Aquatic Toxicology 47 (2000) 243 – 257
www.elsevier.com/locate/aquatox
Long-term toxicity of zinc to bacteria and algae in
periphyton communities from the river Göta A8 lv, based on
a microcosm study
Maria Paulsson, Bo Nyström, Hans Blanck *
Department of Plant Physiology, Göteborg Uni6ersity, Box 461, 405 30 Göteborg, Sweden
Received 20 September 1998; received in revised form 26 February 1999; accepted 4 March 1999
Abstract
This study aims to clarify the relation between pollution-induced community tolerance (PICT), community
structure and net production in periphyton communities exposed to environmentally realistic concentrations of zinc
and to determine levels of no-effect (NEC). Therefore, periphyton communities from a relatively uncontaminated
river (Göta A8 lv, Sweden) were exposed to zinc during 4 weeks in a flow-through aquaria system. PICT was estimated
as the increase in EC50 for the short-term inhibition of photosynthesis and thymidine incorporation. Community
structure was measured as species richness and Bray – Curtis similarity index, and biomass as dry weight and
chlorophyll a. NEC values were estimated as the intercept between a regression line and a control base line. NEC
values for biomass and biomass dependent variables were 0.12 – 0.42 mM which implies that there may be many rivers
affected by zinc. However, the PICT response and marked changes in community species composition were found
only at much higher concentrations (9.7 mM). We hypothesise that this discrepancy in effect concentrations between
biomass-dependent variables and other structure-related variables (including PICT) is due to an interaction between
zinc and phosphorus leading to nutrient depletion. An indirect toxic effect on biomass due to nutrient deficiency
should not be detectable as an increased zinc tolerance. © 2000 Elsevier Science B.V. All rights reserved.
Keywords: Zinc; Toxicity; PICT; Periphyton; Microalgae; Bacteria
1. Introduction
* Corresponding author. Tel.: + 46-31-773 2609; fax: +4631-773 2626.
E-mail address: [email protected] (H. Blanck)
Zinc is a widely used heavy metal and concentrations of total zinc in European rivers range
from nmoles per litre to near hundred mmoles per
litre in the most polluted ones (Whitton et al.,
1982), the latter comprise around 10% of 176
0166-445X/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 6 - 4 4 5 X ( 9 9 ) 0 0 0 1 3 - 2
244
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
European river stations investigated (Stanners
and Bourdeau, 1995). Long-term effects of zinc
on microbenthic communities have generally been
reported at concentrations ranging from 0.05 mg
l − 1 (0.8 mM) to 2.5 mg l − 1 (38 mM) (Williams
and Mount, 1965; Genter et al., 1987; Colwell et
al., 1989; Dean-Ross, 1990; Niederlehner and
Cairns, 1992, 1993; Loez et al., 1995) implying
that long-term effects only can be expected in the
more polluted ones. However, zinc concentrations
as low as 4.2 mg l − 1 (0.06 mM) has also been
reported to affect microbenthos (Pratt et al.,
1987). In most of these studies effects were found
already at the lowest exposure concentration. Noeffect concentrations (NEC) of zinc seems therefore to be open for debate. There is thus a need
for better ways to determine NEC levels to be
able to assess environmental hazard of long-term
zinc exposure.
In a field study of a heavily zinc-contaminated
stretch of the river Dommel in Belgium performed by Admiraal et al. (1999), microbenthic
communities showed high tolerance to zinc, as
determined by short-term tests on photosynthesis
and thymidine incorporation (EC50 around 1 mM
for algae and 0.4 to \ 1 mM for bacteria).
Thymidine incorporation into bacteria indicated
a gradual increase in tolerance with increasing
zinc concentrations from 4 to 102 mM. These
elevated tolerance levels might be due to a longterm selection, favouring tolerant organisms and
leading to a tolerant community as predicted by
the pollution-induced community tolerance
(PICT) concept (Blanck et al., 1988). It states
that exposure of a community to a toxicant
above its effect threshold exerts a selection pressure on sensitive species or individuals leading to
their exclusion. The community will consist of
more tolerant life-forms and the overall tolerance
of the community will increase as a consequence
of this toxicant-induced succession. The induced
tolerance can be detected in a short-term test as
an increase in, e.g. the EC50 of community photosynthesis to that particular toxicant. The tolerance increase is coupled to changes in species
composition as demonstrated in controlled experiments with different toxicants such as arsenate
(Blanck and Wängberg, 1988a,b; Wängberg et al.,
1991), 4,5,6-trichloroguaiachol (Molander et al.,
1990), copper (Gustavson and Wängberg, 1995)
and tri-n-butyltin (Dahl and Blanck, 1996a).
However, it cannot be excluded that other factors
than selection contributed to the high tolerance
found in the river Dommel. Periphyton sensitivity
to zinc varied by a factor of 5–6 between different running waters in the same drainage area
(Paulsson and Blanck, 1995). This could most
likely be attributed to bioavailability or other
environmental factors since there were no major
differences in zinc or other metal pollution between the sites.
To clarify whether or not the zinc tolerance
found in the Dommel was due to zinc exposure
leading to a selection for tolerant life-forms
(PICT), it had to be shown, in controlled experiments, that zinc concentrations found in the
Dommel can induce zinc tolerance and a succession to zinc tolerant species. Therefore, a controlled microcosm experiment was made with
periphyton from the river Göta A8 lv. This river is
relatively unpolluted by metals and nutrients (e.g.
0.02–0.09 mM of zinc) making it possible to
avoid communities already pre-selected for zinc
tolerance. By using a regression design we could
in this way also determine threshold concentrations (NEC) for long-term effects of zinc on periphyton communities unexposed to elevated
concentrations of zinc.
2. Material and methods
Periphyton communities, from the river Göta
A8 lv on the west coast of Sweden (57°46%N,
12°00%E), were exposed to zinc during 4 weeks in
a microcosm system between the beginning of
September and the beginning of October in 1995.
After the 4-week exposure, long-term effects on
photosynthetic and bacterial activity, biomass
(dry weight and chlorophyll a), species composition and tolerance development were measured.
Tolerance was quantified as short-term toxicity of
zinc
on
photosynthesis
and
thymidine
incorporation.
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
2.1. Microcosm system
An indoor flow-through aquaria system, previously described by Blanck and Wängberg (1988a),
was used. Water, with its content of indigenous
microbiota, was continuously taken from the river
at a water depth of about 0.5 m. It was pumped
to 12 22-l glass aquaria with an air-driven teflonmembrane pump (Dominator Maskin AB, Sweden). Before reaching the aquaria the water
passed through a nylon net with a mesh size of 1
mm to prevent larger organisms from entering. A
flow distributor (Blanck and Wängberg, 1988a),
modified to radial symmetry, maintained a continuous water flow of 221 ml min − 1 through each
aquarium, giving a mean residence time of 99
min. Stock solutions of zinc were delivered at a
flow rate of 0.5 ml min − 1 by means of a peristaltic pump (Ismatech IPN 26, Ismatech AB,
Switzerland). The stock solutions were prepared
by dissolving zinc chloride (ZnCl2, \98% purity,
Merck, Germany) in Milli-Q water in a geometric
concentration series giving final estimated concentrations of zinc in the aquaria from 10 − 7.3 to
10 − 4.5 M and a 2 ‰ dilution of the natural water
with Milli-Q water. Four aquaria received no zinc
and were used as controls. All solutions were
stored in darkness. Flow rates of river water and
zinc solutions were regulated daily to get the right
concentration of zinc in the aquaria. Temperature
and pH were also measured regularly, ranging
from 12 to 15°C and 6.1 – 7.1, respectively, during
the 4 week experiment. Algae and bacteria in the
incoming water settled on glass discs (1.5 cm2, 160
per aquarium) arranged vertically along the sides
of the aquaria as described by Blanck and Wängberg (1988a). To obtain a rough surface for the
periphyton to settle on, the glass discs had been
blasted with aluminium oxide with a grain size of
0.1 – 0.2 mm. Before inserted into the aquaria, the
glass discs were cleaned in hot concentrated nitric
acid, rinsed in de-ionised water and finally in 70%
ethanol. A Plexiglas mixing device in each of the
aquaria maintained a continuous mixing of zinc
solutions and incoming water and reduced the
boundary layer between the bulk water and the
periphyton. Each aquarium was illuminated from
above by two fluorescent light tubes (Osram Lu-
245
milux daylight L18 W/12) giving a photon flux
density (PAR) at the surface of around 120 mE
m − 2 s − 1. A timer controlled the light regime and
was adjusted weekly to the actual daylength
which corresponded to a light period between 11
and 13 h.
2.2. Analysis of total zinc in ri6er water
Water samples were taken from each aquarium
after 2, 3 and 4 weeks for analysis of total zinc.
The samples were kept frozen in acid-rinsed (2 M
HNO3) polyethylene bottles until analysis. Fifty
microlitres of HNO3 (concentrated, ultrapure)
and 50 ml of H2O2 (30%, suprapure) were added
to 10 ml of sample followed by a 3-h digestion of
the sample in a Metrohm UV digester (550 W).
After cooling, the samples were analysed with
Differential Pulse Anodic Stripping Voltammetry
(Metrohm VA processor 646, VA stand 647) using a Hanging Mercury Drop Electrode and an
Ag/AgCl as reference. Ten millilitres of Milli-Q
water was added followed by 200 ml of sodium
acetate (3 M, pH 4.7) to increase the ionic
strength and adjust the pH. The sample was deaerated with nitrogen and stirred for 5 min after
which a pre-electrolysis at − 1.3 V for 60 s was
made. The scan rate was 2 mV s − 1 and the
measuring time was 4 min. A standard solution of
zinc nitrate (1 g l − 1, Titrisol, Merck) was used for
making a standard curve from which the total
zinc concentration was calculated. Samples with
high concentrations of zinc were diluted with
Milli-Q water before analysis. All equipment (in
contact with the samples) had previously been
washed with 10% nitric acid.
2.3. Short-term toxicity tests
Short-term toxicity tests on photosynthesis and
thymidine incorporation in periphyton communities were scheduled over 4 days (Table 1). Three
tests were made each day, three treatments including one control. Periphyton glass discs were transported and kept in aquarium water. Tests were
started within 1 h after sampling. Atypical samples were sorted out and the discs cleaned on all
but the rough front surface and put into scintilla-
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
246
microlitres of a H14CO−
3 -solution was added to
each vial after the 2 h pre-incubation with zinc.
The H14CO−
3 -solution was prepared by a 200-fold
dilution of an Amersham CFA stock solution (2
mCi ml − 1, Amersham Laboratories, UK) in
Milli-Q water (adjusted to pH 9 with 0.1 M
NaOH) giving a final activity of 0.25 mCi ml − 1
(9.25 kBq ml − 1) and a concentration of Milli-Q
water of 0.025% in the vials. After 1 h of incorporation 200 ml of formaldehyde was added to terminate the incorporation. To drive off the
remaining inorganic carbon, 200 ml of concentrated acetic acid was added and the samples
dried at 60°C under a stream of air. To enhance
the release of incorporated 14C-carbon the periphyton were soaked in 1 ml of dimethylsulfoxide
(DMSO) at least for 10 min (Filbin and Hough,
1984). After adding 9 ml of a scintillation cocktail
(Ready Safe™, Beckman AB, Sweden), the
amount of 14C-carbon was measured in a liquid
scintillation spectrometer (Beckman, USA). The
activities, as disintegrations per minute (dpm),
were calculated from the counts per minute (cpm)
data, using external standard calibration and automatic quench compensation. The abiotic activity was then subtracted from the obtained dpm
values.
tion vials with the periphyton-covered side pointing upward. The samples were then pre-incubated
in 2 ml of GF/F (Whatman) filtered river water
and 2 ml of a zinc test solution in a geometric
concentration series (10 − 6 – 10 − 3 M) in a temperature-controlled water bath under continuous
shaking and light exposure for 2 h. Controls
without added zinc were also included. Five replicates were used per concentration. The zinc test
solutions were prepared from GF/F filtered river
water and stock solutions of zinc chloride (ZnCl2,
p a, \ 98% purity, Merck, Germany) dissolved in
Milli-Q water (pH adjusted with 0.1 M NaOH if
necessary). The final concentration of Milli-Q water in the scintillation vials was 2‰. Samples for
determination of abiotic activity (n =2) were
treated with 200 ml of 37% formaldehyde and run
together with the concentration series. The photon flux density from fluorescent light tubes (Osram Lumilux Daylight L18W/11) at the bottom of
the incubator was set to 80 mE m − 2 s − 1 and the
temperature adjusted to the temperature in the
aquaria.
2.3.1. Periphyton photosynthetic acti6ity
Periphyton photosynthetic activity was measured as incorporation of radioactively labelled
carbon, added as 14C-bicarbonate. One hundred
Table 1
Sampling schedule for long-term parameters and short-term testsa
Date
Aquaria no.
02/10/95
05/10/95
All
2
8
3
6
1
7
9
4
12
10
11
5
All
06/10/95
07/10/95
08/10/95
09/10/95
a
Nominal zinc
exposure
(log M)
−7.3
−5.7
Control
Control
−6.5
−4.9
−5.3
Control
−6.9
Control
−6.1
−4.5
Photosynthesis
Thymidine
incorporation
Dry weight
Chlorophyll a
Species
–
x
x
x
x
x
x
x
x
x
x
x
x
–
–
x
x
x
x
x
x
x
x
x
x
x
x
–
–
–
–
–
–
–
–
–
–
–
–
–
–
x
–
x
x
x
x
x
x
x
x
x
x
x
x
–
x (fresh)
x
x
x
x
x
x
x
x
x
x
x
x
–
–: No samples taken nor experiments performed; x: Samples taken or experiments performed.
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
2.3.2. Thymidine incorporation into bacteria
Incorporation of 3H-thymidine was used as a
measure of bacterial activity. After 2 h of pre-incubation with zinc, 50 ml of a 3H-thymidine solution was added, giving a final concentration of 20
nM (6.6 mCi) in the vials. The incorporation was
allowed to proceed for 30 min. The 3H-thymidine
solution was prepared from a 3H-thymidine stock
solution (83 Ci mmol − 1, Amersham Laboratories,
UK) and diluted 7.4-fold with filtered (0.22 mm)
Milli-Q water. The incorporation was terminated
by adding 200 ml of formaldehyde. The samples
were then extracted by adding NaOH to a final
concentration of 0.25 M at 60°C for 60 min. This
treatment facilitated the detachment of periphyton from the substrata and the release of labelled
macromolecules. In case of unspecific labelling,
this treatment will also hydrolyse RNA. After
cooling the samples on ice, 100% (w/v) ice-cold
trichloro acetic acid (TCA) was added to a final
concentration of 22%. When over 30 min had
passed the samples were filtered on to nitrocellulose filters (0.2 mm, Sartorius) and the vials rinsed
and stirred thoroughly with 2 ×1.5 ml of ice-cold
TCA (5%). In-between the rinsings the samples
were kept on ice. The filters were washed five
times with 1.5 ml of ice-cold TCA (5%) and twice
with 80% ice-cold ethanol. The steel funnels keeping the filters in place on the filtering manifold
(The International Institutional Agency for 14C
Determination, Water Quality Institute, Denmark), were removed and the filters cleaned twice
more with 80% ice-cold ethanol to make sure that
the non-filtering margins were rinsed as well. Finally, the filters were dried with a hair dryer. The
filters were put into new scintillation vials and 10
ml of a scintillation cocktail (Ready Protein+™,
Beckman Instruments, USA) was added. When
the filters had dissolved, the amount of incorporated 3H-thymidine was measured in a liquid scintillation spectrometer, as previously described for
photosynthetic activity.
2.4. Quantification of community tolerance
Community tolerance to zinc after long-term
zinc exposure was quantified by short-term toxicity tests and expressed as the EC50 estimated
247
through lin-log interpolation between the mean
value of the activities (dpm, n= 5) and log zinc
concentration. The mean value of the control
samples in each test was set to 100% activity.
2.5. Chlorophyll a
The content of chlorophyll a was used as an
indicator of algal biomass. Subsamples (n= 3)
were taken from the same disc population used in
the short-term toxicity tests (Table 1). Periphyton
glass discs (n= 2–8 depending on the amount of
biomass), cleaned on all but the rough surface
side, were put into 1 ml of DMSO (Hiscox and
Israelstam, 1979) and stored frozen until analysis.
All handling of samples were performed in dim
light. The chlorophyll a was extracted at 60°C for
30 min. After cooling the samples, an equal volume of acetone (Shoaf and Lium, 1976) was
added and the extracts analysed spectrophotometrically (Milton Roy Spectronic 3000 Array). The
chlorophyll a concentration was calculated according to equations given by Jeffrey and
Humphrey (1975).
2.6. Dry weight
Samples for determination of periphyton dry
weight were taken from all aquaria at the same
day (Table 1). One Plexiglas plate from the mixing
device was sampled from each aquarium. The
periphyton biofilm was brushed into a specified
volume of river water. A volume of the slurry was
then filtered on to pre-weighed filters (GF/F
Whatman) and dried (100°C) to a constant
weight.
2.7. Analysis of species composition
Species composition was quantified using the
Bray–Curtis similarity index (Bray and Curtis,
1957) based on relative abundance (0–50), and
species richness quantified as the number of different taxa or groups of taxa. One glass disc
(n=3), cleaned on all but the rough surface side,
was put into 5 ml of 70% ethanol and stored at
4°C until analysis. The presence of species or
taxonomically different groups was identified in
248
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
Fig. 1. Short-term effects on thymidine incorporation and photosynthesis in periphyton from a control aquarium (07/10/95) without
added zinc. Nominal zinc concentrations are given. The activity (n = 5) at each concentration is related to the mean control activity
(n= 5). Bars represent standard deviation with the large SD referring to thymidine incorporation. Dotted lines indicate the EC50.
50 randomly chosen fields (diameter of 252 mm)
per disc, using a phase contrast microscope with a
magnification of 1000. This gives a relative abundance of species between 0 and 50. According to
Dahl and Blanck (1996b) the number of new taxa
found, reached a plateau at 35 – 50 counted fields.
Later experiments on Göta A8 lv periphyton
confirm these findings (unpublished data). Fresh
samples (n=1) were taken from all aquaria on 1
day (Table 1) and put into river water. An
overview of these samples was recorded on a
video recorder to facilitate later identification of
species.
2.8. Statistical analysis
No-effect concentration (NEC) values were estimated for photosynthetic activity, bacterial activity, chlorophyll a, dry weight and species
composition which all decreased at a certain concentration of zinc in a concentration dependent
way. The NEC range was determined with a
linear regression technique (Liber et al., 1992;
Dahl and Blanck, 1996a). A linear regression with
a 95% confidence limit was calculated for values
that decreased with increasing zinc concentration.
Close to the NEC intercept, a decision had to be
made whether or not to include a certain data
point. Our criteria for selecting which values to
include in the regression was: (1) a significant
slope (PB 0.05); and (2) a maximum r 2 value. A
mean with a 95% confidence limit was then calculated for the controls. This mean represented a
no-effect baseline. The NEC range was then calculated as the intercept of the regression line with
this baseline mean, including an upper and lower
limit calculated as the intercepts of the 95% confidence limits.
When the effect parameters showed an increase
with time in the controls that could be related to
growth, the similarity between the controls were
tested in an ANOVA (PB 0.05). For parameters
with a significant difference between the controls
(chlorophyll a and photosynthetic activity), the
data were related to the control of the day of the
experiment (Table 1). For comparability the same
approach was used for bacterial activity.
3. Results
Thymidine incorporation was more sensitive to
zinc than photosynthesis (Fig. 1). The EC50 in the
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
Table 2
Nominal and corresponding analysed concentrations of total
zinc in the aquariaa
Concentration of total zinc (mM)
Nominal
Analysed 9SD
0
0.05
0.13
0.32
0.79
2
5
13
32
0.09 90.05
0.11 90.01
0.14 90.08
0.28 90.01
0.62 9 0.05
1.5 9 0.03
3.6 9 0.3
9.7 9 2.7
25 91.7
Deviation factor
(analysed/nominal)
2.2
1.1
1/1.1
1/1.3
1/1.3
1/1.4
1/1.3
1/1.3
a
The analysed concentrations were based on the mean from
three weekly samples. The mean for the controls were based
on 12 measurements.
short-term toxicity test on thymidine incorporation was 10 mM while the EC50 for photosynthesis
was 56 mM. This does not necessarily mean that
bacteria are more sensitive than algae, since the
two methods used are reflecting two different
metabolic processes.
249
Long-term effects of zinc was determined in a
4-week microcosm experiment. The mean concentration of total zinc in the control aquaria was
0.09 mM (Table 2). Added amounts of zinc led to
a factor of 1.3–1.4 lower values than nominal for
the higher concentrations and somewhat varying
values at lower concentrations. The response of
all long-term effect parameters was thus expressed
in relation to analysed concentrations.
Long-term effects on periphyton biomass were
detected at much lower concentrations than shortterm effects (Figs. 2 and 3). The NEC values,
determined as the intercept between the linear
regression with the mean no-effect baseline, were
0.15 mM for dry weight and 0.42 mM for chlorophyll a (Fig. 4). The long-term NEC values for
photosynthetic and bacterial activity (Fig. 5) were
within the same range. It appears that biomass
and biomass-dependent variables (like photosynthetic and bacterial activities) are particularly sensitive to zinc. The ranges of NEC values are given
in Fig. 4.
Several other long-term effect indicators were
less sensitive. The assimilation ratio, i.e. 14C incorporated per hour and chlorophyll a, was not
clearly affected until exposed to 9.7 mM of total
Fig. 2. Periphyton community dry weight after long-term exposure to zinc. The dry weight is expressed in mg cm − 2 of one plate
from the mixing device in each zinc exposed aquarium. The zinc concentrations given are analysed total zinc in the aquaria water.
The horizontal solid line indicates the mean baseline of no-effect. The solid regression line is based on the values between 0.11 and
3.6 mM of zinc (n = 6; P B 0.001). The dotted lines indicate 95% confidence limits for the regression. A NEC value of 0.15 mM was
determined from the intercept.
250
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
Fig. 3. Chlorophyll a content of periphyton communities after long-term exposure to zinc. The relative chlorophyll a content per
glass disc (n =3) at each concentration was calculated as proportion of the mean chlorophyll a of the corresponding control (0.09
mM; n= 3) on the day of sampling (Table 1). The zinc concentrations given are analysed total zinc in the aquaria water. Bars
represent standard deviations. The horizontal solid line indicate the mean baseline of no-effect. The solid regression line is based on
the values between 0.28 and 3.6 mM of zinc (n = 12; PB 0.001). The dotted lines indicate 95% confidence limits for the regression.
A NEC value of 0.42 mM was determined from the intercept.
zinc (Fig. 6). This indicates that there were no
‘direct’ effects of zinc on photosynthesis below
this concentration, which is consistent with the
increase in community tolerance (PICT) to zinc
for both bacteria and algae which occurs between
9.7 and 25 mM (Fig. 7). At the same threshold
(zinc concentration) there was also a marked decrease in community similarity, i.e. a change in
species composition (Fig. 8), all in accordance
with the PICT concept. The increase in tolerance
was not correlated to any expansion of tolerant
algal species but coincided with a general decrease in abundance of all species, especially the
two dominant species Achnanthes minutissima
‘Kützing’ and Synedra sp. D (Fig. 9). Accordingly, this leads to a decreased assimilation ratio
and not to an increased one as can be expected
by the increase in tolerance. Only at much lower
zinc concentrations (0.1 – 1 mM) there was a sign
of succession favouring any algal species. This
mismatch between induced zinc-tolerance and
zinc-induced succession is in conflict with the
PICT concept.
However, the NEC value for species richness
was in the same range as the biomass-dependent
NEC values (Figs. 4 and 10), which indicates a
biomass dependence also for this variable. Most
likely, it was the detection of rare species that
was biomass dependent. The total number of
algal taxa in the entire experiment was 108, while
at maximum only 40 were found in one aquarium. Obviously there where many scarce species
with a random presence in our experimental system.
4. Discussion
Long-term effects of zinc on both algae and
bacteria in river periphyton was observed at 0.12–
0.42 mM. This threshold concentration is presently
exceeded in many European rivers (Fig. 11), indicating that zinc might affect microbiota in these
environments. All biomass-dependent variables
for both algae and bacteria had NEC values in
this range (Figs. 2–5). PICT could not detect
these biomass-related effects. However, there was
a consistent response of several effect indicators
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
251
Fig. 4. NEC ranges of zinc for periphyton long-term parameters. Mean NEC values are given with their lower and upper limits. The
equation for the linear regression (LR) and r 2 are also given.
suggesting ‘direct’ zinc toxicity, but only at higher
concentrations (\ 9.7 mM). Short-term effects on
thymidine incorporation and photosynthesis occurred in the same concentration range (Fig. 1)
and so did the assimilation ratio (14C incorporated per hour and chlorophyll a) (Fig. 6) as well
as community tolerance (Fig. 7). Community species composition changed slightly at low concentrations but was strongly affected only with more
severe zinc stress.
This inconsistency led us to hypothesise that the
impact of zinc might be due to different mechanisms, one ‘direct’ due to zinc toxicity and one
‘indirect’ on nutrient availability. An interaction
with a growth-limiting nutrient like phosphate
would explain our observations. The fact that the
solubility product of zinc phosphate is quite low
(10 − 35..3 M) gives some support to the idea. Zinc
interferes with cellular phosphorus in algae (Bates
et al., 1985; Kuwabara, 1985), and observations
of increased activity of alkaline phosphatase after
zinc exposure (Pratt et al., 1987; Vinter et al.,
1987; Cohen et al., 1991; Chappell and Goulder,
1994; Wong et al., 1995) imply that there might
also be a zinc-phosphorus interaction extracellularly, making phosphorus less available to the
periphyton community. Free zinc ion concentrations were analysed after the addition of different
concentrations of total zinc to Göta A8 lv water
(Fig. 12). With an analysed phosphate concentra-
Fig. 5. a – c: Photosynthetic activity and bacterial activity in periphyton after long-term exposure to zinc. The activity (dpm) per glass
disc (n =5) at each concentration is calculated as the proportion of the mean corresponding control activity (0.09 mM; n =5) of the
short-term tests performed that day (Table 1). a–b: Activities with standard deviations represented by bars; and c: mean activities
(n =5). The horizontal solid lines indicate the mean baseline of no-effect. The solid regression lines are based on the values between
0.11 and 9.7 mM of zinc (n= 35; PB 0.001). The dotted lines indicate 95% confidence limits. The zinc concentrations given are
analysed total zinc in the aquaria water. NEC values of 0.17 mM for photosynthetic activity and 0.12 mM for bacterial activity were
determined from the intercepts.
252
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
Fig. 5.
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
253
Fig. 6. Periphyton assimilation ratio after long-term exposure to zinc. The assimilation ratio is expressed as the mean incorporated
C (dpm; n = 5) per hour and mean mg chlorophyll a (n = 3). The 14C and chlorophyll measurements were made on different
subsamples but from the same sampled population. The assimilation ratio is expressed as the proportion of a corresponding control
of the short-term tests performed that day (Table 1). The zinc concentrations given are analysed total zinc in the aquaria water.
14
tion of 0.2 mM in Göta A8 lv, zinc phosphate can
precipitate at a free zinc ion concentration higher
than 0.05 mM, assuming interaction between zinc
and phosphate only. The corresponding total zinc
Fig. 7. Periphyton community tolerance. The community tolerance was estimated as the EC50 from short-term tests (Fig. 1).
Long-term zinc concentrations given are analysed total zinc in
the aquaria water.
Fig. 8. Species composition of periphyton after long-term
exposure to zinc. Species composition is expressed as Bray –
Curtis similarity indices. Each index represent a mean similarity (n =4 – 9) to a control (0.09 mM), generating four indices
per concentration and three per control. The zinc concentrations given are analysed total zinc in the aquaria water. The
horizontal solid line indicate the mean baseline of no-effect.
The solid regression line is based on the values between 3.6
and 25 mM of zinc (n = 12; PB0.001). The dotted lines
indicate 95% confidence limits for the regression. A NEC value
of 1.8 mM was determined from the intercept.
254
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
Fig. 9. Relative abundance of periphyton algal species after
long-term exposure to zinc. Each value represents the mean of
the relative abundance (0–50) on three discs at each zinc
concentration. Indicated are the two most abundant species,
A. minutissima and Synedra sp. D. The total amount of species
or groups of species presented in the figure is 108. The zinc
concentrations given are analysed total zinc in the aquaria
water.
Fig. 10. Periphyton species richness after long-term exposure
to zinc. Species richness represents the number of algal taxa or
groups of taxa found in any of the 50 fields counted on a glass
disc (n =2 – 3) at each concentration. Zinc concentrations are
based on analysed total zinc in the aquaria water. Bars represent standard deviations. The horizontal solid line indicate the
mean baseline of no-effect. The solid regression line is based
on the values between 0.28 and 25 mM of zinc (n= 17;
P B0.001). The dotted lines indicate 95% confidence limits for
the regression.
concentration for Göta A8 lv water was roughly
estimated to be 0.11 mM through extrapolation.
This is the concentration range where the biomassdependent variables start to respond to zinc stress
(Figs. 2–5). However, according to preliminary
model calculations, a zinc phosphate precipitate is
thermodynamically unlikely in natural waters, due
to the presence of competing ligands. Whether a
kinetic perspective would lead to a dissimilar conclusion remains to be tested. However, there are
circumstantial biological evidence of phosphate
depletion due to zinc exposure, since a stimulation
of alkaline phosphatase together with a reduced
phosphate retention and reduced biomass accumulation was observed in periphyton upon zinc stress
(Pratt et al., 1987).
PICT was originally (Blanck et al., 1988) suggested to discriminate between direct and indirect
effects of toxicants. In this perspective it seems
reasonable to speculate that ‘direct’ toxicity of zinc
to periphyton occurs only at 9.7 mM of zinc or
higher, which is the concentration at which both
bacteria and algae increase their zinc tolerance
(Fig. 7). At the same concentration there is a
drastic drop in community similarity, all in accordance with the PICT concept. However, the
biomass decrease occurs at much lower concentrations which appears to be contradictory to the
reasoning above. Assuming that zinc induces a
phosphorus deficiency (which remains to be conclusively demonstrated), the biomass decrease is
easy to explain. It also follows from the same
reasoning that bacterial and algal activity per unit
area (after long-term exposure) would decrease as
a consequence of the reduced biomass. Note, however, that the reduction of photosynthetic activity
per hour and chlorophyll a content (Fig. 6) reflects
‘direct’ toxicity on photosynthesis and thus is
observed only in the higher zinc concentration
range.
Interestingly enough, algae and bacteria seem to
be very similar in their sensitivity to long-term zinc
exposure. This is the case both for the ‘indirect’
effect (Fig. 5c) where bacteria seem to be dependent on algal production (or suffer just as much
from phosphorus limitation), and for the ‘direct’
effects as indicated by their PICT response (Fig. 7).
With our way of determining species composition a low periphyton biomass would affect the
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
255
Fig. 11. Zinc concentration percentiles (horizontal bars) for running waters in south of Sweden (Johansson et al., 1995) and Europe
(Stanners and Bourdeau, 1995) compared to indirect and direct effects of zinc (vertical bars). Indirect effects refer to our estimated
NEC values for biomass and biomass-dependent variables and direct effects to community tolerance, the NEC for species
composition, the assimilation ratio and EC50 for short-term effects. Numbers indicate the different percentiles.
Fig. 12. Estimated threshold concentrations for zinc phosphate precipitation. Free zinc ion concentrations were analysed after
addition of different concentrations of total zinc to Göta A8 lv water. Shown are the solubility product for zinc phosphate and the
analysed phosphate concentration in Göta A8 lv. Dotted lines indicate a free zinc concentration of 0.05 mM with the estimated total
zinc concentration of 0.11 mM, above which zinc phosphate is estimated to precipitate.
chance to detect low-abundant algae, since a species with few individuals per biomass has a lower
probability of being found if the biomass is reduced. The taxa that disappeared at low concentrations were only the low-abundant ones.
However, there were many low abundant taxa
(Fig. 9) and their presence was occasional, i.e.
they disappeared in one sample at low zinc exposure and could reappear in another sample at a
higher zinc exposure. This appears to affect the
species richness indicator stronger than the Bray–
Curtis index. When looking in some detail at the
zinc-induced succession (Fig. 9) it appears that
some species were favoured by the zinc exposure,
256
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
but only in the low concentration range (0.1 – 1
mM). It is not clear whether this is due to release
of competition or a superior ability of these algae
to thrive at low phosphate levels. PICT is unable
to detect this species shift with the endpoints used
in this study for estimating tolerance, which may
point
to
a
phosphate-deficiency-induced
succession.
In view of these results it seems appropriate to
conclude that the elevated tolerance levels found
in the Dommel could be a PICT response reflecting a selection pressure due to high zinc exposure.
The zinc concentrations (4 – 102 mM) to which
Dommel periphyton was exposed (Admiraal et al.,
1999) were sufficient to cause direct effects on both
algae and bacteria (Figs. 7 and 8). However, it can
still not be excluded that other factors also contributed to the elevated tolerance levels for bacteria and algae in the Dommel. There are high levels
of other metals such as cadmium (0.01 – 7 mM) in
the river (Admiraal et al., 1999), and the communities might therefore be cadmium selected as well.
Cadmium-tolerant communities are likely to show
co-tolerance to zinc. Such co-tolerance or multitolerance to metals is a common phenomenon for
both algae and bacteria (Foster, 1982; Takamura
et al., 1989; Hashemi et al., 1994; Gustavson and
Wängberg, 1995; Pennanen et al., 1996). The presence of high phosphate concentrations in river
Dommel (1–6 mM) and high iron deposits in the
periphyton biofilm (Admiraal et al., 1999) should
offer some protection against zinc toxicity to the
periphyton. This might be reflected in the high
tolerance levels observed in the Dommel, 400 mM
for bacteria and around 1000 mM for algae compared to 75 and 180 mM for bacteria and algae,
respectively, in the Göta A8 lv microcosm study.
In conclusion, the most important impact of
zinc on river periphyton seems to be an indirect
toxicity on biomass and biomass-dependent properties. We have hypothesised that the effects are
mediated by interaction of zinc with phosphate
leading to nutrient depletion. Environments where
primary production is phosphorus limited should
therefore be particularly at risk. In waters with
similar chemistry as Göta A8 lv, 0.1 – 0.4 mM of zinc
should have long-term effects on primary production. This concentration range severely overlaps
the so-called No Risk Area for zinc (van Assche et
al., 1996), with zinc deficiency at less than 1.6 mg
l − 1 (0.02 mM) and zinc toxicity at greater than 50
mg l − 1 (0.8 mM). The toxicity of zinc to aquatic life
therefore seems to be strongly underestimated.
Acknowledgements
Mats Engdahl at the Göteborg Municipal
Drinking Water Administration is deeply acknowledged for making it possible to install the
microcosm system at the river Göta A8 lv. Dr Mats
Kuylenstierna at the Botanical Institute, Göteborg
University, is acknowledged for skilful taxonomic
analysis of periphyton algae. Dr Greg Morrison at
the Department of Sanitary Engineering,
Chalmers University of Technology, is also acknowledged for providing and helping us with the
ASV equipment. The project was funded by the
European Union, Environment and Climate, contract numbers EV5V-CT94-0402 and ENV4CT96-0298 and by the Swedish Natural Science
Research Council, contract number S-AA/FM
01507-315.
References
Admiraal, W., Blanck, H., Buckert-De Jong, M., Guasch, H.,
Ivorra, N., Lehmann, V., Nyström, B.A.H., Paulsson, M.,
Sabater, S., 1999. Short-term toxicity of zinc to microbenthic algae and bacteria in a metal polluted stream. Water
Res. 33 (9), 1989 – 1996.
Bates, S.S., Tessier, A., Campbell, P.G.C., Létourneau, M.,
1985. Zinc-phosphate interactions and variation in zinc
accumulation during growth of Chlamydomonas 6ariabilis
(Chlorophyceae) in batch culture. Can. J. Fish. Aquat. Sci.
42, 86 – 94.
Blanck, H., Wängberg, S.-A, ., 1988a. Validity of an ecotoxicological test system: short-term and long-term effects of
arsenate on marine periphyton communities in laboratory
systems. Can. J. Fish. Aquat. Sci. 45, 1807 – 1815.
Blanck, H., Wängberg, S.-A, ., 1988b. Induced community tolerance in marine periphyton established under arsenate
stress. Can. J. Fish. Aquat. Sci. 45, 1816 – 1819.
Blanck, H., Wängberg, S.-A, ., Molander, S., 1988. Pollutioninduced community tolerance — a new ecotoxicological
tool. In: Cairns, J.J., Pratt, J.R. (Eds.), Functional Testing
of Aquatic Biota for Estimating Hazards of Chemicals.
ASTM STP 988. American Society for Testing Materials,
Philadelphia, PA, pp. 219 – 230.
M. Paulsson et al. / Aquatic Toxicology 47 (2000) 243–257
Bray, J.R., Curtis, J.T., 1957. An ordination of the upland forest
communities of southern Wisconsin. Ecol. Monogr. 27,
325 – 349.
Chappell, K.R., Goulder, R., 1994. Epilithic extracellular-enzyme activity in a zinc-contaminated stream. Bull. Environ.
Contam. Toxicol. 52, 305–310.
Cohen, I., Bitan, R., Nitzan, Y., 1991. The effect of zinc and cadmium ions on Escherichia coli B. Microbios 68, 157–168.
Colwell, F.S., Hornor, S.G., Cherry, D.S., 1989. Evidence of
structural and functional adaptation in epilithon exposed to
zinc. Hydrobiologia 171, 79–90.
Dahl, B., Blanck, H., 1996a. Pollution Induced Community Tolerance (PICT) in periphyton communities established under
tri-n-butyltin (TBT) stress in marine microcosms. Aquat.
Toxicol. 34, 305 – 325.
Dahl, B., Blanck, H., 1996b. Toxic effects of the antifouling
agent irgarol 1051 on periphyton communities in coastal water microcosms. Aquat. Toxicol. 32 (4), 342–350.
Dean-Ross, D., 1990. Response of attached bacteria to zinc in
artificial streams. Can. J. Microbiol. 36, 561–566.
Filbin, G.J., Hough, R.A., 1984. Extraction of 14C-labelled photosynthate from aquatic plants with dimethylsulfoxide
(DMSO). Limnol. Oceanogr. 29, 426–428.
Foster, P.L., 1982. Metal resistances of Chlorophyta from rivers
polluted by heavy metals. Freshwater Biol. 12, 41–61.
Genter, R.B., Cherry, D.S., Smith, E.P., Cairns, J.Jr., 1987. Algal-periphyton population and community changes from
zinc stress in stream mesocosms. Hydrobiologia 153, 261–
275.
Gustavson, K., Wängberg, S.-A, ., 1995. Tolerance induction and
succession in microalgae communities exposed to copper
and atrazine. Aquat. Toxicol. 32, 283–302.
Hashemi, F., Leppard, G.G., Kushner, D.J., 1994. Copper resistance in Anabaena 6ariabilis: effects of phosphate nutrition
and polyphosphate bodies. Microb. Ecol. 27, 159–176.
Hiscox, J.D., Israelstam, G.F., 1979. A method for the extraction of chlorophyll from leaf tissue without maceration. Can.
J. Bot. 57, 1332 – 1334.
Jeffrey, S.W., Humphrey, G.F., 1975. New spectrophotometric
equations for determining chlorophylls a, b, c1 and c2 in
higher plants, algae and natural phytoplankton. Biochem.
Physiol. Planzen (BPP) 167, 191–194.
Johansson, K., Bringmark, E., Lindevall, L., Wilander, A.,
1995. Effects of acidification on the concentrations of heavy
metals in running waters in Sweden. Water Air Soil Pollut.
85, 779 – 784.
Kuwabara, J.S., 1985. Phosphorus-zinc interactive effects on
growth by Selenastrum capricornutum (Chlorophyta). Environ. Sci. Technol. 19, 417–421.
Liber, K., Kaushik, N.K., Solomon, K.R., Carey, J.H., 1992.
Experimental designs for aquatic mesocosm studies: a comparison of the ‘ANOVA’ and ‘regression’ design for assessing the impact of tetrachlorophenol on zooplankton
populations in limnocorrals. Environ. Toxicol. Chem. 11,
61– 77.
Loez, C.R., Topalian, M.L., Salibian, A., 1995. Effects of zinc
on the structure and growth dynamics of a natural freshwa-
.
257
ter phytoplankton assemblage reared in the laboratory. Environ. Pollut. 88, 275 – 281.
Molander, S., Blanck, H., Söderström, M., 1990. Toxicity assessment by pollution-induced community tolerance (PICT),
and identification of metabolites in periphyton communities
after exposure to 4,5,6-trichloroguaiacol. Aquat. Toxicol.
18, 115 – 136.
Niederlehner, B.R., Cairns, J. Jr., 1992. Community response to
cumulative toxic impact: effects of acclimation on zinc tolerance of aufwuchs. Can. J. Fish. Aquat. Sci. 49, 2155 – 2163.
Niederlehner, B.R., Cairns, J. Jr., 1993. Effects of previous zinc
exposure on pH tolerance of periphyton communities. Environ. Toxicol. Chem. 12, 743 – 753.
Paulsson, M., Blanck, H., 1995. The response of periphyton
communities to toxicants — a comparison between a northern and a central European river system. Poster presentation
at the 5th SETAC – Europe Congress, 25 – 28 June, Copenhagen.
Pennanen, T., Frostegård, A, ., Fritze, H., Bååth, E., 1996. Phospholipid fatty acid composition and heavy metal tolerance of
soil microbial communities along two heavy metal-polluted
gradients in coniferous forests. Appl. Environ. Microbiol. 62
(2), 420 – 428.
Pratt, J.R., Niederlehner, B.R., Bowers, N.J., Cairns, J.Jr., 1987.
Effects of zinc on microbial communities. In: Lindberg, S.E.,
Hutchinson, T.C. (Eds.), International Conference on
Heavy Metals in the Environment. CEP Consultants, Edinburgh, pp. 324 – 329.
Shoaf, W.T., Lium, B.W., 1976. Improved extraction of chlorophyll a and b from algae using dimethyl sulfoxide. Limnol.
Oceanogr. 21, 926 – 928.
Stanners, D., Bourdeau, P., 1995. Europe’s Environment: the
Dobriš Assessment. European Environment Agency, Luxembourg.
Takamura, N., Kasai, F., Watanabe, M.M., 1989. Effects of Cu,
Cd and Zn on photosynthesis of freshwater benthic algae. J.
Appl. Phycol. 1, 39 – 52.
van Assche, F., van Tilborg, W., Waeterschoot, H., 1996. Environmental risk assessment for essential elements case study:
zinc. In: International workshop on risk assessment of
metals and their inorganic compounds. International Council on Metals and the Environment. Angers, France, pp.
171 – 180.
Vinter, V., Smid, F., Smrckova, I., 1987. Factors influencing the
activity of cellular alkaline phosphatase during growth and
sporulation of Bacillus cereus. Folia Microbiol. 32, 80 – 95.
Wängberg, S-A, ., Heyman, U., Blanck, H., 1991. Long-term and
short-term arsenate toxicity to freshwater phytoplankton
and periphyton in limnocorrals. Can. J. Fish. Aquat. Sci. 48,
173 – 182.
Whitton, B.A., Say, P.J., Jupp, B.P., 1982. Accumulation of
zinc, cadmium and lead by the aquatic liverwort Scapania.
Environ. Pollut. [B] 3, 299 – 316.
Williams, L.G., Mount, D.I., 1965. Influence of zinc on periphytic communities. Am. J. Bot. 52 (1), 26 – 34.
Wong, Y.S., Tam, N.F.Y., Lau, P.S., Xue, X.Z., 1995. The toxicity of marine sediments in Victoria harbour, Hong Kong.
Mar. Pollut. Bull. 31 (4 – 12), 464 – 470.