1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Appendix S1
Site description
Rifle, Colorado (USA) experimental location
Detailed descriptions of the mineralogy, microbiology, and hydrogeology of the
Integrated Field Research Challenge (IFRC) site at Rifle, Colorado have been presented
elsewhere. The current test plot encompasses an area of 5 m by 16 m with one
background monitoring well (B-04) placed 3.7 m upstream of an injection gallery
positioned perpendicular to groundwater flow. Three staggered down-gradient
monitoring wells were placed at distances of 3.7 m (M16), 7.3 m (M17), and 14.6 m
(M18) from the injection gallery. Acetate and bromide were delivered through a
peristaltic pump for 3 weeks and mixing occurred at the injection point by circulating
groundwater through cross-well mixing tubes. Previous biostimulation work at the site,
a year prior to this experiment, had involved introduction of sodium acetate into the
aquifer in quantities sufficient to achieve a target in situ concentration of 10 mM for
approximately 30 days. The cross-well mixing tubing found to host the selenium-rich
biofilm community after 14 days was isolated for direct exposure to the sun during the
experiment through a combination of enclosure within coolers and obfuscation under
tarps and within sealed boreholes.
Experimental procedures
Groundwater Analysis
The groundwater was collected at different time points during the field experiment and
stored at 4˚C until analysis. Acetate was quantified via ion chromatography using a
Dionex ICS-2100 (Dionex Corp., Sunnyvale, CA) equipped with an AS18 analytical
column. Total selenium, selenate and selenite concentrations were determined using an
ICP-MS (Elan DRC-e, PerkinElmer, Wellesley, MA). Speciation of selenate from selenite
was achieved through ion chromatography with an AS4 ion-exchange column (Dionex,
Sunnyvale, CA) using a 1.8 mM carbonate/1.7 mM bicarbonate buffer. Two isotopes of
selenium (80Se and 78Se) were quantified. The effluent from the column was then
pumped directly into the ICP-MS. The bromide signal (78Br) present in all samples during
acetate addition created an interference with the signal for 80Se. Therefore, only the
data for 78Se is presented.
Oxygen diffusion associated with low-density polyethylene (LDPE) tubing
The LDPE tubing used for cross-well mixing and which hosted the selenium-rich biofilms
was sourced from Cole Parmer (Vernon Hills, IL) and specified as having a gas
permeability (µ) for molecular oxygen (O2) value of 21.5 cm3 cm cm-2 day-1 atm-1 at 25°C.
Elemental mapping of selenium biofilms using energy dispersive x-ray spectroscopy (EDS)
analysis
The EDS analysis was performed using a 20keV accelerating voltage and current of ca.
1.0 nA with a dead time of ca. 30%. EDS data were collected using a Si-Li detector over a
count period of 100 seconds. 4-6 points were analyzed within the field of view (two such
1
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
points are indicated on Figure S5 along with their corresponding EDS spectra). Given a
limited number of points/areas of analysis, quantitative determination of the weight
percentages of the elements present (e.g. selenium) was not possible; however,
qualitative determination of the presence of specific elements was possible.
16S rRNA-based Microbial Community Analysis of Tubing Biofilms
Cross-well mixing tubing. The cross-well mixing tubing was collected from the field upon
cessation of acetate addition. The tubing was then flash frozen and shipped immediately
to the laboratory. Samples were stored in a freezer (-80˚C) until extraction. Fivecentimeter long segments of the tubing were introduced into 15 mL sterile conical tubes
and used to selectively extract two operationally defined fractions of the biofilm:
Loosely bound and tightly bound fractions. Fresh water media (composition described
below) was added to the conical tubes and they were gently vortexed to dislodge
loosely bound cells from the walls of the cross-well mixing tubing, followed by
centrifugation for 15 minutes at 4000 rpm to pellet the cells. The supernatant was
discarded and the rest of the biofilm was scraped off of the cross-well mixing tubing to
collect the tightly bound fraction. The DNA from both fractions was then extracted using
a FastDNA Spin for Soil kit (Q-BIOgene, Solon, OH) and following manufacturer’s
instructions. The DNA was then amplified and sequenced as previously described
(Vrionis et al., 2005).
Cultivation of Se-reducing microorganisms
Groundwater from Rifle aquifer (well D04; 2006) was pumped to the surface, dispensed
into sterile flasks, and shipped overnight at 4oC to the laboratory at UC Berkeley.
Anaerobic enrichment cultures (N=6) were started by inoculating 2 mL Rifle
groundwater into 48 mL synthetic Fresh Water Media (FWM) containing (in g L-1 of milliQ water): NaHCO3 (2.5), NH4Cl (0.25), NaH2PO4  H2O (0.6), KCl (0.1) (Lovley and Phillips,
1988). 10 ml of a vitamin mix and 10 ml of a mineral mix was also added per liter of
media in addition to acetate (20 mM) and selenate (10 mM). Incubations were carried
out in 100 mL serum bottles sealed with thick butyl rubber stoppers and stored at 25 oC
in the dark for ~3 weeks. Following the reduction of selenate to elemental selenium,
enrichment cultures were transferred on two occasions by inoculating a small volume
(~2 mL) of growing culture into fresh media and incubating under the same conditions.
A similar procedure was used to generate selenium-reducing enrichment cultures using
scrapings from the 2006 tubing biofilms as the inoculum. Both inocula (groundwater vs.
biofilm scrapings) yielded enrichment cultures dominated by Dechloromonas like
strains, with groundwater-based enrichments repeatedly showing quasi-pure cultures
closing resembling D. aromatica RCB) and tubing-based enrichments (i.e., scrapings)
showing greater diversity given an identical incubation period (3-weeks followed by
successive (2x) transfers).
16S rRNA-based Microbial Community Analysis of Enrichment Cultures
DNA was extracted from a culture that had been re-subbed on three occasions using a
MoBio Powersoil DNA extraction kit (MoBio, Carlsbad, CA, USA). 16S rRNA genes were
2
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
then amplified using the broad-specificity primers 27F (5’-AGAGTTTGATCCTGGCTCAG3’) and 1492R (5’-GGTTACCTTGTTACGACTT-3’). Reaction mix was made up to 50 µl and
consisted of 36.75 µl sterile water, 5 µl 10x PCR buffer, 4 µl 10 mM dNTP mix, 1 µl of
each primer at 25 µM, 0.25 µl TaKaRa Taq DNA polymerase (Takara Bio, Shiga, Japan)
and 2 µl purified DNA template. The purity of the amplified product was determined by
electrophoresis of 10 l samples in a 1.0% agarose Tris-acetate-EDTA (TAE) gel. DNA was
stained with ethidium bromide and viewed under short-wave UV light. To avoid
contamination, PCR mix was exposed to UV radiation for 30 min before the addition of
the DNA template, Taq and the dNTPs. PCR amplification was performed with an initial
denaturation step at 94oC for 2 min followed by 24 cycles of 92oC (45 s), 50oC (45 s) and
72oC (1 min 45 s) with a final extension step at 72 oC for 10 min. Clone libraries were
constructed using an Invitrogen cloning kit according to the manufacturer’s instructions.
47 clones were sampled using sterile toothpicks and inoculated into 1.8 ml LB media and
incubated overnight at 37oC to grow up cell mass. Clonal 16S rRNA inserts were
sequenced at the UC Berkeley sequencing facility using the T7/M13 primer set. 16S
rRNA gene fragments (typically about 1400 b.p in length) were analyzed against the
NCBI (USA) database using BLAST program packages and matched to known 16S rRNA
gene sequences. Matched sequences were aligned using the Near Alignment Space
Termination (NAST) Greengenes algorithm (http://greengenes.lbl.gov) and imported
into ARB. Phylogenetic trees were inferred using the neighbor-joining method.
Estimation of Se-removal rates and efficiency
Conceptually analogous to our field system, Chung and colleagues (2006) report the
removal efficiency of selenate as Se(0) within a laboratory membrane biofilm reactor as
a function of Se(VI) flux and H2 partial pressure. They report their values in terms of
Se(VI) flux (units of grams of Se per m2 of biofilm/tubing surface area per day) and
percent removal of total selenium from which an accumulation rate in units similar to
those reported here can be calculated (grams of Se(0) per m2 of tubing per day). From
the data presented in our experiment, we can estimate the flux of selenium (as both
selenate and selenite; Figure 1) entering the borehole containing the tubing in which
selenium reduction and concentration as Se(0) were associated with the biofilm
community. This flux was estimated by calculating the saturated surface area of the
well into which selenium-bearing fluids were advecting (0.48 m2 of which only the
upgradient half of the surface area was assumed to contribute, so 0.24 m2). Given an
effective porosity of the Rifle alluvium of 20% (Williams et al., 2011), this yields an
available area for groundwater advection of 0.048 m2, which can be used to calculate a
flow rate into the well bore by multiplying by the average pore water velocity observed
for the Rifle aquifer of 0.50 m day-1 (flow rate=0.024 m3 per day or 24 L per day). As
selenium-bearing fluids advecting into the tubing biofilm well were derived from an
upgradient source (e.g. well B04; Figure S1 and illustrated conceptually in Fig. 4), we
elect to use an average total selenium (Se(VI,IV)= selenate + selenite) concentration of
60 µg/L (i.e., that reported at well B04). Using the flow rate (24 L/day) and the
concentration (60 µg/L), we can calculate a flux of 1440 µg Se(VI,IV) per day. Assuming
that 100% of this flux is pulled into the tubing hosting the biofilm community,
3
133
134
135
136
137
138
139
normalizing the flux of selenium by the tubing surface area (0.1197 m 2) yields a value of
0.012 g Se m-2 d-1. Given our extraction results yielding a Se(0) deposition rate of 9 mg
Se m2 d-1 and the calculated flux of selenium oxyanions of 0.012 g Se m -2 d-1, we
calculate an efficiency of Se removal within the tubing biofilms of 75%. Such a value is
consistent with the values reported by Chung and colleagues (2006) that ranged from
62% to 94% depending upon Se(VI) flux rate and H2 partial pressure.
4