**Behavior and Fate of Imidacloprid in Croatian Olive Orchard Soils Under Laboratory Conditions**

Dalibor Broznić, Jelena Marinić and Čedomila Milin *Department of Chemistry and Biochemistry, School of Medicine, University of Rijeka, Croatia* 

## **1. Introduction**

488 Pesticides in the Modern World - Risks and Benefits

Sparks, D.L. (2003). *Environmental Soil Chemistry* (2nd edition), Academic Press, ISBN 0-12-

Sposito, G. (1982). On the use of the Langmuir equation in the interpretation of "adsorption"

Stevenson, F.J. (1994). *Humus Chemistry. Genesis, Composition, Reactions* (2nd edition), John

Roberts, T.R.; Dyson, J.S. & Lane, M.C.G. (2002). Deactivation of biological activity of

Tadanier, C.J. & Eick, M.J. (2002). Formulating the charge-distribution multisite surface

Tipping, E. (2002). *Cation Binding by Humic Substances*, Cambridge University Press, ISBN 0-

Tunega, D.; Haberhauer, G.; Gerzabek, M.H. & Lischka, H. (2004). Sorption of phenoxyacetic

van Riemsdijk, W.H.; Koopal, L.K.; Kinniburgh, D.G.; Benedetti, M.F. & Weng, L. (2006).

Vasiliadis, B.; Antelo, J.; Iglesias, A.; López, R.; Fiol, S. & Arce, F. (2007). Analysis of the

Venema, P.; Hiemstra, T. & van Riemsdijk, W.H. (1998). Intrinsic proton affinity of reactive

Vidali, R.; Remoundaki, E. & Trezos, M. (2009). An experimental and modelling study on

Villaverde, P.; Gondar, D.; Antelo, J.; López, R.; Fiol, S. & Arce, F. (2009). Influence of pH on

Weng, L.; Temminghoff, E.J.M. & van Riemsdijk, W.H. (2001). Contribution of individual

Weng, L.; Temminghoff, E.J.M.; Lofts, S.; Tipping, E. & van Riemsdijk, W.H. (2002).

Weng, L.; van Riemsdijk, W.H. & Hiemstra, T. (2007). Adsorption of humic acids onto

a sandy soil. *Environ. Sci. Technol.* 36, 4804-4810, ISSN 0013-936X

phenomena. II. The "two-surface" Langmuir equation. *Soil Sci. Soc. Am. J*. 46, 1147-

paraquat in the soil environment: a review of long-term environmental fate. *J.* 

complexation model using FITEQL. *Soil Sci. Soc. Am. J.* 66, 1505-1517, ISSN 0361-

acid herbicides on the kaolinite mineral surface - An ab initio molecular dynamics

Modelling the interactions between humics, ions, and mineral surfaces. *Environ. Sci.* 

variable charge of two organic soils by means of the NICA-Donnan model. *Eur. J.* 

surface groups of metal (hydr)oxides: application to iron (hydr)oxides*. J. Colloid* 

humic acid concentration effect on H+ binding: Application of the NICA-Donnan

copper, lead and cadmium binding by an ombrotrophic peat. *Eur. J. Soil Sci.* 60,

sorbents to the control of heavy metal activity in sandy soil. *Environ. Sci. Technol.* 

Complexation of dissolved organic matter and solubility control of heavy metals in

goethite: effects of molar mass, pH and ionic strength. *J. Colloid Interface Sci.* 314,

656446-9, San Diego, California

521-62146-1, Cambridge U.K.

Wiley & Sons, ISBN 0-471-59474-1, New York

*Agric. Food Chem*. 50, 3623-3631. ISSN 0021-8561

simulation. *Soil Sci.* 169, 44-54, ISSN 0038-075X

*Technol.* 40, 7473-7480, ISSN 0013-936X

*Soil Sci.* 58, 1358-1363, ISSN 1365-2389

*Interface Sci.* 198, 282-295, ISSN 0021-9797

377-385, ISSN 1365-2389

107-118, ISSN 0021-9797

35, 4436-4443, ISSN 0013-936X

model. *J. Colloid Interface Sci.* 339,330-335, ISSN 0021-9797

1152, ISSN 0361-5995

5995

Over the last two decades, the worldwide production and use of pesticides have greatly increased, causing great concern about their fate in the soil environment, as well as their adverse effects on nontarget organisms, including human beings. An important thing to realize it that only a small part of the pesticide doses used reaches its intended target (< 0.1%), while the major part (over 99%) of it is distributed into the ecosystem (Pimentel & Levitan, 1986), where it can cause difficulties through its toxicity to nontarget species, and cause serious environmental problems, such as groundwater contamination, food contamination, and air pollution (Larson et al., 1997; Mathys, 1994). There is also increasing interest in their transformation products, because they can be present at higher levels in the soil than the parent itself. In some instances, transformation products are more toxic, so they represent a greater risk to the environment than the parent molecule. Therefore, it is essential to study the residue and degradation pattern of pesticide in crop, soils and water systematically in order to generate meaningful data from the point of view of plant protection, public health and environmental safety.

In the past few decades, three major groups of insecticides have dominated the market: organophosphates, carbamates and pyrethroids. Nevertheless, pests resistance limited their use and caused a need for the synthesis of a new group that will be effective and nontoxic to the environment and to mammals. The results was "the birth" of neonicotinoids which exhibited high insecticidity and low toxicity to the environment (Maienfisch et al., 2001). But because neonicotinoids are becoming extensively used, both in agriculture and for home use, the chance of their polluting water is still present despite the low application rates.

Imidacloprid [1-(6-chloro-3pyridylmethyl)-N-nitroimidazolidin-2-ylideneamine; IMI] was the first neonicotinoid registered by the United States Environmental Protection Agency (USEPA) for use as a pesticide through its actions as an agonist on the nicotinic acetylcholine receptor (nAChR) (Bai et al., 1991). The mode of action of IMI in the brain is shown in Figure 1. The toxicity of IMI is largely due to interference of the neurotransmission in the nicotinic cholinergic nervous system. Prolonged activation of the nAChR by IMI causes desensitization and blocking of the receptor, and leads to incoordination, tremors, decreased activity, reduced body temperature and death. IMI's favorable selective toxicity to insects versus mammals makes it safer for insect control than other neurotoxins (Tomizawa & Casida, 2003) and enables its diverse use in soil and foliar treatment in different crops, as

Behavior and Fate of Imidacloprid in Croatian Olive Orchard Soils Under Laboratory Conditions 491

including soil texture, organic carbon content (OC), cation exchange capacity (CEC), pH and temperature. Field and laboratory studies have determined that IMI sorption to soil particles increases as the concentration of the insecticide decreases (Cox et al., 1998a; Kamble & Saran, 2005; Oi, 1999). Cox et al. (1997; 1998b; 1998c) have found that the potential for IMI to leach would decrease, as the OC levels and laminar silicate clay content in the soil increase. In contrast, IMI sorption in a calcerous soil was found to decrease with the addition of OC (obtained from peat and tannic acid) (Flores-Cespedes et al. 2002), which increased the mobility and leaching potential of IMI. In the study of Cox et al. (1998a) positive correlation between IMI sorption and CEC have observed, while the effect of soil pH did not significantly contributed to the sorption. The effect of temperature is of special importance in greenhouses, where higher temperatures are used for cultivation (Cox et al., 1997; Fernandez-Bayo et al., 2007; Gonzalez-Pradas et al., 2002; ten Hulscher & Cornelissen, 1996). On the other hand, desorption governs the release of IMI from the soil and several studies have reported irreversible sorption and the occurrence of hysteresis phenomena; i.e. less desorption than predicted by sorption isotherms (Cox et al., 1997; Fernandez-Bayo et al., 2007; Papiernik et al., 2006). This behavior can be attributed to a portion of the sorbed compound that is bound irreversibly to soil surfaces (Celis & Koskinen, 1999; Cox et al., 1997; Cox et al., 1998b). Numerous studies indicated that values for IMI half-lives (DT50) are highly dependent on experimental conditions; namely field or laboratory, (Krohn & Hellpointner, 2002). In fact, DT50s for IMI have been reported ranging from approximately 80 days to 2 years. Examples of laboratory experiments include Krohn & Hellpointner (2002) who reported a DT50 of 156 days, representing the geometric mean value of five studies. In a field experiment, Krohn & Hellpointner (2002) reported a DT50 of 96 days for the 11 bare soils in Northern and Southern Europe. However, lengthier DTs have also been determined from field studies. Mulye (1996) reviewed a two-year field investigation in Germany using IMI and from the study results calculated a DT50 of approximately 2 years, indicating that

Since a variability of pesticide sorption-desorption and persistence can occur among regions and even within the areas with the same geological and climatic characteristics, additional knowledge is needed to improve IMI's applicability in conditions which cover the Croatian climate. Consequently, the objective of this study is to analyze sorption-desorption characteristics as well as persistence of IMI in four soils, representative of northern Adriatic region, namely an island Krk and a coastal Istrian region. For a better understanding of factors governing IMI's behavior and fate in these regions, the relationship between selected soil properties, IMI's concentration and soil sorption-desorption coefficients as well as persistence was determined for the soils among and within regions. Additionally, applicability of

mathematical models to predict IMI's sorption-desorption and degradation was tested.

The behavior and fate of pesticides in the soil environment is controlled by their physicochemical properties and by various complex dynamic physical, chemical and biological processes, among which are the most important sorption-desorption interactions of pesticide molecules with natural sorbents: soil organic matter and soil minerals, as well as degradation processes. These processes directly control the transport, retention and transformation of pesticides within the soil matrix and their transfer from the soil to other environmental compartments, and also determine the efficacy of pesticides in controlling target organisms and their potential for adverse effects on non target organisms (Pimentel &

**2. The behavior and fate of pesticides in the soil** 

the compound would persist in soil.

well as in non-agricultural practice. Thus, IMI is found in a variety of commercial insecticides. Its major manufacturer is Bayer Corporation that markets IMI products with the brand names Admire®, Confidor®, Gaucho®, Premier®, Premise®, Provado®, and Marathon®. However, one of the drawbacks of IMI's usage is a high toxicity in honeybees. In France, between 1994 and 1996, greatly increased mortality in honeybees was noticed when sunflowers were treated with a new pesticide Gaucho®. In addition Maxim & van der Sluijs (2007, 2010) and Suchail et al. (2004) found that IMI at very low doses causes bees mortality and adverse effects on laboratory-conditioned behavioral responses associated with feeding.

Fig. 1. IMI mode of action.

In Croatian coastal regions, IMI is increasingly being used in olive growing areas, including Istria and Kvarner islands, as an effective means of olive fruit fly infestation control. Although used at low dose rates (Capri et al., 2001), it is usually applied more than once during the growing season. Thus, intensive use of IMI, in addition to its high water solubility (510 mg/L, 20 °C) (Tomlin, 2001) might impose a great risk of water resources contamination which is consistent with USEPA statement regarding IMI's potential to leach to groundwater (USEPA, 1993). The review of literature revealed that few reports are available on leaching of IMI in soil (Carbo et al., 2008; Felsot et al., 1998; Gonzalez-Pradas et al., 2002; Gupta et al., 2002; Schmidt, 2010). For these reasons, there is a need for a more complete insight into IMI's fate in the soil (USEPA, 1993).

Among the processes that determine the behavior and fate of pesticides in the soil/water environment, sorption-desorption and degradation processes are the key processes affecting pesticide persistence, transport and bioavailability determining the amount of pesticide that can reach the target organism and that can be volatilized, or leached. Information regarding the sorption and desorption characteristics of IMI are essential for predicting its fate within the soil environment (Cox et al., 1998a; Cox et al., 1998b; Cox et al., 2001; Fernandez-Bayo et al., 2007; Liu et al., 2006; Ping et al., 2010). Capri et al. (2001) and Cox et al. (1998c) investigated the effect of selected soil properties on the sorption of IMI using the batch equilibrium technique. Their results indicated that IMI retention was highly dependent on the amount of the pesticide applied and on the physicochemical properties of the pesticide, but the key factor governing pesticide sorption-desorption were soil characteristics,

well as in non-agricultural practice. Thus, IMI is found in a variety of commercial insecticides. Its major manufacturer is Bayer Corporation that markets IMI products with the brand names Admire®, Confidor®, Gaucho®, Premier®, Premise®, Provado®, and Marathon®. However, one of the drawbacks of IMI's usage is a high toxicity in honeybees. In France, between 1994 and 1996, greatly increased mortality in honeybees was noticed when sunflowers were treated with a new pesticide Gaucho®. In addition Maxim & van der Sluijs (2007, 2010) and Suchail et al. (2004) found that IMI at very low doses causes bees mortality and adverse effects on laboratory-conditioned behavioral responses associated with feeding.

In Croatian coastal regions, IMI is increasingly being used in olive growing areas, including Istria and Kvarner islands, as an effective means of olive fruit fly infestation control. Although used at low dose rates (Capri et al., 2001), it is usually applied more than once during the growing season. Thus, intensive use of IMI, in addition to its high water solubility (510 mg/L, 20 °C) (Tomlin, 2001) might impose a great risk of water resources contamination which is consistent with USEPA statement regarding IMI's potential to leach to groundwater (USEPA, 1993). The review of literature revealed that few reports are available on leaching of IMI in soil (Carbo et al., 2008; Felsot et al., 1998; Gonzalez-Pradas et al., 2002; Gupta et al., 2002; Schmidt, 2010). For these reasons, there is a need for a more

Among the processes that determine the behavior and fate of pesticides in the soil/water environment, sorption-desorption and degradation processes are the key processes affecting pesticide persistence, transport and bioavailability determining the amount of pesticide that can reach the target organism and that can be volatilized, or leached. Information regarding the sorption and desorption characteristics of IMI are essential for predicting its fate within the soil environment (Cox et al., 1998a; Cox et al., 1998b; Cox et al., 2001; Fernandez-Bayo et al., 2007; Liu et al., 2006; Ping et al., 2010). Capri et al. (2001) and Cox et al. (1998c) investigated the effect of selected soil properties on the sorption of IMI using the batch equilibrium technique. Their results indicated that IMI retention was highly dependent on the amount of the pesticide applied and on the physicochemical properties of the pesticide, but the key factor governing pesticide sorption-desorption were soil characteristics,

Fig. 1. IMI mode of action.

complete insight into IMI's fate in the soil (USEPA, 1993).

including soil texture, organic carbon content (OC), cation exchange capacity (CEC), pH and temperature. Field and laboratory studies have determined that IMI sorption to soil particles increases as the concentration of the insecticide decreases (Cox et al., 1998a; Kamble & Saran, 2005; Oi, 1999). Cox et al. (1997; 1998b; 1998c) have found that the potential for IMI to leach would decrease, as the OC levels and laminar silicate clay content in the soil increase. In contrast, IMI sorption in a calcerous soil was found to decrease with the addition of OC (obtained from peat and tannic acid) (Flores-Cespedes et al. 2002), which increased the mobility and leaching potential of IMI. In the study of Cox et al. (1998a) positive correlation between IMI sorption and CEC have observed, while the effect of soil pH did not significantly contributed to the sorption. The effect of temperature is of special importance in greenhouses, where higher temperatures are used for cultivation (Cox et al., 1997; Fernandez-Bayo et al., 2007; Gonzalez-Pradas et al., 2002; ten Hulscher & Cornelissen, 1996). On the other hand, desorption governs the release of IMI from the soil and several studies have reported irreversible sorption and the occurrence of hysteresis phenomena; i.e. less desorption than predicted by sorption isotherms (Cox et al., 1997; Fernandez-Bayo et al., 2007; Papiernik et al., 2006). This behavior can be attributed to a portion of the sorbed compound that is bound irreversibly to soil surfaces (Celis & Koskinen, 1999; Cox et al., 1997; Cox et al., 1998b). Numerous studies indicated that values for IMI half-lives (DT50) are highly dependent on experimental conditions; namely field or laboratory, (Krohn & Hellpointner, 2002). In fact, DT50s for IMI have been reported ranging from approximately 80 days to 2 years. Examples of laboratory experiments include Krohn & Hellpointner (2002) who reported a DT50 of 156 days, representing the geometric mean value of five studies. In a field experiment, Krohn & Hellpointner (2002) reported a DT50 of 96 days for the 11 bare soils in Northern and Southern Europe. However, lengthier DTs have also been determined from field studies. Mulye (1996) reviewed a two-year field investigation in Germany using IMI and from the study results calculated a DT50 of approximately 2 years, indicating that the compound would persist in soil.

Since a variability of pesticide sorption-desorption and persistence can occur among regions and even within the areas with the same geological and climatic characteristics, additional knowledge is needed to improve IMI's applicability in conditions which cover the Croatian climate. Consequently, the objective of this study is to analyze sorption-desorption characteristics as well as persistence of IMI in four soils, representative of northern Adriatic region, namely an island Krk and a coastal Istrian region. For a better understanding of factors governing IMI's behavior and fate in these regions, the relationship between selected soil properties, IMI's concentration and soil sorption-desorption coefficients as well as persistence was determined for the soils among and within regions. Additionally, applicability of mathematical models to predict IMI's sorption-desorption and degradation was tested.

#### **2. The behavior and fate of pesticides in the soil**

The behavior and fate of pesticides in the soil environment is controlled by their physicochemical properties and by various complex dynamic physical, chemical and biological processes, among which are the most important sorption-desorption interactions of pesticide molecules with natural sorbents: soil organic matter and soil minerals, as well as degradation processes. These processes directly control the transport, retention and transformation of pesticides within the soil matrix and their transfer from the soil to other environmental compartments, and also determine the efficacy of pesticides in controlling target organisms and their potential for adverse effects on non target organisms (Pimentel &

Behavior and Fate of Imidacloprid in Croatian Olive Orchard Soils Under Laboratory Conditions 493

*OC*

potential for losses through leaching.

nonlinearity), usually in the range 0<*1/n*≤1.

impact the linear sorption.

equation:

sorption.

*sor / des D*

*f* (2)

*e Fe q = K γ* (3)

(4)

*OC <sup>K</sup> K = ×100*

where *OC f* represents the percentage of the OC content in the soil. High *KOC* values (greater than 1000) indicate a tendency for the pesticide molecule to be sorbed by soil particles rather than remain in the soil solution (McCall et al., 1980). Since pesticide bond mainly to soil OC, the division by the percentage OC in soil makes the sorption coefficient a pesticide-specific property, independent of soil type. Sorption coefficients less than 500 indicate a considerable

This linear model is adequate if the sorption sites are of the same nature and in great amount to accommodate the chemical as the concentration increases. But in many cases, due to the heterogeneity of the soil, deviations from the linear sorption model are predictable and are effectively observed for pesticides (Delle Site, 2001; Wauchope et al., 2002). Two other nonlinear sorption isotherm models, the Freundlich and the Langmuir model, are frequently used when the amount of contaminant retained by the soil is abundant enough to

The Freundlich isotherm is derived by assuming a heterogeneous surface with a nonuniform distribution of heat of sorption over the surface and it is represented by the

*sor / des sor / des 1 / n* 

where *KFsor/des* is the Freundlich partition coefficient (coefficient of sorption-desorption capacity)(mg/kg)(mg/L)1/n) and *1/n* is the Freundlich's exponent (coefficient of

Sorption isotherm data could also be fitted to the Langmuir model given by Equation 4, with the assumption that soils have a finite number of sorption sites of uniform energy.

In Equation 4, *qmaxsor/des* designates the maximum amount of sorbed IMI per unit mass of soil (mg/kg) to form a monolayer and *KLsor/des* is the constant which depends on the enthalpy of

Although sorption processes of pesticides are usually characterized by a partition mechanism, in many cases, significant deviations between sorption and desorption isotherms have been observed. Typically, desorption is strongly delayed or hindered relative to the sorption (Huang & Weber, 1997; Lesan & Bhandari, 2003). In this phenomenon, called hysteresis, the Freundlich exponent *1/nsor* for desorption can be greater than the *1/ndes* measured for desorption at a constant *γe* concentration (Huang et al., 2003). Sorption-desorption hysteresis can usually be explained by irreversible chemical binding, sequestration of a pesticide molecule into specific components of the organic matter, or entrapment of the pesticide into microporous structures or into the organic matter matrix (Pignatello & Xing, 1996). The extent of sorption-desorption hysteresis can be quantified for each pair of sorption and desorption isotherms using the hysteresis coefficient *H* (Cox et al., 1997). This coefficient is calculated using Freundlich exponent (coefficient of nonlinearity,

*sor / des e*

*q =* 

*sor / des sor /des max L e sor / des L e*

*q K γ*

*1+K γ*

Levitan, 1986). Once a pesticide molecule is bound to soil particles in the soil, the main processes in the soil cause its loss and transformation. Significant losses of pesticides can occur during application, with the amount of loss affected by the nature of the pesticide, formulation, atmospheric conditions, method of application, and application characteristics. High vapor pressure, photodegradability, and weak sorption by the soil contribute to losses of pesticides after application (Navarro et al., 2007). Retention does not affect the amount of pesticide present in the soil, but can decrease the amount available for transport; whereas transformation reduces the amount of pesticide present in the soil. Transport processes include leaching, surface runoff, volatilization, and uptake by plants. Figure 2 shows the main processes of pesticides inactivation in the soil.

Fig. 2. The behavior and fate of pesticides in the environment.

#### **2.1 Sorption-desorption processes**

Sorption-desorption is a dynamic process in which molecules are continually transferred between the bulk liquid and solid surface (Koskinen & Harper, 1990). Sorption is the binding of the pesticide molecules by the surface of the treated soil, whereas desorption implies detachment of the molecules to the liquid medium. The ability of pesticides to sorb on soils and their tendency to desorb are the most important factors affecting soil and water contamination.

Several sorption models have been developed to describe, quantify and explain the sorptive process of pesticides on soils. The simplest one is the linear model depicted by the equation:

$$
\eta\_e^{\text{sor/des}} = \mathbf{K}\_D^{\text{sor/des}} \mathbf{y}\_e \tag{1}
$$

where *qe sor/des* is the sorbed pesticide amount (mg/kg), γ*<sup>e</sup>* is the equilibrium concentration in solution (mg/L) and *KDsor/des* is the sorption-desorption partition coefficient (L/kg). Because *KDsor/des* values for pesticides are soil specific and the *KDsor/des* of one pesticide can differ considerably from soil to soil or with depth in a soil profile, the more widely accepted partition coefficient normalized to the fraction of OC content in the soil, *KOC* was used. The *KOC* was determined using the equation:

Levitan, 1986). Once a pesticide molecule is bound to soil particles in the soil, the main processes in the soil cause its loss and transformation. Significant losses of pesticides can occur during application, with the amount of loss affected by the nature of the pesticide, formulation, atmospheric conditions, method of application, and application characteristics. High vapor pressure, photodegradability, and weak sorption by the soil contribute to losses of pesticides after application (Navarro et al., 2007). Retention does not affect the amount of pesticide present in the soil, but can decrease the amount available for transport; whereas transformation reduces the amount of pesticide present in the soil. Transport processes include leaching, surface runoff, volatilization, and uptake by plants. Figure 2 shows the

main processes of pesticides inactivation in the soil.

Fig. 2. The behavior and fate of pesticides in the environment.

*sor/des* is the sorbed pesticide amount (mg/kg),

Sorption-desorption is a dynamic process in which molecules are continually transferred between the bulk liquid and solid surface (Koskinen & Harper, 1990). Sorption is the binding of the pesticide molecules by the surface of the treated soil, whereas desorption implies detachment of the molecules to the liquid medium. The ability of pesticides to sorb on soils and their tendency to desorb are the most important factors affecting soil and water

Several sorption models have been developed to describe, quantify and explain the sorptive process of pesticides on soils. The simplest one is the linear model depicted by the equation:

*sor / des sor / des* 

solution (mg/L) and *KDsor/des* is the sorption-desorption partition coefficient (L/kg). Because *KDsor/des* values for pesticides are soil specific and the *KDsor/des* of one pesticide can differ considerably from soil to soil or with depth in a soil profile, the more widely accepted partition coefficient normalized to the fraction of OC content in the soil, *KOC* was used. The

*e De q = K γ* (1)

*<sup>e</sup>* is the equilibrium concentration in

γ

**2.1 Sorption-desorption processes** 

*KOC* was determined using the equation:

contamination.

where *qe*

$$K\_{\rm OC} = \frac{K\_D \, ^{sor/des}}{f\_{\rm OC}} \times 100 \,\tag{2}$$

where *OC f* represents the percentage of the OC content in the soil. High *KOC* values (greater than 1000) indicate a tendency for the pesticide molecule to be sorbed by soil particles rather than remain in the soil solution (McCall et al., 1980). Since pesticide bond mainly to soil OC, the division by the percentage OC in soil makes the sorption coefficient a pesticide-specific property, independent of soil type. Sorption coefficients less than 500 indicate a considerable potential for losses through leaching.

This linear model is adequate if the sorption sites are of the same nature and in great amount to accommodate the chemical as the concentration increases. But in many cases, due to the heterogeneity of the soil, deviations from the linear sorption model are predictable and are effectively observed for pesticides (Delle Site, 2001; Wauchope et al., 2002). Two other nonlinear sorption isotherm models, the Freundlich and the Langmuir model, are frequently used when the amount of contaminant retained by the soil is abundant enough to impact the linear sorption.

The Freundlich isotherm is derived by assuming a heterogeneous surface with a nonuniform distribution of heat of sorption over the surface and it is represented by the equation:

$$
\eta\_e^{\text{sor/des}} = \mathbf{K}\_F^{\text{sor/des}} \,\chi\_e^{1/n} \tag{3}
$$

where *KFsor/des* is the Freundlich partition coefficient (coefficient of sorption-desorption capacity)(mg/kg)(mg/L)1/n) and *1/n* is the Freundlich's exponent (coefficient of nonlinearity), usually in the range 0<*1/n*≤1.

Sorption isotherm data could also be fitted to the Langmuir model given by Equation 4, with the assumption that soils have a finite number of sorption sites of uniform energy.

$$q\_e^{\text{sor/des}} = \frac{q\_{\text{max}}^{\text{sr/des}} K\_{\text{L}}^{\text{sr/des}} \mathcal{Y}\_{\text{s}}}{1 + K\_{\text{L}}^{\text{sr/des}} \mathcal{Y}\_{\text{s}}} \tag{4}$$

In Equation 4, *qmaxsor/des* designates the maximum amount of sorbed IMI per unit mass of soil (mg/kg) to form a monolayer and *KLsor/des* is the constant which depends on the enthalpy of sorption.

Although sorption processes of pesticides are usually characterized by a partition mechanism, in many cases, significant deviations between sorption and desorption isotherms have been observed. Typically, desorption is strongly delayed or hindered relative to the sorption (Huang & Weber, 1997; Lesan & Bhandari, 2003). In this phenomenon, called hysteresis, the Freundlich exponent *1/nsor* for desorption can be greater than the *1/ndes* measured for desorption at a constant *γe* concentration (Huang et al., 2003). Sorption-desorption hysteresis can usually be explained by irreversible chemical binding, sequestration of a pesticide molecule into specific components of the organic matter, or entrapment of the pesticide into microporous structures or into the organic matter matrix (Pignatello & Xing, 1996). The extent of sorption-desorption hysteresis can be quantified for each pair of sorption and desorption isotherms using the hysteresis coefficient *H* (Cox et al., 1997). This coefficient is calculated using Freundlich exponent (coefficient of nonlinearity,

Behavior and Fate of Imidacloprid in Croatian Olive Orchard Soils Under Laboratory Conditions 495

*a*, and *b*, express the quantitative partition between the two compartments, where a + b is

The fast degradation in the first compartment occurs when the pesticide is in the soil-water phase and readily available for microorganisms. In the second compartment the pesticide is sorbed to soil particles. Degradation is, therefore, controlled by the rate of desorptiondiffusion into the soil-water phase. The partition between the two compartments depends on the pesticide sorption properties and soil characteristics. These characteristics suggest that a single DT50 may not be sufficient as an index of persistence. Beulke & Brown (2001) recomended using DT90 as a risk index to indicate the persistence, where the DT90 represents the time for 90% of the initial residues to dissipate; whereas Grover et al. (1997) and Wolt

The proposed metabolic pathway of IMI in the soil is shown in Figure 3. Two main routes of metabolism responsible for the degradation of IMI were identified. The first step is hydroxylation of the imidazolidine ring leading to the mono- and dihydroxylated compounds, followed by loss of water to yield the olefinic compound. The second important degradation step starts with dehydrogenation of the imidazolidine ring to form desnitro-

*1 2 -k t -k t C(t) = ae + be* (8)

approximately equal to *C*0 (mg/kg):

**2.2.1 Metabolites** 

(1997) used both DT50 and DT90 as indices of persistence.

metabolit with further oxidation to 6-chloronicotinic acid (6-CNA).

Fig. 3. The proposed metabolic pathway of IMI in the soil.

*1/n*) estimated from the sorption and desorption isotherms and it can be expressed by following equation:

$$H = \frac{1 \int n^{\text{des}}}{1 \int n^{\text{ors}}} \tag{5}$$

where, *1/nsor* and *1/ndes* are Freundlich coefficients of nonlinearity for sorption and desorption, respectively. The lower the value of *H* is, the stronger the soil will sequestrate the pesticide molecule. Value *H* = 1 indicates that the hysteresis is insignificant and the sorption is reversible.

#### **2.2 Degradation processes**

Concern about the persistence of pesticides in soils has led to increased efforts to identify the nature, mechanisms, and factors affecting degradation processes, to identify the degradation products, and to predict persistence. Soil is an ideal medium for supporting degradation reactions of pesticides, which include photochemical, chemical, and biological reactions (Chen et al., 2005; Kuhad et al.; Ward & Singh).

Many published degradation studies assume that the degradation of pesticides follows simple first-order degradation kinetics ( Baskaran et al., 1999; Calderon et al., 2004; Krohn & Hellpointner, 2002; Sarkar et al., 2001) which is represented with the following mathematical equation:

$$C(t) = ae^{-k\_1 t} \tag{6}$$

where *C* is the amount of pesticide remaining at the time *t* (mg/kg), *a* is initial amount of pesticide degraded through one *1st* order process, *t* is the time (days) and *k* is degradation rate constant (1/days). First-order kinetics is advantageous for use in modeling as a constant degradation rate and allows estimation of a pesticide's half-life, DT50 (the time at which the concentration reaches half the initial concentration), which can be estimated according to equation:

$$DT\_{50} = \frac{\ln 2}{k} \tag{7}$$

DT50 values are important in understanding the potential environmental impact of a pesticide. In fact, a molecule which degrades quickly has a low DT50 value and thus the impact of this species on the environment is reduced if the degradation products are harmless. On the contrary, the environmental impact of species with a high DT50 value can be substantial even if the molecule is only moderately toxic. Gavrilescu (2005) classified persistence of pesticides according to the DT50 value into three groups, where pesticides with DT50 < 30 days are non-persistent pesticides, compared to pesticides with DT50 > 100 days which are persistent.

However, deviations from the first-order degradation of pesticides have been reported. Typically, a fast initial degradation is followed by a gradual decrease in the degradation rate and eventually a very slow degradation. The gradual change in degradation rate may be better described by using two rate constants instead of one (Beulke & Brown, 2001; Henriksen et al., 2004; Ma et al., 2004; Sanchez et al., 2003). The two-compartment model (Equation 8) describes the degradation process as shared between two different compartments, where degradation proceeds at different rates (*k1* and *k2*). The two constants, *a*, and *b*, express the quantitative partition between the two compartments, where a + b is approximately equal to *C*0 (mg/kg):

$$C(t) = ae^{-k\_1t} + be^{-k\_2t} \tag{8}$$

The fast degradation in the first compartment occurs when the pesticide is in the soil-water phase and readily available for microorganisms. In the second compartment the pesticide is sorbed to soil particles. Degradation is, therefore, controlled by the rate of desorptiondiffusion into the soil-water phase. The partition between the two compartments depends on the pesticide sorption properties and soil characteristics. These characteristics suggest that a single DT50 may not be sufficient as an index of persistence. Beulke & Brown (2001) recomended using DT90 as a risk index to indicate the persistence, where the DT90 represents the time for 90% of the initial residues to dissipate; whereas Grover et al. (1997) and Wolt (1997) used both DT50 and DT90 as indices of persistence.

#### **2.2.1 Metabolites**

494 Pesticides in the Modern World - Risks and Benefits

*1/n*) estimated from the sorption and desorption isotherms and it can be expressed by

*1/n H =* 

where, *1/nsor* and *1/ndes* are Freundlich coefficients of nonlinearity for sorption and desorption, respectively. The lower the value of *H* is, the stronger the soil will sequestrate the pesticide molecule. Value *H* = 1 indicates that the hysteresis is insignificant and the

Concern about the persistence of pesticides in soils has led to increased efforts to identify the nature, mechanisms, and factors affecting degradation processes, to identify the degradation products, and to predict persistence. Soil is an ideal medium for supporting degradation reactions of pesticides, which include photochemical, chemical, and biological

Many published degradation studies assume that the degradation of pesticides follows simple first-order degradation kinetics ( Baskaran et al., 1999; Calderon et al., 2004; Krohn & Hellpointner, 2002; Sarkar et al., 2001) which is represented with the following mathematical

where *C* is the amount of pesticide remaining at the time *t* (mg/kg), *a* is initial amount of pesticide degraded through one *1st* order process, *t* is the time (days) and *k* is degradation rate constant (1/days). First-order kinetics is advantageous for use in modeling as a constant degradation rate and allows estimation of a pesticide's half-life, DT50 (the time at which the concentration reaches half the initial concentration), which can be estimated according to

> *50 ln2 DT =*

DT50 values are important in understanding the potential environmental impact of a pesticide. In fact, a molecule which degrades quickly has a low DT50 value and thus the impact of this species on the environment is reduced if the degradation products are harmless. On the contrary, the environmental impact of species with a high DT50 value can be substantial even if the molecule is only moderately toxic. Gavrilescu (2005) classified persistence of pesticides according to the DT50 value into three groups, where pesticides with DT50 < 30 days are non-persistent pesticides, compared to pesticides with DT50 > 100

However, deviations from the first-order degradation of pesticides have been reported. Typically, a fast initial degradation is followed by a gradual decrease in the degradation rate and eventually a very slow degradation. The gradual change in degradation rate may be better described by using two rate constants instead of one (Beulke & Brown, 2001; Henriksen et al., 2004; Ma et al., 2004; Sanchez et al., 2003). The two-compartment model (Equation 8) describes the degradation process as shared between two different compartments, where degradation proceeds at different rates (*k1* and *k2*). The two constants,

*des sor*

*1/n* (5)

*<sup>1</sup> -k t C(t) = ae* (6)

*k* (7)

following equation:

sorption is reversible.

equation:

equation:

days which are persistent.

**2.2 Degradation processes** 

reactions (Chen et al., 2005; Kuhad et al.; Ward & Singh).

The proposed metabolic pathway of IMI in the soil is shown in Figure 3. Two main routes of metabolism responsible for the degradation of IMI were identified. The first step is hydroxylation of the imidazolidine ring leading to the mono- and dihydroxylated compounds, followed by loss of water to yield the olefinic compound. The second important degradation step starts with dehydrogenation of the imidazolidine ring to form desnitrometabolit with further oxidation to 6-chloronicotinic acid (6-CNA).

Fig. 3. The proposed metabolic pathway of IMI in the soil.

Behavior and Fate of Imidacloprid in Croatian Olive Orchard Soils Under Laboratory Conditions 497

through a 2-mm sieve prior to use. They were selected on the basis of their texture (mechanical composition), pH values, OC content and CEC. The soils have never been treated with IMI, as verified by analyzing its residues in the soil. Selected physicochemical

pH CEC*<sup>a</sup>*

Krk I sandy clay 21.12 7.12 25.67 (±0.74) 2.98 (±0.90) 0.55 Krk II sandy loam 15.38 6.88 14.01 (±0.63) 1.93 (±0.23) 0.42 Istria I clay loam 34.28 4.76 34.19 (±0.99) 5.65 (±0.45) 1.30 Istria II clay 47.21 6.35 49.16 (±0.31) 16.21 (±0.09) 1.91

The mechanical composition of the soil samples was determined by sedimentation using the "pipet method" (Kroetsch & Wang, 2007). Soil samples pH values were measured in a soil + deionised water and in a soil + 0.01 M calcium chloride suspension (1:2.5, w/v). The MP 220 laboratory pH meter (Metler Toledo, Germany) was used for pH determination in aqueous phase. Hydrolitic acidity (HA) was determined by the Kappen method (Hendershot et al., 2007 ), CEC was measured using ammonium replacement (Sumner & Miller, 1996), while Na, K, Mg and Ca were analyzed by Atomic Absorption Spectrophotometer (Perkin Elmer Analyst, USA). The OC content of the soils was determined spectrophotometrically (Cary

100 Bio WINUV, Varian, Australia) by dichromate method (Darrel & Nelson, 1996 ).

In the present study, the IMI sorption by soils was quantified using the standard batch equilibrium method (OECD, 2000). The predetermined mass of each soil (5 g), in triplicate, was equilibrated with 25 mL of aqueous solutions of IMI by shaking in an rotary agitator (Unimax 1010, Heidolph, Germany) at 20 (±1)° C for 48 h to achieve equilibrium. The equilibrium time was determined according to previous sorption kinetics studies of the IMI sorption (Capri et al., 2001; Nemeth-Konda et al., 2002). Initial insecticide solutions, in the concentration range of 0.1, 0.25, 0.5, 1, 2.5, 5, and 10 mg/L respectively, were prepared in the background 0.01 M calcium chloride and 100 mg/L mercury chloride solution from stock IMI solutions prepared in HPLC-grade acetonitrile. Calcium chloride solution was used as background electrolyte in order to minimize ionic strength changes and to promote flocculation. Mercury chloride was added to the pesticide solution as a biocide to prevent any microbial activity during the sorption experiment. After equilibration, the suspensions were centrifuged at 4000 rpm for 30 min at 20 (±1) °C (BR4i Multifunction, Thermo electron corporation, France) to separate the liquid and solid phases. After filtration through a polypropylene hydrophilic filter of 0.45 µm (Whatman, Puradisc 25 TF, USA) the aqueous phase was analyzed by High Performance Liquid Chromatography (HPLC) using a Thermo Separation Products (Spectra System, USA) liquid chromatographic system, as described in the section 3.6. Blank samples without soil were also prepared in the same way and used to account for possible losses due to the volatilization and sorption of IMI to the cuvette walls. The average system losses were shown to be consistently lower than 3.4% of the initial

(cmol/kg)

HA*<sup>b</sup>* (cmol/kg) OC*<sup>c</sup>* (%)

properties of the tested soils are given in Table 1.

(%)

*<sup>a</sup>* cation exchange capacity; *<sup>b</sup>* hydrolitic acidity; *<sup>c</sup>* organic carbon content.

Table 1. Physicochemical properties of the tested soils.

**3.3 Batch sorption-desorption experiments** 

Soil Textural classes Clay

The main metabolites of IMI which have been identified in the soil include IMI-urea, 6- CNA, and 6-hydroxynicotinic acid (Rouchaud et al., 1996), which ultimately degrades to CO2 (Scholz & Spiteller, 1992). For instance, depending on the soil type, IMI labeled with imidazolidin-14C had a maximum mineralization to CO2 of 8.8% or 14% after incubation for 12 weeks (Anderson, 1995; as reviewed in Mulye, 1996). In soils, when conditions were anaerobic and without light exposure, IMI was found to be readily decomposed, resulting in desnitro-IMI as the main transformation metabolite (Heim et al., 1996; as reviewed in Mulye, 1996). The desnitro-IMI produced under dark, anaerobic conditions has been found to be more persistent than its parent compound (Fritz & Hellpointner, 1991; as reviewed in Mulye, 1995). The major transformation products resulting from incubation under nonsterile, aerobic conditions and light exposure were desnitro-IMI, IMI-urea, 6-CNA and an unknown compound. Both desnitro-IMI and IMI-urea are highly water soluble, with solubility of 180 – 230 g/L and 9.3 g/L at 20°C, respectively (Krohn, 1996a, 1996b; as reviewed in Mulye, 1996), which is much higher than IMI's solubility, while 6-CNA has been found to be more toxic to honey bees than IMI itself.
