Introduction

Miracle Farm Blueprint

Organic Farming Manual

Get Instant Access

Modern agriculture relies heavily on herbicides for the control of weeds in crops and pastures to maximize yields and economical benefits to sustain an increasing world population. The introduction of herbicide-resistant traits in several crops, such as glyphosate-resistant (GR) soybean, maize and canola, has further increased herbicide consumption worldwide (Cerdeira & Duke, 2006). United States consumed roughly 200 million kg in 2001, with glyphosate representing 20 % of the total. Glyphosate is, undoubtedly, the most used herbicide worldwide (Woodburn, 2000). In Argentina, where GR soybean accounts for almost 90 % of planted soybean, it was estimated that 160 million l of glyphosate were used with this crop in 2004, representing 37 % of the total herbicide consumed in agriculture (Altieri & Pengue, 2006; Pengue, 2004).

The environmental fate of herbicides is a matter of recent concern given that only a small fraction of the chemicals reach the target organisms (Pimentel, 1995), leading to potential impacts of residual herbicides in soil and water have on human, animal and crop health. Bunce (1993) wrote in 1993 "It is useful to keep in mind the concept that a pollutant is a substance in the wrong place, at the wrong time, or in the wrong amount". While herbicides are very important to agriculture, under certain circumstances they may act as pollutants that can deteriorate soils, ground waters and surface waters. While most herbicides are not intentionally applied onto soil, they can enter the soil environment from 1) direct interception of spray by the soil surface during early season or post-harvest applications, 2) runoff of the herbicide from vegetation and 3) leaching from dead plant material. The herbicide concentration may vary from a few ^g to mg per kg soil, as most of the applied chemical is retained within the top 5 cm of soil. This chapter will present aspects of the behavior of herbicides in soils, focusing on soil retention and microbial degradation as main factors controlling persistence. The potential impact of herbicides on non-target soil microbes, including their processes and interactions, will be also discussed. Adsorption to soil is of critical importance for the regulation of herbicide persistence and mobility throughout the environment because sorption processes control the amount of herbicide present in the soil solution. These processes are dependent on several factors related to soil characteristics such as mineral composition, organic matter content, soil solution chemistry, and chemical characteristics of the herbicide. Soil-bound herbicide or residues are temporarily inactivated, which prevents harmful effects on soil biota but also makes them less bioavailable for microbial degradation because most microbes may not be able to utilize herbicides in the sorbed state (Ainsworth et al., 1993). Soil biochemical and biological processes are critical for ecosystems functioning, as microbes have key roles in organic matter transformations, nutrient cycling and degradation of organic pollutants, including pesticides (Beck et al., 2005). Biological degradation mediated by microbial enzymes is the main route for pesticides detoxification in soils (van Eerd et al., 2003). Most isolated herbicide-degrading microorganisms belong to bacterial species, but fungi are also well-known for their capacity to degrade complex substrates, and may be more important than present isolation approaches have suggested (Smith & Collins, 2007). Differential toxicity of herbicides to soil microorganisms may alter community structure, including potential increases in plant or animal pathogens. Herbicides may also cause changes in microbial community function and concomitant impacts on soil health and ecosystem processes. Even though functions may appear unaltered, due to species redundancy in soil, the extinction of resistant species may compromise the continuity of such processes.

The enormous variety of herbicides commercially available today makes it impossible to review all of them. Thus, this work will focus on some of the herbicides most used in the (semiarid) Pampa region of Argentina and worldwide (i.e., glyphosate, 2,4 dichlorophenoxyacetic acid, metsulfuron-methyl), based on our own research data.

2. Factors influencing the fate of herbicides in soil 2.1 Physicochemical interactions with soil

Soil is one of the main regulators of herbicide mobility in the environment. Many chemical and biological processes that determine the retention or transport of herbicides take place on the soil surface. These processes include adsorption phenomena, chemical degradation, and biological degradation. While all these processes are interrelated, occurring in parallel (Cheng, 1990), it is important to first understand adsorption since it regulates the bioavailability of herbicides in the environment, i.e. the ability to be used by microorganisms and thus be biodegraded (Laor et al., 1996; Boesten, 1993; Martins & Mermoud, 1998). Adsorption determines the quantity of herbicide that is retained on the soil surface and therefore is one of the primary processes that affect the transport of these compounds in soils. Thus to relate bioavailability and microbial ecology it is helpful to understand this primary process. Soils are complex assemblies of solids, liquids, and gases. A typical mineral soil contains 50% solid material (45% mineral and 5 % organic matter) and 50% pore space. The mineral particles in the soil are distributed into three sizes: sand, silt, and clay. Between the solid components of soil is space forming pores that plays a major role in movement of water, solutes and air. The adsorption processes depend fundamentally on the composition and properties of the solid component as well as the physicochemical characteristics of the herbicide. The solid component is formed mainly by primary and secondary minerals and by organic matter. These materials provide the specific sites for herbicide adsorption. Their properties and behavior have been treated extensively (Dixon & Weed, 1990; Greenland & Hayes, 1978).

Important characteristics of herbicides include: structure of the compound (including functional groups), water solubility, vapor pressure, octanol-water partitioning constant (Kow), and acidity. Table 1 shows the structures and some physicochemical properties of glyphosate, 2,4-D and metsulfuron-methyl.

The distribution of an herbicide in the soil depends on partitioning between the soil solution and the solid phase (Figure 1). The chemical is partitioned between the soil solution and the solid phase. The proper term for this process is adsorption equilibrium, which can be written to describe the interaction between any herbicide and any soil component as follows:

Where S represents a surface site of soil, H(aq) the herbicide in soil solution and SH the herbicide attached to the surface site. Surface sites where the herbicide can be adsorbed are numerous and varied in soils. These sites are provided by soil minerals (clays, Fe and Mn oxides, etc) as well as by organic matter. Equation (1) gives an idea of the general process involving adsorption, but it does not specify the mechanism by which it occurs, which are varied in the complex soil system (formation of surface complexes, electrostatic interactions, hydrophobic interactions, ion exchange, etc.).

Adsorption

Herbicide (H)

Soil solid phase (S)

Fig. 1. Distribution of an herbicide in soil

Herbicide (H)

Soil solid phase (S)

Adsorption

Fig. 1. Distribution of an herbicide in soil

Defining the bioavailability of an herbicide requires an understanding of the strength of its 1) interaction with a particular soil and 2) concentration of herbicide in the soil solution. This can be known by using adsorption isotherms. An adsorption isotherm shows the relationship between the herbicides concentration in the soil solution (C, correspond to H(aq) in Equation (1) and the amount adsorbed (q, correspond to SH in Equation (1)) at constant temperature and after equilibrium was reached (Stumm, 1992). As an example, Figure 2 shows adsorption isotherms of the herbicide metsulfuron methyl (MM) on different soils of the semiarid pampean region of Argentina (Zanini et al., 2009). Although isotherms with 30 different soils were measured in that study, the figure presents the results for three selected soils. These soils are characterized by having rather similar specific surface areas (SSA) and clay contents (% clay), but rather different total organic carbon (TOC) content. The physical and chemical characteristics of the 30 soils are shown in Table 2.

Herbicide

Chemical Structure

pKa

Water Solubility

(Pa)

Me

L-1 (pH9, 20°C)

1.8 (pH 5) 0.018 (pH 7) 0.0002 (pH9, 25°C)

OH

2.3 5.3 10.9

1.157 wt% in water at 25°C

-4.1

(Dichlorophenoxy) acetic acid (2,4-D)

O-CH,-COOH

Cl

(20°C-nonionised, est.)

3.2

1x10-5 (20°C)

Table 1. Structure and some physicochemical properties of the selected herbicides (Roberts, 1998).

Table 1. Structure and some physicochemical properties of the selected herbicides (Roberts, 1998).

As stressed by Sparks (2003), isotherms are only descriptions of macroscopic data and do not definitively prove a reaction mechanism. Mechanisms must be gleaned from molecular investigations, e.g. the use of spectroscopic techniques. However, the fit of experimental data with theoretical and/or empirical equations for adsorption isotherms is very useful in determining some parameters that provide information on the strength of soil-herbicide interaction, which will give an idea of the bioavailability of the herbicide in a particular soil. There are several adsorption isotherms equations applied to soils and sediments (Haws et al., 2006; Hinz, 2001). In this chapter, only the simplest and most widely applied equations are discussed.

2.1.1 Linear equation

The linear, or partitioning equation is expressed as (Pateiro-Moure et al., 2009; Cooke et al., 2004):

where K is the partition coefficient and q and C as defined above. The parameter K provides a measure of the ratio of the amount of material adsorbed to the amount in solution. The higher the value of Kd, the greater the affinity of the herbicide for the surface, resulting in lower bioavailability. The problem with the application of this equation is that linear behavior of the system in the range of concentrations of interest must be proved. If experimental data do not show a linear response in all the concentration range, the use of K¡ values obtained from linear regression will cause over- or underestimation of the true behavior in the non-linear ranges. Calculating K¡ with only a pair of values (C, q) may not be very useful to evaluate bioavailability across a range of environmentally relevant concentrations. It is recommended to perform an adsorption isotherm in the range of concentrations of interest, to test for linearity. Since adsorption of hydrophobic organic pollutants has been shown to be well correlated with the organic carbon content of soil and relatively independent of other soil properties, K¡ is sometimes expressed on the basis of TOC (Laor et al., 1996):

Q.Q1TOC

where TOC is expressed in % units.

Most experimental data do not respond to the linear equation; the most common models that describe non-linear adsorption isotherms are the Freundlich equation and the Langmuir equation.

115 it s

Fig. 2. Binding isotherms at pH=6 for samples with different TOC%. 4.02% (empty squares), 2.18% (filled circles), 0.98% (filled triangles). From Zanini et al. (2009).

2.1.2 Freundlich equation

The Freundlich equation is perhaps the most widely applied model in environmental soil chemistry to describe nonlinear sorption behavior (Valverde-Garcia et al., 1998; Kibe et al., 2000). It is an empirical adsorption model (Stumm, 1992; Sparks, 1986) and it can be written as:

Where, Kf is the distribution coefficient and n is a correction factor. The lines of Figure 2 have been drawn according to the Freundlich equation. The fitting parameters are present in Table 2 and discussed below. It is important to note that when n = 1 Equation (4) becomes Equation (2) and Kf = K;. In addition, when C is equal to unity the distribution coefficient gives the amount adsorbed at that concentration.

While researchers have often used the Kf and 1/n parameters to make conclusions concerning mechanisms of adsorption, and have interpreted multiple slopes from

Freundlich isotherms as evidence of different binding sites, such interpretations are speculative (Sparks, 2003). This is especially true in very complex and heterogeneous systems such as those formed by soil particles. For these systems, fitting of experimental data with isotherm equations should only be used for comparative purposes and to give some interpretation of the shape of the isotherms. Comparison of parameters should be performed with caution. It is necessary to be sure that the Kf values present the same units. The best way to avoid mistakes is to compare different sets of experimental data made under the same conditions and with isotherms performed in the same units of concentration. As in the case of the linear equation, parameters derived from the Freundlich equation should not be used to predict for behavior outside of the range of experimental data.

2.1.3 Langmuir equation

This model has been employed in many fields to describe sorption on colloidal surfaces (Zanini et al., 2006; Xi et al., 2010). The Langmuir adsorption equation can be written as:

Where KL is a constant related to the binding strength, b is the maximum amount of herbicide that can be adsorbed (monolayer coverage) and q and C were defined previously. This equation has several assumptions that Langmuir (1918) made in its development. Most of these assumptions are not valid for the heterogeneous surface found in soils. However, many researchers used this model to describe adsorption on soils (Gimsing et al., 2007; Ketelsen & Meyer-Windel, 1999). As with Kf above, KL is useful for comparative purposes but they do not provide information on reaction mechanisms. Some researchers fit the experimental data with both Langmuir and Freundlich equations to compare methodological approaches (Campbell & Davies, 1995; Martínez-Villegas et al., 2004).

2.2 Isotherm parameters and soil properties

In order to understand the bioavailability of an herbicide it is important to know the factors that affect its adsorption on soil. A good approach is to perform adsorption isotherms under different experimental conditions, and then relate the parameters of the isotherm to the soil properties. This will be demonstrated for a series of data on MM adsorption on the 30 different soils of the semiarid pampean region of Argentina listed in Table 2. As indicated above, the Freundlich equation was applied to this set of data. Table 2 shows the parameters Kf and n for all soils. All these soils are subject to similar farming practices (no till and production of the same kind of crops), thus the quality of the soil organic matter is expected to be similar, and the adsorptive differences among soils should be mainly given by differences in TOC. In most soils 1/n is lower than 1 and thus their isotherms are L shaped (Hunter, 2002) (Table 2). This kind of shape was also found by Pusino et al. (2003) for the adsorption of primisulfuron on soils, suggesting that the affinity of surface sites for MM is decreasing as the surface is becoming populated with MM. It may also suggest a decrease in vacant adsorption sites as MM concentration increases.

It is necessary to be careful when Kf values are compared. If the values of 1/n for the different soils are equal or similar, Kf values can be directly compared, and large Kf mean a strong herbicide-soil interaction. However, if the values of 1/n are rather different the comparison is not straightforward.

Values of Kf and 1/n were used to calculate the adsorption of MM at different equilibrium concentration for the 30 analyzed soils. From these calculations, plots relating adsorbed amounts with a given soil property can be constructed. For example, Figure 3 a shows the adsorbed amount at equilibrium concentration of 10 mg l-1 as a function of TOC. A positive and significant relationship between q and TOC is observed in the Figure. Although not shown here, this positive and significant relationship was found for all the studied concentrations (10, 20, 30 and 40 mg l-1). The results show that TOC is a very important factor that affects MM adsorption in the entire range of MM concentrations investigated. This is known for other herbicides (Kah & Brown, 2006; Weber et al., 2002). However, Cramer et al. (1993) found no clear relationship between adsortion of metsulfuron methyl and soil organic matter in Colorado soils, and the adsorbed amount showed only a weak correlation with organic matter content.

Soils

TOC %

Sand %

Clay %

SSA m2 g-1

PH

K f

1/n

R2

1

0.98

53.7

28.4

8.3

7.50

0.16 (0.01)a

0.95 (0.03) a

0.99

2

1.28

64.1

25.3

3.4

5.94

0.24

0.04)

0.91 (0.05)

0.99

3

1.29

51.9

38.9

12.0

6.51

0.22

(0.03)

0.98 (0.04)

0.99

4

1.40

48.5

33.9

5.2

6.05

0.54

0.08)

0.74 (0.05)

0.95

5

1.43

56.5

28.2

4.5

6.90

0.24

0.09)

1.02 (0.10)

0.96

6

1.44

57.4

33.4

4.9

6.05

0.61

0.06)

0.70 (0.03)

0.98

7

1.58

54.5

33.4

5.4

6.51

0.22

0.08)

0.86 (0.11)

0.94

8

1.76

44.7

41.8

7.3

6.19

0.78

(0.12)

0.55 (0.05)

0.97

9

1.82

45.8

39.8

8.4

6.55

0.36

0.07)

0.86 (0.05)

0.98

10

1.88

43.7

44.2

7.9

6.47

0.74

(0.16)

0.68 (0.07)

0.96

11

1.91

50.3

41.8

4.6

6.04

0.96

0.29)

0.40 (0.09)

0.86

12

1.94

42.7

39.4

4.6

6.46

0.54

0.08)

0.74 (0.05)

0.97

13

2.06

47.2

39.2

6.1

6.77

0.20

0.07)

1.02 (0.10)

0.97

14

2.07

44.0

38.3

5.8

6.30

0.53

(0.13)

0.88 (0.08)

0.98

15

2.10

46.4

38.9

4.8

6.51

0.31

0.07)

0.85 (0.07)

0.97

16

2.18

52.7

28.2

4.6

6.90

0.78

0.07)

0.69 (0.07)

0.98

17

2.46

52.2

35.9

3.7

6.14

0.76

0.27)

0.77 (0.10)

0.94

18

2.50

42.5

36.7

10.6

6.58

0.60

(0.12)

0.98 (0.06)

0.99

19

2.56

30.7

49.7

7.7

6.59

1.18

(0.12)

0.55 (0.06)

0.93

20

2.59

32.6

44.0

5.6

6.08

0.91

(0.12)

0.69 (0.04)

0.91

21

2.75

36.1

40.9

5.6

5.80

0.84

0.21)

0.65 (0.08)

0.97

22

2.88

43.0

35.4

5.2

6.44

0.56

(0.10)

0.92 (0.06)

0.98

23

2.93

31.2

48.6

5.2

7.80

0.36

(0.11)

0.80 (0.09)

0.96

24

3.07

47.2

33.7

4.3

6.36

1.07

0.25)

0.67 (0.07)

0.96

25

3.10

45.3

33.9

4.4

7.10

0.61

0.14)

0.69 (0.07)

0.98

26

3.34

50.4

30.6

9.0

6.51

0.85

(0.16)

0.89 (0.06)

0.98

27

3.91

46.5

31.2

5.5

6.10

1.10

0.32)

0.79 (0.09)

0.99

28

4.02

52.0

28.6

6.1

6.65

0.96

0.05)

0.83 (0.02)

0.99

29

4.62

61.4

25.3

4.0

5.86

0.80

(0.15)

0.93 (0.06)

0.99

30

4.85

44.3

32.8

6.0

8.03

1.35

0.22)

0.77 (0.05)

0.98

Values within brackets correspond to standard error.

Table 2. Selected physical and chemical properties of the studied soils.

Values within brackets correspond to standard error.

Table 2. Selected physical and chemical properties of the studied soils.

In order to investigate the effects of soil inorganic compounds on the adsorption of MM, the adsorbed amount was also plotted as a function of clay percent (Figure 3 b). There is no significant correlation indicating that inorganic compounds are not important on adsorption. The lack of interaction with inorganic compounds is not always the case for the adsorption of sulfonylureas. Pusino et al. (2003), for example, reported that inorganic solids such as amorphous Fe oxides and Al3+ and Fe3+ exchanged montmorillonites were active in adsorbing primsulfuron. The absence of important amounts of Fe oxides and smectites exchanged with trivalent cations in the studied soils (Blanco & Stoops, 1993) might explain the weak effect that inorganic components have on the adsorption of MM. The above discussion highlights the variable behavior of MM among soils. These differences may result from variation in the properties of the inorganic compound, organic matter, or other soil properties such as pH. Another important factor to take into account is pH, especially if the herbicide has acid or basic groups. Figure 4 shows the adsorption isotherms of MM on soil at pH 4, 6 and 8. MM adsorption decreases as the pH increases, in agreement with the general trend observed for sulfonylureas (Hay, 1990). This figure shows that changes in pH can affect the adsorption of MM. This behavior is usually explained in terms of charge development at the surface of soil particles and speciation of the herbicide in aqueous solutions as a function of pH (Berglöf et al., 2003). Since the surface charge of soil particles becomes more negative as the pH increases, the adsorption of the negatively charged MM species becomes less favored by increasing pH as a consequence of electrostatic repulsion. In addition, although the adsorption of the neutral MM species should not be affected by electrostatics, its concentration decreases with increasing pH, also causing less favorable adsorption with higher pH.

Fig. 3. (a) q (at 20 mg l-1 equilibrium concentration) as a function of TOC % for data at pH=6.(b) q (at 20 mg l-1 equilibrium concentration) as a function of the clay content for data at pH=6.

Fig. 3. (a) q (at 20 mg l-1 equilibrium concentration) as a function of TOC % for data at pH=6.(b) q (at 20 mg l-1 equilibrium concentration) as a function of the clay content for data at pH=6.

Fig. 4. Binding isotherms at (filled circle) pH=4, (empty square) pH=6, (filled triangle) pH=8. Lines have been drawn according to the Freundlich isotherm. From Zanini et al. (2009).

Fig. 4. Binding isotherms at (filled circle) pH=4, (empty square) pH=6, (filled triangle) pH=8. Lines have been drawn according to the Freundlich isotherm. From Zanini et al. (2009).

Some of the parameters obtained from adsorption isotherms are useful for an indirect estimation of the mobility of herbicides in soils. This can be obtained from the groundwater ubiquity score, GUS (Gustafson, 1989) defined as:

where GUS is a dimensionless index, ti/2 is the herbicide half-life in soil and Koc was defined previously (Equation 3). According to Oliveira Jr. et al. (2001) herbicides with GUS < 1.8 are ranked as non-leachers, those with GUS > 2.8 are leachers, whereas those with 1.8 < GUS < 2.8 are considered transitional. It must be remarked that t1/2 values should be measured for the specific soil under study because it may change greatly from soil to soil (Juhler et al., 2008; Bedmar et al., 2006).

In summary, it can be stated that adsorption depends on the physicochemical characteristics of the herbicide and the particular soil properties. In order to understand the processes that affect the bioavailability of an herbicide it is necessary to perform adsorption isotherms with the soils under study. Since adsorption and bioavailability change greatly from one soil to another, literature data can help to understand the general behavior of an herbicide, but they cannot give specific information about the behavior on a particular soil. Another important conclusion is that the mobility of an herbicide cannot be assessed only by knowing its physicochemical data. It is also very important to consider the presence of microorganisms in the soil system, as they can significantly affect the value of t1/2 in Equation 6.

2.3 Spatial distribution of microbial populations

As discussed earlier, soil is characterized by the heterogeneity in physicochemical and structural characteristics that provide many different micro-habitats for microbial life. The distribution of soil microorganisms varies both horizontally and vertically and from the micro site (few millimeters) to the regional (kilometers) scale. The abundance and diversity of most soil organisms are highest in the top 0-10 cm of soil and decline with depth in parallel with organic matter contents, the source of energy, nutrients and carbon (C) for the vast majority of soil microorganisms. Most microorganisms are located surrounding the water layer attached to soil particles within micro-aggregates. Pallud et al. (2004) estimated that soil bacteria occupy only 0.1 mm3 of the 500 mm3 of pores in 1 cm3 of soil. Moreover, different physiological groups present at the same density in soil might have significantly different microscale spatial distributions (e.g., forming clumps of cells in a few "hot spots" versus evenly spread out across soil particle surfaces).

Biodegradation of herbicides constitutes a clear example of the importance of understanding the spatial distribution of soil microorganisms. Spatial variability in both glyphosate

mineralization and general soil microbial characteristics, was observed even across small areas (decimeter scale) within a single field in two Norwegian sandy loam soils (Stenr0d et al., 2006), reflecting the importance of soil physicochemical parameters controlled by surface topography. Similarly, Vieuble-Gonod et al. (2005) reported that potential for 2,4-D mineralization was heterogeneous from field to microhabitats. High mineralization potential was not distributed randomly in the soil, but rather as systematic hot spots organized at centimeter scales (Vieuble-Gonod et al., 2005). Most pesticides may represent an occasional source of C and nutrients for soil microorganisms, so they can be completely dissipated from the soil environment by either a single microbial species or the joint action of a microbial consortium. In the later case, the degradation pathway in soil involves the cooperative activity of several strains that possess enzymes that catalyze different degradation steps. This cooperation is only possible with intimate contact between microbial cells and their substrates, as metabolites resulting from one step of the pathway may act as substrate for a different strain. Moreover, the bacterial distribution at the microscale may facilitate spreading of degradative genes located in plasmids or transposons (McGowan et al., 1998; DiGiovanni et al., 1996). Pallud et al. (2004) found that for low abundances of 2,4 D degraders, there was strong spatial isolation within the degraders populations, with less than 2 cells per colonized patch. 2,4-D amendment caused an increase in degrader abundance and concurrent spreading of degraders, reducing the distance between colonized patches, although the number of cells per patch remained low (< 28). They argued that the spatial spreading of bacteria was an ecological strategy that increased the probability of encountering the substrate (2,4-D), and proposed that this was achieved either through active cell movement (chemotaxis) or degradative plasmids transfer to indigenous microbial populations (Pallud et al., 2004). The zone of soil directly influenced by the presence of plant roots, known as rhizosphere, is of particular importance. Plant roots act on microbes essentially through the input of a wide variety of organic compounds (e.g., sugars, amino acids, cellulose, proteins, phenolic acids), known as rhizodeposition, and by providing a surface for attachment, creating an environment that can greatly differ from the surrounding bulk soil. As a consequence, higher microbial biomass and activities are found in the rizosphere as opposed to bulk soil. Several studies have reported that rhizosphere enhances biodegradation of chlorophenoxyacetic acids (Shaw & Burns, 2004; 2005; Merini et al., 2007), metsulfuron-methyl (Ghani & Wardle, 2001) and atrazine (Piutti et al., 2002). Biodegradation pathways and strategies will be discussed in the following section.

2.4 Biodegradation: co-metabolism vs. growth-linked metabolism

Biodegradation is the enzyme-mediated transformation of a xenobiotic by living microbial cells. In soil systems, biodegradation is a fundamental attenuation process for pesticides and is controlled by biotic factors (i.e. microbial activity) and a number of physicochemical processes such as sorption and desorption, diffusion, and dissolution (Chen et al., 2009). Pesticide degradation by microorganisms that are capable of using the chemical as a source of C and energy for growth, is called mineralization. This metabolic strategy results in the complete dissipation of the chemical and its conversion to CO2, water and inorganic elements. In this case, the biomass of the degrading population increases at the expense of the substrate. The rate of change in herbicide concentration in the medium follows the dynamic of the expanding microbial population, i.e., as the herbicide concentration decreases in the solution, growth of the microbial population reaches a plateau at a high cell density. Conversely, the partial transformation of an herbicide by microorganisms that gains no C or nutrients and energy, is called co-metabolism. In most cases, co-metabolism of herbicides involves microbial growth at the expense of a co-substrate that provides C and energy, but the pesticide in itself does not support microbial proliferation. The biomass of the herbicide degrading microbial population and the concomitant rate of herbicide degradation is not affected by the herbicide concentration in solution. Even though an herbicide may be partially transformed by co-metabolism, intermediate metabolites may be completely degraded by other microorganisms in soil.

2.4.1 Metabolic pathway of 2,4-D

One of the most studied herbicide degradation pathways is that of 2,4-D, which can be readily used as a C and energy source by environmental microorganisms. Numerous 2,4-D degrading bacteria have been isolated and characterized (Tiedje et al., 1969; Don & Pemberton, 1981; Kamagata et al., 1997; Muller et al., 2001). Most of these strains are members of genera belonging to the P and y subdivisions of the class Proteobacteria and were isolated from 2,4-D treated environments (Kamagata et al., 1997; Lee et al., 2005). These P and y subgroups carry tfd genes homologous to the canonical genes found in Cupriavidus necator JMP134 (Lerch et al., 2007), the model organism for 2,4-D degradation studies (Figure 5). These genes are located on conjugative plasmids like pJP4 which carries tfdABCDEF (Don & Pemberton, 1981). On this plasmid, tfdA encodes a 2,4-D/ o-ketoglutarate dioxygenase, which transforms 2,4-D into 2,4-dichlorophenol (DCP), while tfdB encodes a dichlorophenol hydroxylase that transforms DCP in 3,5-dichlorocatechol. The tfdCDEF operon encodes enzymes involved in the ortho-cleavage of the aromatic ring and subsequent reactions (Fukumori & Hausinger, 1993a;b; Vallaeys et al., 1996). Zabaloy et al. (unpublished results) recovered several Cupriavidus-like isolates from an agricultural soil in Argentina, able to grow with up to 1.1 mM herbicide as sole C source with complete primary degradation in < 72 h. These isolates harbored tfdA and tfdB genes similar to the canonical degradation genes described by Vallaeys et al. (1996), as determined by PCR and restriction fragment length polymorphism (RFLP). Recently, isolation of 2,4-D degrading bacteria from pristine environments has unveiled the existence of other degradative genes, namely the cadRABKC operon, which are responsible for 2,4-D catabolism in slow-growing Proteobacteria (Kitagawa et al., 2002).

Considerably less information is available regarding 2,4-D degradation by fungi. Donnelly et al. (1993) reported that the basidiomycete Phanerochaete chrysosporium was able to degrade 2,4-D when provided with external nitrogen sources. Vroumsia et al. (2005) screened the ability of ninety strains of filamentous fungi to degrade the herbicide in liquid media, finding that 2,4-D was less accessible to degradation than its metabolite DCP, although both compounds were inefficiently used. The kinetics studies performed on the most efficient strains revealed a one-day lag phase before 2,4-D degradation and no lag phase for DCP. 2,4-D application to agricultural soils triggers specific degradation pathways in existing degrading bacterial populations (Baelum et al., 2006; 2008). Degradation of 2,4-D in soil initially occurs at low rates, as the specific degrading population increases in size. During that 1-3 day period, which corresponds to the lag phase of degraders, 2,4-D degradation probably proceeds by co-metabolism in soil (Vieuble-Gonod et al., 2006; Lerch et al., 2009). Zabaloy et al., 2010 examined aerobic degradation of 2,4-D in soil microcosms treated with environmentally-relevant level (ERL, 5 mg kg-1 soil) and high level (HL, 50 mg kg-1 soil) of 2,4-D after 3 and 14 days of incubation, using the BD Oxygen Biosensor System (BDOBS) . The use of 2,4-D as sole source of C and energy (50 mg l-1) was initially retarded (> 40 h at day 3) and was maximal 2 weeks after treatment (Figure 6). They argued that

2,4-dichlorophenoxyacetic acid

2,4-dichlorophenol

2,4-D/a-ketoglutarate dioxygenase tfdA

2,4-dichlorophenol hydroxylase tfdB

3,5-dichlorocatechol

3,5-dichlorocatechol -1,2-dioxygenase tfdC

2,4-dichloro- c/s,c/s-muconate chloromuconate cycloisomerase tfdD

2-chloro-frans-dienelactone chlorodienelactone isomerase

2-chloro-c/s-dienelactone chlorodienelactone hydrolase tfdE

2-chloromaleylacetate chloromaleylacetate reductase tfdF

2-maleylacetate maleylacetate reductase tfdF

3-oxoadipate

Fig. 5. Metabolic pathway of 2,4-D/a-ketoglutarate dioxygenase acid in Cupriavidus necator JMP134. Adapted from Caspi et al. (2010).

between day 3 and 14 a shift in dominant degrader populations might have occurred in the HL, as opposed to the ERL microcosms, and the increase in use of 2,4-D was reflecting the activity of specific degraders that were favoured by the high dose.

2.4.2 Metabolic pathways of glyphosate

Microbial degradation of glyphosate has been extensively explored and several degrading bacteria, belonging to Arthrobacteriaceae, Bacillaceae, Rhodobacteriaceae, Alcaligenaceae, Pseudomonaceae, Enterobacteriaceae and Rizhobiaceae, have been isolated and characterized (Kononova & Nesmeyanova, 2002). Most degrading isolates posses the capability to use glyphosate as a source of P, once extracellular inorganic phosphate becomes limiting in the environment (McGrath et al., 1997). Bacterial degradation of glyphosate follows either of two metabolic pathways. For example, a C-P lyase catalyzes the breakdown of the C-P bond in Pseudomonas sp. PG2982, releasing inorganic phosphate (Pi) and sarcosine, which is subsequentially mineralized to CO2 and NH4+ (Kishore & Jacob, 1987). Glyphosate dehydrogenase catalyzes the conversion of glyphosate to aminomethylphosphonic acid (AMPA) and glioxylate as primary metabolites in Geobacillus caldoxylosilyticus T20, and a C-P lyase releases Pi from AMPA afterwards (Obojska et al., 2002). Although bacteria are considered as the main biological degrader of glyphosate in soils, fungi also may be important (Singh &Walker, 2006). Krzyko-Lupicka & Orlik (1997)

Fig. 6. Fluorescence curve (i.e., oxygen consumption) in BDODS plate with 2,4-D as sole source of C and energy, in 2,4-D-acclimated soil slurries after 14 days of incubation. Microcosms received 5 and 50 mg kg-1 soil of 2,4-D (ERL and HL). NRFU= normalized relative fluorescence units. From Zabaloy et al., 2010.

reported that the diversity of species isolated from soil diminished in media containing glyphosate, with a predominance of strains of Mucor, Fusarium and Trichoderma. Almost all the tolerant species isolated grew well when glyphosate was used as the unique source of P, but only a few were able to grow on it as the sole C source. Glyphosate also served as a nitrogen source for a Penicillium chyrsogenum strain isolated from soil (Lipok et al., 2003) and as C source for indigenous yeasts isolated from treated and untreated soils (Romero et al., 2004). Two yeast species were identified as active biodegraders (Yarrowia lipolitica and Candida krusei) but failed to uptake the herbicide when phosphate was present, suggesting that glyphosate was important for C and P nutrition in the yeasts (Romero et al., 2004). Glyphosate degradation in soil is predominantly a co-metabolic process, as it is not used as a C and energy source by the vast majority of microorganisms (Forlani et al., 1999; Singh & Walker, 2006). As opposed to degradation in pure-culture experiments, glyphosate degradation in soil occurs without a lag phase, further suggesting a co-metabolic process as enzymes were present in soil before the application of the herbicide (Borggaard & Gimsing, 2008). Moreover, no adaptation to glyphosate degradation has been observed in soils with long histories of herbicide treatment (Gimsing et al., 2004). Lancaster et al. (2009) reported that the total amount of 14CO2 evolved from glyphosate was reduced with repeated herbicide applications compared to a single application, which proved that biodegradation was not enhanced (i.e., no evidence of accelerated degradation) and was probably the result of a co-metabolic process. It is not well-known which metabolic pathway prevails in soils (Borggaard & Gimsing, 2008) although most isolated bacteria (Liu et al., 1991; Kishore & Jacob, 1987) and fungi (Sailaja & Satyaprasad, 2006) possess the sarcosine pathway. Even though AMPA is the main metabolite detected in soil, this could be attributed to the rapid mineralization of sarcosine and the persitance of AMPA in the environment.

2.4.3 Biodegradation of metsulfuron-methyl

Many studies have focused on the environmental fate and behavior of metsulfuron-methyl, but research on microbial degradation is still scarce. Zanardini et al. (2002) isolated a

Pseudomonas strain that degraded metsulfuron through co-metabolism and Boschin et al. (2003) studied the degradation pathway in the common soil fungus Aspergillus niger. Yu et al. (2005) isolated a Curvularia sp. capable of using the herbicide as a sole source of C and energy and studied several features of herbicide degradation in pure culture and in soils. Vázquez et al. (2008) isolated several filamentous fungi able to grow with metsulfuron as sole source of C and energy. Only Penicillium and Trichoderma strains were able to complete their life cycle in metsulfuron-containing medium. Trichoderma strains showed the best capacity to grow using the herbicide and were selected to perform tolerance assays; all could grow from spores in minimal medium containing metsulfuron and two showed heavy sporulation with increasing concentrations (up 1 x 10-2 ppm) (Vázquez et al., 2009). He et al. (2007) inoculated wheat rhizosphere with a highly effective metsulfuron-degrading Penicillium sp., previously isolated from treated soil, reporting that the inoculation enhanced the degradation of the target herbicide. Regardless of the pure-culture experiments that showed microbial degradation of metsulfuron methyl, limited mineralization of this herbicide in soil has been reported (Pons & Barriuso, 1998; Andersen et al., 2001). Ghani & Wardle (2001) studied 14C-labeled metsulfuron-methyl in soil-plant microcosms and reported that 42% of the applied metsulfuron was respired or incorporated into microbes in the planted treatment while 36% was used in the unplanted system by day 131. They argued that greater utilization of metsulfuron in the planted microcosm would have been influenced largely by a greater microbial biomass in that treatment. Despite the positive rhizosphere effect on herbicide mineralization, more than 50% was still present in soil even four months after an application at recommended rates.

3. Approaches to link bioavailability and biodegradation

Bioavailability is influenced by a variety of factors, including physical characteristics of the sorbent (e.g., particle shapes, sizes, and internal porosities), chemical properties of the sorbates and sorbents, and biological factors (e.g., microbial density and degradative capacity). Generally, sorbed compounds are assumed to be less accessible to attached or suspended microorganisms, which preferentially or exclusively utilize herbicides in the aqueous phase. In this view, herbicide is available for biodegradation only after desorption, followed by diffusion into solution. The sorbed fraction remains protected from microbial attack as a result of: 1) physical sequestration of the herbicide in the organo-chemical matrix, 2) chemical stabilization in the sorbent surface, and/or 3) reduction of aqueous-phase concentrations to levels that do not sustain microbial growth (Ainsworth et al., 1993; Lerch et al., 2009). However, some other investigations revealed that silica-sorbed 2,4-D (Park et al., 2001) and soil-sorbed atrazine (Park et al., 2003) can be directly utilized by degrading bacteria. Park et al. (2001) proposed two plausible explanations for the observed enhanced bioavailability of silica-sorbed herbicide: 1) attached biomass is able to access adjacent elevated concentrations of herbicide before complete dilution in the liquid phase; 2) attached cells are capable of higher metabolic rates. Similarly, soil organic matter is implicated in sorption processes and therefore, affects the availability of herbicides to degrading microbes (Benoit et al., 1999). This section will briefly present recent research approaches that have been successfully used to link bioavailability and degradation of herbicides. Benoit et al. (1999) studied the degradation of 14C ring-labeled 2,4-D and two chlorophenols, adsorbed on different organic materials (wood chips, straw, lignin, humic acids) and aluminum oxide in soil incubations. They observed that mineralization of these compounds, when incubated in direct contact with soil, varied greatly according to the nature of the sorbent, but was generally higher in more humified organic matter (humic acid) than in less transformed organic matter (wood, lignin and straw). However, separation of chemical-sorbent associations from soil during incubation in polyamide bags with non-decomposed and composted straw showed higher mineralization levels than in direct touch with soil for all compounds, despite slower mineralization rates. They proposed that straw-associated microorganisms actively degraded 2,4-D and chlorophenols. Schnurer et al. (2006) tested the effect of surface sorption on the bioavailability of glyphosate by adding goethite to an organic soil, and using respirometric and attenuated total reflectance Fourier transform (ATR-FTIR) spectroscopy approaches. Addition of goethite reduced the negative effects of glyphosate on microbial respiration, as surface sorption reduced toxic effects of the herbicide or its metabolites. However, ATR-FTIR data showed that sorbed herbicide was bioavailable and was degraded despite the reduction in soil respiration in the presence of glyphosate. Hermosin et al. (2006) evaluated the bioavailability of organoclay-based formulations of 2,4-D for bacterial degradation in pure culture and leaching potential in soil columns. They observed that the rate of mineralization of 14C-2,4-D from the organoclay complexes was related to the rate of release from the complexes, suggesting that desorption into the aqueous phase was the limiting step for biodegradation. The organoclay-based formulations reduced the leaching losses of 2,4-D in soil columns, and the amount of herbicide leached was considerably less than the amounts of 2,4-D mineralized. They concluded that these formulations slowly released 2,4-D, reducing risk of leaching losses in soil, while maintaining accessibility for bacterial degradation. S0rensen et al. (2006) studied sorption and biodegradation of 14C-labeled glyphosate and the phenoxyacid herbicide 4-chloro-2-methylphenoxy-acetic acid (MCPA) in soil and subsurface samples from a sandy agricultural site and a clay rich till in Denmark. These authors observed that MCPA sorbed to a minor extent and was mineralized rapidly in most samples, except in the deepest layers at both sites, and no relation was found between sorption and mineralization for this herbicide. Interestingly, the highest extent of mineralization of MCPA occurred in the top soil which coincided with the largest sorption and lowest desorption. Conversely, samples which showed higher sorption and low desorption exhibited no or reduced mineralization of glyphosate indicating a limited glyphosate bioavailability. Lerch et al. (2009) assessed the link between bioavailability and mineralization of 13C-2,4-D over a 6-month study, by using stable isotope probing (SIP) coupled to fatty acid methyl ester (FAME) to study degrader populations. These authors reported that the proportion of readily available (water extracted) as well as potentially available (solvent extracted) herbicide residues decreased rapidly to less than 1.2 % of the initial amount added to soil at day 8, which corresponded to the period of highest biodegradation activity. From day 8 onwards, labelled C-2,4-D was present in the form of non-extractable residues (NER), which nontheless were biodegradable. The 13C-2,4-D enriched FAME profiles during this period of incubation were similar to those of the populations degrading 2,4-D when it was still available. They proposed that either the degradation of NER was due to the activity of the same specific degraders involved in degradation of available 2,4-D, limited by desorption of 13C-2,4-D in the soil solution, or that specific degraders were present in a dormant state and/or their fatty acids were being recycled by cells from the same taxonomical group. Overall, these studies show that while adsorption-desorption phenomenons affect bioavailability, there is evidence that sorbed herbicides may be accessible for biodegradation.

The above discussion has practical implications, when predicting the potential in situ biodegradation of a certain herbicide. In the field, additional factors other than presence of potential degraders and bioavailability must be considered, including the 1) presence of other contaminants that can compete for adsorption sites and for access to microbial enzymes (Haws et al., 2006); 2) availability of nutrients and co-factors necessary for degraders growth and activity; 3) intrinsic environmental factors (e.g., temperature, oxygen concentration, surface charge, water availability, pH, etc.). The presence and nature of crop residues should also be considered, as they can have great impact on the bioavailability of herbicides in the agroecosystem, helping reduce hazardous effects on soil and water resources (Benoit et al., 1999).

Was this article helpful?

0 0

Post a comment