Effects of the phenylurea herbicide isoproturon on periphytic diatom communities in freshwater indoor microcosms

Environmental Pollution, Vol. 94, No. 2, pp. 141-152, 1996 PII:

SO269-7491(96)00080-2

Copyright 0 1997 Elswier Science Ltd Printed in Great Britain. All rights reserved 0269-7491/96 %15.00+0.00

ELSEVlER

EFFECTS OF THE PHENYLUREA HERBICIDE ISOPROTURON ON PERIPHYTIC DIATOM COMMUNITIES IN FRESHWATER INDOOR MICROCOSMS F. PCrk,” D. Floriqb T. Grollier,” A. Feurtet-Mazel,” M. Coste,c F. Ribeyre,” M. Ricardb & A. Boudou= “Laboratoire d’Ecotoxicologie, Universitk Bordeaux I/CNRS, Avenue des Facultks, 33405, Talence Cedex, France bInstitut d’Am&agement, UniversitC Bordeaux III, Domaine Universitaire, 33405, Talence Cedex, France ‘CEMAGREF, 50 Avenue de Verdun, 33610, Cestas, France

(Received 13 March 1996; accepted 13 June 1996)

Abstract

systems. Several field studies on streams, small lakes and outdoor mesocosms have shown that their productivity can be extremely large, in some cases exceeding that of phytoplankton (Wetzel, 1964; Wetzel, 1983; Goldsborough et al., 1986; Herman & Kaushik, 1986; Thielcke & Ratte, 1994). Periphyton is often an important non-target community given the many contamination sources of natural, and in particular, anthropogenic origin; the resulting structural and functional effects may cause severe damage to higher trophic levels, within the benthic and pelagic trophic networks (Patrick, 1978; De-Noyelles et al., 1982). Over the last decade, several methods using periphytic diatoms as biological indicators have been developed, to assess and indirectly quantify water quality (Descy, 1979; Coste et al., 1991; Round, 1991; Prygiel & Coste, 1993). The toxicological effects of pollutants (heavy metals, organic compounds) on freshwater diatoms have often been restricted to studies based on single species bioassays, with the algae being grown axenically in more or less artificial conditions (Rachlin et al., 1983; French & Evans, 1988; Guanzon et al., 1994). The aim of our ecotoxicological approach was to assess the effects of contaminants on the development of periphytic diatom communities on artificial substrata within indoor microcosms. Each experimental system was based on a mixed biotope (water column and natural sediment), with different species occupying key positions in relation to the potential transfer routesrooted macrophytes, burrowing insect larvae, molluscs, fish, etc. (Boudou & Ribeyre, 1983; Ribeyre, 1993; Odin et al., 1995). This paper investigates the effects of a phenylurea herbicide, Isoproturon or IPU (3-(4-isopropylphenyl)-1, 1-dimethylurea), when added to the water column or sediment compartment as initial contamination source. IPU is a pre- or post-emergence systemic herbicide, commonly used to control annual grasses and broad-leaved weeds in barley, wheat and rye. Its phytotoxic action is based on the inhibition of the photodependent electron transport in thylakoids, at the photosystem II level, by

The toxic eflects of the phenylurea herbicide Zsoproturon -ZPU: (3-(4-isopropylphenyl)-1, I-dimethylurea)-were studied on the colonization of periphytic diatom communities, within indoor microcosms consisting of a mixed biotope (water column and natural sediment) and two biological species-rooted macrophyte cuttings (Elodea densa) and benthic bivalve molluscs (Corbicula fluminea) . The periphyton, essentially composed of diatoms, was collected on artljicial substrata (glass slides) in the upper layers of the water column, after two periods of exposure (34 and 71 days). ZPU was initially added in the water or in the sediment compartment, at two nominal concentrations (Ll and L2 levels) for each contamination source5 and 20 pg litre-’ and 100 and 400 pg kg-’ in sediment (w/w) respectively. The eflects of ZPU on the density and community structure of periphytic diatoms are described. A marked reduction in the diatom density was observed after 34 days exposure to the lower concentration of ZPU in the water (5 pg litre-I). For the L2 levels, the very small number of live cells present did not permit quantiJication of the diatom density. After 71 days, recovery in community parameters occurred for the two contamination levels of the sediment and water column sources. Samples collected in the experimental units contaminated with the L2 levels were dominated by heterotrophic and smaller diatom species, such as Sellaphora seminulum. Data treatment based on factorial discriminant analysis enabled us to distinguish the d@erent contamination conditions, with only 11 species from the 130 taxa identljied. 0 1997 Elsevier Science Ltd. All rights reserved. Keywords: Freshwater ecotoxicology, periphyton, diatoms, indoor microcosm, herbicide, Isoproturon.

INTRODUCTION Within freshwater environments, the ecological importance of periphytic diatoms, notably their role in primary production, varies among different types of 141

F. Pe’rt%et al.

142

binding specifically to the quinone site on the Dl protein, a 32 kDa protein within the thylakoid membrane (Ducruet, 1991; Ponte-Freitas et al., 1991; Arnaud et al., 1994). Very little research has been devoted to the study of IPU bioaccumulation and its toxic effects on primary producers in freshwater systems (Kirby & Sheahan, 1994); there are, to the present authors’ knowledge, no publications on the periphyton. In this study, several complementary criteria were applied to analyse IPU effects on the periphytic diatom communities: cell density, species richness and relative abundance. The effects of the herbicide were studied for two contamination levels of the water column and sediment sources, after 34 and 71 days exposure.

MATERIALS

AND

METHODS

Structure of the experimental systems The basic structure of the freshwater microcosms consisted of a glass tank (12 x 12 x 30 cm), lined with a plastic bag (Plastiluz, alimentary standard with a very small IPU adsorption capacity) and containing natural sediment (5 cm depth), 3.1 litres of dechlorinated tap water, two rooted macrophyte cuttings and two benthic molluscs. The sediment was collected from the banks of the Garonne river, upstream from Bordeaux, France. This sediment was rich in clays (75580%) with little organic carbon content (2%). It was mixed with pure sand (98% silica, granulometry 0.8-l .4 mm, SILAQ, Facture, France) in a proportion of SO/_50 and homogenized by mechanical mixing in order to obtain a substrate adapted to the burying activity of the molluscs. The substrate was then distributed among the experimental units (EUs) using plastic trays of 10.5x10.5x5 cm (Monoplast-BC 50; alimentary standard). The general chemistry of the tap water was: pH = 7.5; conductivity = 40 @.m; HCOs = 23 1.8 mg litre-‘; Cl = 16.0 mg litre-‘; S04= 37.5 mg litre-‘; Ca = 53.5 mg litre-‘; Mg = 12.2 mg litre-‘; NH4 = 0.01 mg litre-‘; NO2 =0.09 mg litre-‘; NO3 = 1.8 mg litre-‘; P04=0.05 mg litree’. The biological samples (plant cuttings and molluscs) and the artificial periphyton substrates were introduced into the EUs 8 days after the water and sediment compartments had been added. This delay was needed to allow the physicochemical conditions to stabilize. The macrophyte species selected was Elodea densa (Hydrocharitacea). A large batch of homogeneous cuttings was supplied from a specialist producer (Ets Quentin, Bouillargues, France) from indoor cultures. The samples for the experiment were obtained by section of the apical part of the main stem, without lateral branches; they were 12 f 0.2 cm in length. Two cuttings were planted in each EU (randomized distribution) by sinking the lower 1.5 cm deep into the sediment. The mollusc species was Corbicula jhminea, the Asiatic clam. Samples were collected from the Canal du Midi, upstream from Toulouse, France; they were stored in the laboratory for 3 weeks before the experi-

ment. A homogeneous batch was then obtained by measuring shell length (1.5-2 cm); two molluscs were introduced into each EU. This filter-feeding bivalve lives buried in the superficial sediment layers, with permanent exchanges with the water column via its inhalant and exhalant siphons (Britton & Morton, 1982). Periphyton came from the sediment (epipelic species) and from the macrophyte cuttings (epiphytic species). Preliminary studies set up in the same experimental conditions showed that periphyton was essentially composed of diatoms. In order to homogenize the initial pools of diatoms in the EUs, the macrophytes were rinsed before planting with dechlorinated tap water to remove epiphyton; each EU received 10 ml of this rinsing water at the beginning of the experiment. Periphyton was collected from cleaned glass slides (area = 19.75 cm2) supported by plastic straws driven into the sediment. The glass slides were inserted perpendicularly on the straws and positioned vertically in the water column at about 10 cm from the sediment surface. Three straws, each carrying two glass slides, were introduced in the EUs at the same time as the organisms. After 34 (tl) and 71 (t2) days exposure, one straw was collected from each EU and two glass slides were pooled for periphyton counting and determination. The last two slides were used for chlorophyll u measurement at the end of the experiment (t2). The EUs were placed in a large water tank (140x65~30 cm) with heating and cooling systems. Temperature was maintained at 20*0.2”C. Light was artificially produced by two neon tubes (Sylvania F36W/GRO) positioned 45 cm above the EUs and operated by timer switches. Average light intensity was the photoperiod was 12: 12 h 35+ 3 PE cm-2 s-l; 1ight:dark. The pH of the overlying water was periodically checked (pHmeter Labo Moderne, Paris) and presented small variations during the 71 days (7.5 i 0.3). Dissolved oxygen concentrations were measured in the water column (Oxymeter Labo-Moderne, OX1 91); IPU effects on the autotrophic species (macrophytes, periphyton) induced a significant decrease via the inhibition of photosynthetic activity. Contamination procedure The experiment was based on a comparative study of the two initial contamination sources (water column and sediment). For the water source, two nominal concentrations were selected: 5 and 20 pg litre-‘, called WLl and WL2 respectively. These levels were chosen in accordance with previous studies on similar experiet al., 1996). IPU mental conditions (Feurtet-Maze1 dosages on field freshwater samples generally indicate concentrations below the drinking water regulation limits (0.1 pg litre-’ in Europe), but in agricultural runoff waters, values of up to 17 pg litree’ have been observed, even though they were short-lived (Kirby & Sheahan, 1994). IPU was added to the EUs from a concentrated aqueous solution at 50 mg litree’, prepared by direct dissolution of technical IPU powder (RP Agro; 99.8% purity) in dechlorinated tap water

EfSects of Isoproturon on periphytic diatoms under magnetic stirring for 72 h, with several ultrasonication steps. Sediment contamination was based on IPU additions from the same concentrated aqueous solution, in order to obtain two contamination levels (SLl and SL2)-100 and 400 pg kg-’ (w/w). Homogeneity of IPU distribution throughout the substratum was obtained by mechanical mixing before addition to the EUs. Weekly IPU dosages on water samples enabled us to measure the evolution of the contamination pressure: decrease of IPU concentrations in the EUs exposed via the water source; IPU transfers from the sediment source, via the release processes. The experimental design was based on four contamination conditions-2 sources x 2 IPU concentrations-plus one control condition, with two replicates, giving 10 EUs set up simultaneously. This protocol was defined within a broader study which included cadmium and mercury compounds (inorganic mercury and methylmercury) and thus justified the use of six control EUs at time zero; some data relating to periphyton will be based on samples taken from these six control units. Periphyton sampling and analysis criteria Periphyton was collected from the glass slides with a razor blade; the algae being scraped from the two surfaces of the slides (39.5 cm2 per slide). Microscopic examination of the scraped slides showed that very few organisms remained attached. Samples were then preserved in 100 ml of 5% formalin solution (Form01 37%, Prolabo). After ultrasonic homogenization (5 min), a subsample was taken (1 ml) and diatoms were counted without identification of the species, using a double Nageotte counting cell. The two cells consisted of a rectangular chamber (100 ~1) containing 80 fields. The periphytic algae were counted in 50 fields, at a total magnification of 200x (Olympus, BH2). Venrick (1978) noted that 40 fields were sufficient when one individual or more were present in each field. Cell density (number of cells cme2) was obtained using the formula [A/(50 x 1.25) x 100000]/79 where A is the total number of individuals counted, 50 is the number of fields, 1.25 is the volume of one field (Al), 100,000 is the initial volume of the sample (pl), 79 is the surface area of two glass slides (cm2). Diatom identification was based on their microscopic siliceous exoskeleton, after removal of the protoplasm. The samples collected from the glass slides were kept for 1 day without mixing in order to allow diatom sedimentation. The residue was treated with an identical volume of boiling hydrogen peroxide (Prolabo, 30%) for 5 min. Immediately after this treatment, the sample was washed thoroughly to remove traces of H202, with repeated centrifugations and re-suspensions in distilled water. An aliquot of the last residue was mounted in a high refractive index medium (Naphrax, Northern Biological Supplies, Ipswich, UK) for microscopic observation at a magnification of 1000x. About 500 diatom frustules were counted per sample; above this number of

143

individuals, the number of species investigated does not increase significantly (Coste, 1978). The diatoms were identified according to the SiiBwasserflora (Krammer & Lange-Bertalot, 1986, 1988, 199la, 199lb). For the Naviculaceae, classification was according to Round et al. (1990). Results were expressed using the relative abundance criterion for each diatom species. From these results, it was possible to determine the species richness N (total number of species detected in each EU) and to calculate two complementary indices-shannonWeaver’s diversity (H> and the regularity (R) indexes. The first takes into consideration the relative abundance of species, increasing as the community becomes more diversified (Shannon & Weaver, 1948). The second represents the relationship between the calculated diversity and optimal theoretical diversity (H_J obtained by assuming that each of the N species in the sample is represented by the same number of individuals Regu(equidistribution): H,,,,, = log2 N, R = H/H,,,. larity varies between 0 and 1 (Pielou, 1966). These indices, which were devised for field studies on running waters, have also been applied to experimental streams (Hudon, 1987; Amblard et al., 1990; Hiirlimann & Schanz, 1993). The average length of each diatom species was taken from the literature (Krammer & Lange-Bertalot, 1986, 1988, 1991a, 1991b); for a small number of species, direct measurements were necessary. Chromatographic analysis of isoproturon in water samples IPU dosages in water samples were obtained according to the protocol validated by the Rhone-Poulenc Research Center (CRIT, Decines, and RP Agro, La Dargoire, France). Samples collected from the EUs were directly analysed after filtration through 0.45 pm DuraporeR filters (Millex) by HPLC (Hewlett-Packard 1050 liquid chromatograph). The autosampler was equipped with a sample loop of 100 ~1 and a 12.5 cm long, 4.6 cm diameter and 5 pm particle size analytical column (Lichrospher 60 RP-C18, Merck). The solvent system delivered 0.9 ml min-’ of a mobile phase consisting of 65:35 (v/v) water:acetonitrile. The column temperature was 30°C. The UV detector operated at 240 nm. Concentrations of IPU in the water samples were determined according to an external calibration curve, based on five concentrations from 0 to 100 pug litre-‘, prepared from a stock solution of IPU in pure acetonitrile (1 g litre-‘). The detection limit of the analytical method was 2 pg litre-‘, without a preconcentration step. Data treatment The results of codified diatom counts were entered in the OMNIDIA program (Lecointe et al., 1993) for calculation of relative abundance of species, diversity (Shannon-Weaver index) and regularity. Multidimensional analysis was carried out using the LADDAD program (versions 89.1 and 92.1). The first stage, based on a hierarchical cluster analysis, enabled one

144

F. P&is

to distinguish, from among the 130 taxons present, those with similar responses to IPU contamination. Those taxons which were highest in this analysis were then subjected to a discriminant analysis (rate of misclassification 12.5%) in order to obtain a global approach to the effects of the different contamination conditions studied on the relative preponderance of the more significant diatom species.

et al. 34 days correspond to approximately 25% of the amounts initially added to the sediment compartments. These transfers of herbicide into the water column represent therefore a secondary source of contamination for the periphytic species. Concentrations are similar to, even greater than, those found in the water source; moreover, they vary very little throughout the 71 days exposure, compared with the decrease in IPU concentration after the initial addition to the water column (Fig. l(A)).

RESULTS Effects of IPU on diatom density Evolution of IPU concentrations in the water column

Results from IPU determinations on water samples corresponding to the Ll and L2 levels of contamination of the water column source are shown in Fig. l(A). Decrease of IPU concentrations was progressive from the nominal values at time zero; the average concentrations measured after 71 days exposure correspond to about 50% of the initial levels. These results are in agreement with estimated half lives from preliminary studies set up in similar exposure conditions, based on a wide range of IPU concentrations in the water (FeurtetMaze1 et al., 1996). This herbicide has a relatively high level of persistence in the aquatic environment, which is confirmed by studies carried out on outdoor microcosms, giving half lives of between 13 and 36 days, depending on the structure of the systems (Merlin et al., 1995). There are several processes that may cause this decrease in IPU concentration in the water column of the EUs: degradation of the herbicide (photolysis, hydrolysis, biodegradation), volatilization, adsorption on the tank walls, transfer to the superficial sediment layers, and bioaccumulation in the organisms (algae, macrophytes, molluscs). In fact, given the physicochemical properties of Isoproturon and the experimental conditions (temperature, pH, size of plant and animal biomass, etc.), it would seem likely that all these different processes contribute to the gradual disappearance of the herbicide into the water column of the EUs (Feurtet-Maze1 et al., 1996). The evolution of IPU concentrations in the water column for the two initial contamination levels of the sediment source is shown in Fig. l(B). IPU was transferred rapidly from the sediment compartment to the water column during the first few days after the EUs were set up. When the macrophyte cuttings, the mollusts and the glass slides were added (time zero), the average IPU concentrations were close to 25 and 5 pg litre-‘, for the two sediment contamination levels. They then increased slightly, until 34 days, with maximum concentrations reaching 35 and 8 pg litre-‘, respectively. During the second phase of the experiment, a slow, progressive decrease was observed, with herbicide concentrations in the water column after 71 days being close to those measured at time zero. These releasing processes are extremely significant; they may be attributed preferentially to diffusion phenomena, via the IPU present in the surface layers of the sediment. Thus the mean quantities of IPU in the water column after

After 34 days (tl), the mean density on the glass slides in the control units was about 30,000 cells per cm2, with a significant scattering of data among the six EUs (Fig. 2). The effect of IPU on this criterion is very marked, giving rise to a decrease in density of between 87 and 96% compared with the control values, after exposure to the Ll contamination level of the two sources. For levels WL2 and SL2, colonization was extremely low and no data could be obtained. At the end of the experiment (t2), the mean density of the periphytic algae in the control units was about half that observed at time t 1. For all the contamination conditions studied, however, these values were much greater: thus, for level Ll, mean densities were comparable with those measured in the control units. In all instances, whatever the fate of IPU concentration in the EU water columns, colonization of the artificial substrata in the second phase of the experiment increased considerably: the very marked effect of the herbicide on this criterion after 34 days exposure did not lead therefore to the total destruction of the algae, especially for conditions WL2 and SL2. Effects of IPU on the diatom community structure

In the samples analysed, 130 taxa of diatoms were identified. Regarding the species richness and the indices of diversity and regularity obtained from samples collected in the EUs after 34 days exposure, there was a difference between the control units and the units contaminated only for level Ll of the water source (Table 1). It should be recalled that because of the very low algal densities observed after contamination by level L2, it was not possible to obtain a representative inventory of the species that had colonized the glass slides at time tl. At the end of the experiment (71 days), there was a marked increase in the number of species compared with time tl, and this was the case for all conditions studied. In the control EUs, this increase was close to 20%, indicating that colonization of the artificial substrates had continued during the second phase of the experiment; the regularity index was highest, indicating optimal conditions for diversity. For the two conditions WL2 and SL2, the average number of species was 20 and 15.5, respectively; these values were lower than those observed in the control units and in those contaminated by level Ll. Values for the diversity index (H) were also lower (1.97 and 1.46, respectively) compared

145

Efects of Isoproturon on periphytic diatoms

the different experimental conditions studied (Fig. 3) showed that, at time t 1, the periphytic communities consisted predominantly of large-sized species; thus, for the control and SLl conditions, 71% of the diatom populations were composed of species greater than 50 pm in length. For condition WLI, 90% of species were longer than 50 pm. However, after 71 days, the average length of more than 50% of the species

EUs. The regularity index also decreased to 0.42 and 0.37, compared with 0.59 for the controls. The determination of the average size of the different species of diatoms led to the setting up of 21 classes according to length, ranging from < 5 to 100 pm. Analysis of the relationship between this criterion and the accumulated relative abundance of the species for

with

3.08 for the control

0-i -8

; ; ; ; ; ; ; ; ; ;; -6

-4

-2

0

2

4

6

8

; ; ;;

Exposure

-8

-6

-4

-2

0

2

4

6

8

; ; ; ;;

;;

; ; ; ; ; ; ; ; ; ; ;,

IO 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70

duration (days)

IO 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48

Exposure

50 52 54 56 58 60 62 64 66 68 70

duration (days)

Fig. 1. (A) Changes in IPU concentrations (pg litreel) during the 71 days of exposure from the water column contamination source (WLl and WL2). (B) Changes in IPU concentrations (pg litre-‘) in the water column of EUs contaminated via the sediment (SLl and SL2). Symbols correspond to the two replicates per condition; lines correspond to the average IPU concentration.

F. Pdrts et al.

146

observed was less than 15 pm, from all sources and contamination levels. Moreover, when the contamination pressure increased, colonization of the glass slides by small species was favoured. Several studies have shown that diatoms that are large in size participate in the initial phases of substrate colonization. They are then replaced by diatoms with relatively small size (Acs & Kiss, 1993). From the 130 taxa identified were selected those with a relative abundance greater than 16%0 in at least one experimental condition, in order to eliminate species whose presence was purely random; thus, 24 species were retained for the comparative study of the diatom community structure. Analysis of their relative abundance according to the different experimental conditions revealed that the length of exposure and the level of contamination of the water column and sediment sources had a marked effect (Fig. 4).

??

At time (ESOL),

tl,

the

three

Navicula

species

Eunotia

soleirolii

aquaedurae

(NAQR) and Gomphonema contraturris (GCTT) predominated in the diatom community in the control units; their mean relative abundance were 33, 21 and 9%, respectively. These species were always present in the

contaminated

units,

whereas

others,

like

Melosira varians (MVAR) or Fragilaria construens var. venter (FCVE), tended to disappear. Certain

on the other hand, such as Gomphonema (GPAR) and Navicula minima (NMIN), which were barely represented in the control units had a relative abundance of close to 10% for condition SL 1. At time t2, the dominant species in the control EUs after 34 days exposure had been replaced by new taxa: Achnanthes minutissima (AMIN), NMIN, Amphora pediculus (APED), with a mean relative abundance of 25, 13 and 1 1%, respectively. species,

par&urn

??

Fig. 2. Density of diatoms, function of the different experimental conditions after 34 (tl) and 71 (t2) days of exposure.Water contamination source_WLl (5 pg litree’) and WL2 (20 pg litre-‘); sediment source-SLI (100 ,ug kg-‘) and SL2 (400 pg kgg’). Symbols correspond to the replicate values per experimental conditions (six EUs were taken into account for the control conditions; black symbols correspond to the two EUs placed in the same tank). Histograms correspond to the average densities.

Table 1. Comparative analysis of the diatom community structure based on three complementary criteria Experimental ~~ ti (34 d)

t2 (71 d)

conditions

_._____ Control WLl SLl WL2 SL2 Control WLl SLl WL2 SL2

N

H

R

27-27-27-l 8-37-20 (26) 24-13 (18.5) 21-27 (24) N.D. N.D. 38-31-35-29-37-55 (37.5) 30-30 (30) 2632 (29) 1624 (20) 15-16 (15.5)

2.78 1.86 3.03 N.D. N.D. 3.08 2.87 2.67 1.97 1.46

0.60 0.44 0.66 N.D. N.D. 0.59 0.58 0.54 0.42 0.37

N, number of total diatom species (six replicates for the controls and two for the contaminated microcosms-average values in brackets-bold values for the two experimental units within the same tank including the IPU contaminated microcosms); H, Shannon-Weaver index (average values); R, regularity index (average values). N.D., non determined (too low diatom densities).

147

EfSects of Isqproturon on periphytic diatoms For the two conditions WLl and SLl, a change in predominance from AMIN to NMIN was shown. APED was no longer present, whereas other taxa appeared, such as Cymbella mesiana (CMES) for condition SLl-t2. For the highest contamination levels (WL2 and SL2), Sellaphora seminulum (SSEM) clearly predominated, with a mean relative abundance greater than 50%. A hierarchical cluster analysis permitted the selection of species whose contribution was greater than the average estimated contribution. Then, a discriminant analysis was carried out on the relative abundance of these selected species. The two dimensional display (Fl xF2) (Fig. 5) shows a clear separation of the different experimental conditions studied, except the WLl-tl condition which appears in superposition with the Control-t1 condition. Information obtained from this global analysis confirmed and clarified those made during the comparative study of the relative abundances (Fig. 4). The Control-t1 and WLl-tl conditions were characterized by the four species ESOL, NAQR, Navicula cryptocephala (NCRY) and MVAR. Data from field studies showed that these species were characteristic of unpolluted or very low polluted freshwater systems, e.g. ESOL and MVAR, or of oligo to mesosaprobic waters, e.g. NAQR and NCRY (Lange-Bertalot, 1979; Van-Dam et al., 1994). GPAR enabled us to distinguish the condition SLl at tl. This facultatively N-heterotrophic taxon is well-known for its ability to resist any major organic charge in the natural environment and was classified by Lange-Bertalot (1979) with the most pollution-resistant species. Results for the control EUs at time t2 were more scattered than at time t 1. The species which were specific to the

controls were NMIN and Achnanthes exigua (AEXG). The species NMIN and Fallacia monoculata (FMOC) characterized the WLI and CMES the SLl condition. NMIN and FMOC are classified by Van-Dam et al. (1994) and Hofmann (1994) with the mesosaprobous group. CMES, on the other hand, was defined as a species living in uncontaminated environments. Nevertheless, its resistance to the low concentrations of IPU in the water could be attributed to the protection afforded to the cells by the gelatinous slime. Navicula vandamii (NVDA) was specific to condition WL2, and SSEM to the SL2 exposure condition. Little information is available concerning the ecology of NVDA. However, SSEM, a facultatively N-heterotrophic taxon, was characteristic of very polluted environments (Lange-Bertalot, 1979; Van-Dam et al., 1994).

DISCUSSION

AND CONCLUSION

This study, based on indoor microcosms, has shown the importance, from a methodological standpoint, of taking into consideration the periphytic compartment within these experimental models. In fact, the diatom community that colonized the artificial supports introduced into the water column of the EUs was very diverse indeed: 130 taxa were identified. Preliminary experiments showed that a period of one month was required in order to obtain a satisfactory colonization of the glass slides. In these experimental conditions, the number of diatom species present on the artificial substrate increased between 34 and 71 days, suggesting that colonization was continuing (Table 1). This kinetics of colonization is clearly relatively slow compared with those generally observed in the natural

1000 900 .800 .700 .600 .500 .400 .300 -200 .loo-on 0

..

,

5

IO

t IS

20

2.5

30

35

40

45

50

55

60

65

70

75

80

8.5

90

95

100

105

Mean length of the diatom cells (pm) Fig. 3. Cumulative relative abundance of the diatom species (average values between 2 replicates), function of the average length of the diatom cells. Control (C), water contamination source (W), sediment source (S). Exposure duration--tl, 34 days; t2, 71 days. Contamination levels-WLl, 5 pg litre- I., WL2, 20 pg Ii’.,,_-‘; SLI, 100 pg kg-‘; SL2,400 Fg kg-‘.

148

F. Pe’r&set al.

environment, where two weeks are usually sufficient to produce algal development on substrates introduced in running waters (Lowe & Gale, 1980; Hoagland et al., 1982). This longer period is linked to the absence of current in the EUs, where the water is mixed only by the inhaling and exhaling water exchanges produced

by the filtering of the bivalves, and by interventions in the water column when samples are collected and physicochemical measurements are taken (pH, dissolved 02, etc.). Current velocity is known to influence the accumulation of diatoms on substrates (Stevenson, 1983).

Control -tl

Control -t2

Fig. 4. Average composition of the diatom communities, in relation to the contamination conditions of the water column (WLl and WL2) and sediment (SLl and SL2) and to the exposure duration (tl = 34 days; t2= 71 days). Species selected have a relative abundance > 16%0 at least in one experimental condition.

Effects of Isoproturon

on periphytic

149

diatoms

toxic effects of IPU; availability of nutrients, which had not been used up by time t 1, since algal development during the first 34 days of the experiment had been low. Studies on photosynthesis-inhibiting herbicides have shown that the treatment of freshwater biotopes may lead to an increase in the level of nutrients present in the water column (Herman & Kaushik, 1986; Gurney & Robinson, 1989); this increase is interpreted as resulting from inhibition of epipelic algal communities, which are important in fixing minerals from the upper sediment layers. At time tl, the diatom community consisted predominantly of large-sized species, both in the control and contaminated EUs (Fig. 3). Community structure (diversity and species inventory) is not strongly affected by contamination at the Ll level (Table 1 and Fig. 4). Only condition SLl can be differentiated from the control only by the presence of the species Gomphonema par&urn. After 71 days exposure, IPU favours species that are small in size (Fig. 3) and that is when herbicide concentrations measured in the water are strongest (WL2 and SL2); IPU exerts a selection pressure which affects the diatom community by eliminating the most sensitive species (Table 1 and Fig. 4). Small diatoms such as Navicula minima and Sellaphora seminulum, whose relative abundance increased considerably in the contaminated EUs, have a high reproduction rate (strategy r); therefore, they are able to grow more rapidly than other species, given that the contamination pressure is fairly selective (Sommer, 1981; Symoens et al., 1988). Moreover, among those species that have been identified as being representative of IPU-contaminated conditions (Fig. 5), some present heterotrophic capacities; this is the case with Gomphonema parvulum and

Measurement of diatom density on samples from the control micrososms revealed an important scattering of data (Fig. 2). There are several phenomena that may account for these variations, in particular the differences in lighting in the water column. Indeed, the growth of the Elodea densa cuttings is quite considerable in the control EUs; between the beginning of the experiments and time 71 days, the biomass had increased by almost a factor of 4. The macrophytes form a fairly efficient and heterogenous light shield. Results indicated a clear decrease in the density of the diatoms on the glass slides in the control EUs between time tl and time t2 (Fig. 2). After 34 days of exposure, the IPU reduced the rate of colonization considerably, for an initial concentration of 5 pg litre-’ in the water column (Fig. 2). After 71 days, the glass slides, which at time tl were slightly colonized (WLl and SLl), or virtually not at all (WL2 and SL2), were characterized by a greater algal density; for condition WLl, the density is comparable with that measured in the control EUs. Thus, the diatoms present initially in the microcosms, which are responsible for the colonization of the artificial supports, are not destroyed by the herbicide, inducing a marked recovery according to the density criterion. Nevertheless, the colonization strategies are affected, probably via a complex series of interactive factors: species reproduction rate, mobility, fixing strategies using threads, stalks, etc. (Round et al., 1990). The increase in diatom density in the contaminated EUs between time t 1 and time t2 may be associated with a combination of several phenomena: decrease in the herbicide concentration in the water (conditions WLl and WL2); selection of taxons which are ‘resistant’ to the

AXIS 2 I sL1-t2 m

CMES

0

NM IhI ? ?

WLI-t2 FMOC

A

‘0

A

A -- c-t2

cl Axis 1

SL2-t2 SSEM 0 NVDA 0 WLZ-t2

AA %PAR NCRY 5L1-t1 NAQR A * +ESOL

c-t1

;@$Ll-tf

exiguo, CMES = Cymhrlla mesrana. ESOL = I:‘unorra .soletrolii; WAR = ~kvnphonrma pan&m; MVAR vanons, NAQR = Navrcula aquaedurae, NCRY = Navicula cryprocephala; NMIK = Navicula mmrmo; FMOC = IGdhxra mrmoculara. SSEM = Sellaphhora .semmulum: NVDA = Nav~~lo vandamrr

= Achnanthes

= Mehrra

Fig. 5. Ordination

A

MVAR +

cl

AEXG

AEXG

of diatom taxa and contamination conditions in the Fl x F2 plan of a discriminant analysis. (+) Controls tl; (0) WLI tl; (A) SLI tl; (A) Controls t2; (0) WLl t2; (m) SLI t2; (0) WL2 t2; (0) SL2 t2.

150

F. PCrPs et al.

especially Sellaphora seminulum (Lange-Bertalot, 1979). Thus, this latter species can be considered as an indicator of IPU pollution. In the present experimental conditions, the species Navicula vandamii is well represented in populations from WL2 contaminated EUs where the IPU concentration was permanently higher than 10 pg litre-‘. It would be interesting to find out the extent to which this species, which is rarely mentioned in the literature, is heterotrophic. Due to the phytotoxic properties of IPU, algae that are able to use a heterotrophic route have an advantage. The differences in the impact of IPU on the various periphytic species may also be due to structural and/or functional properties at the different stages that govern herbicide uptake, its accessibility to the binding sites on the thylakoid membranes, or the presence of biotransformation mechanisms. IPU absorption is based essentially on diffusion transport through the barriers separating the diatoms from the surrounding environment. Thus, the presence of polysaccharides around the cells may restrict the absorption capacity of the herbicide, the gelled capsule constituting an additional barrier against IPU diffusion across the frustule and the plasmic membrane. This may be the case, for example, with the species Cymbella mesiana, which belongs to the Encyonema group, specific to condition SLl (Fig. 5) and which is able to withstand a contamination pressure of 5 pug litre-‘. The tolerance of the algae to this type of herbicide could also be linked to alterations of the binding site on the Dl protein. Blanck and Wangberg (1991) envisage the possibility of substituting amino acids, which would increase the tolerance of the diatoms to the phytotoxic action of the PS II herbicides. IPU can also be metabolized, as described in several ‘resistant’ terrestrial species. Experimental studies on wheat exposed to (r-14C isopropyl-labelled IPU demonstrated an efficient metabolization of the herbicide, essentially based on oxidation followed by conjugation of the hydroxy derivatives produced. Dealkylation, the mechanism generally described in the degradation of phenyl-urea herbicides, was less effective (Cabanne et al., 1987; et al., Durst, 1991; Scarponi, 1993; Bohnenkamper 1994). The present authors have no information on IPU metabolization in freshwater diatom species. Nevertheless, if such biotransformation reactions occur in some species, then given that the rate of decrease in the concentration of IPU in the water column is low, especially for conditions SLl and SL2, these reactions should result in an increased uptake of IPU from the surrounding environment, in order to compensate for the displacement of the chemical equilibria between the extra- and intra-cellular compartments (Cabanne et al., 1987; Burnet et al., 1993). Studies at the cellular level on isolated Elodea densa leaves, based on the herbicide-sensitive rapid enhancement of in-vivo PSI1 fluorescence induction using digital imaging technology, demonstrated that the time-course of IPU accessibility to the Dl protein sites was very short, between 10 min and several hours, depending on IPU concentration in the medium (Grouselle et al., 1995).

It should be stressed also that the results obtained relating to the effects of IPU on the density of diatom populations and on the characteristics of the communities were correlated to the concentration of the herbicide in the water column, whatever the initial contamination source (water column or sediment). For example, the effects observed on cell density at time tl and time t2 (Fig. 2) were very closely dependent on the herbicide concentrations in the water, or, more precisely, on concentrations measured during the phase preceding sample collection (Fig. 1). Thus, in the experimental conditions, the presence of IPU in the sediment compartment would have exerted a very weak direct action on the ability of the diatoms to colonize the artificial substrates. The effects observed would be mainly associated with the water column source, via IPU transfers from the surface sediment layers. Studies carried out in parallel to this one, on trace metals (mercury compounds and cadmium) confirm the potential capacity of the periphytic compartment in indoor microcosms to reveal and quantify the toxic effects of contaminants, both at structural and functional levels. From a methodological point of view, several complementary approaches are currently being tested, in order to analyse the effects of herbicides and heavy metals on a periphytic community already installed on artificial supports and also to study the recovery capacities of the diatom communities after exposure to toxic products. Lastly, it should be mentioned that experimental studies are also underway into an analysis of the actions and interactions between various abiotic factors (temperature, pH, photoperiod) on the effects of contaminants on the dynamics of diatom populations. Abiotic factors play a fundamental role on the chemical fate of contaminants within the aquatic biotopes and on their bioavailability. Laboratory studies, based on factorial designs and specific equipment, should make it possible to define accurately the role of these factors, singly and in combination. It should also help to provide an interpretative analysis of data acquired during field studies.

ACKNOWLEDGEMENTS This work was supported by Rhone-Poulenc, within the research programme “Aquatic Ecotoxicology” set up in 1992. The authors thank G. Freyssinet (RP Scientific Direction, Paris) and P. McCahon (RP Agrochemistry, Sophia-Antipolis) for their comments on the manuscript, and H. Koziol, (Department of Foreign Languages, University Bordeaux II), for the improvement of the narrative style of the manuscript.

REFERENCES Acs, E. & Kiss, K. T. (1993). Colonization processes of diatoms on artificial substrates in the river Danube near Budapest (Hungary). Hydrobiologia, 269/270, 307-3 15.

Eflects of Isoproturon Amblard, C., Couture, P. & Bourdier, G. (1990). Effects of a pulp and paper mill effluent on the structure and metabolism of periphytic algae in experimental streams. Aquat. Toxicol., 18, 1377162. Arnaud, L., Taillandier, G., Kaouadji, M., Ravanel, P. & Tissut, M. (1994). Photosynthesis inhibition by phenylureas -A QSAR approach. Ecotoxicol. Environ. Safety, 28, 121& 133. Blanck, H. & Wangberg, S. A. (1991). Pattern of cotolerance in marine periphyton communities established under arsenate stress. Aquat. Toxicol., 21, l-14. Bohnenkamper, O., Glabgen, W. E. & Haas, M. (1994). Biodegradation de 1’Isoproturon dans les plantes et les sols. In Biodisponibilite des Pesticides et Biomarqueurs d’exposition et de resistance, ed. P. Soler. Groupe Francais des Pesticides, Bordeaux, France, pp. 139-142. Boudou, A. & Ribeyre, F. (1983). Contamination of aquatic biocenoses by mercury compounds: an experimental ecotoxicological approach. In Aquatic Toxicology, ed. J. 0. Nriagu. Wiley, New York, pp. 73-l 16. Britton, J. C. & Morton, B. (1982). A dissection guide, field and laboratory manual for the introduced bivalve Corbiculu Jiuminea. Malacol. Rev., 3, 1-82. Burnet, M. W. M., Loveys, B. R., Holtum, J. A. M. & Powles, S. B. (1993). A mechanism of Chlorotoluron resistance in Lolium rigidum. Planta, 190, 182-l 89. Cabanne, F., Huby, D., Gaillardon, P., Scalla, R. & Durst, F. (1987). Effect of the cytochrome P-450 inactivator l-aminobenzotriazole on the metabolism of Chlorotoluron and Isoproturon in wheat. Pest. Biochem. Physiol., 28, 371-380. Coste, M. (1978). Sur I’utilisation des diatom&es benthiques pour l’appreciation de la qualite biologique des eaux courantes. Doctorate Thesis, University of Besancon, France. Coste, M., Bosca, C. & Dauta, A. (1991). Use of algae for monitoring rivers in France. In Use of algaefor Monitoring Rivers, ed B. A. Whitton, E. Rott, & G. Friedrich. Studia, Innsbruck, Austria, pp. 75-88. De-Noyelles, S. F., Kettle, W. D. & Sinn, D. E. (1982). The responses of plankton communities in experimental ponds to Atrazine, the most heavily used pesticide in the United States. Ecology, 63, 1285-1293. Descy. J. P. (1979). A new approach to water quality estimation using diatoms. Nova Hedwigia, 64, 3055323. Ducruet, J. M. (1991). Les herbicides inhibiteurs du photosysteme II. In Les Herbicides: mode dhction et principes d’utilisation, ed. R. Scalla. INRA, Paris, pp. 799114. Durst, F. (1991). Metabolisation des herbicides. In Les Herbicides: mode dhction et principes d’utilisation, ed. R. Scalla. INRA, Paris, pp. 193-236. Feurtet-Mazel, A., Grollier, T., Grouselle, M., Ribeyre, F. & Boudou, A. (1996). Experimental study of bioaccumulation and effects of the herbicide Isoproturon on freshwater (Elodea densa and Ludwigia natans). rooted macrophytes Chemosphere, 32, 1499-l 5 12. French, M. S. & Evans, L. V. (1988). The effects of copper and zinc on growth of the fouling diatoms Amphora and Amphiprora. Biofouling, 1, 3-18. Goldsborough, L. G., Robinson, G. G. C. & Gurney, S. E. (1986). An enclosure/substratum system for in situ ecological studies of periphyton. Arch. Hydrobiol., 106, 373.-393. Grouselle, M., Grollier, T., Feurtet-Mazel, A., Ribeyre, F. & Boudou, A. (1995). Herbicide Isoproturon specific binding in the freshwater macrophyte Elodea densa-A single cell fluorescence study. Ecotoxicol. Environ. Safety, 32, 254259. Guanzon, N. G., Nakahara, J. H. & Yoshida, Y. (1994). Inhibitory effects of heavy metals on growth and photosynthesis of three freshwater microalgae. Fish. Sci., 60,37%384. Gurney, S. E. & Robinson, G. G. C. (1989). The influence of two triazine herbicides on the productivity, biomass and community composition of freshwater marsh periphyton. Aquat. Bot., 36, l-22.

on periphytic

diatoms

151

Herman, D. & Kaushik, N. K. (1986). Impact of Atrazine on periphyton in freshwater enclosures and some ecological Canadian J. Fish. Aquat. Sci., 43, 1217-1225. consequences. Hoagland, K. D., Roemer, S. C. & Rosowski, J. R. (1982). Colonization and community structure of two periphyton assemblages, with emphasis on the diatoms (Bacillariophyceae). American J. Bot., 69, 188-213. Hofmann, G. (I 994). Aufivuchs diatoms in Seen und ihre Eignung als Indikatoren der Trophie. Biblio. Diutomol., 30, l-241. Hudon, C. (1987). The ecological implications of growth forms in epibenthic diatoms. J. Phycol., 23, 434441. Htirlimann, J. & Schanz, F. (1993). The effects of artificial ammonium enhancement on riverine periphytic diatom communities. Aquat. Sci., 55, 4&64. Kirby, M. F. & Sheahan, D. A. (1994). Effect of Atrazine, Isoproturon, and Mecocrop on the macrophyte Lemna minor and the algae Scenedesmus subspicatus. Bull. Environ. Contam. Toxicol., 53, 12&l 26. Krammer, K. & Lange-Bertalot, H. (1986). Bucilluriophyceae 1. Naviculaceae. Subwasserjora von Mitteleupora, ed. H. Ettl J. Gerloff, H. Heynig & D. Mollenhauer. G. Fisher Verlag, Stuttgart, Germany. Krammer, K. & Lange-Bertalot, H. (1988). Bucillariophyceae 2. Bacillariaceae, Epithemiaceae, Surirellaceae. Stibwasserflora von Mitteleuropas, ed. H. Ettl, J. Gerloff, H. Heynig & D. Mollenhauer. G. Fisher Verlag, Stuttgart, Germany. Krammer, K. & Lange-Bertalot, H. (1991~). Bacilluriophyceae 3. Centrales, Fragilariaceae, Eunotiaceae, SlibwasserJlora von Mitteleuropa, ed. H. Ettl, J. Gerloff, H. Heynig & D. Mollenhauer. G. Fisher Verlag, Stuttgart, Germany. Krammer, K. & Lange-Bertalot, H. (1991 b). Bucilluriophyceae 4. Achnanthaceae. Kritische Ergdzungen zu Navicula (Lineolatae) und Gomphonema. Szibwasserjora volt Mitteleupora, eds H. Ettl, J. Gerloff, H. Heynig & D. Mollenhauer. G. Fisher Verlag, Stuttgart, Germany. Lange-Bertalot, H. (1979). Pollution tolerance of diatoms as a criterion for water quality estimation. Nova Hedwig., 64, 285-304. Lecointe, C., Coste, M. & Prygiel, J. (1993). ‘OMNIDIA’ software for taxonomy, calculation of diatom indices and inventories management. Hydrobiologiu, 269-270, 509-5 13. Lowe, R. L. & Gale, W. F. (1980). Monitoring river periphyton with artificial benthic substrates. Hydrobiologiu, 69, 235-244. Merlin, G., Clement, B., Mermillod, A. C., Yosowigdado, R., Mouriot, D. & Blake, G. (1995). Determination of long term effects of an herbicide on aquatic ecosystems, using outdoor microcosms. In Environmental Science and Vulnerable Ecosystems. SETAC, Brussels, Belgium, p. 165. Odin, M., Ribeyre, F. & Boudou, A. (1995). Cadmium and methylmercury bioaccumulation by nymphs of the burrowing mayfly Hexagenia rigida from the water column and sediment. Environ. Sci. Pollut. Res., 2, 145-152. Patrick, R. (1978). Effects of trace metals in the aquatic ecosystem. American Sci., 66, 185-l 9 1. Pielou, E. C. (1966). The measurement of diversity in different types of biological collections. J. Theor. Biol., 13, 131-144. Ponte-Freitas, A. L., Haddad, G., Tissut, M. & Ravanel, P. (1991). Distribution of Isoproturon, a photosystem II inhibitor, inside wheat leaf fragments. Plant Physiol. Biochem., 29, 67-74. Prygiel, J. & Coste, M. (1993) The assessment of water quality in the Artois-Picardie basin (France) by the use of diatom indices, Hydrobiologia, 2691270, 343-349. Rachlin, J. W., Jensen, T. E. & Warkentine, B. (1983). The growth response of the diatom Naviculu incerta to selected concentrations of the metals: cadmium, copper, lead and zinc. Bull. Tour. Bot. Club, 110, 217-223. Ribeyre, F. (1993). Evolution of mercury distribution within an experimental system ‘water-sediment-macrophytes (Elodea densa)‘. Environ. Technol., 14, 201-214.

152

F. P&s

Round, F. E. (1991). Diatoms in river water-monitoring studies. J. Appl. Phycol., 3, 129-145. Round, F. E., Crawford, R. M. & Mann, D. G. (1990). The Diatoms: Biology and Morphology of the Genera. Cambridge Univ. Press, Cambridge, UK. Scarponi, L. (1993). Metabolisme des pesticides dans les plantes. In Dkgradation des Pesticides dans l’environnement, ed. R. Belamie & C. Joubert. Groupe Fran@s des Pesticides, Lyon, France, pp. 120-135. Shannon, C. E. & Weaver, W. (1948). The Mathematical University of Illinois Press, Theory of Communication. Urbana, IL, USA. Sommer, U. (1981). The role of r- and k- selection in the succession of phytoplankton in lake Constance. Acta Oecolog., 2, 327-342.

Stevenson, R. J. (1983). Effects of current and conditions simulating autogenically changing microhabitats of benthic diatom immigration. Ecology, 64, 15 14-l 524. Symoens, J. J., Kusel-Fetzmann, E. & Descy, J. P. (1988). Algal communities of continental waters. In Vegetation of

et al. Inland Waters, ed. J. J. Symoens. Kluwer, Dordrecht,

The Netherlands, pp. 183221. Thielcke, A. & Ratte, H. T. (1994). The role of periphyton in primary production in outdoor microcosms treated with 3,4-dichloroaniline (DCA). In Freshwater Field Tests for Hazard Assessment of Chemicals, ed. I. R. Hill, F. Heimbath, P. Leeuwangh & P. Matthiessen. CRC Press, BocaRaton, FL, USA, pp. 369376. Van-Dam, H., Mertens, A. & Sinkeklam, J. (1994). A coded checklist and ecological indicator values of freshwater diatoms from the Netherlands. NetherlanA J. Aquat. Ecol., 28, 117-133. Venrick, E. L. (1978). How many cells to count? In Phytoplankton Manual, ed. A. Sournia. UNESCO, Paris, pp. 167180. Wetzel, R. G. (1964). A comparative study of the primary productivity of higher aquatic plants, periphyton and phytoplankton in a large, shallow lake. Znt. Rev. Gesamten Hydrobiol., 49, l-61. Wetzel, R. G. (1983). Periphyton of Freshwater Ecosystems. Junk, The Hague, The Netherlands.