Environment International 26 (2001) 131 ± 135 www.elsevier.com/locate/envint

Effects of atrazine and nicosulfuron on freshwater microalgae Christophe Leboulanger*, FreÂdeÂric Rimet, Mathilde HeÁme de Lacotte, Annette BeÂrard Station d'Hydrobiologie Lacustre, INRA, BP 511, 74203 Thonon les Bains cedex, France

Abstract Growth modifications caused by various concentrations of atrazine and nicosulfuron were monitored in closed and continuous culture of Chlorella vulgaris (chlorophyta), Navicula accommoda (diatomophyta), and Oscillatoria limnetica (cyanophyta). The concentration at which algal growth rate was reduced twofold (EC50) was determined in the three species for both herbicides. Comparatively, the two toxicants were applied at 10 mg/l level in microcosms inoculated with natural phytoplankton from Lake Geneva. The relative abundances of major phytoplanktonic species were measured by algal cell count at the beginning and at the end of each experiment. Atrazine and nicosulfuron have different targets in plant metabolism, respectively, photosystem II (PSII) and acetolactate synthase (ALS), and the expected effects were different. Generally, the cultured phytoplankton exhibited various sensitivities, depending on species or herbicide. In the microcosms, the major taxa of natural phytoplanktonic samples exhibited various patterns, from acute toxicity to growth enhancement. For example, the diatoms inside the community were not affected by atrazine and nicosulfuron, except for Stephanodiscus minutulus that was sensitive to both, and Asterionella formosa that was sensitive only to nicosulfuron. The specific physiology and the relationships among the phytoplanktonic communities have to be carefully considered when one would try to predict the extent of herbicide action on natural phytoplankton using in vitro tests. There is a need to test the toxic effect on various cultured strains, representative of most of the taxonomic composition of natural communities, to take into account the wide range of sensitivities and reaction to herbicide contamination. But this is not enough to give a solid frame when transposing the results to the field, and the use of more ecologically relevant systems is recommended. D 2001 Elsevier Science Ltd. All rights reserved. Keywords: Phytoplankton; Ecotoxicology; Herbicides; Freshwater; Microcosm

1. Introduction Herbicides are widely used in modern agricultural practices, and as most of their residues ends in surface waters, their behavior and effects in rivers, ponds, and lakes are a global concern. Among the living organisms in aquatic ecosystems, phytoplankton communities are key targets for herbicide contaminations because of their ecophysiological similarities with terrestrial plants (i.e., a potent sensitivity of the same metabolic processes) and their function of primary producers (i.e., a change in quality and quantity of primary producers will lead to a global ecosystemic perturbation). On the other hand, freshwater microalgae are often Abbreviations: ALS, acetolactate synthase (EC 4.1.3.18); Atrazine, 2chloro-4-ethylamino-6-isopropylamino-s-triazine; EC50, effective concentration of toxicant reducing by 50% a given parameter; Nicosulfuron, 2[[(4,6-dimethoxypyrimidin-2-yl)amino-carbonyl]aminosulfonyl]-N,N-dimethyl-3-pyridinecarboxamide; NOEC, no observable effect concentration; PSII, photosystem II. * Corresponding author. Fax: +33-450-260-760. E-mail address: [email protected] (C. Leboulanger).

used as test organisms in laboratory ecotoxicological investigations (USEPA, 1989). But there is still a gap between the in vitro results (EC50 and NOEC [no observable effect concentration] are the most common values used in reglementary notices) and their ecological signification, which are questions still to be solved. In this paper, we present results on the effects of two herbicides, atrazine and nicosulfuron on selected species and communities of freshwater phytoplankton, the latter molecule being the most probable substitute for the former, as restrictions on use arise from legislation. If both are used in corn culture, only atrazine is considered to date as a contaminant in freshwater ecosystems for years (e.g., de Noyelles et al., 1982). They also differ in their application methods, chemical characteristics and half-life in water (PMRA-ARLA, 1996), and overall, in their metabolic target in the cells: atrazine inhibits photosystem II (PSII) by irreversible binding to D1 protein, whereas nicosulfuron prevent synthesis of isoleucine, leucine, and valine by inhibition of acetolactate synthase (ALS) (Simpson et al., 1995).

0160-4120/01/$ ± see front matter D 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 6 0 - 4 1 2 0 ( 0 0 ) 0 0 1 0 0 - 8

132

C. Leboulanger et al. / Environment International 26 (2001) 131±135

under 16:8 light:dark cycle (60 mmol photons/m2/s) at 20 ‹ 1°C.

Table 1 Culture media recipes Medium Composition DV L-C Z

Na2CO3 20 mg/l, Ca(NO3)2 40 mg/l, K2HPO4 10 mg/l, MgSO4
7H2O 25 mg/l, Na2SiO3 25 mg/l, trace elements, Fe ± EDTA KNO3 200 mg/l, K2HPO4 40 mg/l, MgSO4 30 mg/l, Ca(NO3)2 30 mg/l, soil and moss extracts, trace elements, Fe ± EDTA NaNO3 467 mg/l, Ca(NO3)2
4H2O 59 mg/l, K2HPO4 31 mg/l, MgSO4
7H2O 25 mg/l, Na2CO3 21 mg/l, trace elements, Fe ± EDTA

All the media were sterilized by autoclaving.

Changes in growth rate and photosynthetic capacities caused by both herbicides were measured in cultured microalgae. Modifications in the taxonomic composition of natural phytoplanktonic communities from Lake Geneva at the beginning and the end of contamination periods, either by atrazine or nicosulfuron, were recorded. We used cultured algae to ensure a comparison with abundant literature for atrazine EC50 and NOEC and provide first insight on nicosulfuron toxicity on phytoplankton. On the other hand, the use of microcosms to contain natural phytoplanktonic communities was chosen to assess the effects of relatively low concentrations of herbicide, generally considered as close to the NOEC for cultured algae. 2. Material and methods 2.1. Phytoplankton cultures Four strains of freshwater microalgae were used: the cyanobacteria Oscillatoria limnetica CCAP 1459/18, the chlorophyceae Pseudokirchneriella subcapitata CCAP 278/4 (formerly Selenastrum capricornutum), and Chlorella vulgaris SHL 108b, and the diatom Navicula accomoda SHL 107. The first two species were purchased from the Culture Collection of Algae and Protozoa (Ambleside, UK) and SHL strains were isolated from Lake Geneva. Phytoplankton strains were cultivated on Z (O. limnetica), L-C (P. subcapitata and C. vulgaris), and DV (N. accomoda) media (recipes in Table 1) in 250-ml borosilicate glass Erlenmeyer,

2.2. Growth measurements Growth was monitored daily by OD650 measurement, after a standardization to chlorophyll-a (chl-a) content and cell density for each species. Growth rate were calculated by best-fitting to the logistic growth rate equation (Hutchinson, 1978). NOECs and EC50s were determinated according to the concentrations that are ineffective and that reduces the growth rate by 1/2, respectively. 2.3. Photosynthetic capacity measurements For the C. vulgaris and N. accomoda strains, the effects of atrazine and nicosulfuron were assessed using 14C incorporation technique according to Lewis and Smith (1983), and a, b, and Pmax were calculated after counting using the spreadsheets of Walsby (1997). 2.4. Outdoor microcosms A set of fifteen 5-l bottles were filled with 2.5-l filtered (0.8 mm) lake water (as nutrient stock) and 2.5-l natural phytoplankton sample on June 1997, and submitted to outdoor natural variations in solar radiant energy (between 475 and 2750 J/cm2/day above the water level) and in temperature (min: 16.5°C, max 20°C) at 1 m depth in the shore lake. A precise description of the apparatus and protocol is given in BeÂrard et al. (1999). The bottles were sampled at the beginning of the experiment for cell determination and counting, then herbicides were added. A set of five bottles received 10 mg/l atrazine, another received 10 mg/l nicosulfuron, and the remaining bottles were kept as controls. Total phytoplankton growth was monitored daily by chl-a fluorescence measurement and the final composition (taxonomy and biomass) was determined at the end of each experiment. 2.5. Species diversity determination Sampling was performed at the beginning and the end of the experiment: 50 ml of algal suspension were removed

Table 2 EC50 and NOEC for atrazine and nicosulfuron on four phytoplanktonic species Atrazine

Nicosulfuron

Species

EC50 (mg/l)

NOEC (mg/l)

EC50 (mg/l)

NOEC (mg/l)

O. limnetica N. accomoda P. subcapitata C. vulgaris

52 104 98 * 42

< 10 < 10 < 10 * < 10

2.4 neo neo neo

0.4 50 neo neo

Values were calculated using the growth rate, m, estimated from the logistic equation of Hutchinson (1978) except for those marked with asterisks, which were calculated on a gross biomass basis. neo: no effect observed.

C. Leboulanger et al. / Environment International 26 (2001) 131±135

133

al., 1958). A correspondence analysis (Thioulouse et al., 1997) between contaminated and control microcosms was made at the end of each experiment. Density of each dominant species in the contaminated species were compared to control, using the Mann ± Whitney nonparametric test (Schwartz, 1963). 3. Results and discussion 3.1. In vitro NOECs and EC50s

Fig. 1. Photosynthesis irradiance curves for C. vulgaris acclimated at low light intensity (30 mmol photons/m2/s). Control (closed circles, bold line) and contaminated by 10 mg/l atrazine (open squares, dashed line) differs significantly on a and b parameters, whereas Pmax is not affected. Photosynthetic activity is expressed as percent of the maximal activity obtained in controls.

from each microcosm and Lugol iodine was added to kill and stain the cells. Cells were allowed to settle during 24 h, after which they were observed under a reverse phase microscope, determinated, and numerated. Dominant species abundance was measured at ‹ 10% precision (LuÈnd et

Fig. 2. Phytoplankton growth in the microcosms monitored using chl-a fluorescence: control (closed circles), atrazine-contaminated (open squares), and nicosulfuron-contaminated (closed triangles) differs significantly at the end of the experiment.

The NOECs and EC50s calculated on each of the four species tested are summarized on Table 2. These values are given for the growth parameter, m, so they can roughly differ from the NOECs and EC50s calculated from the total biomass at a given time. It appeared that EC50s for atrazine were comprised between 40 and 100 mg/l, whereas overall, nicosulfuron had no effect on growth, except for the cyanobacteria O. limnetica. Nevertheless, the effective concentrations for nicosulfuron were around 50 times higher than for atrazine. 3.2. Effects on photosynthesis ± irradiance curves Only atrazine was shown to have an effect on these shortterm experiments (30 min long), and only in one of the

Fig. 3. Taxonomic composition and abundance of major taxa among phytoplanktonic populations inside the microcosms at the beginning (A) and the end of the June 1997 experiment ((B) control; (C) atrazinecontamination; (D) nicosulfuron-contamination). (1) Pennate diatoms; (2) cryptophyceae; (3) chlorophyceae; (4) xanthophyceae; (5) chrysophyceae; (6) centric diatoms; (7) cyanobacteria.

134

C. Leboulanger et al. / Environment International 26 (2001) 131±135

maximal biomass attained were slightly higher and lower than in controls, in atrazine- and nicosulfuron-treated microcosms, respectively. 3.4. Species diversity evolution At the beginning of the experiment, most of the phytoplankton population was dominated by pennate diatoms; this domination increased at the end of the experiment for each set of microcosms, whereas cryptophycea disappeared from all the bottles. The three sets of treatments were visibly different in taxonomic composition (Fig. 3) and a correspondence analysis was made to confirm this impression (Fig. 4). There was also a difference in the sensitivity to both herbicides, as shown on Fig. 5, among species and algal groups. Atrazine seemed to be more toxic than nicosulfuron, especially to chlorophycea, whereas nicosulfuron inhibited mostly diatoms growth.

Fig. 4. Correspondence analysis of the final phytoplanktonic communities in the microcosms. Communities are described in the data matrix as percent of the total density for each algal species. (T) control bottles; (A) atrazinetreated bottle; (N) nicosulfuron-treated bottles.

species tested (C. vulgaris) was affected, both on a and b parameters (Fig. 1). 3.3. Microcosms growth The biomass evolution in the microcosms during June 1997 experiment are shown in Fig. 2. Whereas there was no strong differences between treatments, the three sets were significantly different (Student's test, P =.05). The

Fig. 5. Number of species unaffected, inhibited, or simulated by contaminations, relative to control.

4. Conclusion In this study, some of the results we obtained for several years of experiment on natural phytoplankton populations and in culture were provided. The choice of testing atrazine and nicosulfuron in order to compare respective effects was driven by the changes in agricultural practices. If no direct evidence for the toxicity of nicosulfuron arises from in vitro experiments shown here as from reglementary publications (PMRA-ARLA 1996), one could not conclude firmly about the effects of possible contamination in aquatic systems (significant effect of 10 mg/l in microcosms). The effects of atrazine, with a welldocumented scientific background from years of use (and

Fig. 6. The advantages and limitations of various approaches in ecotoxicology, from tube test to the field. Modified after Zieris (1991).

C. Leboulanger et al. / Environment International 26 (2001) 131±135

abuse) of this molecule, are easier to demonstrate in pure culture as in microcosm experiments. At first sight, and due to the chemical and behavioral characteristics of both molecules in natural waters (i.e., short lifespan and low predicted environmental concentrations, and long life and high levels of contamination already reached, respectively, for nicosulfuron and atrazine), it seems that the ecotoxicological risk of the new molecule is lower than that of the xenobiotic it attend to replace. But we could not support such optimistic conclusions as long as we could not collect a broad range of data from in vitro to field experiment. There are differences in advantages and disadvantages of both approaches (Fig. 6) but there is no means to address accuracy, low cost, reproducibility, and ecological realism. We consider that there is a need to provide users and reglements a more accurate means to evaluate the toxicity of a chemical to algae, not only based on standardized in vitro tests, but also on more relevant, if not standardized, ecosystem-like experiments, such as the one we presented.

Acknowledgments The authors wish to thank Eliane Menthon for technical support in laboratory cultures, Jean-Pierre Bosse for help in performing the pesticide monitoring, and Jean-Claude Druart for assistance in studying algal taxonomy.

135

References BeÂrard A, Pelte T, Druart JC. Seasonal variations in the sensitivity of Lake Geneva phytoplankton community structure to atrazine. Arch Hydrobiol 1999;145:277 ± 95. de Noyelles F, Kettle DN, Sinn DE. The responses of plankton communities in experimental ponds to atrazine, the most heavily used pesticide in the United States. Ecology 1982;63:1285 ± 93. Hutchinson GE. An introduction to population ecology. New Haven: Yale Univ. Press, 1978. Lewis MR, Smith JC. A small volume, short-incubation-time method for measurement of photosynthesis as a function of incident irradiance. Mar Ecol Prog Ser 1983;13:99 ± 102. LuÈnd JW, Kipling C, Le Cren ED. The inverted microscope method estimating algal number and statistical basis estimations by counting. Hydrobiologia 1958;11:143 ± 70. PMRA-ARLA. Decision document: nicosulfuron. Canada: Pesticide Management and Regulation Agency, 1996 (42 pp.) http://www.hc-sc.gc.ca/ pmra-arla/e9601e.pdf. Schwartz D. MeÂthode statistique aÁ l'usage des meÂdecins et des biologistes. Paris: Flammarion Sciences MeÂdicales, 1963. Simpson DM, Stoller EW, Wax LM. An in vivo acetolactate assay. Weed Technol 1995;9:17 ± 22. Thioulouse J, Chessel D, DoleÂdec S, Olivier JM. ADE-4: a multivariate analysis and graphical display software. Stat Comput 1997;7:75 ± 83. USEPA (U.S. Environmental Protection Agency). Algal growth test. Short-term methods for estimating the chronic toxicity of effluent and receiving waters to freshwater organisms. Cincinnati, OH: U.S. Environmental Protection Agency, 1989. pp. 147 ± 74. Walsby AE. Numerical integration of phytoplankton photosynthesis through time and depth in a water column. New Phytol 1997;136: 189 ± 209. Zieris FJ. Means and problems of the ecotoxicological test with artificial aquatic ecosystems. Vehr-Int Ver Limnol 1991;24:2322 ± 5.