WFL Publisher Science and Technology

Journal of Food, Agriculture & Environment Vol.4 (1): 175-179. 2006

www.world-food.net

Ultrastructural modifications in sunflower (Helianthus annuus L.) chloroplasts resulting from the mixture of the herbicides metribuzin and clomazone Nelson Diehl Kruse1*, Ribas Antonio Vidal2 and Michelangelo Muzel Trezzi3 1 Plant Protection Department,Federal University of Santa Maria, Santa Maria, RS, Brazil 97105-900. 2Crop Science Department, Federal University of Rio Grande do Sul, Porto Alegre, RS, Brazil. 3Crop Science Department, CEFET, Pato Branco, PR, Brazil. *e-mail:[email protected], [email protected], [email protected]

Received 5 September 2005, accepted 20 November 2005.

Abstract The effects of the mixture of two or more herbicides can be investigated at the ultrastructural modifications of chloroplasts level. An experiment was carried out with the application of the herbicides metribuzin at 25 g ha-1 and clomazone at 275 g ha-1 separately and in mixture on sunflower plants in post emergence. The material collected from the leaves was prepared to obtain electron micrographs with a transmission electronic microscope. The length, width, chloroplast area and grana-thylakoid system area measurements were obtained from the images. The analysis of the electron micrographs, the length per width ratio and relation grana-thylakoid area per chloroplast area showed that the herbicides applied separately produced moderate damage in the chloroplasts. The herbicides mixture produced severe damage, destructed the grana-thylakoid and ruptured the plastid. The mixture of these herbicides presents a potentialized action compared to the action of each herbicide applied separately. Key words: Herbicide mixture, metribuzin, clomazone, chloroplasts, thylakoid, electron micrographs, oxidative stress and synergism.

Introduction The modifications at the ultrastructural level in the cell, their chloroplasts, mitochondria and other organelles caused by herbicides have been investigated. Such studies have helped to gain knowlegde and improve our understanding about herbicide mechanism of action. The anatomic alterations observed using an electronic transmission microscope completed and confirmed the studies carried out at the biochemical and physiological level. Similarly, the interactions resulting from the mixture of two or more herbicides can be better understood when the assessments include studies of the ultrastructural modifications, such as shown in various studies on the herbicide action mechanism 4, 24, 25, 28. Chloroplast shape and size vary according to plant age, metabolic activity and environmental conditions. The chloroplast is covered by a double membrane called envelope that defines the inside content. The envelope contains the stroma, the plastid matrix that involves the membrane system, the thylakoids, that house the light and photosystems I and II and the set of thylakoids that form the stacked vesicules, the grana 1, 24, 28. The chloroplast structure can be significantly altered by herbicide action. Studies with herbicides that inhibit the photosystem II (PS II) show that the damage in cells can follow a progressive scale of modifications, depending upon the herbicide, the rate and the time lapse between application and sampling. An initial change was observed in the elliptic or discoid shape of the chloroplast to a spherical shape. Further on the starch granules deposited in the stroma and the plastoglobules, set of granules found in the plastids that contain mainly lipids, disappear. The following effect involves the disintegration of the interlamellar grana-thylakoid, with the disorganization of the grana arrangement, followed by swelling of the internal compartment of the grana-thylakoid vesicles, culminating with the entire Journal of Food, Agriculture & Environment, Vol.4 (1), January 2006

disintegration of these expanded compartments and the chloroplast envelope 1, 4, 20, 24, 25, 28. Research involving detailed studies on the effect caused by herbicides have used electronic microscopy techniques. Electron micrographs of rice grass chloroplast treated with atrazine, granathylakoid degradation by the treatment of various species with amitrole, and the description of the mechanisms of action and selectivity of various PS II inhibitors are examples of the use of these techniques in the understanding of herbicide action 1, 4, 22, 24, 25, 28 . The objective of this study was to quantify the effects of mixtures of a PS II inhibitor herbicide (metribuzin) and a carotenoid synthesis inhibitor herbicide (clomazone) at the ultrastructural level. The hypothesis was that the mixture of these two herbicide is synergistic, i.e., the effect is greater than the sum of isolated effects. Materials and Methods Hybrid DKB 11 (Monsanto do Brasil) sunflower plants were grown up to the three pairs of true leaves stage in 25 cm x 40 cm x 10 cm trays containing sieved soil. The physico-chemical analysis of the soil used indicated 290 g kg-1 clay, 19 g kg-1 organic matter, 25 mg L-1 P, 123 mg L-1 K and pH 6.0 (H2O:soil 1:1 vol by wt). The herbicide treatments were applied over trays in post emergence (three replications), metribuzin at rate 25 g ha-1, clomazone at 275 g ha-1 and mixture of both at these same rates. These herbicide rates were chosen to provide damage of around 50% of the sunflower plants in the separate applications of each herbicide. This approach allowed for comparison of the effects caused by the association of the products. Paraffin mineral oil 4 0.5% (by vol) was added to all applications. The application was 175

made at the end of the day with a CO2 pressurized sprayer using a bar with four flat fan nozzles 8002 XR, spaced 50 cm and volume output equivalent to 200 L ha-1. Air temperature was 27°C and relative humidity was 67%. During the three days after application, sunshine duration was 7 h 30 min, 6 h 54 min and 5 h 6 min, respectively, with photosynthetic photon flux density measured with a Li-Cor 1600 porometer of 500 µmol m-2 s-1 (average of three days). Plants remained in plastic greenhouse when the average of maximum temperature was 30°C, average minimum temperature was 18°C. When the first visible herbicide symptoms appeared on the leaves, three days after application, leaves were randomly collected at the end of the day. Leaf cubes approximately 0.5 mm x 0.5 mm x 1.0 mm were sampled from each treatment and were immersed in primary fixing solution, a mixture of 2.5% (by vol) glutaraldehyde and 2% (by vol) formaldehyde in 0.1 M sodium phosphate buffer pH 7.2 15 at room temperature, for 24 h. The samples were washed in phosphate buffer of the same osmolarity and pH (3 hour-long washings) and transferred to a mixture (1:1 by vol) of 2% (by vol) osmium tetroxide and 0.8% (by vol) potassium ferricyanide (K3Fe(CN)6) in 1 M sodium phosphate buffer, pH 7.2 6 for 12 h in the dark at room temperature. After washing in distilled water for one hour, the material was dehydrated in increasing series of acetone (30, 50, 70, 90 and 100% (by vol), 15 min each) with the addition of 2% (by vol) uranium acetate in 70% (by vol) acetone at room temperature. Spurr low viscosity epoxi resin 27 was used as blocking medium. The pure resin was added drop by drop every 10 min to acetone containing the dehydrated samples until a 1:1 (by vol) mixture was obtained. Samples were kept in this mixture for 24 h in a rotating mechanical agitator. Samples were then transferred to a Spurr resin and acetone mixture at a ratio of 3:1 (by vol) for 24 h and finally to pure resin for 72 h under constant agitation. Samples were polymerized in pure resin inside gelatin capsules in an oven at 70°C for 18 h. Ultrafine 70 to 90 nm thick sections were cut (Leica Ultracut UCT ultramicrotome) and the sections were mounted on copper grids with a single oval hole (1 mm by 2 mm), previously covered with Butvar-B-98 film (terpolymer of polyvinyl butyl, polyvinyl alcohol and polyvinyl acetate; 0.25% (by vol) in chloroform). Sections were contrasted with a modified Thiery reaction 6, 29 followed by lead solution 19 for 5 min. The electron

micrographs were obtained with an transmission electron microscope (TEM - Jeol JEM 1200 EX-II) at 80 kV. Chloroplast length, width and area, and grana-thylakoid system area were measured using the computer software “Jandel SigmaScan” on the images obtained from the negatives. This approach allowed for selection of the grana-thylakoid system by difference grey color intensity between the system and the chloroplast stroma. Chloroplast length per width ratio, and the percentage of grana-thylakoid area per chloroplast area were calculated. Three images were obtained from treatments from prepared ultrafine sections. Statistical analysis consisted of analysis of variance and the means comparison test (LSD test, at 5% level of significance). The variable “grana-thylakoid system area” was transformed by

x + 1 7.

Results The electron micrographs of the sunflower leaf cells with no herbicide treatment are shown in Fig.1. Fig. 1A shows a cell with a chloroplast with elliptic or disk-shaped form, surrounded by several mitochondria. The grana-thylakoid system, the stroma that surrounds this system and the envelope, and the membrane that defines the chloroplast, are evident. These forms correspond to chloroplasts in fresh tissue with normal photosynthesis and metabolic activity 3, 21, 23, 26. Fig. 1B shows a chloroplast with the membrane system, the thylakoids, which in the shape of stacked vesicles form the grana, joined together by the interlamellar thylakoids or stroma thylakoids. Two starch granules are also visible and so are several plastoglobules, indicating that collected sunflower leaves were at high photosynthesis and metabolic activity. The large number of mitocondria in the cell (Fig. 1A) reinforces this high photosynthesis and metabolic activity 1, 3, 26. Another important aspect in Fig. 1 is the shape of the chloroplasts. Both are elliptic or disk-shaped, which is referred to in the literature as the expected shape of chloroplasts in photosynthetic active tissues 5, 25, 26. Chloroplasts from leaves treated with metribuzin are shown in Fig. 2. This treatment, with 25 g ha-1 metribuzin, caused visible ultrastructural changes. The electron micrograph shows a cell where several chloroplasts have modified shape, approaching more to a circular or spherical than a discoid shape (Fig. 2). The grana-thylakoid system spatial arrangement was also altered,

Figure 1. Electron micrographs of sunflower (Helianthus annuus) leaf cells; A: C = chloroplast, M = mitochondria; B: E = envelope, G = stacked grana, GA = starch granules, PG = plastoglobules, S = stroma, T = thylakoid. UFRGS, Porto Alegre, RS, 2001. 176

Journal of Food, Agriculture & Environment, Vol.4 (1), January 2006

becoming concentric (Fig. 2), very different from the placing seen in the chloroplasts without herbicide treatment in Fig. 1. The treatment with 275 g ha-1 clomazone also affected the chloroplast shape, from discoid to spherical or regular (Fig. 3). Starch granules and the plastoglobules were not present, which means a more intense effect than the treatment with metribuzin. This is confirmed with modified shape of the grana-thylakoid system (Fig. 3) when compared with the treatment without herbicide (Fig. 1) and in relation to the metribuzin treatment (Fig. 2). The membranes assumed more linear than concentric orientation, located in parallel to only one plane. More pronounced damage and near entire plastid destruction were detected in the treatment with the mixture of metribuzin and clomazone herbicides (Fig. 4). The chloroplasts of the leaves treated with this mixture were also altered in shape, from spherical to circular or irregular. Chloroplasts did not present starch granules or plastoglobules, and their grana-thylakoid system was completely destroyed (Fig. 4). The length and width of the chloroplasts are a quantitative description of the organelle (Table 1). For plants non-treated with herbicides, the average length of the chloroplasts was much greater than their width, characterizing their elliptic shape. The length per width ratio expresses the differences between the two dimensions and is a good measure of the herbicide treatment effect on the chloroplast shape. In untreated plants, the length was, on the average, 4.33 times greater than the width (Table 1). Analyzing the treatments together in Table 1, the length and width do not differ among the treatments. The variation among the chloroplasts produced high coefficients of variation for the variables, which certainly influenced the F value. On the other hand, when these dimensions were expressed as a ratio, variation was reduced and the alteration in the chloroplast shape was evident. The herbicide treatments did not differ in this ratio, as the alteration in the chloroplast shape is noticed even at low rates. In other words, the discoid shape is one of the first alterations noticed and even greater damage will not alter the chloroplast shape more, unless when there is complete destruction. The chloroplast area is a variable related to the chloroplast size that, although quite variable, allows a comparison with the alterations that the herbicide may cause (Table 2). It was not observed significant differences among treatments. If there is no plasmalemma rupture, the cell size will not change in response of herbicide treatment. In this study, the evaluations were performed during a period immediately before the complete tissue

Figure 2. Electron micrograph of sunflower (Helianthus annuus L.) leaf cells after treatments with the herbicide metribuzin (25 g ha-1) ; C = chloroplast. UFRGS, Porto Alegre, RS, 2001.

Figure 3. Electron micrographs of sunflower (Helianthus annuus L.) leaf cells after treatments with the herbicide clomazone (275 g ha-1). UFRGS, Porto Alegre, RS, 2001.

destruction. The area of the grana-thylakoid system can be serving as criterion to assess the herbicide effect on the damage to the grana-thylakoid membranes (Table 2). Only the treatment with mixture of metribuzin and clomazone statistically differed from the others, indicating the complete destruction of the membrane system of organelle. Although there was no statistic difference among the treatments without herbicide, metribuzin at 25 g ha 1 and clomazone at 275 g ha-1, it can be observed a decrease in grana-thylakoid system area in response to herbicide treatments.

Figure 4. Electron micrographs of sunflower (Helianthus annuus L.) leaf cells after treatment with the mixture metribuzin (25 g ha-1) plus clomazone (275 g ha-1) herbicides; A: C = chloroplast; RE = ruptured envelope; B: M = mitochondria. UFRGS, Porto Alegre, RS, 2001. Journal of Food, Agriculture & Environment, Vol.4 (1), January 2006

177

Table 1. Mean length and width, and length per width ratio of sunflower chloroplasts in response to the application of metribuzin and clomazone herbicides separately and in mixture. UFRGS, Porto Alegre, RS, 2001.

stress. The metribuzin rate should then be sufficient to produce more reactive chlorophyll than the carotenoids can neutralize. Damage is partial when the rate used is lower than that necessary to Length Width Treatment Length per width ratio overcome carotenoid neutralization (µm) (µm) capacity 9-11 as confirmed with in the effects Treatment without herbicide 5.3 1.2 4.33 a* -1 of metribuzin applied separately in this Metribuzin 25 g ha 3.6 2.2 1.10 b study (Fig. 2 versus Fig.1). Clomazone 275 g ha-1 2.8 3.4 1.26 b Metribuzin + clomazone 3.2 2.9 1.12 b Several studies have confirmed alterations F test (treatments) 1.75 1.64 86.96 at chloroplast level in plants treated with Probability 0.235 0.256 < 0.0001 herbicides in a pattern similar to that LSD1 0,05 2.73 2.36 0.55 obtained here with clomazone (Fig. 3) 20, 22. 2 3 CV (%) 38.6 51.3 15.0 The greater intensity of the effects in the * Means followed by same letter within a column do not differ by LSD test at 5% significance level. LSD = least square difference. Coefficient of variation. treatment with clomazone in relation to metribuzin may be attributed to two causes. Table 2. Mean chloroplast area, grana-thylakoid system area and the grana-thylakoid First, the clomazone herbicide rate used area per chloroplast area (GTA per CA) ratio of sunflower chloroplasts in suggested different sunflower sensitivity response to the application of the herbicides metribuzin and clomazone to the products tested, and it was much separately and in mixture. UFRGS, Porto Alegre, RS, 2001. more sensitive to metribuzin than to clomazone12, 13, 17. The second cause is the Grana-thylakoid GTA per CA ratio Chloroplast clomazone mechanism of action. Treatment system area1 (µm 2) (%)2 area (µm 2) Although clomazone final action is similar Treatment without herbicide 6.3 1.56 a*(1.5) 24.3 a to that of metribuzin (oxidative stress) its Metribuzin 25 g ha-1 4.1 1.34 a (0.8) 19.4 ab initial action is quite different and consists 5.0 1.30 a (0.7) 14.7 b Clomazone 275 g ha-1 Metribuzin + clomazone 6.1 1.00 b (0.0) 0.0 c of inhibiting the carotenoid biosynthesis F test (treatments) 0.25 7.64 16.00 route. With the reduced carotenoid level, Probability 0.86 0.01 0.001 there is no neutralization of the reactive 3 LSD 0,05 6.72 0.27 8.55 4 4 chlorophyll normally formed (high CV (%) 66.5 11.1 31.1 luminosity and/or low temperature) and * Means followed by same letter within a column do not differ by LSD test at 5% significance level. Values transformed by x + 1 ; originals mean are in parenthesis. reactive forms of oxygen are produced Grana-thylakoid system area divided by chloroplast area times 100. causing lipid peroxidation and membrane LSD = least square difference. destruction. In the final analysis, the end Coefficient of variation. damage is the same. Herbicides can also present other effects, besides the fact that the enzyme inhibited A further analysis of damage to the grana-thylakoid membranes by clomazone has not been detected yet. Clomazone action also can be done by the grana-thylakoid area per chloroplast area includes the inhibition of chlorophyll phytilization, the last step (GTA per CA) ratio in Table 2. This percentage is a measure of the in chlorophyll synthesis, where the tetrapyrrole ring links to the net herbicide effect by normalizing the natural chloroplast size phytol tail, which causes its reduction or absence in the treated variation. In other words, chloroplasts that are smaller also have tissues (bleaching) 2, 8, 10. proportionally less area in the grana-thylakoid system, so the The alterations caused by clomazone treatment are situated at ratio corrects it. Results showed that metribuzin at 25 g ha-1 caused the start of the sequence of predicted ultrastructural alterations. minimum damage in the system, with GTA per CA ratio no This process may be explained by the fact that the rate used was statistically different from the treatments without herbicide. dimensioned not to cause total damage. The rate used was about Clomazone at 275 g ha-1 caused a damage statistically different four times less than the rate needed to destroy sunflower seedlings from the treatment without herbicide, but not different from the (unpublished data). Higher rates completely inhibited carotenoid metribuzin treatment. To the mixture of herbicides treatment the synthesis and the oxidative stress usually produced would GTA per CA ratio was zero, expressing destruction of the probably destroy completely the membranes. membrane system, statistically different from the others. The treatments with only clomazone or metribuzin caused effects that could be classified as moderate damage (Figs 2 and 3), based Discussion on the ultrastructural alteration sequence proposed by Ashton The damage caused by the metribuzin herbicide results from its and Crafts 24. The change from discoid shape to spherical shape, mechanism of action as PS II inhibitor. The link of this herbicide in both herbicide treatments confirms this statement (Table 1). to D1 protein, a component of the PS II, blocks the way where The grana-thylakoid system was also altered, although at a quinone B should link to give continuity to the electron flow magnitude that can be considered moderate, because in the among the photosystems 16. The interruption in the electron flow treatment with metribuzin at the GTA per CA ratio was equivalent generates reactive chlorophyll, a free radical form that accounts to that without herbicide and the clomazone treatment. for membrane peroxidation. The carotenoids produced by the More pronounced damage and near to complete plastid plant react with the reactive chlorophyll, dissipating its surplus destruction were detected in the treatment with the mixture of of energy in the form of heat, protecting the plant against oxidative 1

2

1 2 3 4

178

Journal of Food, Agriculture & Environment, Vol.4 (1), January 2006

metribuzin and clomazone herbicides (Fig. 4). The chloroplasts of the leaves treated with this mixture were also altered in shape, from spherical to circular or irregular (Table 1) and did not have starch granules or plastoglobules. Also their grana-thylakoid system was completely destroyed (Table 2). The next step in the progressive sequence of effects would be the envelope rupture and complete loss of cell function (Fig. 4). The membrane system called thylakoid and their stacked vesicles, the grana, were destroyed directly because of the mechanism of action of the herbicides. While metribuzin blocks the photosynthesis electron transport and produces highly reactive forms of oxygen, with peroxidation of the membrane lipids, clomazone inhibits the carotenoid synthesis that acts in the dissipation of the surplus energy that produces these free radicals11 . Therefore, the mixture of the herbicides metribuzin and clomazone at rates that separately only caused moderate damage is much more powerful with metribuzin producing reactive oxygen forms and clomazone reducing the natural defense of the plant to free radical formation. The metribuzin at 25 g ha-1 was insufficient to overcome the natural defenses when separate, but when in mixture with 275 g ha-1 clomazone, it overcame them because the latter acts to reduce these defenses. The reactive chlorophyll level caused by metribuzin before insufficient to generate fatal oxidative stress, now with fewer neutralizing carotenoids because of clomazone action, could destroy the grana-thylakoid system and begin rupture of the internal and external membranes (envelope) that cover the chloroplast. Our findings at ultrastructural level confirm the hypothesis that the joint action of PS II inhibitor herbicides and carotenoids inhibitor herbicides results in an effective synergistic interaction. Conclusions The mixture of herbicides metribuzin and clomazone produced severe damage, destroyed the grana-thylakoid and ruptured the plastid. The mixture of these herbicides has a greater action compared to the action of each herbicide separately. Acknowledgement We thank Dr. Nereu Streck and Dr. Dalvan Reinhert for reviews and useful comments to this paper. References 1

Molas, J. 2002. Changes of chloroplast ultrastructure and total chlorophyll concentration in cabbage leaves caused by excess of organic Ni(II) complexes. Environ. Exp. Bot. 47:115-126. 2 Buchanan,B.B., Gruissem, W. and Russell, L.J. 2001a. Biochemistry and Molecular Biology of Plants. Part 1: Compartments. American Society of Plant Physiologists, Rockville, MD, 1367 p. 3 Buchanan,B.B., Gruissem, W. and Russell, L.J. 2001b. Biochemistry and Molecular Biology of Plants. Part 3: Energy flow. American Society of Plant Physiologists, Rockville, MD, 1367 p. 4 Fayez, K.A. 2000. Action of photosynthetic diuron herbicide on cell organelles and biochemical constituents of the leaves of two soybean cultivars. Pestic. Biochem. Physiol. 66:105-115. 5 Kim, S., Han, S. and vanden Born, W.H. 1997. Effect of chlorsulfuron on assimilate transport: ultrastructural implications. Weed Sci. 45: 470-473. 6 Weber, M. 1996. The existence of a special exine coating in Geranium robertianum pollen. Int. J. Plant Sci. 157:195-202. 7 Riboldi, J. 1995. Cadernos de matemática e estatística: análise de variância. Instituto de Matemática - UFRGS, série B, nº 27, 105 p. Journal of Food, Agriculture & Environment, Vol.4 (1), January 2006

8

Scott, J.E., Weston, L.A., Chappell, J. and Hanley, K. 1994. Effects of clomazone on ipp isomerase and prenyl transferase activities in cell suspension cultures and cotyledons of solanaceous species. Weed Sci. 42:509-516. 9 Devine, M., Duke, S.O. and Fedtke, C. 1993a. Physiology of Herbicide Action. Cap. 7: Herbicidal inhibition of photosynthetic electron transport, PTR Prentice Hall, Englewood Cliffs, NJ, 441 p. 10 Devine, M., Duke, S.O. and Fedtke, C. 1993b. Physiology of Herbicide Action. Cap. 8: Other herbicidal interactions with photosynthesis. PTR Prentice Hall, Englewood Cliffs, NJ, 441 p. 11 Devine, M., Duke, S.O. and Fedtke, C. 1993c. Physiology of herbicide action. Cap. 9: Oxygen toxicity and herbicidal action. PTR Prentice Hall, Englewood Cliffs, NJ, 441 p. 12 Johnson, D.H. and Talbert, R.E. 1993. Imazaquin, chlorimuron, and fomesafen may injure rotational vegetables and sunflower (Helianthus annuus). Weed Technol. 7:573-577. 13 Lyon, D.J. and Anderson, R.L. 1993. Crop response to fallow applications of atrazine and clomazone. Weed Technol. 7: 949-953. 14 Wrischer, M., Ljubesic, N. and Devide, Z. 1992. Ultrastructural studies of degradational processes in amitrole-damaged photosynthetic membranes. J. Struct. Biol. 108:1-5. 15 Roland, J.C. and Vian, B. 1991. General preparation and staining of thin sections. In Hall, J.L. and Hawes, C. (eds). Electron Microscopy of Plant Cells. Academic Press, Londres, pp. 1-66. 16 Tietjen, K.G., Kluth, J.F., Andree, R. et al. 1991. The herbicide binding niche of photosystem II - a model. Pestic. Sci. 31:65-72. 17 Halstead, S.J. and Harvey, R.G. 1988. Effect of rate and carrier on clomazone movement off-site. Weed Technol. 2:179-182. 18 Towers, G.H.N. and Arnason, J.T. 1988. Photodynamic herbicides. Weed Technol. 2:545-549. 19 Hanaichi, T., Sato, T., Iwamoto, T., Malavasiyamashiro, J., Hoshiro, M. and Mizuno, N. 1986. A stable lead by modification of Sato method. J. Electron Microsc. 35:304-306. 20 Bartels, P.G. 1985. Effects of herbicides on chloroplast and cellular development. In Duke, S.O. (ed.). Weed Physiology. V 2.CRC Press, Inc., Boca Raton, pp.63-90. 21 Freeman, T.P., Duysen, M.E. and Gulya, T.J. 1985. Ultrastructural changes in sunflower chloroplasts following inoculation with Pseudomonas syringae pv. tagetis. Am. J. Bot. 72:707-714. 22 Kenyon, W.H., Duke, S.O. and Vaughn, K.C. 1985. Sequence of effects of acifluorfen on physiological and ultrastructural parameters in cucumber cotyledon discs. Pestic. Biochem. Physiol. 24:240-250. 23 Fahn, A. 1982. Plant Anatomy. Cap.2: The Cell. 3 ed. Pergamon Press, Oxford, 544 p. 24 Ashton, F.M. and Crafts, A.S. 1981. Mode of action of herbicides. 2 ed. Wiley-Interscience, Nova York, 525 p. 25 Pallett, K.E. and Dodge, A.D. 1980. Studies into the action of some photosynthetic inhibitor herbicides. J. Exp. Bot. 123:1051-1066. 26 Casadoro, G. and Rascio, N. 1979. Plastid ultrastructural features in the various tissues of sunflower leaves. Cytobios 24:157-166. 27 Spurr, A.R. 1969. A low-viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res. 26:31-34. 28 Hill, E.R., Putala, E.C. and Vengris, J. 1968. Atrazine-induced ultrastructural changes of barnyardgrass chloroplasts. Weed Sci. 16: 377-380. 29 Thiéry, J.P. 1967. Mise en évidence des polysaccharides sur coupes fines en microscopie électronique. J. Microsc. 4:987-1018.

179