted that our
254
sampling procedure actually allows the transfer to solution of most, if not all, the water-soluble
255
boron initially present in the collected dusts. In reverse, the slightly acidic rainwater pH makes
256
unlikely a significant removal of dissolved boron by adsorption onto the collected dusts.
257 258 259 260
Figure 1: Evolution of B concentrations and isotopic compositions (‰) of atmospheric particulate deposition (Ap) and Total dissolved deposition (Aw) in 2012 and 2013. Gray bands correspond to the time of local fertilizer spreading.
261
11 Environment ACS Paragon Plus
Environmental Science & Technology
262
Fertilizers and soils – The δ11B values of the three fertilizers analyzed in this study range from -
263
13.5 (F2) to +0.5‰ (F3) but these values may vary from one manufacturer to the other and even over
264
time because of the industrial process as well as the fluctuations of the boron market. The soil
265
untreated with boron-rich fertilizer S1 shows lower boron concentration and lower δ11B (80 mg.kg-1; -
266
7.8‰) than the fertilized soil S2 (160 mg.kg-1; -3.6‰).
267
268
Discussion
269
•
Budgets of boron atmospheric inputs During 2012 and 2013, the inputs of atmospheric dust deposition (Ap) range from 1.3 µg.m-
270 271
2
.d-1 in winter 2013 to 8.0 µg.m-2.d-1 in June 2012 with an mean annual value of 3.9 ± 2.2 µg.m-2.d-1
272
(±1SD, Table 1). Based on the measured boron concentrations and dust deposition rates, we
273
calculate a yearly-integrated particulate boron deposition of 15.5 g.ha-1.yr-1 with δ11B = -0.3‰ in
274
2012 and 10.4 g.ha-1.yr-1 with δ11B = +3.2‰ in 2013. A seasonal cyclicality of the boron dust
275
deposition rates was observed (higher during summer months and lower during winter ones), which
276
is mostly driven by the cyclicality of the dust deposition rate rather than by seasonal changes of the
277
boron concentrations (Fig. 1). The monitoring of the intensity of the rainfall events were used to
278
calculate the flux of the boron dissolved deposition. The inputs of atmospheric dissolved boron (Aw)
279
are comparable in amount to dust deposition, ranging from 1.8 µg.m-2.d-1 in August 2012 to 9.7 µg.m-
280
2
281
yearly-integrated dissolved boron deposition of 17.2 g.ha-1.yr-1 with δ11B = +21.7‰ in 2012 and 14.9
282
g.ha-1.yr-1 with δ11B = +13.4‰ in 2013. Although the rates of dissolved boron are comparable
283
between the two years, we observe a significant discrepancy of δ11B values, which is discussed later.
284
By comparison with previous studies, the atmospheric input of dissolved boron in Montiers is similar
.d-1 in July 2012 with a mean annual value of 4.8±2.0 µg.m-2.d-1 (±1SD, Table 1). We calculate a
12 Environment ACS Paragon Plus
Page 12 of 28
Page 13 of 28
Environmental Science & Technology
285
to the one in Clermont-Ferrand (5.5 µg.m-2.d-1)
286
ecosystems by a factor about 3.5 (1.4 µg.m-2.d-1) .
14
but significantly higher than other forest
12
287
From our data, we determine that dust depositions contribute to 45% of the total boron
288
atmospheric inputs to the Montiers forest. This is consistent with the value determined in a previous
289
work in which dust depositions account for half of the total boron atmospheric inputs . The reason
290
for this discrepancy has to be found in the unusual high boron concentrations measured in Ap
291
samples in Montiers.
13
292
293
•
Origin of boron in atmospheric dust
294
We observe a general decrease of the boron concentrations in dust with increasing silicon,
295
aluminum ones indicating that aluminosilicate minerals are not the main boron carriers. To the
296
opposite, boron concentrations globally increase with the complement to 100% of the sum of the
297
major cations analyzed (rock forming cations, except C), which complement is expected to be mostly
298
composed of organic compounds. In addition, correlations (r>0.75) are observed between B/Zr and
299
Na/Zr, P/Zr, Cu/Zr, Mo/Zr and Zn/Zr wt ratios (fig. S1). Here, Zr is considered as an immobile element
300
mostly derived from minerals and serve to compare elements without a firm correction of the
301
organic compounds. In particular, the lower δ11B values observed in this study correspond to the
302
sample showing the extreme Na concentrations and not Cl ones. Since boron is usually present in
303
fertilizers as sodium borate and supplied in pair with molybdenum, our results therefore point to
304
fertilized-derived organic materials as main boron carriers in dusts in Montiers. Quartz being
305
abundant in the dust samples and never showing important boron concentrations, we deduce that
306
the boron-bearing phases should be characterized by very high boron contents, likely exceeding
307
1,000 mg.kg-1. This value is unusual in geological material except in evaporite, tourmaline or
308
palygorskite that are not observed in our samples.
13 Environment ACS Paragon Plus
Environmental Science & Technology
309
Examination of boron isotopes indicates that at least three distinct components or physico-
310
chemical processes are necessary to explain our data (Fig. 2). One component, noted A, is principally
311
defined by dust samples collected in the second half of 2012 and November 2013. It is characterized
312
by a moderate boron concentration (400 mg.kg-1)
315
but with contrasted δ11B values, one +12‰ for the component noted C has no equivalent amongst the fertilizer samples analyzed in
317
this study. The component B is defined by most of the dust samples collected in 2012, except
318
November, and the early months of 2013 whereas the component C is defined by the summer
319
months of both years and December 2013. No clear relationships have been observed between δ11B
320
values and major or trace element concentrations.
321
These data tend to indicate that the B-rich fertilizers widely used in the region of Montiers
322
dope with boron the cultivated plants and likely the top soil organic and mineral phases, in particular
323
those for which boron is known to have great chemical affinity like clay minerals and iron oxides39,40.
324
This conclusion is also supported by shift in boron concentration and isotopes observed in the
325
fertilized top soil S2. It has to be noted that the fertilized soil S2 has been sieved at 50 µm, a
326
granulometric size coarser than found in atmospheric dust. We expect that the finest granulometric
327
fraction of S2 is composed of material with higher reactive surface and therefore higher boron
328
concentration. The component A would then be most representative of dust from unfertilized soils.
329
During growth, plants absorb boron without significant isotopic fractionation12 such that boron
330
isotopes in plants are sometimes used to unravel their origin . It is then expected that the organic
331
compounds present in fertilized soils have δ11B values close to the fertilizer or the mean of the
332
fertilizers spread. We therefore suggest that the component B would reflect mostly fertilized organic
333
compounds.
37
14 Environment ACS Paragon Plus
Page 14 of 28
Page 15 of 28
334 335 336 337
Environmental Science & Technology
Figure 2: δ 11B vs 1/[B] for atmospheric dusts (Ap). Dust deposited in 2012 appear mainly influenced by the boron-rich endmember A, whereas dusts deposited in 2013 tend to be influenced by the three endmembers. Endmember C, the less boron concentrated endmember, would then reflect less contaminated soils
338
However, neither the fertilizers F1, F2 and F3 nor the top soil samples analyzed in this study
339
allow to better constrain the component C characterized by high boron concentrations and high δ11B.
340
Since boron-rich fertilizers are mostly used for the culture of oleaginous products (colza and
341
sunflower) and beets, the inter-annual variations of boron concentrations and isotopes in Ap might
342
reflect crop rotation and possible changes of fertilizers and manufacturers. Note that if synthetic
343
fertilizers general present low δ11B values, some can present high values (>+20‰)40, which might be
344
consistent with endmember C. Accordingly, dusts deposited in 2012 might be mainly influenced by
345
the boron-rich source, whereas dusts deposited in 2013 would be influenced by a more balanced
346
contribution of the three sources. Contribution of the source C (high concentrations and higher δ11B
347
values) seems to be more important during the summer months and possibly at end of the year,
348
coincident or just after fertilizer spreading (fig. 1 and 2). Identification of the boron sources in the
349
dust collected in Montiers is therefore not straightforward and requires more extensive field
350
investigations including collection of larger sample size to allow chemical and isotopic analyses of
351
separate mineral and organic phases and a larger sample set, with spreading schedule, of the boron-
352
rich fertilizers used in the region. Table 1: B concentration (µg L-1), flux (µg m-2 d-1) and isotopic composition (‰) of total dissolved deposition (Aw) and dust deposition (Ap). Analytical errors (±2SD) are 1% for boron concentrations, and 0.5‰ for boron isotope analyses.
353 354 355 356
Origin of boron in dissolved atmospheric inputs
357
•
358
Assuming that all the sodium in dissolved atmospheric inputs derives from seasalt dissolution,
359
which is the upper limit, the maximum seasalt contribution to rainfall chemistry can be calculated for
360
each element X using the following equation:
15 Environment ACS Paragon Plus
Environmental Science & Technology
[] =
[]
[ ]
Page 16 of 28
× [ ] 1
361
Where [X]mar is the concentration of the element X in Aw derived from seasalt dissolution;
362
([X]/[Na])seawater is the concentration ratio of X and Na in seawater and [Na]Aw is the concentration of
363
Na in Aw. The relative seasalt contribution to the total atmospheric deposition of dissolved Ca, Mg,
364
SO4 and B is illustrated in Fig. S2. It is calculated that 33 to 100% of the total dissolved Mg can be
365
explained by dissolution of seasalts whereas only a small contribution is calculated for Ca and SO4,
366
about 12 % and 8 %. Most of the dissolved Ca deposition occurs in spring and summer months coeval
367
with the increase in dust deposition rate, which may reflect the dissolution of carbonate dust. SO4 is
368
mainly emitted as gas and fine particles by industrial, urban and agricultural activities41. From these
369
observations, we conclude that the rainwaters collected in Montiers in 2012-2013 have chemical
370
compositions consistent with a marine air mass with variable geologic and anthropic contributions,
371
which relative magnitudes depend on the elements. For boron, the contribution of seasalt
372
dissolution does not exceed 29% of the total dissolved boron during winter and decrease down to 8%
373
during summer. Since seawater is characterized by a high and invariable δ11B value (≈+40‰), the
374
calculated relative seasalt contributions are inconsistent with the lower δ11B values observed in
375
winter and the highest ones in summer, indicating that the greater seasalt injection to the
376
atmosphere during winter months is unlikely the major control of dissolved boron. Additionally, the
377
boron enrichment factor (EFsw) in each rainwater sample, defined as the B/Na chemical ratio
378
normalized to the B/Na seawater ratio, is systematically above unity (3.4 ≤ EFsw ≤ 80.5, Fig. S3).
379
Noteworthy, the seasonality of EFsw (lower values in winter and higher ones in summer) is driven by
380
seasonal variations of the sodium concentrations, in line with previous studies
381
these observations contradict the hypothesis of boron degassing after seawater bubble bursting or
382
any physico-chemical reaction at the seasalt particle surface as proposed earlier10,16 because mass
383
balance would lead Esw factors to be alternately higher and lower than unity depending on the
16 Environment ACS Paragon Plus
14,42
. In Montiers,
Page 17 of 28
Environmental Science & Technology
384
relative boron and sodium depletion in the air masses. Alternatively, possible explanations may
385
include incorporation of either mineral, biogenic or anthropogenic materials with B/Na ratio above
386
the seawater value, gradual boron depletion by rainfalls of the air masses as well as effects of
387
temperature variations on the boron partition between atmospheric vapor and liquid phases.
388
Based on SO4/B and NO3/B ratios in rain waters, previous studies have identified the leaching
389
of coal burning ashes and the spread of fertilizers as possible sources of boron to the
390
atmosphere15,21,24,43. Our samples are globally consistent with this conclusion, as shown in Fig. S4
391
and by the correlation observed between EFsw and (NO3+SO4)/Na (Fig. S3). Compared to the available
392
boron dataset in rainwaters, our samples appear particularly rich in NO3 suggesting that atmospheric
393
dissolved boron in Montiers is more affected by fertilizer spreading than by leaching of coal ashes.
394
However, the low δ11B values are frequently, but not always, associated with low boron
395
concentrations, which is inconsistent with a simple addition of anthropogenic boron to the
396
atmosphere. The combination of low boron concentration with low δ11B has been earlier observed
397
and explained by evaporation of gaseous boric acid from the seawater and subsequent integration in
398
rains of isotopically fractionated boron. Boron isotopes in rains are then modeled by a Rayleigh-like
399
distillation process:
400
=
Eq 2
401
With R and R0 being the isotopic ratios of the rainwater and the original atmospheric gaseous boron
402
reservoir, respectively. ƒ is the mass fraction of residual gaseous
403
fractionation factor between the gaseous reservoir and the rainwaters. Note that α actually
404
integrates a series of reactions including evaporation/condensation and interactions with aerosol
405
surface and thus may vary from one site to the other. Using the empirical α value (α=1.031)
406
empirically determined by Rose-Koga et al.17 and assuming the coastal rains from Brest (Western
407
France) as representative of the initial air mass, the model can explain most of the boron data
408
measured in Montiers. However, a Rayleigh-like model alone cannot explain neither the existence of
10
17 Environment ACS Paragon Plus
B and α is the boron isotopic
Environmental Science & Technology
409
boron-rich rainwaters with low δ11B values (Fig. 3) nor the systematic difference of the mean δ11B
410
values measured in 2012 and 2013 because it would imply important variations of the atmospheric
411
conditions (temperature, relative humidity)17,45, which are not observed.
412
Finally, we propose a model of boron sources in the dissolved atmospheric deposits
413
combining a condensation of gaseous boron, most likely originating from seawater evaporation, a
414
small contribution of seasalts dissolution and episodic inputs of anthropogenic boron, mostly
415
deriving from the nearby fertilizer spreading. Condensation of marine-derived gaseous boron would
416
be the dominant process during summer months when the higher temperatures promote the
417
evaporation of boron. This will lead to high δ11B values and moderate boron concentration. In winter
418
months, the dissolution of seasalts may also partially contribute to the dissolved boron deposition
419
thus maintaining δ11B values around +10‰. Fertilizer spreading episodically affects the rain
420
chemistry by adding significant amount of boron with an isotopic composition close to +10‰. Taken
421
from published data that anthropogenic sources are characterized by boron concentrations ranging
422
from 20 µg.L-1 to 50 µg.L-1 and δ11B from 0‰ to +10‰ 7,14,15, a mass balance calculation leads to an
423
global anthropogenic contribution below 10% of the total dissolved boron with the exception of
424
fertilizer spreading periods, when it may reach 25%. Additionally, the summer δ11B maximum in Aw is
425
delayed by 1 or 2 months compared to Ap. The difference of residence time in the atmosphere
426
between particulate (few days) and gaseous boron (about one month)1010 combined to the time of
427
fertilizer transfer from soil to atmospheric dust deposits may explain this delay.
428 429 430 431 432 433 434 435
Figure 3: δ11B values plotted as a function of B concentration (µg L-1) in rainwaters of Montiers (this study), Brest (West France) 7, Clermont Ferrand (Center France) 7, Dax (South-West France)7, Orleans (Center France) 7, Paris (Center France) 15, Guiana 14, Reunion 45, Nepal 18, Quebec 19. This figure also displays the two mechanisms controlling boron in Aw: a Rayleigh distillation model (α = 1.031, after Rose-Koga et al.17) using coastal rains as the origin of air masses and a secondary contribution (biogenic/anthropogenic). The latter has been roughly evaluated to contribute up to 10% to the overall Aw boron budget. The dashed lines illustrate trends corresponding to the Rayleigh model using extreme rainwater values measured in Brest.
436
18 Environment ACS Paragon Plus
Page 18 of 28
Page 19 of 28
Environmental Science & Technology
Environmental implications
437
•
438
With regard to the critical physiological role of boron, this study suggests that the atmospheric
439
boron fluxes need to be considered to establish conditions of a sustainable management of
440
ecosystems. In particular, an advance of this work is to recognize that the boron input by dust
441
deposition to forest ecosystem about equals the boron dissolved flux by rainwaters. Additionally,
442
boron concentrations and isotopes point to a significant impact at regional scale of fertilizers on the
443
boron deposition by dust. By comparison, the boron deposition by dust represents about 1/10 of
444
boron recycled each year by the vegetation of the nearby Vosges mountains12.
445
Beyond the evidence of a regional fertilizer imprint on boron atmospheric depositions, this study
446
emphasizes the transfer of contaminants through clay-size dusts. Although dust deposition
447
constitutes a small flux at the global scale compared to rivers or aquifers it could affect a wide area.
448
The schematic boron transfer in the atmosphere presented in Fig. 4 could therefore serve as a
449
framework for geochemical models including a more detailed description of the physico-chemical
450
properties of the gas/water/particles exchanges.
451 452 453 454
Fig 4: Conceptual model of both dissolved boron (Aw) and particulate boron (Ap) dynamics in the atmosphere in the forest ecosystem of Montiers. Aw seem to mostly derive from condensated gaseous boron with a contribution (