Bacterioplankton in the Seine River (France): impact of the Parisian urban effluent

TLDR: Bacterial abundance and biomass were studied in April, July, and October 1989 at 13 stations along 300 km of the course of the river Seine, including Paris and its suburbs as mentioned in this paper.

Abstract: Bacterial abundance and biomass were studied in April, July, and October 1989 at 13 stations along 300 km of the course of the river Seine, including Paris and its suburbs. Monthly investigations were carried out at five stations downstream from Paris where the river receives the effluent of an important waste water treatment plant (Acheres). As a result of an input of allochthonous bacteria from the effluent of the plant, an increase in bacterial abundance and biomass was observed below Acheres (from about 5 × 109 to 15 × 109 cells L−1 and from 100 to 750 μg C L−1). This was followed by a rapid decrease. The allochthonous bacteria comprised a high proportion of large bacteria, which disappeared at a much higher rate than the small bacteria (0.0366 vs. 0.0125 h−1). Paradoxically, these large bacteria grew at a rate twice that of the smaller cells in culture experiments (0.129 vs. 0.065 h−1 in June and 0.118 vs. 0.071 h−1 in October). These large bacteria must therefore be subjected to intense losses (graz...

Full text

56
Bacterioplankton
in the Seine River (France): impact of the Parisian urban
effluent'
JOSETTE GARNIER^
Laboratoire
d'écologie, U.A. 258, Centre national de la recherche
scientifique.
École Normale Supérieure,
46,
rue
D'Ulm,
75005 Paris, France
...^
AND
PIERRE
SERVAIS AND GILLES BILLEN
Groupe
de microbiologie des milieux
aquatiques.
Université Libre de Bruxelles, Campus de la Plaine,
CP
221, BD du Triomphe, 1050 Bruxelles, Belgique
Received
February 28,
1991
Revision
received August
12,
1991
Accepted
August
15,
1991
GARNIER,
J., SERVAIS, P., and BILLEN, G.
1992.*feacterioplankton
in the Seine River (France): impact of the Parisian
urban
effluent. Can. J.
Microbiol.
38: 56-64.
Bacterial
abundance and biomass were studied in
April, July,
and October
1989
at 13 stations
along
300 km of the
course
of the river Seine, including Paris and its suburbs. Monthly investigations were carried out at five stations
downstream
from Paris where the river receives the effluent of an important waste water treatment plant (Achères).
As
a resuit of an input of allochthonous bacteria from the effluent of the plant, an increase in bacterial abundance
and
biomass was observed
below
Achères (from about 5 x
lO'
to 15 x
lO' cells
L
"
' and from 100 to 750
/ig
C L
"
').
This
was followed by a rapid decrease. The allochthonous bacteria comprised a high proportion of large bacteria, which
disappeared
at a much higher rate than the
small
bacteria (0.0366 vs. 0.0125 h
"
'). Paradoxically,
thèse
large bacteria
grew
at a rate twice that of the smaller cells in culture experiments (0.129 vs. 0.065 h ' in June and
0.118
vs.
0.071
h '
in
October).
Thèse
large bacteria must therefore be subjected to intense losses (grazing,
sédimentation,
etc.). Higher
rates
of discharge of the river, i.e., a shorter
résidence
time of the water masses, appeared to transmit the Achères
signal
farther, leading to faster transport of the bacterial populations.
Key words:
bacterioplankton ecology, size fractions, river
ecosystem." '
GARNIER,
J., SERVAIS, P., et BILLEN, G. 1992. Bacterioplankton in the Seine River (France): impact of the Parisian
urban
effluent. Can. J. Microbiol. 38 : 5664.
Les
variations temporelles et spatiales des abondances et biomasses bactériennes ont été étudiées dans
le
fleuve fran
çais,
la Seine, sur un secteur long de 300 km. La zone d'influence de la station d'épuration d'Achères, à l'aval de Paris
a
été plus particulièrement analysée. En raison d'un apport massif de bactéries allochtones par les effluents d'Achères,
il
se produit à l'aval de
la
station un accroissement important tant des abondances (d'environ 5 x
10**
à 15 x
lO'
cellules
L"') que des biomasses (de 100 à 750
/ig
C L '). Les valeurs diminuent ensuite rapidement. Les bactéries
allochtones
se composent d'une proportion importante de bactéries de grande taille qui disparaissent
plus
rapidement
que
les petites (0,0366 vs. 0,0125 h '). Comme le montrent les expériences en culture, ces bactéries de grande taille
croissent
pourtant à un taux environ deux fois plus élevé que les petites (0,129 vs. 0,065 h ' en juin et 0,118 vs.
0,071
h ' en octobre). Ces grosses bactéries doivent donc être soumises à des processus de pertes importantes (brou
tage,
sédimentation ...). 11 apparaît que le signal de la station d'épuration d'Achères est observé d'autant plus loin
que
le débit de la rivière est élevé, le transport de la population bactérienne étant plus rapide.
Mots
clés : bactérioplancton, écologie de la taille, écosystem fluvial.
[Traduit
par la rédaction]
Introduction
The dégradation
of organic matter is a major process in
the
oxygen balance of river Systems that has been studied
for
a long time, particularly when there is organic enrich
ment
from
wastewater
discharges. The approach, first
proposed
by Streeter and Phelps (1925) and which
still
prevails
in most
récent
studies (Lesouëf and André 1982),
does
not take into account the
complex
dynamics of
plank
tonic
bacteria but
only
considers a firstorder kinetics of
organic
matter
dégradation
with respect to organic matter
concentration.
Methods
and concepts developed in the
last décade
for
studying
biomass and activity of bacterioplankton have
led
'Publication
No. 3 of the French program
PIRENSEINE
(Centre
national
de la recherche scientifique).
^Présent
address: Groupe de microbiologie des milieux aqua
tiques.
Université Libre de Bruxelles, Campus de la plaine, CP
221,
BD
du Triomphe, 1050 Bruxelles, Belgique.
Prinled
in Canada / imprimé au Canada
to
a better understanding of carbon metaboiism in marine
Systems
(Williams 1981; Azam et al. 1983; Sherr et al. 1988),
in Iakes
(Riemann and S0ndergaard
1986;
Simon and Tilzer
1987;
Garnier 1989), and in estuaries (Admiraal et al. 1985;
Relexans
et al. 1988; Riemann et al. 1990). Only few stud
ies
of microbial ecology were devoted to rivers with an eco
systemic
view (Wissmar et al. 1981; Healey et al. 1988).
A
comprehensive study of major hydrodynamic, chemical,
and
biological processes involved in the functioning of the
ecosystem
of the river Seine is in progress. The goal is to
establish
a
model
of oxygen balance and organicmatter
dégradation,
specifically taking into account the dynamics
of
the bacterial compartment.
As
a part of this study, we investigated the temporal and
spatial
variations of planktonic bacteria in the river Seine.
This
paper analyzes
thèse
variations in response to the
impact
of a waste water treatment plant (Achères). Partic
ular
emphasis is given to the
différent
behavior of small and
large
bacteria in the area surrounding the effluent outfall

GARNIER
ET AL.
57
Seine
Poses
(Pk 847)
Oise
Porcheville
>(Pk750)
Méricourt V///
^(Pk
765) Achères
(Pk708)
Paris (Pk
645)
r
Manie
Eure
Seine
Montereau
(Pk543)
Yonne
FiG.
1. Study area showing the main station locations.
of
the plant. We
présent évidence
that large allochthonous
bacteria
brought into the river by the effluent of the plant
are
an active component of the microbial community.
Study
area
The
watershed of the Seine River has a temperate rainfall pat-
tern
with an oceanic influence downstream. The mean tendency
of
the
hydrographie
conditions in the
last
50 years is characterized
by
the highest rates of discharge occurring in winter (monthly mean
at
Paris: 550
m' s~'
in February) and the
lowest
in summer
(50 m'
s~' in August), increasing from October onwards. How-
ever,
during the period studied
(April
1989 - March 1990), discharge
was
highest in April 1989, decreased in May, and remained
low
until
March 1990 because of dry meteorological conditions.
The
bacterioplankton of the Seine River was studied for a
distance
of 300 km between Montereau, located at the confluence
with
the river Yonne, and Poses, 150 km from the mouth of the
Seine
estuary (Fig. 1). About 60 km downstream from Paris, the
river
receives an urban effluent from the waste water treatment
plant
of Achères. About 80% of the waste water produced by the
10
million inhabitants of Paris and its suburb reaches this plant;
70"%
is treated by the conventional activated sludge process, before
being
discharged into the river; the remaining
30%
is discharged
without
treatment. The study has therefore
also
focused on the
reach
downstream from Paris (Andrésy-Méricourt), at the effluent
outfall
of the water-treatment plant of Achères.
Material
and methods
Sampling
program
Three
sampling programs were established. In April,
July,
and
October
1989, the course of the river was studied at
13
stations
between
Montereau and Poses. In addition, five stations, one
upstream
from the outfall of the Achères effluent and four
downstream,
were investigated almost every month, and weekly
investigations
were carried
eut,
from May to October, at a
médian
station
of the reach (Porcheville).
The
stations are named by a kilometric unit, Pk, used by the
Financial
Agency of the "Seine Normandie" Bassin. Pk is set at
1000
at Honfleur, the mouth of the estuary, and then decreases
upwards:
Poses is located at Pk 847 and Montereau at Pk 543.
The
water-treatment plant of Achères is at Pk 708.
At
each station, before
samphng,
the tubing was abundantly
rinsed
with the river water. To ensure that the samples were
représentative,
at the two stations where the effluent was shown
to yield
to heterogeneity of the river (Pk
713
and Pk 721), 10 L
of
water was taken from two depths (0.5 and 5 m) at three points
along
a transverse section of the river and mixed (Chesterikoff et al.
1991).
At the other stations, the 10 L was pumped in the middle
of
the river at two depths (0.5 and 5 m) and mixed.
Bacterial
enumeration
Samples
(20 mL) for direct epifluorescence microscopic counts
were
preserved in
2%
formalin immediately following collection
from
the river or from the culture (Hobbie et al. 1977); they were
stored
at 4°C until analysis (within 1-2 months).
A
subsample of 200-500
;^L
was brought to 2 mL with 0.22
iim
filtered
water (Millex, 0.22
jim),
stained for 4 min with acridine
orange
at a final concentration of 2.5 x 10
"*
g mL ', and
filtered
at a low vaccum
(less
than 100 mbar (1 bar = 100 kPa))
through
a
black
polycarbonate membrane (Nuclepore, 0.22
/xm
porosity,
25 mm diameter) placed on a cellulose
acétate
filter
(Millipore,
0.45
/im
porosity). The membrane was mounted
between
a
slide
and a
glass
cover slip, with a
nonfluorescence
immersion oil
(Olympus).

58
CAN.
J.
MICROBIOL.
VOL. 38, 1992
Pk
(km)
FiG.
2. Longitudinal variations of (a) bacterial abundance
(lO'
cells
L~')
and (b) biomass
(/ig
C
L~')
between Montereau (Pk 546)
and
Poses (Pk 846) in April (A), July (•), and October 1989 (•). Pk, in
kilomètres
(see text).
Counting
was performed with a Leitz microscope (Laborlux)
fit-
ted
with a
3-Lambda-Ploemopak
illuminator for epifluorescence
and
a
100-W
Hg
lamp.
A
100
x Fluotar
oil
immersion objective
was
used at a total magnification of
1000 x.
Bacteria
were counted on
12
fields on one
préparation
or on 8
fields
when
préparations
were duplicated. Bacteria (250-500) were
counted
for each
préparation.
The bacteria were classified by four
shape
criteria (coccoids, rods, ellipsoids, and vibrio
like),
and within
each,
six or seven size classes were distinguished. Measurements
were
made during the counting with an eyepiece graticule and
calculations
of biovolumes were based on the formula for
géométrie
shapes.
Biovolumes were converted into a carbon unit, using a
conversion
factor that varies in relation to the
cell
biovolume, from
4
X 10"" g C
,im"'
for smaller bacteria (0.026
/^m')
to
1.3
x
10
g C
^^m '
for larger ones (>0.4
/j.m^)
(Simon and Azam
1989).
Culture
experiments
In
June and October 1989, at the Porcheville station, river-water
culture
experiments were performed according to Ammerman et at.
(1984).
However, the inoculum was filtered through a
2-^tm
Nuclepore
membrane filter to eliminate grazers and diluted
10-fold
into
0.22
tim
filtered river water. Considering the size of the bac-
teria,
the filtration of the inoculum through 2
nm
was assumed
to
be the best compromise to separate the bacterial populations
from
the grazers community (see Fig. 5). The culture was incubated
in
the dark at in situ
température
(20 and 15°C in June and
October,
respectively). About 10 samples were taken during 20 h

GARNIER
ET AL.
59
16
.
12
4
.
«
0
8
16
%
12
.
I
I I
I
I I
1
2 3
10-4-89
T
r
209-89
244-89
T—r
10-10-89
I
I I
1
2 3
I
I I
1
2 3
19««9
8-12 89
Station
5-7-89
6-2-90
1 I
I
1
2 3
1
2 3
23-8
«9
-I—r
24-340
FiG.
3. Longitudinal and seasonal (day-month-year) variations of bacterial abundance
(lO'
cells L"') around the water-treatment
plant
of Achères (Pk 708). Station 1, upstream from Achères, la Frette (Pk 706); station 2, downstream from Achères (Pk
713);
station
3,
Poissy (Pk 721); station 4, Porcheville (Pk 750); station 5, Méricourt (Pk 765). Pk, in
kilomètres.
TABLE 1.
Average profiles of total bacterial abundance and biomass around the perturbed zone
of
the water-treatment plant of Achères (Pk 708). Distribution of the relative abundance and
biomass
of
small
( <
1 fim)
and large ( >
1
m) bacteria
Total
bacterial abundance
Biomass
(Pk) xlO'
cells L"'
%
>
1 /im
lig
C
L-'
%
<1
;im "/o >1 ;tm
706
5.4
82.5 17.5
107.9 61.8 38.2
713
11.2
69.7
30.3
489.5 32.9
67.1
721
9.7 73.3
26.7 322.3
44.7
55.3
750 6.0
85.9
14.1
106.1
68.6
31.4
765 4.9
88.4
11.6 83.3
72.2
27.8
for détermination
of bacterial abundance and biomass as described
above.
Results
Spatial
variations of bacterial abundance and biomass
The
three longitudinal profiles studied in
April, July,
and
October along
the 300 km separating Montereau from Poses
followed
the same
gênerai
pattern (Fig. 2); they
ail
showed
a
sharp increase both in abundance and biomass in the area
where
the river receives effluent from the waste water treat-
ment
plant of Achères. This was followed by a marked
decrease,
so that values close to those found upstream from
Paris
were observed from Méricourt (Pk 765) to Poses
(Pk
847).
Similar
observations were made at shorter
sampUng
inter-
vais
of time, for the zone perturbed by the effluent of the
waste
water treatment plant,
although
the impact of Achères
appears
to vary in intensity (Fig. 3).
The
abundance of attached bacteria was estimated
only
in
April 1989, at the highest rate of discharge. The attached
bacteria
represented 23% of the total abundance, as an aver-
age
(SD = 10), over the longitudinal profile. The abundance
of
attached bacteria is higher when there is elevated
suspended
matter (Healey et al. 1988), which is the case at
high
discharge rates; therefore, the proportion of attached
bacteria
observed here at the highest discharge rate can be
considered
as maximum.
An
expected linear relationship existed between abun-
dance
and biomass for most of the data (Fig. 4). However,
biomass
values were higher for the stations located just
downstream
from Achères (Pk 713 and 721), where larger
bacteria
were observed.
The
shift in size, as affected by the plant effluent, is clearly
shown
by the appearance of a second peak in size distribu-
tion
towards higher values
(>
1
nmïn
greatest dimension),
whereas
upstream from the plant and farther downstream,
the
size distribution was unimodal (at a size < 1
^tm)
(Fig. 5).
Thèse
observations allowed two size classes of bacteria to
be
distinguished, i.e., small bacteria (greatest dimension
<
1
^m)
and large bacteria
(>
1
^m).
An
increase or a decrease in total bacterial abundance and
biomass
was related to an increase or a decrease in abun-
dance
and biomass of the two size classes (Table 1). Small
cells
dominated the bacterial populations at
ail
the stations
and
represented on average at
least
69.7% of the total abun-
dance.
However, the proportion of large bacteria increased
at
the station downstream from the waste water treatment
plant
of Achères (Table 1). Regarding the biomass, the pro-

portion
of
small
bacteria represented at
least
60% above the
water
treatment plant of Achères but decreased notably
down
to 32.9% just
below.
As a
whole,
for both abundance
and
biomass, proportions identical with those found above
Achères
were again observed at the
médian
station of the
reach,
Porcheville, Pk 750 (Table 1).
The
discharge of the waste water treatment plant pre-
sented
little variation during the course of the year, from
25
to 32
m^
s"', and represented from 4 to 60% of the
discharge
of the river. The abundance of large bacteria, just
below
the water treatment plant, increased rapidly with
increasing
percentage of discharge, showing that
thèse
larger
bacteria
were concentrated under
low
discharge of the river
(Fig.
6). This concentration effect was not observed for
small
bacteria.
Temporal
variations of abundance and biomass
At
Porcheville, abundance and biomass, respectively,
varied
from 3.6 x
lO'
to 13 X
lO' cells L"'
and from
69
to
294,1g
C L"' from
April
1989 to March 1990
(Fig.
la). The decrease in both abundance and biomass,
from
13 X
lO'
to 5 x
lO' cell
L"' and from 294 to
100/ig
C
L~'
respectively, was associated with a con-
sidérable réduction
in discharge rate, from 936 to 287
m'
s"'
(Fig. la). Although both the rates of discharge and
their
amplitude of variations were relatively low from May
1989
to March 1990, the increases in bacterial populations
coincided
most of time with augmentations in discharge
rates.
Similar coïncidences were
also
observed between bac-
teria
and discharges upstream from the water treatment plant
of
Achères (Fig.
Ib).
Growth
and disappearance rates of small and large bacteria
Water
culture experiments conducted with Porcheville
water
show a distinct exponential growth for small- and
large-sized
bacteria (Fig. 8). Although
thèse expérimental
conditions
do not lead to the
détermination
of an in situ
growth
rate of the populations, calculations show that the
large
bacteria were able to growth at rates about twice
(0.129
and 0.118
h"' in June and October, respectively) that of
small
ones (0.065 and 0.071 h"' in June and October,
respectively).