Arch. Environ. Contam. Toxicol.

31, 84-90 (1996)

ARCHIVES

OF

Environmental Contamination a n d Toxicology

1996 Springer-Verlag New York Inc.

Effects of Organic Solvents and Solvent-Atrazine Interactions on Two Algae, Chlorella vulgaris and Selenastrum capricornutum
A. E1 Jay Institut National de la Recherche Agronomique, Station d'Hydrobiologie Lacustre, BP 511, 74203 Thonon Cedex, France Received: 31 July 1995/Revised: 2 January 1996

Abstract. When assessing pesticides toxicity on test organisms, the use of organic solvents is frequently required to formulate solutions of pesticides having low or moderate water solubility. These solvents may influence the results of toxicity tests. This study examined the effects of ethanol and dimethyl sulfoxide (DMSO) on two species of algae, Chlorella vulgaris and Selenastrum capricornuturn. The interactives effects of these organic solvents with various concentrations of atrazine were tested to analyse how these solvents can affect the toxicity of atrazine. Toxicity was measured as the change in chlorophyll (a) content (estimated via fluorescence) in cultures of the test organism that were incubated over a 96-h period at standard conditions of temperature and light. In the absence of atrazine, ethanol was toxic to both algae, yielding significant inhibition of chlorophyll (a) content at concentrations as low as 0.05%. Morever, S. capricornuturn was less sensitive to ethanol than was C. vulgaris. DMSO did not show any toxic effects on either chlorophycea. At concentrations up to 0.5%, DMSO interacted additively with atrazine for both chlorophycea. An additive response was also observed with ethanol towards S. capricornuturn. For C. vulgaris, ethanol interacted antagonistically at most of atrazine concentrations and gave few additive and synergistic interactions. An additive response of an atrazine-solvent mixture indicate the inherent toxicity of the atrazine. Since DMSO generally gave additive responses for both algae, it can be considered as an adequate organic solvent to use in bioassays.

The presence of herbicides in aquatic systems, due to run-off from agricultural fields, has lead to extensive study of their effects on aquatic non target organisms (Goldsborough and Robinson 1984; Mayashich et al. 1986; E1-Dib et al. 1989; Frangois and Robinson 1990; Kasai et al. 1993). Atrazine, a herbicide commonly used in the United States (De Noyelles et al. 1982) and other developed countries, has been widely studied in ecotoxicological bioassays (Maule and Wright 1984; Fran9ois and Robinson 1990; Abou-Waly et al. 1991). Atrazine, like many pesticides with low or moderate water solubility, is often

dissolved in organic solvents prior to addition into experimental systems. These organic solvents can also make their way into the environment as industrial wastes and components of pesticide formulations. In laboratory bioassays, these organic solvents can impose stress on test organisms (Bowman et al. 1981). Ethanol (Maule and Wright 1984; Yarden el al. 1993), methanol (Franqois and Robinson 1990; Abou-Waly et al. 1991; Shehata et al. 1993), dimethyt sulfoxide (O'Neal and Lembi 1983; Kasai et al. 1993) and acetone (Parasher et al. 1978) are solvents often used in bioassays. However, relatively few reports have been published on the comparative toxicity of solvents on tests organisms. Although the US Environmental Protection Agency has set allowable limits for solvent use in aquatic toxicity tests at 0.05 % for acute tests and 0.01% for chronic tests (USEPA 1975), such guidelines do not exist for most microbial bioassays. Morever, the choice of solvent and the final concentration of the solvent used in bioassays differs among investigators, and are generally greater than those values, due to problems associated with the use of small volumes of test compounds, toxicant solubility and other technical limits. In the literature, several authors do not specify the solvent (E1-Dib el al. 1989) or the concentration of solvent (Shehata et al. 1993) used in their experiments, thus limiting the comparison of data. Some studies have assessed the toxic effects of solvents on various species of cyanophycea (Stratton et al. 1980; Stratton and Corke 1981; Stratton 1987) and pointed out that the effects of the solvent differed both with the species studied and on the concentration of solvent. Pesticide and solvent can interact in complex manner giving synergistic, antagonistic or additive responses, depending on the solvent, the solvent concentration, the pesticide and the specie considered. Such interactions have been reviewed on fungi (Burrel and Corke 1980) and algae (Stratton and Corke 1981; Stratton et al. 1982). Ideally, the solvent to be used in bioassays should not show toxic effects nor demonstrate interactive effects with the tested pesticide. The goal of this work was to study the toxicity of ethanol and DMSO and to analyse how these solvents affect the toxicity of atrazine, on two chlorophycea, C. vulgaris and S. capricornuturn. These species have been used as the test organisms because

Solvent-Atrazine Interactions on Algae

85

they are widely employed in ecotoxicological assays (Mingazzini 1993; Hatakeyama et al. 1994). Ethanol and dimethyl sulfoxide were tested because they are widely used solvents for pesticide dissolution.

Table 1. Data analysis procedure atrazine concentration(~lg/I)

[ 10

30

50

100

//j/"
Materials and Methods

Stocks Cultures
Two species of chlorophycea were used in this study. One species, Chlorella vulgaris, was isolated from Lake Geneva by M. Feuillade in 1980. The second organism, Selenastrum capricornutum (Printz), was obtained from the U.S. Environmental Protection Agency laboratory in Corvallis, Oregon. Axenic stock cultures of both algae were incubated in 500 ml flasks, equipped with a sterile air bubbling system, and containing 250 ml of algal assay media (USEPA 1971) that was modified by increasing the concentration of NaHCO3 and K2HPO4 to 15 mg/L and 12 mg/L, respectively. Both stock cultures were incubated at 21 + / - IC under continuous illumination of 100 IxE'm 2. sec-~.

Net pest cid~ lille(

Chlorophyll (a) Analysis

The chlorophyll (a) concentration of the axenic exponentially growing stock culture of each species was determined spectrophotometrically on a 90% acetone extract according to Strickland and Parsons (t968). These same stock cultures were serially diluted and the fluorescence of each dilution was measured with a Turner Model 111 Fluorometer. The fluorometer was equipped with a blue light source, blue excitation lamp and red emission filters. Aperture sizes and attenuation filters were adapted so that measured fluorescence was between 20 and 60% of full scale. These values of fluorescence were then related to the chlorophyll (a) concentrations calculated for a set of serial dilutions of the acetone extract. The r 2 for each of these calibration curves exceeded 0.99. The fluorescence of each test tube was determined on day 4 after inoculation and the chlorophyll (a) content estimated from the appropriate calibration curve.

Test Chemicals and Stock Solutions

The atrazine (2-chloro-4-ethylamino-6-isopropyl-amino-s-triazine, 98.5% pure) was supplied by Greyhound Chemical Service, Inc., Merseyside, UK. Ethanol (95% pure) and dimethyl sulfoxide (DMSO, 99.9% pure) were obtained from Aldrich-Chimie S.A.R.L., France. Because of the conditions required by the bioassay, stock solutions of atrazine ranging in concentrations from 10 to 400 mg/L were necessary. Since the maximal solubility of atrazine in water in only about 30 mg/ L, these stock solutions were prepared by dissolution either in ethanol or DMSO.

Data Analysis Bioassay Procedure

All bioassays experiments were conducted in glass culture tubes (20"125 mm) that contained 20 ml of algal assay medium. Prior to inoculation, these tubes, filled with medium, were autoclaved at 120C for 30 rain. A known volume of the exponentially growing stock culture (either C. vulgaris or S. capricornutum) was added to provide an initial chlorophyll (a) content of about 5 txg/L. Appropriate amounts of solvent and atrazine were then added to provide the desired concentration of solvent and atrazine in the test tubes. The first experiment (Experiment A), was conducted to determine the toxicity of each solvent for both test organisms. A serie of tests tubes received different volumes of the appropriate solvent (0, 10, 20, 40, 100 and 200 ~xl) so that the final concentrations were 0.05, 0.1, 0.2, 0.5, and 1% (v/v). This set of tubes also served as the controls for the experiment B (see below). There were four replicates for each solvent concentration. Experiment B was designed to test solvent-atrazine interactions. Atrazine concentrations of 0, 5, 10, 30, 50, and 100 jxg/L and 0, 10, 30, 50, 100, and 200 ixg/L were tested towards C. vulgaris and S. capricornutum, respectively. Each concentration of atrazine was tested with 0.05, 0.1, 0.2, 0.5, and 1% of ethanol and DMSO. Ten ~1 of atrazine stock solutions was added to each test tube so that the final atrazine concentrations cited above were obtained. This resulted in a solvent concentration of 0.05% (v/v). Additional volumes of the appropriate solvent were then added to provide the higher solvent concentrations (0.1, 0.2, 0.5, and 1%). The tubes were then placed in a rotative incubator for 96 h at 21C under continuous illumination. The effects of solvent and/or atrazine on algal growth was determined by measuring the percent inhibition of chlorophyll (a) relative to an appropriate standard. The percent inhibition was defined as: [(chlorophyll (a) content in the test tube)/ (chlorophyll (a) content in the standard)]*100. In experiment A, the percent inhibition was calculated for each solvent concentration using an initial control containing only the growing culture. The percent inhibitions for the conditions defined by the shaded area of Table 1 are those for experiment A. For experiment B, percent inhibition (which correspond to net pesticide effect) was calculated for each solvent-atrazine mixture using solvent controls of the same concentration of solvent as in the atrazinesolvent mixture. These solvent controls are same tubes as tested in Experiment A (see Table 1). For example, the tube containing only solvent at 0.05% concentration served as the control for all atrazine concentrations mixed with 0.05% solvent. The percent inhibitions for the conditions defined by the unshaded area of Table 1 are those for experiment B. Stratton et al. (t982) suggested the data be analyzed using both a screening test and a graphical verification technique. In the screening test, for each atrazine concentration, the percent inhibition for the various solvent concentrations were compared statistically to percent inhibition for the 0.05% solvent concentration using the Student's ttest (P = 0.05). If the statistical comparisons showed no significant difference, the interaction was additive. If the statistical comparison showed a significantly greater or less inhibition than that expected for

86

A. E1 Jay

Chorella vuIgaris
ethanol I0 DMSO

Selenastrum capricornutum
ethanol DMSO

51

23456

123456 12345 concentration of solven

111111 liiii1111
123 5

1=control; 2=0.05%; 3 = 0 . l % ; 4=0.2%; 5=0.5%~ 6=1% Fig. 1. Chlorophyll (a) content, expressed as percentage of the control (0% of solvent), in ChlorelIa vulgaris and Selenastrum capricornutum treated with ethanol and DMSO. Bars represent confident level, at 0.05. A star above a bar means that it differs significantly from the control (Student test)

additive interaction, synergisticor antagonistic interaction are observed, respectively (Stratton et al., 1982). A graphical verification technique was provided by plotting percent inhibition versus atrazine concentrations for each solvent-organism mixture. According to Stranon et al. (1982), when all solvent concentrations gave the same responses this is indicative of an additive interaction between atrazine and the solvent. When the percent inhibition is greater or lower than the control for different concentrations of solvents, the curves would denote synergism and antagonism, respectively.

and should give the same inhibition when calculated using the appropriate solvent control. This would be indicative of an additive solvent-pesticide interaction. Percent inhibitions significantly greater or less than that obtained for the lowest solvent concentration are indicative of synergism or antagonism, respectively. The screening test summarized in Table 2 indicated that atrazine and DMSO interacted in an additive manner at all concentrations of DMSO regardless of the atrazine concentration, with the exception of 0.5% and 1% levels of DMSO when interacting with 5 ~xg/L of atrazine. In these cases synergism was observed. Stratton et al. (1982) recommended that at least one adjacent solvent level must give a response statistically similar to that observed for lowest solvent concentration to ensure that the conclusion regarding additive interaction was correct. This is the case for DMSO concentrations up to 0.2% for 5 ~g/L of atrazine and for all concentrations of DMSO for the others atrazine concentrations. The graphical verification technique (Figure 2) confirmed these results except for the interaction between 5 Ixg/L and 1% DMSO. The graphical technique showed an additive response while the screening technique indicated a synergism for this mixture. Data of the screening test for S. capricornutum are presented in Table 3. Atrazine and DMSO interact in an additive manner at all the solvent levels regardless to the atrazine concentration except for the 1% DMSO level with 200 txg/L of atrazine. In this case, the interaction is antagonistic. The graphical technique confirmed these results.

Atrazine-Ethanol Interactions
Results for C. vulgaris are outlined in Table 4 and Figure 3. Ethanol and atrazine interacted in an additive manner at ethanol levels of 0.05 and 0.1% at all atrazine concentrations, with the exception of the two lowest atrazine concentrations (5 and 10 ixg/L) at 0.1% ethanol which exhibited synergistic interactions (Table 4). Synergistic interactions were also observed for ethanol concentration greater than 0.1% when interacted with 10 Ixg/L of atrazine. Antagonistic interactions were observed for all the other cases. The graphical verification technique (Figure 3) gave the same general responses. The curves for 0.05 and 0.1% ethanol, above a 10 lxg/L concentration of atrazine, are similar to one another and demonstrate an additive interaction between the two compounds. Data falling above or below this group of curves denote synergism or antagonism, respectively. Data for S. capricornutum are presented in Table 5. Atrazine and ethanol interact in an additive manner at all concentrations of ethanol regardless of the atrazine concentration with three exceptions: antagonistic interactions were observed at 30 and 200 ixg/L atrazine with 1% ethanol and synergistic interactions were observed at 30 Ixg/L atrazine with 0.2% ethanol. These results were also confirmed by the graphical technique.

Results

Experiment A: Solvent Toxicity

The results of toxicity tests performed with solvents in the absence of atrazine are outlined in Figure 1. Overall, DMSO was less toxic for both species than was ethanol. C. vulgaris was unaffected over the range of DMSO concentrations tested. For C. eapricornutum only the 1% level of DMSO caused a significant inhibition, but this inhibition was low. Ethanol was toxic for both species at all concentrations tested. However, C. vulgaris appeared to be more sensitive to ethanol than was S. capricornutum. Indeed, the lowest concentration of ethanol caused a 63% inhibition of the chlorophyll (a) content for C. vulgaris while this inhibition was only 6% for S. capricornutum.

Experiment B: Solvents-Atrazine Interactions: Atrazine-DMSO Interactions

Results using the screening test and the graphical verification techniques of Stratton et al. (1982) for water-insoluble pesticides are outlined in Table 2 and Figure 2 for C. vulgaris. Theoretically, the percent inhibition values calculated for each of the solvent levels used should be statistically the same, since all test systems contain the same concentration of pesticide

Discussion
It is usually necessary to use organic solvents in pesticide bioassays although the actual solvent chosen and the concentra-

Solvent-AtrazineInteractionson Algae

87

' T
;g

+ 0,05% --*--0,I0% 0,20% ---x--0,50% --1%

C. vulgaris "DMSO

~--/-/'

I 20

L 40

I 60

I 80

100

-20

atrazine concentration (~tg/1)

Fig. 2. Effect of atrazine at various DMSO concentrations towards C. vulgaris. Chlorophyll (a) content is expressed as percent inhibition of the control containing the appropriate level of solvent

Table 2. Interactions between DMSO and atrazine toward chlorophyll (a) content of C. vulgaris
atrazine concentration (pg/I)

Table 3. Interactions between DMSO and atrazine toward chlorophyll (a) content of S. capricornutum
atrazine concentration (pg/I) )

lo 13o I =o 110o12oo
10.05 L ~ ~ ~< l . ~ ! ~ i l ti ~ ;/ ~ :ii ):~~~ii; ;~ :~ ] ~~ ~ ~ 1 ;! I :~:,"~'. ~::~:,~;~ :!i /i i :~. %~; : N I f , ~ , ~. g ~ ~ ~-:" ~ " :./~t /
-~-" ~- + ~ 2 J ,~ "44:::. -->x R

0.05 o~ o 0.1 0.2 0.5 1

I
[

0.1

. ~ ~1;: ~'.- *- ~,I.

~'" + ~ " ":" "- +

~:#~ ~',:'::." ,. "."

I o.s

03

I]iiii

!IiN!i

ii{iNiiiiNNI

ch

Tables entries (net atrazine effect) are mean percent inhibition values calculated from chlorophyll (a) content in appropriate ethanol solvent controls. ~ These values do not differ significantly from that calculated for 0,05% and are indicative of an additive solvent herbicide interaction. ~] These values are significantly lower (P=0.05) than that calculated at 0.05% ethanol and are indicative of an antagonistic solvent-herbicide interaction. ~ These values are significantly greater (P=0 05) than that calculated at 0.05% ethanol and are indicative of a synergistic solvent-herbicide interaction.

Tables entries (net atrazine effect) are mean percent inhibition values calculated from chlorophyll (a) content in appropriate ethanol solvent controls N These values do not differ significantly from that calculated for 0.05% and are indicative of an additive solvent herbicide interaction. [ ~ These values are significantly lower (P=005) than that calculated at 005% ethanol and are indicative of an antagonistic solvent-herbicide interaction. ~ These values are significantly greater (P=0.O5) than that calculated at 0.05% ethanol and are indicative of a synergistic solvent-herbicide interaction.

tion employed are often restricted by problems of pesticide solubility, pipetting accuracy and total test volume (Stratton 1985). It is generally recommended that non-toxic solvent be used for bioassays. In the literature, data on solvent toxicity towards aquatic invertebrates and fish are abundant (Le Blanc and Surprenant 1983; Majewski et al. 1978). However few data are available on the comparative toxicity of solvents towards algae. Moreover, in ecotoxicological work, some authors did not specify the solvent or the concentration of solvent they have used (E1-Dib et al. 1989; Shehata et al. 1993). Atrazine is a commonly used herbicide in agriculture. Even though atrazine has a moderate solubility of about 30 mg/L in

water, its dissolution is quite difficult and time consuming. As our stock solutions of atrazine ranged from 10 to 400 rag/L, we were induced to use solvents. Ethanol and DMSO were chosen because of their widespread use in bioassays and their high solvent efficiencies. The first experiment, designed to determine the effects of ethanol and DMSO in the absence of atrazine, on C. vuIgaris and S. capricornutum, confirmed that the results of the bioassay test differed depending on the solvent, the concentration of solvent and the species considered. C. vulgaris and S. capricor-

88

A. El Jay

Table 4. Interactions between ethanol and atrazine toward chlorophyll (a) content of C. vulgaris
atrazine concentration (IJg/I)

Table 5. Interactions between ethanol and atrazine toward chlorophyll (a) content of S. capricornutum
atrazine concentration (IJg/I)

5 g

10

30

50

100

...... o 4 ;
g 0.05 0.1
o E

O.05
0.1 ~::~O~:.~i~i~i

ii
.*,.~ ! ! ~
:::::::::::::::::::::::::::

ii
........... ~

..............
46.7

I!IIN!~IIIilI~jI~;NN, i~IIIN~NN~Ni I!INi~

iiiiNNiiilNN!iiiliiiiiNNiiiNNNiiiliNNiii

oo
~J ~d

0.2
0.5

-10.2
,..........,

10.2

22.2 --5.0

0.2 0.5 1

3.6 0.0 -ii!i~"ti~ii

-40.3 -40.9

8.7 -35.0

-42.5

..............................

!~!~::~!~'~!~j~i!~!!! .'.~!N~!I~ N."..'

Tables entries (net atrazine effect) are mean percent inhibition values calculated from chlorophyll (a) content in appropriate ethanol solvent controls. These values do not differ significantly from that calculated for 0,05% and are indicative of an additive solvent herbicide interaction. These values are significantly lower (P=0.05) than that calculated at 0.05% ethanol and are indicative of an antagonistic solvent-herbicide interaction. These values are significantly greater (P=0.05) than that calculated at 0.05% ethanol and are indicative of a synergistic solvent-herbicide interaction.

Tables entries (net atrazine effect) are mean percent inhibition values calculated from chlorophyll (a) content in appropriate ethanol solvent controls. ~iiii These values do not differ significantly from that calculated for 0.05% and are indicative of an additive solvent herbicide interaction. ] These values are significantly lower (P=0.05) than that calculated at 0,05% ethanol and are indicative of an antagonistic solvent-herbicide interaction !~ These values are significantly greater (P=0.05) than that calculated at 0.05% ethanol and are indicative of a synergistic solvent-herbicide interaction

80 T
i

+0.05%
+ 0 . 1 0 %

C. vulgaris e t h a n o l

60

~ - 0.20%
- x - 0.50%

40 + ! ~20 -9

= 80
-20

.... 60 80

100

-40

-60

atrazine concentration (gg/1)

Fig. 3. Effect of atrazine at various ethanol concentrations towards C. vulgaris. Chlorophyll (a) content is expressed as percent inhibition of the control containing the appropriate level of solvent

nutum were relatively unaffected by D M S O while ethanol had

inhibitory effects at concentrations as low as 0.05%. Moreover, S. capricornutum was less sensitive to ethanol than C. vulgaris. The differences in alga sensitivity to these solvents could be due to variations in the biochemistry and physiology of the organisms tested. For example, the extraction of pigments by solvents from S. capricornutum is more difficult than for C. vulgaris, thus suggesting that cell walls of S. capricornutum might be more resistant to destruction by solvents (personal work). Ethanol Was shown to be equal in toxicity to D M S O when

tested on C. eugamatos (Hess 1980) and less toxic than D M S O towards C. pyrenoidosa. However, D M S O was shown to be less toxic than ethanol towards an other strain of C. pyrenoidosa (Stratton et al. 1988). Parasher et al. (1978) showed that concentrations of acetone greater than 3% can cause membrane destruction in C. pyrenoidosa cells. Stratton and Corke (1981) reported that A. inaequalis was relatively unaffected by acetone concentrations up to 1% whereas this solvent stimulated-photosynthesis of A. variabilis at concentrations below 1%. Based upon this first set of experiments, D M S O appeared to be a solvent of choice in bioassays involving growth of C.

Solvent-AtrazineInteractionson Algae

89

vulgaris and S. capricornutum. While none of the DMSO concentrations gave toxic responses, it is still preferable to use the lowest possible concentration in bioassays. Ethanol should be avoided because of its high toxicity towards both algae. To ensure that solvent toxicity or no-toxicity can be modified when they are mixed with another toxicant, both solvents were tested in interaction with atrazine. Each solvent was interacted with atrazine to determine the effect of solvent type, and solvent concentration on the atrazinesolvent interactions of both green algae. Solvent concentrations that elicit antagonistic or synergistic interactions are not suitable for bioassays purposes as they mask the pesticide inherent toxicity (Stratton et al. 1981). In this study, the majority of interactions observed were additive or antagonistic with few cases of synergistic interactions. However, the only solvent and solvent concentrations that could be used are those that induce additive responses. It is recommended that the <<bioassayists>> use the lowest concentration of solvent that produced additive interaction regardless of the atrazine concentration being tested. The results of this study allows the use of the 0.05% level of DMSO for C. vulgaris and for S. capricornutum and also the 0.05% level of ethanol in the case of S. capricornutum (Tables 2-5). When tested with C. vulgaris, ethanol gave inconsistent data and thus should be avoided as solvent in bioassays with this species. The choice of solvent may alter the solvent-pesticide interaction pattern noted for a given pesticide bioassay system (Stratton 1985). DMSO mostly elicited additive responses for both algae; ethanol also elicited additive responses in most cases for S. capricornutum, but gave divergent and inconsistent responses with C. vulgaris. The reasons for these observations are probably related to the solvent's mode of action in biological systems, but also to the individual organism's biochemical characteristics. Solvent concentration was also a significant factor in the observed interactions patterns. The actual concentration of solvent at which synergism and antagonism occurred was dependent upon both the solvent type and the atrazine level used. Reasons for the effects of solvent concentrations on interaction responses can only be elucidated by further research on the mode of action behind solvent-pesticide interactions. It is possible that organic solvents induce membrane damage and apparently a threshold level must be reached before the organism is altered in such a way as to become more or less sensitive to a second toxicant (Stratton 1985). Below this threshold, the solvent does not affect the toxicity of the pesticide and an additive response is observed. Synergism could result from a solventinduced increase in cell permeability and subsequent increase in pesticide uptake. Antagonism could result from a solventinduced disruption of membrane structure and transport systems (Stratton 1985). These hypotheses have to be verified by further research. Stratton and Corke (1981) showed, during experiments on acetone-atrazine interaction towards A. cylindrica, that increasing in acetone concentrations can affect membranes integrity and thus alter membrane permeability. The effect of test solvent on inherent atrazine toxicity was dependent upon the type of solvent. DMSO was the least toxic solvent tested but usually induced the greatest atrazine toxicity values when tested on both chlorophycea (Tables 2-5). The screening test and the graphical verification technique generally gave the same results, but it is recommended that

both methods be used to get sufficient indication to assure achievement of the correct conclusion. Studies on solvents effects and solvent-pesticide interactions are essential to ensure that the results obtained from ecotoxicological assays are due to the pesticide and not to the solvent. Such studies must be conducted for each alga species and for each solvent. The least toxic solvent should be chosen and subjected to evaluation using an interaction analysis technique. The dimethyl sulfoxide (DMSO) is a promising solvent to test in further experiments on other algae as it generally gave the least toxic effects. It would be useful to develop a data base containing the solvent effects on different algae so that bioassay data can be easily compared. It would allow a rapid choice of the adequate solvent to use in the experiment considered.

Acknowledgments. This study has been partially supported by the

Agence de l'Eau Rh6ne M6diterran6e Corse of France. I gratefully acknowledge the Environmental Protection Agency for providing the Selenastrum capricornutum species free of charge, and M. Feuillade for the Chlorella vulgaris species. I thank my colleagues for their corrections. I also warmly thank Dr. G. W. Stratton for his advices and his constructive comments. I particularly thank Dr. A. Cassel for his kind assistance in the revision of this paper.

References
Abou-Waly H, Abou-Setta MM, Nigg HN, Mallory LL (1991) Doseresponse relationship of Anabaena flos-aquae and Selenastrum capricornutum to atrazine and hexazinone using chlorophyll (a) content and HC uptake. Aquatic Toxicol 20:195-204 Algal Assay Procedure Bottle Test (AAP) (1971) In: Environmental Protection Agency (EPA), National Eutrophication Research Program, Corvallis, Oregon 24 pp Bowman M, Oller WL, Cairns T (1981) Stressed bioassay systems for rapid screening of pesticide residues. Part I: evaluation of bioassay systems. Arch Environ Contain Toxicol 10:9-24 Bunell RE, Corke CT (1980) Interactions of the solvent acetone with the fungicides benomyl and captan in fungal assays. Bull Environ Contain Toxicol 25:554-561 Butler GL, Deason TR, O'Kelly JC (1975) The effect of atrazine, 2, 4-D, methoxychlor, carbaryl and diazinon on the growth of planktonic algae. Phycol J 10:371-376 DeNoyelles F, Kettle WD, Sinn DE (1982): The responses of plankton communities in experimental ponds to atrazine, the most heavily used pesticide in the united states. Ecology 63(5):1285-1293 E1-DibMA, Shehata SA, Abou-WalyHF (1989) Response of freshwater alga Scenedesmus to triazine herbicides. Water Air and Soil Pollution 48:307-316 Francois DL, Robinson GGC (1990) Indices of atrazine toxicity in Chlamydomonas geitleri Ettl. Aquatic Toxicol 16:205-228 Goldsborough LG, Robinson GGC (1984) A simple bioassay for photosystem-II inhibitors in water using in vivo chlorophyll fluorescence. Weed Res 24:351-358 Hains, JJ (1985) Practical considerations for routine chlorophyll measurements: Precautions and comparison of extraction methods. J Freshwater Ecol 3(2):175-179 Hatakeyama S, Fukushima S, Kasai F, Shiraishi H (1994) Assessment of herbicides effects on algal production in the Kokai River (Japan) using a model stream and Selenastrum bioassay. Ecotoxicology 3:143-156 Hess FD (1980) Chlamydomonas algal bioassay for detecting growth inhibitor herbicides. Weed Sci 28:515-520 Kasai F, Takamura N, Hatakeyama S (1993) Effects of simetryne on growth of various freshwater algal taxa. Environ Pollut 79:77-83

90

A. E1 Jay

LeBlanc GQ, Surprenand DC (1983) The acute and chronic toxicity of acetone, dimethylformamide, and triethylene glycol to daphnia magna (Straus). Arch Environ Contam Toxicol 12:305310 Majewski HS, Klaverkamp JR Scott DP (1978) Acute lethality and sub-lethal effects of acetone, ethanol, and propylene glycol on the cardiovascular and respiratory systems of rainbow trout (Salmo gairdneri). Water Res 12:217-221 Mayasich, JM, Karlander ER Terlizzi Jr DE (1986) Growth responses of Nannochloris oculata Droop and Phaeodactylum tricornutum Bohlin to the herbicide atrazine as influenced by light intensity and temperature. Aquatic Toxicol 8:175-184 Maule A, Wright SJL (1984) Herbicide effects on the population growth of some green algae and cyanobacteria. Journal of Applied Bacteriol 57:369-379 Mingazzini M (1993) Comparison of different methods for quantitative toxicity measurements on natural waters using S. capricornutum. Wat Res 27(6):1055-1062 O'Neal SW, Lembi CA (1983) Effect of simazine on photosynthesis and growth on filamentous algae. Weed Science 31:899-903 Parasher CD, Ozel M, Geike F (1978) Effects of hexachlorobenzene and acetone on algal growth: physiology and ultrastructure. Chem Biol Interactions 20:89-95 Shehata, SA, E1-Dib MA, Abou-Waly HF (1993) Effects of triazines compounds on freshwater algae. Bull Environ Contam Toxicol. 50:369-376 Stratton GW (1985) The Influence of Solvent Type and Solvent-Pesti-

cide Interactions in Bioassays. Arch Environ Contam Toxicol 14:651-658 Stratton GW (1987) Toxic effects of organic solvents on the growth of blue-green algae. Bull Environ Contain Toxicol 38:1012-1019 Stratton GW (1988) Method for determining toxicant interaction effects towards microorganisms. Toxic Assess 3:345-353 Stratton GW, Corke CT (1981) Effect of acetone on the toxicity of atrazine towards photosynthesis in Anabaena. J Environ Health B16(1):21-33 Stratton GW, Burrell RE, Corke CT (1982) Technique for identifying and minimizing solvent-pesticide interactions in bioassays. Arch Environ Contam Toxicol 11:437-445 Stratton GW, Burrell RE, Kurp ML, Corke CT (1980) Interactions between the solvent acetone and the pyrethroid insecticide permethrin on activities of the blue-green alga Anabaena. Bull Environ Contam Toxicol 24:562-569. Strickland JDH, Parsons TR (1968) Pigment analysis In: A practical handbook of seawater analysis, edited by J.D.H. Strickland and T.R. Parsons. Fisheries Research Board Canada, Ottawa, Canada, 1968, 185-192 U.S. Environmental Protection Agency, Committee on Methods for Toxicity Tests with aquatic organisms (1975) Methods for acute toxicity tests with fish, macroinvertebrates and amphibians. US Environ Protection Agency, Ecol Res Ser EPA-660/3-75-009, National Water Quality Laboratory, Duluth, MN Yarden O, Freund M, Rubin B (1993) Dunaliella salina: a convenient test organism for detection of pesticide residues in water and soil. Fresenius Envir Bull 2:31-36