Subscriber access provided by Nanyang Technological Univ
Letter
Use of Stable Isotope Signatures to Determine Mercury Sources in the Great Lakes Ryan Lepak, Runsheng Yin, David P. Krabbenhoft, Jacob M. Ogorek, John F. DeWild, Thomas M. Holsen, and James P. Hurley Environ. Sci. Technol. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.estlett.5b00277 • Publication Date (Web): 12 Nov 2015 Downloaded from http://pubs.acs.org on November 17, 2015
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 Letters 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 Letters
1 1
Use of Stable Isotope Signatures to Determine Mercury Sources
2
in the Great Lakes
3 4
Ryan F. Lepak1, Runsheng Yin1,2, David P. Krabbenhoft3, Jacob M. Ogorek3,
5
John F. DeWild3, Thomas M. Holsen4 and James P. Hurley*1,5,6
6 1
Environmental Chemistry and Technology Program, University of Wisconsin-Madison, Madison, WI 53706, USA
7 8 9 10 11 12
2
State Key Laboratory of Ore Deposit Geochemistry, Institute of Geochemistry, Chinese Academy of Sciences, Guiyang 550002, China 3
13 14 15
4
16 17 18
Department of Civil and Environmental Engineering, Clarkson University, Potsdam, NY 13699, USA
5
Department of Civil and Environmental Engineering, University of Wisconsin-Madison, Madison, WI 53706, USA
19 20 21 22
United States Geological Survey, Wisconsin Water Science Center, Middleton, Wisconsin 53562, USA
6
University of Wisconsin Aquatic Sciences Center, Madison, WI 53706, USA
23 24
25
26
* corresponding author
27
James P. Hurley, University of Wisconsin-Madison, Environmental Chemistry and
28
Technology Program, Madison, WI 53706;
[email protected]; 608-262-0905.
ACS Paragon Plus Environment
Environmental Science & Technology Letters
Page 2 of 28
2 29
Abstract
30
Sources of mercury (Hg) to Great Lakes sediments were assessed with stable Hg isotope
31
ratios using multi-collector inductively coupled plasma mass spectrometry. An isotopic
32
mixing model based on mass dependent and mass independent fractionation (δ202Hg and
33
∆199Hg), identified three primary Hg sources to sediments: atmospheric, industrial and
34
watershed-derived. Results indicate atmospheric sources dominate in Lakes Huron, Superior,
35
and Michigan sediments while watershed-derived and industrial sources dominate in Lakes
36
Erie and Ontario sediments. Anomalous ∆200Hg signatures, also apparent in sediments,
37
provided further validation of the model. Comparison of ∆200Hg signatures in predatory fish
38
from three lakes reveal that bioaccumulated Hg is more isotopically similar to
39
atmospherically derived Hg than a lake’s sediment. Previous research suggests ∆200Hg is
40
conserved during biogeochemical processing and odd MIF is conserved during metabolic
41
processing so it is suspected even MIF is similarly conserved. Given these assumptions, our
42
data suggest that in some cases, atmospherically derived Hg may be a more important source
43
of MeHg to higher trophic levels than legacy sediments in the Great Lakes.
ACS Paragon Plus Environment
Page 3 of 28
Environmental Science & Technology Letters
3 44
INTRODUCTION Mercury (Hg), a persistent global pollutant, is transported atmospherically and enters
45 46
freshwater systems through direct atmospheric deposition (wet and dry), through watershed
47
inputs, or directly through point-source discharge.1-3 In aquatic ecosystems, microbially-
48
mediated pathways can transform Hg to methyl mercury (MeHg), a more toxic form that is
49
biomagnified in aquatic food webs.4 Bioaccumulated MeHg is a potential threat to aquatic
50
ecosystems and to human health through fish consumption. Anthropogenic activity has
51
largely altered sources of Hg and our understanding of the impacts to the aquatic
52
environment continues to emerge.4 The use of multi-collector inductively coupled plasma
53
mass spectrometry (MC-ICP-MS) to accurately resolve and measure isotopic ratios of the
54
seven natural stable isotopes of Hg (196Hg, 198Hg, 199Hg, 200Hg, 201Hg, 202Hg and 204Hg) has
55
enhanced our understanding of both sources and biogeochemical processes.5,6 Specifically,
56
determination of mass dependent (MDF) and mass independent (MIF) fractionation patterns
57
provides critical information useful for management of Hg contamination in valued aquatic
58
resources. Mass dependent fractionation, identified by delta notation, is calculated by utilizing a
59 60
standard bracketing solution equation (1): δxxxHg(‰) = {(xxxHg/198Hgsample)/( xxxHg/198HgNIST-3133)－1}×1000 (Equation 1)
61 62 63
where xxxHg is the isotope of interest, NIST-3133 is the chosen normalization solution, and 198
Hg is the denominator.7 MDF can occur either during kinetic processes or from
64
equilibrium exchange of Hg. Processes affecting MDF include: microbially-mediated
65
reactions (e.g., reduction, methylation, and demethylation); abiotic chemical reactions (e.g.,
ACS Paragon Plus Environment
Environmental Science & Technology Letters
Page 4 of 28
4 66
photo-reduction and chemical reduction); and, physical processes (e.g., volatilization,
67
evaporation, adsorption, and dissolution).5,6
68
An advantage of Hg isotope geochemistry for understanding sources and
69
transformation processes is use of mass independent fractionation (MIF)8,9 data signified by
70
capital delta (∆) calculable by:
71
∆xxxHg ≈ δxxxHg - (δ202Hg * β).
(Equation 2)
72
MIF describes the difference between a β scaled δ202Hg values and measured δxxxHg values.
73
The scaling factor, β, is an independent isotope-specific constant determined by the
74
theoretical laws of MDF.7 Specific processes resulting in odd MIF, a result of nuclear volume
75
and magnetic isotope effects,10,11 include photo-reduction of aqueous Hg (II), photo-
76
degradation of MeHg, elemental Hg evaporation, and equilibrium Hg-thiol complexation.5
77
More recently, MIF anomalies of even isotopes (e.g., 200Hg) have been reported in elemental
78
gaseous Hg(0) and atmospheric precipitation (as divalent Hg (II)). This has led to the
79
hypothesis that atmospheric photochemical oxidation of Hg alone causes MIF for even
80
number Hg isotopes.5,9 To date, large variations (~10‰) of δ202Hg, ∆199Hg, and ∆200Hg
81
(~1‰) have been reported. 9,12-15 We propose that these three isotopic markers of MDF and
82
MIF are effective tracers of biotic and abiotic fractionating processes that operate to differing
83
degrees in various environmental compartments (e.g., atmosphere, watershed, and within
84
lake), thus resulting in an ability to discriminate among mercury sources to receiving points,
85
as well as infer important Hg geochemical transformation processes.5
86 87
The Laurentian Great Lakes are presently the focus of a large scale ecosystem restoration effort by the U.S. and Canada (e.g., Great Lakes Restoration Initiative16). The
ACS Paragon Plus Environment
Page 5 of 28
Environmental Science & Technology Letters
5 88
lakes currently range from oligotrophic conditions in the western Great Lakes (Superior,
89
Michigan, Huron) to oligo-mesotrophic (Ontario) to mesotrophic (Erie) in the Eastern Great
90
Lakes. However, specific embayments (i.e. Green Bay, Lake Michigan; Saginaw Bay, Lake
91
Huron) and western Lake Erie exhibit eutrophic conditions.16 Previous Hg research has
92
focused on loading (atmospheric and tributary inputs) and transformation of Hg within the
93
lakes, including biological uptake at river mouths, particle partitioning and sedimentation,
94
methylation in both benthic waters and sediments, and MeHg bioaccumulation in pelagic
95
food webs.1,2,17-19 For the two lakes (Superior and Michigan) where enough data has been
96
collected to support a mass balance, those calculations support the notion that atmospheric
97
deposition is by far, the largest Hg source.1,19,20
98 99
The application of Hg studies on the Great Lakes has increased dramatically in the past few years, particularly for providing a better understanding of atmospheric deposition
100
signatures, cycling within the watershed, and observations of isotopic distribution in
101
fish.8,9,14,21,22 Here, we extend these recent studies and propose the use of Hg isotope
102
signatures to infer relative contributions of mercury from key end member sources to the
103
Great Lakes. Specifically, we hypothesize that the use of δ202Hg, ∆199Hg, and ∆200Hg
104
measurements on sediment samples from across the Great Lakes can be used to derive a tri-
105
linear mixing model that provides estimates of the relative amounts of mercury derived from
106
atmospheric deposition, watershed runoff, and point-source (near field) Hg sources. The
107
results from our Hg source model are then extended to recent results from Hg isotope
108
measurements in predatory fish to ascertain the relative importance of our end-member
109
sources to the food web.
ACS Paragon Plus Environment
Environmental Science & Technology Letters
Page 6 of 28
6 110 111
MATERIALS AND MEHODS
112
Sample Collection. Surface sediment samples were collected using a ponar dredge sampler
113
from the U.S. Environmental Protection Agency’s R/V Lake Guardian in 2012-2014.
114
Sediment was subsampled (0-2 cm; n=58), promptly stored, and frozen in either clean Teflon
115
vials or acrylic centrifuge tubes. Surface sediment (0 to 2 cm) collected from each lake
116
integrate different time periods ranging from 3 years (Western Lake Erie) to 220 years
117
(offshore Lake Huron) as sedimentation rates range from 0.0043 to 0.88 g cm-2 yr-1.23 These
118
temporal differences in sediment accumulation time periods for the top 2 cm is a limitation of
119
our model. Upon return to the lab, samples were lyophilized, homogenized by a ball mill, and
120
stored in borosilicate vials. Lake trout samples from Lakes Superior and Ontario were
121
collected by Clarkson University during USEPA monitoring cruises in 2006. Each sample
122
represents a homogenized composite of five equally portioned samples of fish axial tissue.
123
Upon receipt in the laboratory, samples were lyophilized, homogenized by a ball mill, and
124
stored in borosilicate vials
125
Total Hg analysis. HgT in sediments was analyzed by a direct combustion system with
126
atomic absorbance detection system24 at the USGS Mercury Research Lab. Standard
127
reference material (SRM) recoveries (IAEA SL 1) were within 90 – 110%, detection limit
128
was 1.38 ng g-1 dry weight, and coefficients of variation of triplicate analyses were less than
129
10%. HgT analyses of fish tissues were performed by Clarkson University using methods
130
based on US EPA 7473 and atomic absorbance detection.25 Lake Superior Fish Tissue (dried,
ACS Paragon Plus Environment
Page 7 of 28
Environmental Science & Technology Letters
7 131
NIST 1946) SRM was used to validate accuracy with recoveries of 90-110%; relative
132
differences between replicates were less than 10%.
133
Hg isotope analysis. Sediment samples were digested in 5 mL of aqua regia (3;1 ratio of HCl
134
and HNO3, v:v) in a water bath (95 °C, 120 mins). Solutions were diluted with Milli-Q water
135
to a Hg concentration of 0.5 to 2.0 ng mL-1 and to less than 20% acid. Two SRMs, NIST
136
2711, and MESS–1, were used in the isotopic analyses both to conform to previous research
137
and to establish a SRM with a concentration similar to the samples measured here (recoveries
138
were 90 - 100%; supplemental section).26, 27 Fish samples were digested in 5 mL of
139
concentrated nitric acid in a water bath (95 °C, 180 mins) and subsequently diluted to 20%
140
acid with 5% bromine monochloride. For fish analyses, two SRMs, DORM-2, and IAEA-
141
407 were used in the isotopic analyses to both conform to previous research, again choosing
142
SRMs with concentrations similar to the samples measured in our study (recoveries were 90 -
143
100%; supplemental section). Differences between Hg concentration in the bracketing
144
solutions (NIST SRM 3133) and samples were less than 10%. Hg isotope ratios were
145
determined using a Neptune Plus MC-ICP-MS coupled with an Apex-Q nebulizer and cold
146
gas phase introduction system (Table S1) housed at the Wisconsin State Lab of Hygiene.7
147 148
RESULTS AND DISCUSSION
149
Development of a Source Apportionment Model for Great Lakes Sediments
150
Our analyses of total Hg (HgT) in sediments reveal a wide concentration range across
151
Great Lakes sediment, which is consistent with previous studies.29 Lowest HgT
152
concentrations were observed offshore in Lakes Huron and Superior, higher concentrations in
ACS Paragon Plus Environment
Environmental Science & Technology Letters
Page 8 of 28
8 153
Western Lake Erie and in Lake Ontario (Fig. 1A; Table S2). Regional increases in Hg
154
concentration relative to offshore were apparent in Lake Michigan (Green Bay) and Lake
155
Superior sediment (Thunder Bay and near the St Louis River).
156
The use of Hg isotopes in sediments for Hg source apportionment studies are
157
typically based on binary and ternary mixing models with well-defined end-
158
members.13,14,26,28 Prior mercury research on the Great Lakes that did not employ Hg isotopes
159
have pointed to three possible Hg sources to sediments: local industrial sources, atmospheric
160
deposition, and watershed loading.2,12,14,22 Here, we demonstrate a mixing model based on
161
literature-supported end-member Hg isotope composition to support source apportionment
162
estimates across the Great Lakes.
163
Industrial sources. Polluted industrial sources have been shown to exhibit higher δ202Hg
164
values (-1 to 0‰) and insignificant MIF (∆199Hg~0).26,27 Within the Great Lakes, Hg derived
165
from industrial sources will be strongly partitioned to particles or dissolved organic carbon
166
(DOC) prior to entering the lakes. While Hg associated with DOC may undergo many
167
complex reactions that dramatically alter Hg signatures, Hg associated with particles most
168
likely deposit near their source.30 For this reason, limited water column processing may occur
169
to particle associated Hg, therefore conserving the odd MIF of the source Hg.31,32 We chose
170
NIST 3133 as our end-member because currently in the literature no relatable published
171
industrial signature has been identified for the Great Lakes system. A location in Lake
172
Ontario, elevated in HgT concentration (327 ng g-1), is quite similar in isotopic composition
173
to NIST 3133 (δ202Hg -0.14 ± 0.03‰; ∆199Hg 0.03 ± 0.01‰). Our chosen end-member is
174
similar to an industrial sourced end-member in another isotopic study.26
ACS Paragon Plus Environment
Page 9 of 28
Environmental Science & Technology Letters
9 175
Watershed soils. Hg in terrestrial soils is primarily a result of atmospheric Hg deposition and
176
subsequent evasion equilibrium, resulting in negative Hg-MIF (∆199Hg
