Content uploaded by Karsten - Pedersen
Author content
All content in this area was uploaded by Karsten - Pedersen on Jun 21, 2014
Content may be subject to copyright.
Journal
of
General Microbiology
(1992),
138,
369-376.
Printed in Great Britain
369
Incorporation
of
CO,
and introduced organic compounds by bacterial
populations in groundwater from the deep crystalline bedrock
of
the
Stripa mine
KARSTEN PEDERSEN"
and
SUSANNE EKENDAHL
Department
of
General and Marine Microbiology, University
of
Goteborg, Carl Skottsbergs gata
22,
S-413
19
Goteborg, Sweden
(Received
20
May 1991; revised 19 August 1991; accepted
8
November 1991)
~~
The incorporation of C02 and assimilation of introduced organic compounds by bacterial populations in deep
groundwater from fractured crystalline bedrock has been studied. Three depth horizons
of
the subvertical borehole
V2 in the Stripa mine, Sweden, 799-807m, 812-820m and 970-1240m, were sampled. The groundwaters,
obtained from fracture systems without close hydraulic connections, were anoxic and had the following
physicochemical characteristics: pH values of 9.5,9.4 and
10.2;
Eh
values of
+
205,
+
199 and
-
3
mV;
sulphide,
0,106
and 233
p~;
Cog-,
158,-
and 57
p~;
CH4, 245,170 and 290
pl
I-';
and
N2,
25,31 and 25 ml
I-'.
Biofilm
reactors, each containing a series of parallel glass surfaces, were connected to the groundwaters issuing from these
depth horizons at flows of approximately
1
x
1O-j
m
S-'
during
two
periods
of
two
and
four
months. There were
from
1.8
x
lo3
to 1.2
x
lo5
bacteria
per
ml groundwater and from 1.2
x
lo6
to 7.1
x
lo6
bacteria
per
cm2 of
colonized test surface. These results imply that the populations of attached bacteria are several orders of
magnitude greater than those of unattached bacteria in bedrock fractures
with
flowing groundwater. The
incorporation of 4C02,
[
14C]formate, (U-14C]lactate,
[U-14Clglucose
and ~-(4,5-~Hlleucine by the bacterial
populations was demonstrated using microautoradiographic and liquid scintillation counting techniques. The
measured
C02
incorporation reflected the
in
sifu
production of organic carbon from C02. Incorporation of
formate followed that of
C02
and indicated the presence of bacteria able to substitute formate for C02, e.g.
methanogenic bacteria. The presence of sulphate-reducing bacteria
is
suggested by the observed incorporation
of
lactate
by
up
to
74%
of the bacterial populations. The recorded uptake of glucose indicates the presence of
heterotrophic bacteria other than sulphate-reducing bacteria. Up to
99
%
of
the populations incorporated leucine,
showing that major fractions of the populations were viable.
Introduction
In recent years there has been an increasing interest in
the microbiology of fractured rock. One reason is that
many countries are seriously considering using fractured
rock as the final repository for nuclear waste, at depths
ranging from a few tens of metres for low and inter-
mediate waste, up to more than a kilometre for high-level
waste. There is considerably less information and
experience on depths below a few hundred metres than at
shallow depths (Ghiorse
&
Wilson, 1988; West
et
al.,
*
Author
for
correspondence. Tel.
031 418700;
fax
031
826790.
Abbreviations
:
AODC, acridine orange direct count; DOC, dissolved
organic carbon; LSC, liquid scintillation counting; MARG, micro-
autoradiography
;
SRB,
sulphate-reducing bacteria.
1982, 1985).
A
more detailed analysis of microbial
interactions with nuclear waste reveals complex and
difficult research areas (West
et al.,
1985), made more
difficult by the classical problem
of
sample disturbance
and contamination during the drilling of boreholes.
Further, the investigations have to be aimed at assessing
the long-term safety of nuclear waste disposal, where
time scales range from hundreds to several millions
of years due to the very long half-lives of many
radionuclides.
Dissolution and transport by the groundwater are the
most important dispersion mechanisms for radionuclides
eventually released from the waste. There are currently
large national research programmes in Canada, Switzer-
land and Sweden for the study of radionuclide transport
in crystalline rock.
In
addition an international field
0001-6941
O
1992
SGM
370
K.
Pedersen and
S.
Ekendahl
research programme organized by OECD/NEA in the
Stripa research mine, Sweden, has entered its final
phase
.
The presence of bacteria can influence the transport of
radionuclides in different ways.
Free-living bacteria
constitute mobile suspended particles which may have a
radionuclide-sorbing capacity higher than that of the
surrounding rock (Strandberg
et al.,
1981
;
Beveridge
&
Fyfe, 1985
;
Pedersen
&
Albinsson, 1991). Radionuclide
transport will then proceed faster with, than without
bacteria. If, on the other hand, the majority of the
bacteria are growing in biofilms on fracture surfaces,
transport of radionuclides may be reduced. Finally,
bacterial production of complexing agents and other
metabolites may affect speciation and thus mobility of
radionuclides independently of whether the bacteria are
attached or not.
The relevance of different mechanisms for radio-
nuclide transport by bacteria can only be evaluated with
a comprehensive knowledge of the ecology and physi-
ology of the bacterial populations that may inhabit a
nuclear waste repository and its surroundings. Identifi-
cation of their nutritional strategies is an important step
in this understanding. Autotrophic bacteria may provide
organic matter for heterotrophic organisms if
H2
and
C02
are available for the autotrophs, e.g. for acetogenic
bacteria (Fuchs, 1986; Wood
&
Ljungdahl, 1991),
methanogenic bacteria (Belyaev
et
al.,
1983; Belyaev
&
Ivanov, 1983; Godsy, 1980) and several species of
sulphate-reducing bacteria (SRB) (Fauque
et
al.,
1991).
Fermentative or respiratory utilization of geological
deposits of organic material are other possible expla-
nations as to why heterotrophic bacteria have been found
in deep geological formations (Chapelle
et
al.,
1987,
1988; Hicks
&
Fredrickson, 1989; Pedersen
&
Ekendahl,
1990).
In this study, we have assayed the total numbers of
bacteria in the groundwater and on surfaces exposed to
slowly flowing groundwater from a borehole called
V2,
which has been well characterized during the geochemi-
cal investigations of the Stripa groundwaters (Nord-
strom
et
al.,
1985). The nutritional responses of the
unattached and attached bacterial populations were
studied by measurement of their incorporation of C02
and different radiolabelled organic compounds using
microautoradiographic (MARG) and liquid scintillation
counting (LSC) techniques. Incorporation of C02 was
used to assay the
in
situ
production of organic carbon
from
C02.
Methods
Description
of
-study site.
The Stripa mine (central Sweden; 15"5'E,
59'40") has been a deep underground research facility since 1976
when the iron ore was mined out. The ore consisted of a quartz-banded
haematite and occurred in a lepatite formation. Adjacent to the lepatite
is a large body of medium-grained granite through which borehole V2
runs. It is a subvertical shaft with a diameter of 76 mm and runs from
one of the deepest drifts of the mine, 410 my down to a depth of 1240 m.
It was drilled to 860 m in 1977 and continued to 1240 m in 1981. A
number of sampling depths of this artesian borehole were closed off
with packers made of inflatable 76 mm rubber tubes and connected to
the drift with 6 mm Teflon tubing. The sampling depths used in this
study were 799-807 m, 812-820 m and 970-1240 m below ground.
There were approximately 2 fractures per m of the borehole, sealed or
coated mainly with chlorite, epidote and calcite (Nordstrom
et al.,
1985). Because of silicate weathering, the pH
of
the groundwater
approaches 10.
Water and gas analysis.
The pH and the
Eh
were measured
in situ
in
the mine with a PHM Autocal pH meter (Radiometer), a GK2421C
combined pH electrode and redox electrode PK1401. Sulphate was
measured turbidimetrically with BaC1, (Franson, 1985), sulphide with
an iodometric method (Franson, 1985) and the CO3- content with a
coulometer (model 501 1
C02
Coulometer) (Huffman, 1977). These
procedures were repeated after a 30 d interval.
Gas pipettes (100 ml) were connected to the flowing groundwaters,
left for 17 h and closed. The dissolved gases of the waters were
extracted by degassing the samples at
<
40 Pa, after which the total gas
was collected in a burette fitted with a septum by step-wise pumping
with mercury and compressed air. The gas composition was analysed
with a Perkin-Elmer gas chromatograph supplied with two columns
[Porapak N 80-100 mesh 4 m
x
1/8" (0.32 cm) and Molecular Sieve 5A
60-80 mesh: 2'
(61
cm)
x
1/8"
(0.32
cm)], a thermal conductivity
detector and a flame ionization detector. Carbon monoxide and carbon
dioxide were converted to methane before detection. The carrier gas
used was argon except for the hydrogen analysis, when helium was
used. This procedure was repeated after a 30 d interval.
Attachment and growth
of
bacteria.
Biofilm reactors (Pedersen, 1982;
Pedersen
et al.,
1986) were connected to the flowing groundwaters from
V2 at flows of approximately
1
x
m
s-'.
Microscope slides,
60
x
24
x
1 mm, were heated in a muffle furnace at 475
"C
for
4
h and
used as hydrophilic substrata for the attachment and growth of
bacteria.
Total number
of
bacteria.
Acridine-orange-stained direct counts
(AODC) were used to determine the total number of unattached
bacteria as described by Pedersen
&
Ekendahl (1990). Surfaces from
the biofilm reactors were rinsed with a
1%
(w/v) NaCl solution in a
surface rinse as described by Pedersen
et al.
(1986)
to
remove
unattached bacteria, stained with acridine orange, rinsed with
deionized water, dried and the stained bacteria counted.
Precision
of
the AODC method.
The frequency distribution of the
number of bacteria counted per microscope field follows a Poisson
distribution (Hallbeck
&
Pedersen, 1990). This means that the
precision of the mean of the counted bacteria is only dependent on the
number of bacteria counted. One filter then predicts the sample mean
with a precision of 5% if 400 or more bacteria are counted (Niemela,
1983).
Variability
of
the total number
of
bacteria determined
on
different
sampling occasions and
in
different samples.
The total numbers of
unattached bacteria per ml flowing groundwater were counted on four
different occasions, 17 September 1987, and 17 April,
1
June and 1
October 1990. The total number of attached bacteria was counted after
two different experimental periods: after 56 d attachment and growth
between 21 February and 17 April 1990, and after 11 7
d
between
8
June
and 1 October 1990. Each serum bottle with a sample as described
below constituted an independent sample of the flowing groundwaters
from the different sampling depths and the results from all sampled
Bacterial populations in deep groundwater
37 1
bottles were used to assay the random variability between samples in
this investigation.
Incorporation
of
C02
and organic compounds by unattached bacteria.
Ten-millilitre volumes of groundwater, sampled on
1
October 1990,
were added to
100
ml sterile serum bottles with aluminium crimp-
sealed butyl rubber stoppers under a 100% nitrogen atmosphere via
hypodermic syringes mounted directly on the tubings from the different
sampling depths. A number of different 14C-
or
3H-labelled organic
compounds and Na2
l4c03
(Amersham Sweden) were subsequently
added in
1
ml oxygen-free portions to assay the nutritional responses of
the populations studied. The radioactive concentrations of the
*4C-labelled organic compounds were adjusted to 14 MBq per ml
sample. The sodium carbonate was adjusted to 73 MBq per ml to
compensate for the isotope dilution caused by the Cog- and C02
present in the groundwaters. The following final concentrations
and specific activities were used: 38 pki-Na214C03 (1.92 GBq
mmol-l);
7
p~-[*~C]formate (2.05 GBq mmol-I); 2.6 pM-[U-'*c]-
lactate (5.7 GBq mmol-l); 1.6 p~-[U-~~C]glucose (9.1 GBq mmol-l)
and
6
n~-~-[4,5-~H]leucine (5.63 TBq mmol-l). The samples were
incubated at
10
"C for 6 h, after which formalin was added to a Anal
concentration of 2% (v/v). Controls for abiotic adsorption of the
labelled compounds were achieved by addition of formalin (2%, v/v)
together with the labelled compounds at sampling, and were processed
like the other samples. Control counts were subtracted from sample
counts. Subsamples of
2
x
2.5 ml (970-1240 m samples) or
5
ml (799-
807 and 81 2-820 m samples) were filtered through Nuclepore filters
(pore size 0.2 pm, 13 mm), rinsed three times with
1
ml portions of
Milli-Q filtered water (pore size 0-2 pm) and subsequently placed in
10
ml scintillation cocktail (Ready-Safe, Beckman) and their radio-
activity measured in
a
Beckman LS 3801 scintillation counter.
Incorporation
of
C02
and organic compounds
by
attached bacteria.
Groundwater, sampled on
1
October 1990, was filtered in 20ml
volumes under a
100%
nitrogen atmosphere through Dynagard hollow-
fibre syringe filters (pore size 0.2 pm), into sterile
50
ml
polypropylene
centrifuge tubes with lids (Nunc). Labelled compounds were added as
above. Microscope slides from the biofilm reactors (experimental
period 8 June-1 October 1990,
11
7 d) were transferred under a 100%
nitrogen atmosphere to tubes with corresponding filtered groundwater,
one slide per tube. The Na214C03 was added after this step. The
microscope slides were incubated with the equivalent controls as
described above, rinsed with the surface rinse (Pedersen
et al.,
1986),
cut into four pieces with a diamond knife, placed in
20
ml scintillation
cocktail (Ready-Safe, Beckman) and their radioactivity measured.
There was no difference in the counting efficiency between pieces
placed individually in scintillation vials
or
when all the pieces were
counted in a single vial.
MARG
studies
of
unattached bacteria.
The MARG procedure
followed was the MARG-E method described by Tabor
&
Neihof
(1982). Briefly, residual
2
x
2.5 ml(970-1240 m)
or
5
ml(799-807 and
8 12-820 m) volumes from the sample bottles used for the incorporation
studies, including the control samples, were filtered through Nuclepore
filters (pore size 0.2 pm) and rinsed three times with
1
ml portions of
filtered
1
%
(w/v)
NaCl in Milli-Q water. The filters were transferred to
clean microscope slides previously dipped in Kodak NTB-2 autoradio-
graphic emulsion, placed in a waterchilled PVC container (10
"C),
moved to a desiccator after solidifying and left for exposure under
vacuum over silica gel for
3
d at 4 "C. A bacterium was scored to be
active, incorporating the labelled compound,
if
at least three silver
grains were located at a maximal distance of 1 pm away from the
bacteria labelled with
3H
and
3
pm away from the bacteria labelled
with
I4C.
The light-sensitive part of the autoradiographic work was performed
in a stainless steel dark
box
(0.6
x
0-45
x
0.35 m), with a top lid and two
neoprene seals through which material inside the box could
be
manipulated in complete darkness. A water bath inside the box,
connected to a heater-circulator bath on the outside, was used to
maintain the NTB-2 autoradiographic emulsion at
46°C.
The PVC
container was chilled inside the box with a heat exchanger, flushed
with tap water.
MARG
studies
of
attached bacteria.
The MARG procedure followed
the procedure described for unattached bacteria, with the following
modifications. The microscope slides with the attached bacteria were
rinsed and dipped directly into NTB-2 emulsion, allowed to gel, and
then exposed as above.
Estimate
of
the lower limit
of
detection
of
the
MARG
method.
The
facultative anaerobe
Psardomonas
jhorescens
(CCUG-25085) and a
SRB, isolated from the
860
m level in a borehole called KASO2 and the
680 m level in
KLXOI,
respectively (Pedersen
&
Ekendahl, 1990), were
used to estimate the lower limit of detection of the MARG method. The
bacteria were cultured as described by Pedersen
&
Ekendahl (1990)
with the addition of 0.2 to 20 n~-~-[4,5-~H]leucine, 2.3 pt~-[U-l~C]-
lactate or 1*4p4-[U-!4C]glUCo~,, incubated for 30 min to 24 h before
sampling and processed
as
for the LCS and MARG studies with
unattached bacteria.
The composition
of
the groundwaters
Tables 1 and 2 show the major parameters and the gas
content of the groundwaters in the Stripa borehole
V2.
The sulphate, sulphide, carbonate and conductivity data
differed between the sampling depths, indicating that
the groundwaters were obtained from fracture systems
with no close hydraulic connections. The temperatures
were measured earlier with a borehole sond during the
geological well logging programme (Nordstrom
et al.,
1985). The waters were chilled to 10°C when flowing
from the sampling depths up to the drift and this was the
temperature at which all incubations were done. The
flow was measured as ml min-' and converted to cm
s-l
over the surfaces in the biofilm reactors.
Numbers
of
bacteria
The numbers of bacteria counted on different occasions
in groundwater and on surfaces exposed to flowing
groundwater from the different sampling depths are
shown in Table
3.
There were
10-
to 100-fold more
bacteria in the groundwater from 970-1240 m depth than
from the other two depths, but this difference
was
not
reproduced on the surfaces exposed to the flowing
groundwaters.
The random variability of the total
number of unattached and attached bacteria in samples
from the same depth examined, on one occasion, ranged
between
6
and 45% of the mean. The bacteria on the
surfaces were distributed in uneven patterns
in
clusters,
indicating that they had grown on the surfaces rather
than just attached randomly from the passing waters.
312
K.
Pedersen and
S.
Ekendahl
Table
1.
The major parameters
of
the groundwaters
of
the Stripa borehole
V2
A
detailed description
of
the Stripa groundwaters is given by Nordstrom
et
al.
(1985).
Values in parentheses are standard deviations
(04).
_____~
Sampling
depth
Eh
Temp.*
sot-
S2-
coj-
Conductivity
Flow
(m) PH (mV>
CC)
(PM)
(PM)
(PM)
(pS
cm-l) (m
s-*)
~~
799-807
9.5
+
205
18
52
(10)
<
0.01
(-)
158
(-)
425
(-)
1.4
x
10-3
8
1
2-820
9.4
+
199
18 1433
(6)
106
(-)
50
(12)
1640
(-)
2.8
x
10-3
97&1240
10.2 -3 26 520
(0)
233 (8)
57
(5)
1180
(-)
0.5
x
10-3
*
The water had been chilled to
10
“C
when it reached the drift
(410
m).
Table
2.
The content
of
nitrogen, hydrogen and carbon-containing gases and the total volumes
of
gas extracted from the samples
of
the groundwaters
of
the Stripa borehole
V2
Values in parentheses are standard deviations
PA).
799-807 25000 (27)
<
10 <1
32
(44)
245
(14)
0.3 (47)
<0.1
2-4 (25)
2-7
(1)
8 12-820 31
000
(39)
<
10
<1
11
(6) 170 (25) 0-6 (12) co.1 3.4
(44)
970-1 240 24500 (3)
<
10 <1 10
(0)
290
(5)
2-9 (2)
<0*1
Table
3.
The total number
of
unattached bacteria (ml-I) in groundwater, and attached bacteria
on
surfaces (cm-’) exposed to
jowing groundwater
from
three sampling depths
of
the Stripa borehole
V2,
799-807
m,
812-820
m
and
970-1240 m, measured
on
drflerent occasions
970- 1240
m
799-807
m
8 12-820
m
10-5
x
NO.
of
10-5
x
NO.
of
10-5
x
NO.
of
N*
bacteria?
SD
(%>
bacteria?
SD
(%>
bacteria?
SD
(”/,>
Groundwater sampling
1
October
1990
6
0.054
26
0.01
8 45
1.2
12
8
June
1990 1 0.240
0-047
2.3
-
18
April
1990
1
0.036
0.0
16
1.6
-
17
September
1987 1 0.097
0-06
1
2.3
-
-
-
-
-
-
-
Surfaces exposed to
flowing groundwater
1
October
1990,
117
d
6
12.0
30
71.0 38
59.0
31
17
April
1990, 56
d
3
10.0
6 72.0 20 22.0 42
~~ ~
*
N
is the number
of
independent samples.
Populations are per ml or per cm2.
Estimate
of
the lower limit
of
detection
of
the
MARG
method
Fig.
1
shows the relation between the percentage
of
bacteria scored to be active using the
MARG
method
and the mean number
of
disintegrations per minute
(d.p.m.) per bacterium
of
the samples. The lowest
radioactivity that resulted in active bacteria was
d.p.m. per bacterium and
80%
of
the bacteria were
scored active at d.p.m. per bacterium. This corre-
sponds to 0-1-1
x
mol
14C
per bacterium and
0.21-
2.1
x
10-19
mol
3H
per bacterium and these are the
minimum amounts
of
the isotopes per bacterium that can
be detected with the
MARG
method.
A
comparison of
the data in Table
4
shows that several samples
demonstrated significant incorporation of the
4C-
labelled Na2C03, formate and glucose using LSC, but
bacteria to be scored active in the
MARG
procedure
Bacterial populations in deep groundwater
373
Table
4.
Incorporation
of
14C
and
3H
from
Cot
and labelled organic compounds
by
unattached bacteria
in
groundwater and attached bacteria on surfaces exposed to flowing
groundwater
from
three sampling depths
of
the Stripa borehole
V2
and the percentage
of
the population scored to be active, using the
MARG
method,
in
incorporating the labelled
compounds
Unattached bacteria Attached bacteria
Percentage Percentage
Labelled Depth
1014
x
MOI
active
1014
x
MOI
active
isotope cm-* bacteria
compound (m) isotope ml-l bacteria
co2* (1°C)
799-807
8 12-820
970-1 240
Formate
('"C)
799-807
8
12-820
970- 1240
Lactate
('"C)
799-807
8 12-820
970-1240
Glucose
(
"C)
799-807
8 12-820
970-1240
Leucine
(3H)
799-807
8 12-820
970-1 240
312
66
88
11
3
-
26
10
68
19
19
66
1
5
5
5
5
4
6
-
16
34
6
5
8
-
55
23
9
-
1
200
4
380
29
81
5
100
13
700
92
400
123
4 800
2 600
160
290
280
11
24
74
77
99
38
-,
Not detected.
*
The data have been corrected for isotope dilution caused by the measured COj- and COz contents of
the groundwaters (Tables
1
and
2).
80
60
10.' lo-?
lo-'
10"
D.p.m.
per bacterium
Fig.
1.
Relation between the percentage of bacteria scored to
incorporate the labelled compounds using the MARG method and the
mean number
of
d.p.m. per bacterium in cultures with
Pseudomoms
Juorescens
(CCUG-25085)
(A)
and a
SRB
(W),
amended with
0.2-20
nM-L-[4,5-3H]leucine,
2.3
pM-[U-14C]lactate or
1.4
pM-[u-'4c]-
glucose between
30
min and
24
h before sampling.
could not be detected. This is explained by the higher
sensitivity of the LSC technique compared to the
MARG technique on the assumption that the bacteria
incorporated too little 14C
(<
0.1
x mol per cell)
to be scored active using the MARG technique, but
sufficient 14C to give a measurable contribution using
LSC.
Incorporation
of
C02
and organic compounds by attached
bacteria
The incorporation of 14C and
3H
from labelled com-
pounds by unattached (ml-l) and attached (cm-*)
bacterial populations after
6
h (from
1
October 1990),
measured with the LSC (mol 14C and
3H
ml-l or cm-')
and MARG (percentage of bacteria scored to be active)
techniques is shown in Table
4.
The LSC results are
presented as mol isotope atoms incorporated because the
metabolic pathways of the organic compounds used are
unknown for the populations studied and stoichiometric
calculations back to mol organic compounds might
therefore be misleading. For instance, many SRB split
lactate into acetate, which is incorporated via the TCA
pathway, and to C02, which is expelled from the cell.
The content of dissolved organic carbon (DOC) in the
borehole
V2
groundwater has previously been deter-
mined to be
1.1
mg
1-'
(0.4-4 mg 1-', eight measure-
ments) (Nordstrom
et al.,
1985), which is in agreement
374
K.
Pedersen and
S.
Ekendahl
with the range of what has been found in
22
other
deep groundwaters from crystalline bedrock (average
2 mg 1-l) (Lakksoharju, 1990). A significant part of this
DOC
consists of fulvic acids (Pettersson
et al.,
1990). The
amounts of 4C-labelled organic compounds used were
between
0.23
(lactate) and
0.32
(formate) mg
1-l.
The
presence of sufficient
in
situ
concentrations of unlabelled
molecules of the corresponding organic compounds to
give significant misleading data due to isotope dilution is
thus unlikely.
There was significant incorporation of C02 in all
samples, except for the attached bacteria at 799-
807 m depth (Table 4), indicating the
in
situ
production
of organic carbon from C02. Incorporation of formate
followed that of
C02
except for the unattached bacteria
at 970-1240 m depth and indicated the presence of
bacteria able to substitute formate for
COz.
The lactate
and glucose incorporation demonstrated the presence of
heterotrophic bacteria. The incorporation of lactate by
the attached bacteria dominated over glucose at all
depths and gave MARG responses up to 74%. Leucine
was incorporated by up
to
99% of the populations, which
showed the major fractions of the populations studied
were viable.
Discussion
The study by Nordstrom
et al.
(1985) showed that the
salinity profile of the borehole
V2
is heterogeneous,
which is typical for groundwaters in crystalline bedrock
that have not been intruded by saline waters such as that
from the sea. The different conductivities, carbonate,
sulphate and sulphide concentrations of the sampling
depths (Table 1) reflect this heterogeneity and indicate
that the groundwaters came from fracture systems
without close hydraulic connections. The two lower
sampling depths had a considerably higher salinity than
the upper depth, indicating that these waters are old (in
excess of 20000 years using conventional
I4C
measure-
ments; cf. Nordstrom
et al.,
1985). The 799-807m
groundwater is probably mixed with surface water via
the mine, which may have diluted the salinity above
810 m (Nordstrom
et al.,
1985). The sulphide content of
the two lowest depths indicates that these habitats were
anaerobic, and that facultative
or
obligate anaerobic
bacteria should be expected.
The
physical and chemical parameters and the flows
measured have not differed significantly since the bore-
hole
V2
was drilled in 1978 (Nordstrom
et
al.,
19-85>;
In
addition, each sampling depth exhibited an amount of
bacteria per volume of water that differed little between
sampling occasions from 1987 to 1990 (Table 3). This
implies that the bacterial populations studied were in
steady states with their environments and that the
fractures of the Stripa bedrock constitute stable habitats
for bacteria. The possibility of contamination of the
groundwater during drilling cannot be excluded at this
point, but it can also be argued that bacteria probably
inhabited this environment long before the mine was
constructed. A population of bacteria, slowly migrating
vertically at a rate of
0.1
to
1
m a year in the present
chemical and physical gradients and with the ground-
water movements, would need 1000 to 10000 years to
reach a depth of lo00 m. This is rapid in relation to the
geological age of many million years of the Stripa
bedrock.
Bacteria, irrespective
of
whether they are migrating,
contaminant or indigenous populations, need carbon and
energy sources to survive. The incorporation of
C02
(Table
4)
reflects an
in
situ
production of organic carbon
from the
C02
which might be sufficient to sustain
autotrophic and also heterotrophic populations with
organic matter. The energy source for autotrophy could
be hydrogen migrating from the earth’s crust. Hydrogen
has been detected elsewhere in the Stripa ground waters
(borehole
V1;
Carlsson
et al.,
1983) but not in borehole
V2
(Table
2),
probably as a consequence of its utilization
by resident bacterial populations. Heterotrophic incor-
poration of
C02
during fermentative or respiratory
utilization of geological deposits of organic material is
another possible explanation for the recorded uptake of
There was a significant content
of
methane
in
the
groundwaters (Table 2), which may have two different
origins. Methanogens can use
COz
and formate as
carbon sources, and use
C02
as terminal electron
acceptor in their energy metabolism, producing methane
(Fuchs, 1990; Ormeland, 1988). The small but significant
incorporation of formate supports the presence of
methanogenic bacteria. Another methane source might
be of a geological origin, of the type proposed for the
Siljan deep gas project area, approximately 100 km north
of Stripa. Large gas and oil deposits, formed as a result of
an ancient collision of a meteorite with the earth, are
postulated to lie several kilometres below ground.
Incorporation of lactate by the attached bacteria was
substantially greater than that of glucose at all depths
and gave MARG responses in all samples (Table 4). The
deep groundwaters of the Stripa mine site are anoxic
and depleted of nitrate and nitrite. The only avail-
able electron acceptor for respiration was sulphate
(Nordstrom
et al.,
19853 and such an environment will be
selective for fermenting bacteria and SRB. Propionate-
producing bacteria are among the few bacteria known to
ferment lactate without involving a respiratory chain;
however, they thrive in nutrient-rich habitats like
cheese, on skin and in mud but not in oligotrophic
co2.
Bacterial populations in deep groundwater
375
environments like the Stripa deep groundwater. Con-
sequently, it is likely that the bacteria that utilized lactate
in the anaerobic incubations were the SRB. This suggests
that
SRB
constitute a substantial part of the bacterial
populations in the fractures of the Stripa crystalline
bedrock, as has been reported for other deep geological
formations (Olson
et al.,
1981
;
Pedersen
&
Ekendahl,
1990). Sulphate reduction by
SRB
will increase the
634S
isotopic content of a groundwater due to the preference
for
32S
by sulphate reducers (Widdel, 1988) and also
result in increasing amounts of sulphide. Fontes
et al.
(1989) found high
S34S
values in borehole V2 and they
postulated the presence of viable populations of
SRB.
The 970-1240 m depth revealed the highest
S34S
value
during their investigations; this depth had the highest
lactate incorporation (9.2
x
lo-'*
mol cm-*), with
74%
of the bacteria actively incorporating lactate and the
highest sulphide content in our study (233
p~)
(Tables 1
and
4).
Our data confirm the hypothesis proposed by
Fontes
et al.
(1989).
Glucose utilization is a constitutive metabolic path-
way, common among mixotrophic bacteria (Kuenen
&
Bos,
1989) and most fermenting, as well as respiring,
heterotrophic bacteria. There was incorporation of
glucose, which was below the detection limit for active
bacteria
(<
0.1
x
10-l6 mol per cell) in several samples
analysed with the MARG method. Since only a few
SRB
are known to utilize glucose, the observed uptake
probably indicates the presence of heterotrophic
bacteria.
Leucine incorporation is virtually specific for bacteria
provided low (nanomolar) concentrations are used
(Kirchman
et
al.,
1985) and is used by many bacteria
during growth for protein synthesis (Kirchman
et al.,
1985). Leucine can also be used as a carbon and energy
source and can be fermented by proteolytic clostridia via
the Stickland reaction. High percentages of the popu-
lations, up to 99%
of
the attached bacteria at 812-820 m
depth, incorporated leucine. The leucine incorporation
showed that major parts of the studied populations were
viable.
A fracture in crystalline bedrock is made of two
surfaces which are wavy and rough. They are in contact
with each other at some points but are at a distance from
each other at others. The openings in the fractures are
potential channels for groundwater. Recently several
model studies have been made on flow and transport in
fractures with variable apertures (Moreno
et a/.,
1988;
Tsang
et al.,
1988). The results indicated that consider-
able channelling is to be expected in such fractures and
that there is a tendency for some pathways to carry much
more water than others. In a limited mass of rock one or a
few channels will dominate flow, radionuclide transport
and transport of nutrients for bacteria. Assuming a mean
channel width of 0.1 mm (Moreno
et
ai.,
1985), our
results imply that there would be from
4
x
lo3 to 8
x
lo5
more attached than unattached bacteria in a channel
after
4
months of contact with Stripa borehole
V2
groundwater flowing at 0.5-2.8
x
m
s-l
(Table
3).
The average hydraulic conductivity,
K,
has been deter-
mined to be m
s-'
or less in fractured rock
of
Stripa
(Carlsson
et al.,
1983) but it will be considerably higher
in
individual channels (Neretnieks, 1990).
K
is a function of
the injection flow rate, the injection excess head, the
length of the injection interval and the radius of the
borehole (Andersson
et al.,
1989). The flows used here
were probably even higher than in a channel with a high
conductivity
;
instead the experiment time was very short
in relation to the time a channel will be open for flowing
groundwater and bacteria.
The availability of energy and nutrients over time for a
biofilm is flow dependent and will determine whether a
biofilm will develop
in situ
and how many bacteria can be
maintained. The slower the flow, the slower the
development rate of a biofilm down to a limit where the
bacteria can no longer grow or maintain a non-growth
metabolism. This limit
is
probably very low for bacteria
in an oligotrophic environment like deep groundwater,
and will select for bacteria with advanced morphological
and physiological mechanisms to survive a very limited
availability of nutrients (Kjelleberg
et ai.,
1987). Assess-
ing the influence of groundwater microbiology
on
the
long-term safety
of
nuclear waste disposal involves time
scales ranging from hundreds to many millions of years,
thus there is practically no time limit for even the slowest
developing biofilm to reach a steady state. The presence
of attached bacteria might retard transport of radio-
nuclides from a nuclear waste repository unless they
produce complexing agents and other metabolites that
affect speciation and thus mobility of radionuclides in a
contrary way. The possibility of such
in
situ
production
by bacteria in fractured bedrock will be an important
task for future research, aimed
to
assess the influence of
groundwater microbiology on the long-term safety of
nuclear waste disposal.
This research was supported by the Swedish Nuclear Fuel and Waste
Management
Co.
The collaboration of our mine guide, Birger
Ekstrand, and the director of the mine, Gunnar Ramqvist, at the Stripa
mine has been a pleasant experience. We would like
to
thank
Dr
Amanda Goodman and
Dr
Malte Hermansson for valuable comments
on the content of the manuscript.
References
ANDERSON,
P., ANDERSSON, P.,
GUSTAVSON, E.
&
OLSSON,
0.
(1989).
Investigation of
flow
distribution in a fracture zone at the Stripa
mine, using the radar method, results and interpretation.
SKB
Technical Report
89-33.
Swedish Nuclear Fuel and Waste Manage-
ment
Co.,
Box
5864,
S-10248
Stockholm.
376
K.
Pedersen and
S.
Ekendahl
BELYAEV,
S. S.
&
IVANOV, M. V. (1983). Bacterial methanogenesis in
underground waters. In
Environmental Biogeochemistry (Proceedings
of’the 5th International Symposium of Environmental Biogeochemistry),
pp. 273-280. Edited by
R.
Hallberg. Stockholm: Liber Tryck.
BELYAEV,
S. S.,
WOLKIN,
R.,
KENEALY, W.
R.,
DENIRO, M. J.,
EPSTEIN,
S.
W.
&
ZEIKUS, J. G. (1983). Methanogenic bacteria from
the Bondyuzhshoe oil field
:
general characterization and analysis of
stable-carbon isotopic fractionation.
Applied and Environmental
Microbiology
45,
691-697.
BEVERIDGE,
T.
J.
&
FYFE, W.
S.
(1985). Metal fixation by bacterial cell
walls.
Canadian Journal
of
Earth Science
22,
1893-1898.
CARLSSON, L., OLSON, T., ANDREWS,
J.,
FONTES,
J.-C., MICHELOT,
J. L.
&
NORDSTROM, K. (1983). Geochemical and isotope character-
ization of the Stripa ground waters
-
progress report.
Stripa Project
Technical Report
83-01. Swedish Nuclear Fuel and Waste Manage-
ment Co., Box 5864, s-10248, Stockholm.
CHAPELLE, F. H., ZELIBOR,
J.
L. J., GRIMES, D. J.
&
KNOBEL, L.
L.
(1987). Bacteria in deep coastal plain sediments of Maryland: a
possible source of Cot to ground water.
Water Resources Research
CHAPELLE,
F.
H., MORRIS, P. B., MCMAHON, P. B.
&
ZELIBOR, J.
L.
(1988). Bacterial metabolism and the
613C
composition of ground
water, Floridan aquifer system, South Carolina.
Geology
16,
FAUQUE, G., LEGALL, J.
&
BARTON, L. L. (1991). Sulfate-reducing and
sulfur-reducing bacteria. In
Variations in Autotrophic Life,
pp.
271-337. Edited by
J.
M. Shively
&
L. L. Barton. London: Academic
Press.
FONTES, J. C., FRITZ, P., LOUVAT, D.
&
MICHELOT, J. L. (1989).
Aqueous sulfates from the Stripa ground water system.
Geochimica et
Cosmochimica Acta
53,
1783-1790.
FRANSON, M. A. H. (1985).
Standard Methods for the Examination of
Water and Waste Water,
16th edn. Washington, DC: AWWA-
Fucm, G. (1986). COz fixation in acetogenic bacteria: variations on a
theme.
FEMS Microbiology Reviews
39,
181-213.
Fucm, G. (1990). Alternative pathways
of
autotrophic C02 fixation.
In
Autotrophic Bacteria,
pp. 365-382. Edited by H. G. Schlegel
&
B.
Bowien. Berlin
:
Springer-Verlag.
GHIORSE, W.
C.
&
WILSON, J. T. (1988). Microbial ecology of the
terrestrial subsurface.
Advances in Applied Microbiology
33,
107-172.
GODSY, E. M. (1980). Isolation of
Methanobacterium bryantii
from a
deep aquifer by using a novel broth-antibiotic disk method.
Applied
and Environmental Microbiology
39,
1074-1 075.
HALLBECK,
L.
8i
PEDERSEN, K. (1990). Culture parameters regulating
stalk formation and growth rate of
Gallionella ferruginea. Journal of
General Microbiology
136,
1675-1680.
HICKS,
R.
J.
&
FREDRICKSON,
J.
K. (1989). Aerobic metabolic potential
of microbial populations indigenous to deep subsurface environ-
ments.
Geomicrobiology Journal
7,
67-78.
HUFFMAN, E. D. W. (1977). Performance of
a
new automatic carbon
dioxide coulometer.
Microchemical Journal
22,
567-573.
KIRCHMAN, D., K’NEES, E.
&
HODSON,
R.
(1985). Leucine incorpor-
ation and its potential as a measure of protein synthesis by bacteria in
natural aquatic systems.
Applied and Environmental Microbiology
49,
KJELLEBERG,
S.,
HERMANSSON,
M.,
MARDEN, P. &JONES,
G.
W. (1987).
The transient phase between growth and nongrowth of heterotrophic
bacteria, with emphasis on the marine environment.
Annual Review
of’
Microbiology
41,
2549.
KUENEN, J. G.
&
Bos, P. (1989). Habitat and ecological niches of
chemolitho(auto)trophic bacteria. In
Autotrophic Bacteria,
pp. 53-96.
Edited by H.
G.
Schlegel
&
B.
Bowien. Berlin: Springer-Verlag.
LAKKSOHARJU, M. (1990). Colloidal particles in deep Swedish granitic
ground water.
SKB
Progress Report
90-37. Swedish Nuclear Fuel and
Waste Management Co., Box 5864, S-10248 Stockholm.
MORENO, L., NERETNIEKS, I.
&
KLOCKARS, C.-E. (1985). Analysis of
some laboratory tracer runs in natural fissures.
Water Resources
Research
21,
951-958.
23,
1625-1632.
117-121.
APHA-WPCF.
599-607.
MORENO, L., TSANG,
Y.
W., TSANG, C. F., HALE,
F.
V.
&
NERETNIEKS,
I. (1988). Flow and tracer transport in a single fracture. A stochastic
model and its relation to some field observations.
Water Resources
Research
24,
2033-2048.
NERETNIEKS,
I.
(1990). Soluble transport in fractured rock
-
appli-
cations to radionuclide waste repositories.
SKB Technical Report
90-38. Swedish Nuclear Fuel and Waste Management Co., Box 5864,
S-10248 Stockholm.
NIEMELA,
S.
(1983). Statistical evaluation of results from quantitative
microbiological examinations.
NMKL Report
no. 1, 2nd edn.
Uppsala: Nordic Committee on Food Analysis.
NORDSTROM, D. K., ANDREWS, J. N., CARLSSON, L., FONTES, J.-C.,
FRITZ, P., MOSER,
H.
&
OLSON, T. (1985). Hydrogeological and
hydrogeochemical investigations in boreholes
-
final report of the
phase 1 geochemical investigations of the Stripa ground waters.
Stripa Project Technical Report
85-06. Swedish Nuclear Fuel and
Waste Management Co., Box 5864, S-10248 Stockholm.
OLSON,
G.
J., DOCKINS, W.
S.
&
MCFETHERS, G. A. (1981). Sulfate-
reducing and methanogenic bacteria from deep aquifers in Montana.
Geomicrobiology Journal
2,
327-340.
ORMELAND,
R.
S.
(1988). Biogeochemistry of methanogenic bacteria.
In
Biology
of
Anaerobic Microorganisms,
pp. 641-705. Edited by
A. J. B. Zehnder. New York: John Wiley.
PEDERSEN, K. (1982). Method for studying microbial biofilms in
flowing-water systems.
Applied and Environmental Microbiology
43,
6-1 3.
PEDERSEN, K.
&
ALBINSSON,
Y.
(1991). Effect of cell number, pH and
lanthanide concentration on the sorption of promethium by
Shewanella putrefaciens. Radiochimica Acta
54,
9 1-95.
PEDERSEN, K.
&
EKENDAHL,
S.
(1990). Distribution and activity of
bacteria in deep granitic ground waters of southeastern Sweden.
Microbial Ecology
20,
37-52.
PEDERSEN, K., HOLMSTROM, C., OLSSON, A.-K.
&
PEDERSEN, A. (1 986).
The influence of species variation, culture conditions, surface
wettability and fluid shear on attachment and biofilm development
of marine bacteria.
Archives of Microbiology
145,
1-8.
PETTERSSON, C., EPHRAIM, J., ALLARD, B.
&
BOREN, H. (1990).
Characterization of humic substances from deep ground waters in
granitic bedrock in Sweden.
SKB Technical Report
90-29. Swedish
Nuclear Fuel and Waste Management Co., Box 5864, $10248
Stockholm.
STRANDBERG,
G.
W., STARLING, E., SHUMATE, 11.
&
PARROTT, J.
R.
(1981). Microbial cells as biosorbents for heavy metals: accumu-
lation of uranium by
Saccharomyces cerevisiae
and
Pseudomonas
aeruginosa. Applied and Environmental Microbiology
41,
237-245.
TABOR, P.
S.
&
NEIHOF,
R.
A. (1982). Improved microautoradiographic
method to determine individual microorganisms active in substrate
incorporation in natural waters.
Applied and Environmental Micro-
biology
44,
945-953.
TSANG,
Y.
W., TSANG, C. F., NERETNIEKS, I.
&
MORENO, L. (1988).
Flow and tracer transport in fractured media. A variable aperture
channel model and its properties.
Water Resources Research
24,
WEST, J. M., CHRISTOFI, N.
&
MCKINLEY, I. G. (1982). Microbes in
deep geological systems and their possible influence on radioactive
waste disposal.
Radioactive Waste Management Nuclear Fuel Cycle
3,
WEST,
J,
M., CHRISTOFI, N.
&
MCKINLEY, I. G. (1985). An overview of
recent microbiological research relevant to the geological disposal of
nuclear waste.
Radioactive Waste Management Nuclear Fuel Cycle
6,
79-95.
WIDDEL,
F.
(1988). Microbiology and ecology of sulfate- and sulfur-
reducing bacteria. In
Biology of Anaerobic Microorganisms,
pp.
469-585. Edited by A. J.
B.
Zehnder. New York: John Wiley.
WOOD, H.
G.
&
LJUNGDAHL, L.
G.
(1991). Autotrophic character of the
acetogenic bacteria. In
Variations in Autotrophic Lge,
pp. 201-250.
Edited by J. M. Shively
&
L. L. Barton. London: Academic Press.
2049-2060.
1-15.
