Environmental Pollution 65 (1990) 25 !

-267

Interaction Between Components of Electroplating

Industry Wastes. Influence of the Receiving Water on the
Toxicity of the Effluent

A. Le Du, D. Dive*
INSERM U!46, Domaine du CERTIA, BP 39, 59651 Villeneuve d'Ascq Cedex.
France
&
A. Philippo
Service Eau-Environnement. lnstitut Pasteur de Lille, I rue du Professeur Calmette,
BP 245. 59019 Lille Cedex. France

(Received 16 August 1989; revised version received 15 January 1990;

accepted 19 January 1990)

ABSTRACT

A copper-cadmium-nickel-zinc mixture was assessed in seven different river

waters to study metal toxicity to the ciliate proto:oan Colpidium campylum.
the interactions occurring between metals, and the influence of the receiving
water on toxicity. In the range of concentrations tested, which are
representative of eleetroplating industry wastes, the main part of the toxicity
can be attributed to copper and to cadmium-copper synergy. A classification
of waters, based on a principal component analysis ( PCA ). was used to
examine the main parameters of the water, which can affect the toxicity of
metal mixtures. It appears that the mineralization of the water, more than the
total organic carbon (TOC). is an important parameter for the expression of
toxicity. A strategy for the estimation of ecotoxicological hazard assessment.
based on a simplified factorial experiment is proposed. It enables one to study.
in a two-step bioassay, the toxicity of an effluent, the influence of river water
on its toxicity, and the effects of contact time and dilution. By applying PCA
to data from very different waters, it may be possible to estimate the
ecotoxicological risk associated with the discharge of an effluent, on the basis
of the chemistry of the receiving water.

* To whom correspondence should be addressed.

251
Environ. Pollut. 0269-7491/90/$03"50 <~) 1990 Elsevier Science Publishers Ltd. England.
Printed in Great Britain
252 A. Le Du, D. Dive, A. Philippo

INTRODUCTION

Wastes from the electroplating industry contain a mixture of toxic metals

(mainly Cu z +, Cd 2 +, Ni 2 +, Zn 2 + and Cr 3 +). Chemical analysis can only
produce an inventory of the toxicants present in the effluent. Extrapolation
of chemical data to ecotoxicological hazard, based on the measurement of
the toxicity of each metal, tested alone by bioassay, depends on the
assumption that the toxic effects of metals are additive. This condition is
rarely found with toxicants present in such wastes. For example, synergy has
been observed between cadmium and copper by Sprague and Ramsay
(1965), Baldry et al. (1977), Stebbing and Santiago-Fandino (1983), Dive et
al. (1988, 1989a)and antagonism between, respectively, cadmium and zinc by
Eisler and Gardner (1973), Thorpe and Alke (1974), Nordberg et al. (1979),
Negilski et al. (1981), Attar and Maly (1982), Pyne et al. (1983), Gingrich et al.
{1984), Dive et al. (1983, 1989a) and cadmium and nickel (Dive et al., 1989a).
More complex interactions occur, as with the antagonistic effect of trivalent
chromium on the copper-cadmium synergy (Dive et al., 1989a). When
effluents are discharged, interactions may also occur with components
TABLE I
Chemical Specifications of the Seven Rivers Studied

Parameters Sler Go- Couesnon Jerse)' Orne Lai'.e Moselle Sarre

Resistivity (Q cm" ') 5 260 3 230 1 960 2 270 2 150 1 510

pH 7,3 7.0 7.2 7-6 8.0 7-6 7.9
TAC (°F) 2 5 10 13 19 12 18
Hardness (~F) 7 8 16 18 23 45 27
Ca 2÷ (mg litrc- ') 16 20 42 60 86 156 88
Mg 2+ (mg litrc -I) 6.7 7.5 12.8 7"1 5'3 13.4 12.0
K + (mglitre -I) 2 9.2 9"8 6"1 4-1 5-2 3"8
Na ÷ (mg litre -t) 12 29.9 47.8 30.4 14.7 97.2 31.1
NH4, (mg litre-') <0.1 <0.1 <0.1 <0.1 <0.1 0.44 0-15
CI- (mg litre-I) 25 45 66 53 30 295 83
SO~- (rag litre-i) 3 17 69 21 23 98 127
NO~" {rag litrc-t) <0.05 <0-05 0.97 0.51 0-09 0"29 0.09
NO;" {rag litrc-') 61-4 34.7 21.1 22.3 26.0 13-0 6-6
PO~- {mg litre-') <0.1 I <0.1 0.5 0.5 2.1 1-7
HCO;" {rag litre-~) 22 61 122 159 232 156 219
F¢ (mg litr¢ -~) <0.02 <0-02 <0.02 <0.02 <0.02 <0.02 <0.02
Mn (mg litrc-t) <0.05 <0-05 <0-05 <0-05 <0.05 103 <0.05
Cr (total) (tlg litre-') <5 <5 <5 <5 <5 24* 5*
Cu (fig litre-~) <5 <5 <5 <5 <5 8"9 <5
C o (t~g l i t r e - ~) < ! < ! < 1 < 1 < 1 < 1 < 1
Ni {pg litre" t) < 10 < 10 < 10 < 10 < 10 < 10 < 10
Zn (#g litre- ') 58 42 79 34 27 85 230

* Results in hexavalcnt chromium.

 Influence of receiving water on electroplating waste toxicity 253

of river waters. Some of these interactions are well known, such as the
antagonistic effect of calcium upon metal toxcity (Tabata, 1969; Winner,
1985; Michibata, 1981; Dive et al., 1989a), and the complexing action of
organic matter (Houba & Remacle, 1982) or humic substances (Dive et al.,
1982; L'Hopitault & Pommery, 1982; Huljev & Strohal, 1983; McKnight et
al., 1983). The modulation of metal toxicity by pH was also demonstrated by
Jeffrey and Pfister (1982), O'Keefe and Hardy (1984), Peterson et al. (1985)
and Winner (1985). But these studies cannot account for the multiple
interactions which may occur between toxic metals present in electroplating
industry wastes and surface waters.
The purpose of the present paper was to study the modulation of the
toxicity of a copper-cadmium-nickel-zinc mixture by surface waters. The
experiments were performed with the ciliate protozoan Colpidium
campylum. Protozoa are well adapted for such studies, because they are very
sensitive to metals, and are representative of a part of natural food chains.
We examined the interactions occurring between metals and, by a principal
component analysis, we tried to determine the main chemical factors which
influence the toxicity of a mixture of metals.

MATERIAL AND METHODS

Chemical analysis of waters

The rivers studied were selected according to the ground characteristics and
the industrial and urban environment to which they were exposed. We tried
to collect, in each type of environment, two different rivers, in order to verify
any relationship between water and metal-mixture toxicity. Water samples
were collected in plastic bottles, filtered on a 0.45-#m filter and stored at 4°C.
All chemical determinations were carried out with usual methods. The
main components of mineralization such as CI-, NO~', SO~-, HCO~, etc.,
were quantified by colorimetric methods developed on Segmental Flow
Analysis (SFA). Other elements such as Ca, Mg, Mn, Zn were analysed using
flame atomic spectrometry (AAS) or graphite furnace AAS for Cr, Cu, Cd
and Ni. The determination of total organic carbon (TOC) was obtained after
chemical oxidation and infra-red (IR) detection of carbon dioxide.
Table I shows the results ofchemical analyses of the different river waters.
Three classes can be distinguished. The first one (Ster Goz, Couesnon)
includes very soft waters. The second class (Jersey, Orne and Laize) includes
intermediate waters, with moderate hardness and significant alkalinity. The
third class (Moselle, Sarre) includes mineralized river waters, with significant
hardness, calcium-rich, and obviously submitted to a significant enrichment
in sodium, chlorine, sulphates, phosphates and zinc.
254 A. Le Du, D. Dive, A. Philippo

Metal solutions

Metal solutions were made with ultrapure metal salts (Aldrich, 99.99%
purity as CuSO4, 5H20; CdCIe; NiCI2; ZnSO~, 7H20) in ultrapure water
(13-15MD., noticed MQ water). They were freshly prepared for each
experiment. The concentrations were selected so as to avoid a complete
inhibition of the growth of the ciliates. The levels studied were: 0-160, 0-50,
0-80, 0-800 #g litre- t respectively, for Cu 2 +, Cd 2 +, Ni 2 + and Z n ' +. Metals
were added in each flask as 10 #1 of concentrated solutions.

The bioassay

The bioassay used the ciliate protozoan Coipidium campyl,~m according to

the protocol defined by Dive et al. (1989b), Vasseur et al. (1988). The
bioassay is based upon the measurement of the growth of the protozoan
exposed to a toxic medium compared to that of a control population. The
growth criterion selected for study is the number of generations (NG)
measured after 24 h of incubation:
NG = iog~z~(N/ N o)
where N O is the number ofciliates at inoculation time and N the number of
ciliates at the end of the experiment (after 24 h of incubation).
Monoxenic stock cultures ofciliates were prepared from axenic cultures,
in a mineral medium (NaCI 14.5 mg; CaCI 2 2H20 10"7 mg; NaNO3 4-5 mg;
MgSO4 7H20 75.7 mg; Na,SO.~ 39-5 mg; N a H C O 3 13"5 mg; H 2 0 1000 ml;
pH 7-8) in presence of0-1 mg ml - t of A erobacter aerogenes (Sigma A6634) as
a bacterial food.
Experiments were conducted in 30-ml polystyrene crystal screw-capped
flasks. The working volume was 5ml. A. aerogenes was added at a
concentration of 0.125 mg m l - t. The inoculum was 500cells m i - t and a
correction for the ionic concentrations in the water was performed to take
into account ions introduced with the protozoan inoculate. Populations
were scored after 24 h of incubation at 28°C and the number of generations
calculated. Each metal combination was tested in the reference medium
(MQ water), and in the surface waters tested, at a concentration of 80%,
because of the volume needed for protozoan inoculate (0-75ml) and
bacterial food (0"25 ml). Thus, by considering river water as a variable, its
impact on effluent toxicity could be estimated. Each treatment was
performed in duplicate and each complete experiment (which represented 64
different treatments) was repeated three times with different stock cultures.
Water-metal mixtures were prepared 24 h, or immediately before the test,
and all flasks were inoculated with the same stock culture, so that we could
study the impact o f contact time between metals and water.
 Influence of receiving water on electroplating waste toxicity 255

Interpretation method

Multiple regression analysis, according to Bois et al. (1986) and Vasseur et al.
(1988), was used to interpret the results. The value of "y" (number of
generations measured after 24h) was modelled as a function of the
experimental variables (metals, water and time of contact metal-water) by
polynomial equation calculated by the least-squares method and validated
by variance analysis. The form of the model is:
.P = bo + blxt + b~.ri + . . . + boxixj...
where .~ is the calculated value of NG, xt, xi are the experimental variables,
and bi, b~, b o are the regression coefficients.
As two levels are tested for each variable, the polynomial is linear and
coefficients can be interpreted directly:
bi > 0 indicates that xj stimulates the growth of the protozoan
b~ < 0 indicates that x~ exerts a toxic effect on growth
b o indicates a significant interaction between xl and xj:
b o > 0 indicates an antagonism (when b~ and b~ are of opposite signs or
are both negative). When b~ and bj are positive a non-additive
stimulatory effect of the association of x~ and x~ is indicated.
b~ < 0 indicates a toxic synergy between x~ and x~.
Graphical representations, calculated from the polynomials found with the
different water samples, allow one to see the influence of the metals and the
water by comparison with the effect observed in MQ water.
A principal component analysis was performed to search for any
relationships existing between water chemistry and the toxicity of the metals.

RESULTS

Toxicity of the metals and interactions with river waters

On the basis of the polynomial models obtained with each surface water, we
determined the effects of the different experimental variables upon the
growth of Colpidium campylum. The results are summarized in Table 2 and
show that copper, cadmium and nickel, at the concentrations used, are toxic
in almost all surface waters and in MQ water. In contrast, zinc exerts a
stimulatory effect upon growth of the ciliate. Only one interaction--the toxic
synergy between copper and cadmium--is constantly significant. The other
interactions seem to be more erratic.
256 A. Le Du, D. Dive, A. Philippo

TABLE 2
Influence of the Different Experimental Variables on the Growth Response of Colpidium
campylum, Based on b~ Signs of the Polynomial Model

Factors Ster Goz Couesnon Jersey Orne Lai=e Moselle Sarre

Oh 24h Oh 24h Oh 24h Oh 24h Oh 24h Oh 24h Oh 24h

C u z* Tox Tox Tox Tox Tox "Fox Tox Tox Tox Tox Tox Tox Tox
Tox
C d =" Tox Tox Sti Sti Tox Tox Tox Tox Tox Tox Tox Tox Tox
Tox
Ni z * Tox Tox Tox Tox Tox Tox Tox Tox Tox
Tox
Z n =* Sti Sti Sti Sti Sti Sti Sti Sti Sti Sti Sti Sti Sti
Sti
Cu-Cd Syn Syn Syn S y n Syn Syn Syn Syn Syn Syn Syn Syn
Syn
Cu-Ni Ant Ant
Ant
Cu-Zn Syn Ant Syn
Cd-Ni Ant Ant Ant
Cd-Zn Syn Ant Ant Ant
Ni-Zn Ant Ant Ant
Cu-Cd-Zn Ant Ant Ant
Cu-Ni-Zn
Water Sti Sti Sti Sti Sti Tox Tox Tox Tox Tox Sti Tox Tox
Water-Cu Ant Ant Ant Ant Ant Syn Ant Syn Syn Syn
Water-Cd Ant Ant Ant Ant Ant Ant Ant Ant Ant
Water-Ni Ant Syn
Water-Zn Syn Syn Syn Syn
Water-Cu-Cd Ant Ant Ant
Water-Cu-Zn Syn

Tox, toxicity; Sti, stimulating effect on the growth; Syn, synergistic effect; Ant, antagonism.
Absence of indication means that the variable has no signilicant effect on the growth response
(no effect of one variable, no interaction between two or more variables). These variables were
not retained in the polynomial model, on the basis of variance analysis on b~ contributions.

The river waters alone stimulate (Ster Goz, Couesnon, Jersey) or inhibit
(Orne, Laize, Moselle, Sarre) the growth of C. campylum. Two important
interactions can be noted between water and metals, respectively, with
copper and cadmium. With copper, an antagonism is observed when the
water stimulates growth, and a toxic synergy when the water is inhibitory to
the protozoan. With cadmium, when an interaction is observed, it is an
antagonism whatever the influence of the water on growth.
Some effects or interactions are transient, showing an evolution of the
toxicity of the metal mixture during contact with the waters.
The graphical representations of the results obtained with copper,
cadmium, cadmium--copper mixtures and the total metal mixture are shown
in Figs 1--4.
The antagonism between water and copper is particularly clear for the
Ster Goz and Couesnon rivers, and toxic synergy for the Moselle and Sarre
 Influence of receiving water on electroplating waste toxicity 257

Jl
i,
I

Fig. I. Action ofcopper on the population growth of C. campyh,m in different river waters
tested by comparison with MQ water. Values calculated from the polynomial models. Square
hatch indicatcs toxic effect of metal in MQ water after 0 h (MQ O) or 24 h (MQ 24) of contact
between metals and MQ water before starting the experiment. Diagonal hatch indicates toxic
effect of the metal in river water lifter 0 h (river name O) or 24 h (river name 24) of contact
between metals and river water before starting the experiment. Shading indicates stimulation
of growth.

5,

7
H ZZ m am
_ 2
~2
IIII
II
II
II

II
II
0 - -
o o ~ ~ o o-~~-oo~.~oo-~~o o~

.J ,,,I

Fig. 2. Action of cadmium on the population growth of C'. campy/.m in different river
waters tested by comparison with M Q water. Values calculated from the polynomial models.
For explanation of symbols see Fig. I.
258 A. Le Du, D. Dive, A. Philippo

:t
0 . . . . .
o o ~ o o ~ o o ~ o o ~ o o ~ o o g ~ o o ~

Fig. 3. Action of a copper-cadmium mixture on the population growth of C. campylum in

different river waters tested by comparison with MQ water. Values calculated from the
polynomial models. For explanation of symbols see Fig. !.

i' 3

i, !

Fig. 4, Action of the four metals mixture on the population growth of C. campylum in
different river waters tested by comparison with MQ water. Values calculated from the
polynomial modcls. For explanation of symbols see Fig. I.
 Influence of receiving water on electroplating waste toxicity 259

rivers. With intermediate rivers, all three possible types of response can be
observed (non-interaction, synergy or antagonism). Note that the
interactions decrease with the duration ofcontact between metals and water
(Fig. 1).
Cadmium, at the concentration used in the experiments, is slightly toxic
(Fig. 2) even in MQ water. All surface waters tested show an antagonism
with cadmium, and in some of them (Jersey, Orne) a higher growth rate was
observed in the presence of cadmium.
Despite its low toxicity when present alone, cadmium exerts a strong
synergistic action with copper (Fig. 3). The decrease of the toxicity of the
copper-cadmium mixture in surface waters, compared to MQ water, can be
related to different causes which can be seen from the results in Table 2. In
the case of the Ster Goz and Couesnon rivers, the decrease of toxicity can be
related to the copper-water antagonism. In the case of Jersey river, both
copper-water and cadmium-water antagonisms are involved in the decrease
of the metal-mixture toxicity. With the Orne, Laize and Moselle rivers, the
water acts on the toxic synergy occurring between copper and cadmium, in
the case of the Sarre river, despite the synergistic effect observed between
copper and the water, a decrease in growth is observed in the presence of the
copper-cadmium mixture, in the absence of a triplicate interaction.
With the mixture of four metals (Fig. 4), an antagonism is observed with
TABLE 3
Results of the Principal Component Analysis on the Chemical Data of
the Seven Rivers

Variables Axis I Axis 2 Axis 3 Axis 4

Hardness 0"849 3 0-039 7 0.012 3 0-088 5

Ca 0-7865 0.0603 0-0103 0.1353
Mg 0.6399 0.0936 0-0206 0.2195
K 0.0034 0"8069 0-0125 0-0083
Na 0-658 1 0-0346 0.3063 0"0000
CI 0"701 7 0.013 ! 0-282 8 0-002 1
SO,, 0.8044 0"000 ! 0.064 7 0"121 0
NO 2 0"0294 0.7488 0.011 1 0.001 8
NO3 0"741 9 0.0648 0"1566 0"0187
PO4 0.684 5 0.134 5 0.002 5 0.004 2
HCO3 0"344 0 0-001 5 0-399 2 0-247 3
Zn 0.3056 0.063 1 0.255 1 0-371 2
TOC 0.028 2 0-838 8 0-083 3 0.025 9
Contribution of axes
to total variation (%) 50.6 22.3 12.4 9.6

Correlations squared between variables and axes. Only correlation

squared > 0.5 were considered significant and are underlined.
260 A. Le Du, D. Dive, A. Philippo

all the river waters except for Moselle (24h) and Sarre, for which a
synergistic effect was found. We can notice that the effluent is generally less
toxic than the Cu--Cd mixture in river and MQ waters tested. So it is
probable that Zn and Ni are low antagonists to the Cu--Cd toxicity, whereas
the triple interaction term does not appear systematically in the models.

Relationship between type of water and toxicity of metals, studied by principal

components analysis (PCA)

Consideration of the significant correlations between chemical parameters

and axes, and the contribution of each axis to total variation (Table 3), shows
that only two axes must be retained for the description of fiver waters. The
first one corresponds to the general mineralization of the waters, the second
one to the total organic carbon, potassium and nitrite richness. Figure 5
shows the position of the different rivers with respect to these two axes.
In order to study the influence of the two axes defined by means of PCA,
we selected three levels for each axis, represented by different rivers. Ster
Go?., Jersey and Moselle rivers were selected as representative of the axis
No. 1, and Ster Goz, Couesnon and Jersey for axis No. 2. Each axis, which

• Jersey

Orne

Couesnon Axis I

Laize Moselle

$arre

Ster Goz

Fig. 5. Distribution o f the rivers studied as a function o f axes I and 2 o f the PCA.
 Influence of receiving water on electroplating waste toxicity 261

E~

g ./.j
m /'/ .,~ •

i
// /,
m // .,,"
///" ..'.o_
• / .//
0 ~e
Z / /..
/"

i p,./ s.
f ~E
°

o
~,-,_. s.
!
!
P
/
i ~.o
!
i
!
!
o

Z !// /. /./.
/
/

N
o
i
g,
Z
0
.~_
262 A. L e Du, D. Dive, A. Philippo

t~
e..

[-
o~
g~
m

,'7/
r~.~

,,1.I
"XI
~A 4. A ~ "
O
,$ ~, -
Z
//.i I
U
•-. i.~
Y I f
•r. O
o
N
Ji~/"
II I . ] l

J o

Z
i .r"
I
!
I
I
I
¢
I 0
)1"
0
A T ,h ,4 :
Z

'///
/I/ t,.;'r"

I"
Z

0
[-

rz

t~
 Influence of receiving water on electroplating waste toxicity 263

represents river water characteristics, was considered as one variable in a

new factorial experiment, by using the data corresponding to experiments
performed at time 0 h. Multiple regression analysis allowed interactions
occurring between axes and metals to be determined. Results are
represented graphically in Figs 6 and 7. It may be seen that axis No. 1, i.e.
general mineralization has a significant influence on the toxicity of copper, of
the effluent, and of the cadmium--copper mixture. Toxicity of copper and of
the effluent increases when mineralization of the water increases. The
phenomenon of copper--cadmium synergy is important for the two less
mineralized waters, but in the Moselle river, the mixture of the two metals is
no more toxic than that ofcopper alone. Toxicity ofcadmium is very slightly
influenced by axis No. 1. It can be noted that the toxicity of the total metals
mixture is of the same order of magnitude or lower than that of copper
alone. Axis No. 2 does not exhibit any interaction with copper, cadmium or
the mixture of metals (Fig. 7). This indicates that, in the range of
concentrations of TOC present in the rivers, and in the range of amounts of
metals tested, this chemical parameter does not have a significant influence
on toxicity.

DISCUSSION

The present work enables us to discuss the origin of the toxicity of

electroplating wastes, and the consequences of the discharge of these
emuents into surface waters.
Our results show that, in the range of concentrations tested, all metals are
toxic to Colpidium with the exception of zinc, but that copper can be
considered to be the main source of toxicity in the effluent. When it is
present, even at a non-toxic concentration to ciliates, cadmium may exert a
detrimental influence, because of its toxic synergistic effect with copper. This
synergy appears to be the second most important factor in the toxicity of
electroplating industry wastes, as was previously pointed out by Sprague
and Ramsay (1965), Baldry et al. (1977), Stebbing and Santiago-Fandino
(1983), Dive et al. (1988). In contrast, the polynomial models obtained during
this work show that some other factors, emphasized in previous studies
carried out in standardized conditions of medium, are not significant when
experiments are performed in river waters and MQ water. For example, the
strong antagonism existing between cadmium and zinc, or the toxicity of
nickel and its antagonism with cadmium, are lost (Dive, 1988). it can be
noticed that the toxicity of the effluent is lower than that of the
copper-cadmium mixture. Antagonisms occurring between cadmium and
nickel or zinc would account for this decrease in toxicity. As cadmium alone
264 A. Le Du, D. Dive, A. Philippo

is weakly toxic at the concentration used in our experiment, only the

interactive effect is significant. In previous work (Dive et al., 1988), we have
pointed out the strong antagonism occurring between trivalent chromium
and copper or copper-cadmium mixtures. It is well kown that the
residence time of trivalent chromium is short in water (Klein, 1975), by
comparison with those of copper and cadmium. Toxicity of electroplating
wastes may vary as a function of the disappearance of trivalent chromium
from the water. Hence, addition of this metal in our model would be useful to
estimate the toxicity of the effluent at the discharge point.
The observation of transient effects of some metals during our study
shows clearly that the toxicity of a mixture depends on the duration of its
contact with water. This is the result of a modification ofchemical equilibria.
Only chemical studies of the speciation of metals would account for the
evolution of toxicity with time.
The use of factorial experiments enables us to assess the role of the
receiving waters and to propose a new strategy to estimate the ecotoxico-
logical hazard of effluents. During routine tests, the lack of any reference
medium, which is non-interacting with the effluents results in a loss of
information about the impact of receiving waters upon the effluent's toxicity.
This was shown by using fractioncd factorial experiments in which MQ water
was not included. In such cases, only copper toxicity was dear, but other
results were more erratic, and the role of the receiving water cannot be
defined (results not shown). So a realistic study of the ecotoxicologicai effect
of effluents must include a reference medium consistent with the biology of
the organism, and showing a minimal interaction with the toxicants tested.
This experimental approach, which was complex in the case of the present
study, can be simplified for extensive controls. The following protocol is
proposed:
(1) A preliminary test is performed both in a low mineralized standard
medium (to minimize interactions between medium and effluent), and in the
river water sampled upstream of the discharge. This bioassay enables the
maximum concentration ofeffluent that can be used for the definitive study,
to be determined.
(2)In place of a precise bioassay carried out to determine an LCso of the
effluent, a factorial experiment is planned and the effluent is taken as the
toxic experimental variable. The factors tested in such experiments would
include:
(a)The receiving water river tested at two levels (0 and the maximum
possible concentration, according to the specificity of the bioassay
protocol and to the possible toxic effect of the water).
(b) The effluent tested at three levels (0 and two toxic dilutions).
 Influence of receivingwater on electroplating waste toxicity 265

(c) The time of contact between effluent and fiver water, tested at two
levels.

Such an experiment would give information on the toxicity of the effluent,

the possible interaction with the river water, the effect of dilution and of the
duration of contact between effluent and the receiving water.
The determination of interactions occurring between water and toxicants
enables us to introduce the notion of'Pouvoir Tampon Ecotoxicologique de
l'Eau" or 'PTE', which corresponds to the capacity of the water to detoxify
substances discharged, by various mechanisms (complexation, precipitation,
adsorption etc.). Our study shows that the "PTE' can be exceeded by the
effluent, as was found for the Moselle and Sarre. In any case, the detection of
a synergistic toxic effect between an effluent and a river water is an alarm
signal and an argument against discharge. The use of simplified factorial
experiments for toxicity measurement ofeffluents is largely warranted by the
information they provide on the river water-toxicant interactions, which are
of primary importance in hazard assessment.
The use of PCA and the introduction of axes as variables in factorial
experiments allowed us to study the relationship existing between the
chemical characteristics of water and ecotoxicological hazard. Surprisingly,
it appears that the rivers with a high mineralization are more sensitive to
copper and to the emuent than are low mineralized waters. In contrast, the
TOC content of the water does not influence substantially the toxicity of the
metal mixture. It is possible that a high mineralization of water results in the
occupation of fixation sites present in organic molecules and prevents the
complexation of metals which thus are bioavailable for organisms.
The set of rivers used during this study is limited, particularly for the range
of variation of TOC. It may be interesting to extend this study to a large
number of waters to improve the accuracy of the PCA. With the protocol
mentioned above, it would be possible to estimate hazard for a large variety
of waters, and to quantify the influence of the axes defined by PCA on the
expression of toxicity. Therefore, the results would allow one to predict the
potential toxicity of an effluent on a new river water, on the basis of its
position with respect to the PCA axes. A routine bioassay could then be used
to verify this prediction. If the results of this assay deviate significantly from
the prediction, this would signify that an important parameter, influencing
the toxicity of effluent, was not taken into account in the chemical analyses.
However, it is clear that, before being accepted, this protocol must be tested
on different effluents and waters, with standard well-known bioassays, on
crustaceans, fish or luminescent bacteria, and compared with current
protocols, in order to verify the validity of information obtained by this
approach.
266 A. Le Du, D. Dive, A. Philippo

ACKNOWLEDGEMENTS

This work was supported by a grant from the Environment Ministry, No.
87140.

REFERENCES

Attar, E. N. & Maly, E. J. (1982). Acute toxicity of cadmium, zinc and cadmium-
zinc mixtures to Daphnia magna. Arch. Environ. Contain. Toxicol., I I, 291-6.
Baldry, M. G. C., Hogarth. D. S. & Dean, A. C. R. (1977). Chromium and copper
sensitivity and tolerance in Kliebsiella (Aerobacter aerogenes). Microbios Lett.,
4, 7-16.
Bois, F., Vaillant, M. & Vasseur, P. (1986). Multiple regression analysis of toxic
interactions: Applications to the Microtox test and general comments. Bull.
Environ. Contain. Toxicol., 36, 707-14.
Dive, D., Pommery, N., Lalande, M. & Erb, F. (1982). Cadmium complexation by
humic substance: Chemical and ecotoxicological study with ciliate protozoan
Coipidium campylum. Can. Tech. Rep. Fish. Aquat. Sci., 1163, 9-21.
Dive, D., Vasseur, P., Bel, C. & Danoux, C. (1983). Cadmium-zinc and
cadmium-selenium interactions. A comparative study with Coipidium
campylum and Phowbacterium phosphoreum. J. Protozool., 30, 65A-66A.
Dive, D., Vasseur, P., Hanssens, O. & Gravil, P. J. (1988). Studies on interactions
between components ofelectroplating wastes. Can. Teeh. R,,p. Fish. ,4quat. Sci.,
1607, 23-34.
Dive, D., Hanssens, O., Gabriel, L. & Benga Bengomme, A. (1989a). Studies on
interactions between components of electroplating industry wastes. Influence
of nickel and of calcium concentrations in the water on interactions occurring
between cadmium, copper, chromium and zinc. The Science of the Total
Environment, 87/88, 355-64.
Dive, D., Robert, S., Bel, C., Bonnemain, H., Fourmaux, M. N., Angrand, E. El
Bouhouti, R., Guery, L., Brun, L., Murat, M., Le Du, A., Demarque, Y. &
Hanssens, O. (1989b). Measurement of the growth inhibition of a ciliate
protozoan: Colpidium campyhtm Stokes. Hydrobiologia, ! gg] i 89, 181-8.
Eisler, R. & Gardner, G. R. (1973). Acute toxicity to an estuarine teleost of mixtures
of cadmium, copper and zinc salts. J. Fish Biol., 5, 131--42.
Gingrich, D. J., Petering, D. H. & Shaw III, C. F. (1984). Zinc and cadmium
metabolism in Euglena gracilis: metal distribution in normal and zinc deficient
cells. Mar. Environ. Res., 14, 89-102.
Houba, C. & Remaclc, J. (1982). Factors influencing toxicity of cadmium to
Tetrahymena pyriformis: particulate on soluble form and degree of complex-
ation. Environ. Pollut. A, 28, 35-43.
Huljev, D. & Strohal, P. (1983). Physico-chemical processes of humic acid-trace
elements interactions. Mar. Biol., 73, 243-6.
Jeffrey, A. T. & Pfister, R. M. (1982). Effects of pH, temperature and Eh on the
uptake of cadmium by bacteria and an "artificial sediment. Bull. Environ.
Contam. Toxicol., 28, 697-704,
 Influence of receiving water on electroplating waste toxicity 267

Klein, D. H. (1975). Fluxes, residence times and sources of some elements to Lake
Michigan. Air, Water, Soil Pollut., 4, 3-8.
L'Hopitault, J. C. & Pommery, J. (1982). Le pouvoir complexant de diff~rentes
fractions de mati~re humique vis-a-vis des pesticides et des m~taux. Sci. Eau, !,
85-92.
McKnight, D. M., Feder, G. L., Thurman, E. M., Wershaw, R. L. & Westall, J. C.
(1983). Complexation of copper by aquatic humic substances from different
environments. Sci. Total Environ., 28, 65-76.
Michibata, H. 098 !). Effect of water hardness on the toxicity ofcadmium on the egg
of the teleost Ory=ias latipes. Bull. Environ. Contam. Toxicol., 27, 187-92.
Negilski, D. S., Ashanullah, M. & Mobley, M. C. (1981}. Toxicity of zinc, cadmium
and copper to the shrimp Callianassa autraliensis. II: Effect of paired and triad
combinations of metals. Mar. Biol., 64, 305-9.
Nordberg, A. T., Spratt, M. P. & Dorn, H. C. (1979). Trace elements interactions. J.
Lab. Ciin. Med., 98, 463-81.
O'Keeffe, D. H. & Hardy, J. K. (1984). Cadmium uptake by the water hyacinth:
Effects of solution factors. Environ. Pollut. A, 34, i 33-47.
Peterson, R., Metcalf, J. L. & Le Ray, S. (1985). Uptake of cadmium by eggs and
alevins of Atlantic salmon (Salmo salar) as influenced by acidic conditions. Bull.
Environ. Contain. Toxicol., 34, 359-68.
Pyne, C. K., lftode, F. & Curgy, J. J. (1983). The effects of cadmium on the growth
pattern and ultrastructure of the ciliate Tetrahymena pyrifi~rmis, and the
antagonistic effect of calcium. Biol. Cell., 48, 121-32.
Sprague, J. B. & Ramsay, B. A. (1965). Lethal levels of mixed copper-zinc colutions
for juvenile salmons. J. Fish. Res. Bd Canada, 22, 425-32.
Stebbing, A. R. D. & Santiago-Fandino, V. J. R. (1983). The combined and separate
effects of copper and cadmium on the growth of Campanularia flexuosa
(Hydrozoa) colonies. Aquat. Toxicol., 3, 183-93
Tabata, K. (1969). Studies on the toxicity of heavy metals to aquatic animals and the
factors to decrease the toxicity. Bull. Tokai Reg. Fish. Res. Lab., 58, 203-61.
Thorpe, V. J. & AIke, P. S. (1974). Toxicity bioassays of cadmium on selected
invertebrates, and the interaction of cadmium and zinc on the freshwater
shrimp Paratya tasmaniensis Rick. Austr. J. Freshwater Res., 25, 97-104.
Vasseur, P., Dive, D., Sokar, Z. & Bonnemain, H. (1988). Interactions between
copper and some carbamates used in phytosanitary treatments. Chemosphere,
17(4), 767-82.
Winner, R. W. (1985). Bioaccumulation and toxicity of copper as affected by
interactions between humic acid and water hardness. Water Res,, 19, 449-55.