ORGANIC POLLUTANTS: An Ecotoxicological Perspective - Chapter 14 doc

257 14 The Ecotoxicological Effects of Herbicides 14.1 INTRODUCTION Herbicides constitute a large and diverse class of pesticides that, with a few excep- tions, have very low mammalian toxicity and have received relatively little attention as environmental pollutants. Much of the work in the eld of ecotoxicology and much environmental risk assessment has focused on animals, especially vertebrate animals. There has perhaps been a tendency to overlook the importance of plants in the natural world. Most plants belong to the lowest trophic levels of ecosystems, and animals in higher trophic levels are absolutely dependent on them for their survival. The direct environmental effects caused by herbicides appear to have been very largely upon plants. Mostly, these have been on target weed species, but sometimes upon nontarget species as well, due to spray drift. A few have had toxic effects upon animals. Paraquat and other bipyridyl herbicides have appreciable mammalian tox- icity and have been implicated in poisoning incidents involving lagomorphs on agri- cultural land (see De Lavaur et al. 1973, Shefeld et al. 2001). The mode of action of these compounds was discussed in Chapter 13. Dinoseb and related dinitrophenols, which act as uncouplers of oxidative phosphorylation in mitochondria, are general biocides that are little used today because of their hazardous nature. The mode of action of these compounds was also discussed in Chapter 13. Also, the herbicide 2,4,5,T, a constituent of Agent Orange, has sometimes been contaminated with sig- nicant amounts of the highly toxic compound tetrachlorodibenzodioxin (TCDD) (see Chapter 7, Section 7.2.1). There have been some reports of herbicides disturbing the metabolism of micro- organisms in soil. For example, dichlobenil was shown to increase the rate of CO 2 formation from glucose in soil (Somerville and Greaves 1987). Herbicides are, in general, readily biodegradable by vertebrates and are not known to undergo substantial biomagnication in food chains. Their principal use has been for weed control in agriculture and horticulture, although they have also been employed as defoliants in forests (e.g., in the Vietnam War), controlling weeds in gardens, on roadside verges, and in water courses, and as management tools on estates and nature reserves. This chapter will be mainly concerned with their impact on the agricultural environment where they have been extensively used. Brief men- tion will be made of their wider dispersal in the aquatic environment. © 2009 by Taylor & Francis Group, LLC 258 Organic Pollutants: An Ecotoxicological Perspective, Second Edition 14.2 SOME MAJOR GROUPS OF HERBICIDES AND THEIR PROPERTIES The following brief account identies only major groups of herbicides not mentioned elsewhere in the text, and is far from comprehensive. Their mode of action is only dealt with in a supercial way. From an ecotoxicological point of view, there has not been as much concern about their sublethal effects upon plants as there has been in the case of mammals, and there has not been a strong interest in the development of biomarker assays to establish their effects. The major concern has been whether weeds, or nontarget plants, have been removed following herbicide application—a rather easy matter to establish as plants are fairly sedentary. For a more detailed account of herbicide chemistry and biochemistry, see Hassall (1990). The phenoxyalkane carboxylic acids are among the most successful and widely used herbicides. They act as plant growth regulators and produce distorted growth patterns in treated plants. Compounds such as 2,4-D, MCPA, and mecoprop (Figure 14.1) are used as selective herbicides to control dicotyledonous weeds in monocotyledonous crops such as cereals and grass. They are formulated as water- soluble potassium or sodium salts or as lipophilic esters, and are frequently sprayed in combination with other types of herbicides having different modes of action and patterns of weed control. They are applied to foliage, and are not soil acting. Ureides (e.g., diuron, linuron) and triazines (e.g., atrazine, simazine, ametryne) all act as inhibitors of photosynthesis and are applied to soil (see Figure 14.1 for structures). They are toxic to seedling weeds, which they can absorb from the soil. Some of them (e.g., simazine) have very low water solubility and, consequently, are persistent and relatively immobile in soil (see Chapter 4, Section 4.3, which also mentions the question of depth selection when these soil-acting herbicides are used for selective weed control). Sulphonylurea herbicides such as chlorsulfuron and sulfometuron are also soil acting, have effects upon cell division, and are highly phytotoxic. Indeed, they can be toxic to plants when present in soil at levels low enough to make chemical analysis difcult. Carbamate herbicides constitute a relatively diverse group. Some, such as barban (Figure 14.1), are applied to foliage, whereas others (e.g., chlorpropham) are soil acting. The latter type have effects upon cell division. Other important herbi- cides or groups of herbicides include glyphosate, aminotriazole, chlorinated benzoic acids (e.g., dicamba), and halogenated phenolic nitriles (e.g., ioxynil, bromoxynil). 14.3 IMPACT OF HERBICIDES ON AGRICULTURAL ECOSYSTEMS Since World War II, herbicides have come to be widely used in agriculture and hor- ticulture in the developed world. Frequently, they have been used in “cocktails” con- taining several ingredients of contrasting modes of action, thus giving control over a wide range of weed species. The effectiveness of the application of herbicides together with cultivation of the land is evident in many agricultural areas of Western Europe and North America, where few weeds are seen. It is easier to control plants, which are stationary, than to control mobile insect or vertebrate pests. By the same token, it is also easier to judge the population effects of control measures (e.g., use of © 2009 by Taylor & Francis Group, LLC © 2009 by Taylor & Francis Group, LLC The Ecotoxicological Effects of Herbicides 259 pesticides) than in the case of mobile animals. Weed species have been very effec- tively controlled over large areas of agricultural land. In Britain, concern has been expressed about the near extinction of certain once-common farmland species that are of botanical interest, for example, corn cockle (Agrostemma githago) and pheas- ants eye (Adonis annua). O 2,4-D MCPA Mecoprop CH 2 COOH Cl Cl O CH 2 COOH CH 3 Cl Cl Cl CH(CH 3 ) 2 SO 2 NH C Cl Chlorsulfuron (c) Sulfonyl urea O NH CH 3 CH 3 Cl 1 4 2 3 5 6 4 1 2 3 3 2 1 Cl O CHCOOH CH 3 Cl CH 3 N CO Cl NH OCH 3 CH 3 N Linuron Isoproturon CO NH Cl CH CH 3 CH 3 O Chlorpropham C O NH CH 3 CH 3 N Diuron CO NH N N 1 4 2 3 6 5 N OCH 3 CH 3 N N N N H H isoPr Atrazine N Et SCH 3 N N N N H H Et Ametryne N Et Cl N N N 4 6 3 1 2 5 H H N Et Simazine N Et Cl O CH 2 CH 2 Cl C O Barban 1' 1 2 3 4 2' 3' 4' C C NH (a) Phenoxyalkane carboxylic acids (b) Ureides (d) Triazines (e) Carbamates fIgure 14.1 Structures of some herbicides. 260 Organic Pollutants: An Ecotoxicological Perspective, Second Edition Ecologically, such a large reduction of weed species represents a major change in farmland ecosystems and may be expected to have knock-on effects on other species. Certain problems have come to light with the investigation of the status of birds on farmland. In one study, the Game Conservancy investigated the reasons for a severe and continuing decline of the grey partridge (Perdix perdix) on farmland in Britain. The study commenced in the late 1960s and established that the decline was closely related to increased chick mortality (Potts 1986, Potts 2000; also Chapter 12 of Walker et al. 2006). The high chick mortality was largely explained by a short- age of their insect food (e.g., sawies) due, in turn, to the absence of the weeds upon which the insects themselves feed. An effect at the bottom of the food chain led to a population decline further up. It is worth reecting that such an effect by herbicides could not have been forecasted by normal risk assessment (see Chapters 14 and 15). The herbicides responsible are in general of very low avian toxicity, and ordinary risk assessment would have declared them perfectly safe to use so far as partridges and other birds are concerned. Subsequent work has shown that partridge popula- tions can continue to survive on agricultural land if headlands are left unsprayed, thereby allowing weeds to survive, weeds that will support the insects on which young partridges feed. This study helped to ring the alarm bells about possible other indirect effects of the wide use of herbicides in agriculture. More recently, further evidence has been gained of the reduction in populations of insects and other arthropods on farmland that may relate, at least in part, to the removal of weeds by the use of herbicides. A study of farmland birds in Britain established the marked decline of several spe- cies in addition to the grey partridge, which may be the consequence of the indi- rect effects of herbicides and other pesticides (Crick et al. 1998, and Chapter 12 of Walker et al. 2006). Species affected include tree sparrow (Passer montanus), turtle dove (Streptopelia purpur), spotted ycatcher (Musciapa striata), and skylark (Alauda arvensis). A study is currently in progress to attempt to establish the cause of these declines. Recently, controversy about the possible side effects of herbicides used on agri- cultural land has intensied with the development of genetically modied (GM) crops. Some GM crops are relatively insensitive to the action of herbicides, thus permitting the application to them of unusually high levels of certain herbicides. The advantage of increasing the dose, from the agricultural point of view, is the control of certain difcult weeds. From an ecotoxicological point of view, though, increas- ing dose rates of herbicides above currently approved levels raises the possibility that this may cause undesirable ecological side effects. It is very important that any such change in practice is rigorously tested in eld trials as part of environmental risk assessment before approval for marketing is given by regulatory authorities. Such new technology, based on GM crops, should only be introduced if it is shown to be environmentally safe. One problem that has arisen with the use of herbicides in agriculture is spray or vapor drift. When ne spray droplets are released, especially if applied aerially, they may be deposited beyond the target area due to air movements to cause damage there. In the rst place, this is a question of application technique. Herbicides, like other pesticides, should not be applied as sprays under windy conditions. In most © 2009 by Taylor & Francis Group, LLC The Ecotoxicological Effects of Herbicides 261 situations, herbicides are not applied aerially because of the danger of drift. Where herbicides have appreciable vapor pressure, there may be problems with vapor drift. Under hot conditions, volatile herbicides may go into the vapor state, and the vapor may drift farther than the spray droplets. This happened with early volatile ester formulations of phenoxyalkanecarboxylic acids (Hassall 1990). Nowadays, formula- tions are of less volatile esters, or as aqueous concentrates of Na or K salts, which are of low volatility. Spray drift of herbicides can result in damage to crops and wild plants outside the spray area. The cause of such damage can be hard to establish with highly active herbicides (e.g., sulfonylureas) where their phytotoxic concentrations are low enough to make chemical detection difcult. 14.4 MOVEMENT OF HERBICIDES INTO SURFACE WATERS AND DRINKING WATER As discussed earlier (Chapter 4, Section 4.2), pesticides have a very limited tendency to move through soil proles into drainage water because of the combined effects of adsorption by soil colloids (important for herbicides such as simazine, which have relatively high K ow ), metabolism (important for water-soluble and readily biodegrad- able herbicides such as 2,4-D and MCPA), and in some cases volatilization. In real- ity, however, there are complications. In the rst place there may be runoff from agricultural land into neighboring water courses after heavy rainfall. Soil colloids with adsorbed herbicides can be washed into drainage ditches and streams. There is an additional problem with certain soils high in clay minerals (Williams et al. 1996, and Chapter 4, Section 4.2). During dry periods these soils shrink and develop deep cracks. If heavy rains follow, free herbicides located near the soil surface and col- loids contaminated with adsorbed herbicides can be quickly washed down into the drainage system without passing through the soil prole. In the Rosemaund experi- ment, the herbicides atrazine, simazine, isoproturon, triuralin, and MCPA were all detected in drainage water following heavy rain. The respective maximum con- centrations in μg/L (ppb) were 81, 68, 16, 14, and 47 (Williams et al. 1996). These levels were reached following normal approved rates of application of the herbicides and raise questions about possible effects on aquatic plants growing in receiving waters. As mentioned elsewhere (Chapter 10, Section 10.3.4), the level of carbofuran found during the same study was sometimes high enough to kill freshwater shrimp (Gammarus pulex) used as a monitoring organism (Matthiessen et al. 1995). Since this study was undertaken, surveys have been carried out that provide more information on the levels of herbicides in British rivers. In one study, a number of different herbicides were detected in the Humber rivers (House et al. 1997). Several triazines were found in the rivers Aire, Calder, Trent, Don, and Ouse, the most abun- dant of them being simazine and atrazine. The results for simazine showed peaks in spring and again in early autumn of 1994 for the Trent and Aire, the autumn peak coinciding with the rst major storm of the year (Figure 14.2). The maximum level recorded for atrazine was > 8 μg/L in the river Calder in spring 1994. This was high enough to be toxic to phytoplankton and algae but was not sustained. It was not regarded as high enough to be toxic to aquatic invertebrates or sh. Phenyl ureas and © 2009 by Taylor & Francis Group, LLC 262 Organic Pollutants: An Ecotoxicological Perspective, Second Edition phenoxyalkanoic acids were also detected. Concentrations were generally low, but levels of the following herbicides were detected up to maximum values (μg/liter), which are given in parentheses. Diuron (< 8.7) Chlortoluron (< 0.67) Mecoprop (< 8.2) 350 300 200 Discharge (m 3 /s) Discharge (m 3 /s) 250 150 100 50 11/04/95 20/02/95 01/01/95 12/11/94 23/09/94 04/08/94 15/06/94 26/04/94 07/03/94 0 Date (a) 11/04/95 20/02/95 01/01/95 12/11/94 23/09/94 04/08/94 15/06/94 26/04/94 07/03/94 Date (b) 0 0.50 0.40 0.30 0.20 0.10 0.45 0.35 0.25 0.15 0.05 0 0 50 100 150 200 250 0.1 0.2 0.3 Concentration (μg/L) Concentration (μg/L) 0.4 0.5 0.6 FIGURE 14.2 Atrazine levels in the Humber River area. Comparison of the concentration of simazine and river discharge over one annual cycle for (a) River Trent at Cromwell Lock, and (b) River Aire at Beale. S, Simazine concentration; D, river discharge. (From House et al. [1977]. With permission.) © 2009 by Taylor & Francis Group, LLC The Ecotoxicological Effects of Herbicides 263 These high levels were sporadic and transitory. However, some of them were high enough to have caused phytotoxicity, and more work needs to be done to establish whether herbicides are having adverse effects upon populations of aquatic plants in areas highlighted in this study. It should also be borne in mind that there may have been additive or synergistic effects caused by the combinations of herbicides found in these samples. For example, urea herbicides such as diuron and chlortoluron act upon photosynthesis by a common mechanism, so it seems likely that any effects upon aquatic plants will be additive. Similarly, simazine and atrazine share a com- mon mechanism of action. With the acceptable concentrations of herbicides in drinking water being taken to very low levels by some regulatory authorities (e.g., the EC), there has been inter- est in very low levels of atrazine present in some samples of groundwater and in drinking water. This nding illustrates the point that mobility of pesticides becomes increasingly evident as sensitivity of analysis improves. 14.5 SUMMARY Contamination of agricultural land by herbicides is an example of the complexity of pollution in the real world. A wide variety of compounds of diverse structure, chemi- cal properties, and mechanism of action are used as herbicides. Important groups of herbicides include phenoxyalkane carboxylic acids, ureides, triazines, and carba- mates. They are often applied as mixtures of compounds with contrasting properties and modes of action. Very few of them have appreciable toxicity to animals, so initial toxic effects are mainly restricted to plants. Being generally biodegradable, they do not usually undergo signicant biomagnication with movement along food chains. The successful use of herbicides and associated cultivation procedures has greatly reduced the populations of weed species in many agricultural areas, sometimes bringing species of botanical interest to near extinction. Intensive weed control in cereal farming has been shown to cause the reduction of certain insect populations, and consequently also of the grey partridge, whose chicks are dependent on insect food. The reported decline of some other insectivorous birds on agricultural land in Britain may have a similar cause. The introduction of GM crops with high tolerance to herbicides may lead to increases in dose rates of herbicides on agricultural land with attendant ecotoxicological risks. Signicant levels of herbicides have also been detected in rivers, although these are usually transitory. Heavy rainfall can move herbicides from agricultural land to nearby ditches and streams due to runoff, and in soils that are high in clay, percola- tion of water occurs through deep ssures with consequent movement into neigh- boring water courses. Such events under extreme weather conditions are likely to have contributed to the pulses of herbicide contamination observed in some rivers. Questions have been asked about possible effects of such episodic pollution on popu- lations of aquatic plants. © 2009 by Taylor & Francis Group, LLC 264 Organic Pollutants: An Ecotoxicological Perspective, Second Edition FURTHER READING Ashton, F.M. and Crafts, A.S. (1973). Mode of Action of Herbicides—Describes the mode of action of major types of herbicides. Hassall, K.A. (1990). The Biochemistry and Use of Pesticides—Includes a readable account of the biochemistry of herbicides. Potts, G.R. (1990). The Partridge—An authoritative account of the factors responsible for the decline of the grey partridge on agricultural land. Williams, R.J. et al. (1996). Report of the Rosemaund Study—Describes the movement of herbicides through cracks and ssures in heavy soil into neighboring water courses fol- lowing heavy rainfall. © 2009 by Taylor & Francis Group, LLC . pollution on popu- lations of aquatic plants. © 2009 by Taylor & Francis Group, LLC 264 Organic Pollutants: An Ecotoxicological Perspective, Second Edition FURTHER READING Ashton, F.M. and Crafts,. invertebrates or sh. Phenyl ureas and © 2009 by Taylor & Francis Group, LLC 262 Organic Pollutants: An Ecotoxicological Perspective, Second Edition phenoxyalkanoic acids were also detected environment. © 2009 by Taylor & Francis Group, LLC 258 Organic Pollutants: An Ecotoxicological Perspective, Second Edition 14. 2 SOME MAJOR GROUPS OF HERBICIDES AND THEIR PROPERTIES The following