Lemna minor and M. aquaticum were grown inside a controlled-environment room in open 500-ml beakers, in sterilized Steinberg medium 3 at 25 ± 2°C, illuminated with continuous cool white fluorescent lighting (100–135 µE · m−2 s−1 equivalent to 8,500–1,0000 lux, checked regularly with a Digital Lux meter LX 101OB, Sinometer Instruments). A parallel culture of M. aquaticum was cultivated on standard 32 artificial sediment (5% dried sphagnum peat, 74% quartz sand, 20% kaolin clay, and 1% CaCO3) in 500-ml flower pots without drain holes, and watered to saturation with semiconcentrated Steinberg solution (1:1 v/v Steinberg medium: distilled water).
Three tests were run nonsimultaneously: standard 7-d L. minor growth inhibition test in semistatic conditions with test solution renewal after 72 h; 10-d M. aquaticum growth inhibition semistatic test (test solution renewal every 72 h) in a sediment-free system; and 10-d M. aquaticum sediment contact test in a water-free system.
The L. minor test was run according to the standard protocol 3: Colonies consisting of two to four visible fronds were transferred from the inoculum culture to a Steinberg medium and randomly assigned to the test vessels (200-ml beakers with 150-ml test solution). Each test vessel contained a total of nine fronds, three replicates per treatment, and six per control. Total fresh weight of gently dried fronds was measured before placing the fronds into treatment and control vessels. Total frond area was determined in each control and treatment vessel at the beginning of the test by image analysis (digital photographs were analyzed with Adobe Photoshop CS3 software). A randomized design for location of the test vessels was applied to minimize the influence of spatial differences in light intensity and temperature. The fronds were then carefully moved to freshly prepared test solutions of corresponding atrazine concentration or pure Steinberg medium (controls) after 3 d. The number of fronds, total frond area, and total fresh weight per test vessel were counted and measured at the end of the test. Technical-grade atrazine (CAS 1912-24-9, purity 98%, manufacturer Oxon SpA) was provided by Galenika Fitofarmacija. A preliminary range-finding test was run in a logarithmic series of five nominal atrazine concentrations ranging from 0.1 to 10,000 µg/L in three replicates per treatment and three per control. A definitive test was set in the following nominal concentration series: 80, 160, 320, 640, and 1,280 µg/L as three replicates per treatment and six per control.
The M. aquaticum 10-d growth inhibition test in a water column (sediment-free system) was run in semistatic conditions. At the beginning of the test, the plants from 21 ± 3-d-old precultures cultivated in Steinberg medium were cut into the whorls needed for testing (two to four maximum per shoot). The whorls (the parts of a plant consisting of a nodium and two adjoining parts of internodi) used for the test showed no signs of side shoots. The cut whorls were then collected in a glass vessel in water for randomization. Before weighing, the whorls were dried gently. The fresh weight of each whorl was in the range of 25 ± 6 mg. Three whorls per replicate, three replicates per treatment, and six per control were placed into the test vessel (500-ml beakers with 250-ml test solution), while a thin (3 mm) polystyrene floating cover with holes was used to keep the plants in an upright position. The plants were transferred to freshly prepared, corresponding atrazine solutions on the third, sixth, and ninth day of the test. A preliminary, range-finding test was run in a logarithmic series of five nominal atrazine concentrations ranging from 0.1 to 10,000 µg/L; the definitive test was set in the nominal concentration series: 10, 20, 40, 80, and 100 µg/L. The test duration was 10 d, and exposure conditions were the same as for the L. minor test.
The M. aquaticum contact sediment test (in a water-free system) was run in static conditions according to Feiler et al. 12. Three whorls per replicate, three replicates per treatment, and six per control were planted in 80 g of spiked and control sediment per test vessel at premarked positions (1–3), closed with translucent lids with openings for aeration. During the exposure period the plants were watered with semiconcentrated Steinberg solution and test vessels were randomized every 48–62 h. Artificial sediment, 65 g, was spiked with a 15-ml solution of technical-grade atrazine in a preliminary, range-finding test with a logarithmic nominal concentration range of 0.1, 1, 10, 100, 1,000, and 10,000 µg/L, calculated to sediment concentrations of 0.000023, 0.00023, 0.0023, 0.0231, 0.231, and 2.31 µg/g. The nominal concentration range for the definite test was 0.5, 1, 2, 4, 8, 10, and 16 mg/L atrazine, which makes 0.12, 0.23, 0.46, 0.92, 1.85, 2.31, and 3.69 µg/g sediment.
The test duration was 10 d and laboratory conditions were the same as for the L. minor and M. aquaticum water-column (sediment-free) tests. The whole plants were weighed again after the test to calculate specific growth rate and inhibition and yield effect, but a number of other growth parameters were also recorded, such as total length of the shoots (the sum of the main and side shoots), total length of the roots (the sum of the main and side roots), total weight of the main and side shoots and the main and side roots, total content of chlorophyll a in shoots (as proxy to biomass), and concentration of chlorophyll a per gram of shoot. The root-to-shoot ratio based on length and weight was also calculated.
The experimental setup was similar to that described for the toxicity tests. The difference was that after 72 h of static exposure of L. minor and M. aquaticum, and after 7 d of static-renewal exposure in the second experiment with L. minor to a series of atrazine concentrations, the plants were transferred to clean Steinberg medium for the recovery phase of the experiment. In the recovery phase the medium was completely renewed every 72 h.
The experiments were run nonsimultaneously: the recovery test with L. minor after short-term exposure to atrazine (72-h exposure followed by a 6-d recovery phase in clean Steinberg medium); the recovery test with M. aquaticum after short-term exposure to atrazine (72-h exposure followed by a 12-d recovery phase); and the recovery test with L. minor after longer exposure to atrazine (7-d exposure followed by a 5-d recovery phase in clean Steinberg medium). Myriophyllum aquaticum was exposed for 72 h to the following range of nominal concentrations: 40, 80, 160, 320, and 640 µg/L of atrazine, while L. minor was exposed for 3 and 7 d to the following nominal atrazine concentrations: 80, 160, 320, 640, and 1,280 µg/L.
The average relative growth rate (RGR) in L. minor toxicity tests and recovery experiments was calculated on the basis of change in the logarithms of frond numbers, total frond area, and fresh weight over time (expressed per day) in the controls and each treatment group. In all the tests with M. aquaticum, relative growth rates for each plant were calculated from the measured total plant fresh weights to enable calculation of the arithmetic mean of the RGR per test and control vessel using the following equations
where RGRi-j is the average specific growth rate from time i to j, Ni is the measurement variable in the test or control vessel at time i, Nj is the measurement variable in the test or control vessel at time j, and t is the time period from i to j.
The percentage of inhibition of growth rate (Ir) was calculated for each test concentration (treatment group) according to the following equation
where %Ir is the percentage of inhibition in average specific growth rate, RGRc is the mean value for RGR in the control, and RGRr is the mean value for RGR in the treatment group.
The effect on yield was assessed on the basis of number of fronds, fresh weight biomass, and total frond area for L. minor test and total plant fresh weight for M. aquaticum tests, according to the following equation.
where %Iy is the percentage of reduction in yield, bc is the final biomass minus starting biomass for the control group, and bt is the final biomass minus starting biomass in the treatment group.
Chlorophyll a concentrations were measured at the end of each toxicity test and recovery experiment. Individual shoots in the case of M. aquaticum and total fronds from each test vessel in the case of L. minor were incubated in 7 ml methanol for 24 h in the dark. The samples were centrifuged for 15 min at 800 rpm and chlorophyll a was measured (Beckman DU-65 Spectrophotometer) in supernatant at two wavelengths, 653 and 666 nm.
The content of chlorophyll a – Ca in the extract (mg) was calculated as:
where A666 and A653 are the absorbance at 666 and 653 nm, respectively.
And as chlorophyll a concentration – C – mg/g of plant material as
where V (ml) is the solvent volume, m (g) is the plant material, and R is the dilution factor.
Analytical verification and water quality analyses
Temperature, dissolved oxygen content, oxygen saturation, electrical conductivity, and pH were measured electrochemically using the multiparameter instrument Ino Lab 3 (Wissenschaftlich–Technische Werkstatten), while biological and chemical oxygen demand, total organic carbon, total suspended solids, nitrates (NO3), and surfactants were analyzed using the portable multiparameter analyzer Pastel UV (Secomam).
Atrazine concentrations were checked in initial solutions of the M. aquaticum definitive test in water (sediment-free) system. The gas chromatography–mass spectrometry system (Agilent Technologies 7890A/5975C MSD) was used for quantification. Preconcentration of water samples was performed with Superclean ENVI-18 cartridges based on the operating procedure described by the U.S. EPA 33. The cartridges were conditioned with a 1:1 mixture of ethylacetate and methylene chloride, then with methanol, and finally with water. The components retained were eluted with ethylacetate and methylene chloride. Anhydrous sodium sulfate was added in eluates to remove residual water. The organic phase thus obtained was evaporated to dryness under a slow stream of nitrogen. The dry residue was dissolved in 1 ml of a 1:1 mixture of hexane:ethylene chloride. The sample of 1 µl was injected into a gas chromatography–mass spectrometry system (splitless) equipped with DB-5 MS column (30 m × 0.25 mm × 0.25 µm). The injector temperature was 250°C; the carrier gas was helium with a flow-rate of 1 ml/min. The initial temperature was 70°C, which was maintained for 2 min. The temperature was then increased to 150°C at 25°C/min, to 200°C at 3°C/min, and finally to 280°C at 20°C/min, which was maintained for 1 min. The selected ion monitoring mode was used for quantitative analysis of atrazine (m/z 200, m/z 215, m/z 73). The recovery of the method for atrazine analysis at the lower concentration level of 3 µg/L was 116% (RSD = 5.4%, n = 5) and for the higher concentration level of 90 µg/L it was 102% (RSD = 2.9%, n = 3). For nominal concentrations of 1, 91, and 100 µg/L, the measured values in initial solutions were 0.8, 103, and 96 µg/L, respectively. For initial solutions with nominal concentrations of 4 and 16 mg/L prepared for atrazine-spiked sediment contact test with M. aquaticum, significantly lower values were obtained by measurement: 2.1 and 5.4 mg/L, respectively.
Statistical analysis and presentation of the data
Mean, standard deviation (SD), and coefficient of variation (CV, %) were calculated for specific growth rates, yield, and concentration of chlorophyll a per test treatment in all toxicity tests and recovery experiments. No observable effect concentrations (NOEC) and lowest observable effect concentrations (LOEC) were calculated by comparing the treatments and controls using Dunnett's procedure, which integrates one-way analysis of variance (ANOVA) and t test (or t test with Bonferroni's adjustment in case of nonequal number of replicates per treatment) as a post-hoc test 34. The significance was assigned uniformly at p = 0.05. Statistical power of the test was assessed by the minimal significant difference (MSD) and sensitivity by the percentage of decrease compared to the corresponding control (MSD, %). Inhibitory concentrations causing 50 and 25% inhibition versus control (IC50 and IC25) based on both growth rate and yield were calculated using the linear interpolation method. A linear interpolation method (method of curve fitting using linear polynomials, developed by the U.S. EPA) for calculating inhibition concentrations in sublethal toxicity tests has been incorporated into TesTox software 34 used in the present study.