Assessment of the environmental fate of and their ecotoxicological impact on soil microorganisms : the case of chlorpyrifos, isoproturon and tebuconazole Véronika Storck

To cite this version:

Véronika Storck. Assessment of the environmental fate of pesticides and their ecotoxicological impact on soil microorganisms : the case of chlorpyrifos, isoproturon and tebuconazole. Life Sciences [q-bio]. Université de Bourgogne, 2016. English. ￿tel-02801785￿

HAL Id: tel-02801785 https://hal.inrae.fr/tel-02801785 Submitted on 5 Jun 2020

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés.

Distributed under a Creative Commons Attribution - ShareAlike| 4.0 International License

University of Burgundy, Dijon (France) Doctoral School of Environment and Health (ED E2S), Dijon (France)

National Institute of Agronomical Research (INRA), Dijon (France)

DOCTORAL THESIS

by

Veronika Storck

Assessment of the environmental fate of pesticides and

their ecotoxicological impact on soil microorganisms

The case of chlorpyrifos, isoproturon and tebuconazole

Intended viva: 5th December 2016

Jury members:

Supervisor: Dr. Fabrice Martin-Laurent INRA, Dijon, France

Co-supervisor: Dr. Marion Devers-Lamrani INRA, Dijon, France

Reviewers: Dr. Edward Topp Agriculture and Agri-Food, London, Canada Dr. Kathrin Fenner Eawag, Dübendorf, Switzerland

Examiners: Dr. Isabelle Batisson CNRS, Clermont-Ferrand, France Dr. Stéphane Pesce IRSTEA, Lyon, France

Jury president: Prof. Jean Guzzo University of Burgundy, Dijon, France

The research of this thesis was carried out at the National Institute of Agronomical Research (INRA) in Dijon (France), at the Catholic University of the Sacred Heart in Piacenza (Italy) and at Enoveo srl. in Lyon (France).

The research was financially supported through the PhD studies of Veronika Storck at the Doctoral School of Environment and Health (ED E2S) at the University of Burgundy funded by the French Ministry of Education and Research (MESR) (grant number 2013-31), and through the European Marie Curie project ‘Love-to-hate’ (FP7 – PEOPLE – 2012 – IAPP) funded by the European Commission (grant number 324349).

INRA of Dijon, France

Assessment of the environmental fate of pesticides and their ecotoxicological impact on soil microorganisms

The case of chlorpyrifos, isoproturon and tebuconazole

Veronika Storck

born 5th May 1988 Essen, Germany

Supervisor: Dr. Fabrice Martin-Laurent (INRA, Dijon, France)

Co-supervisor: Dr. Marion Devers-Lamrani (INRA, Dijon, France)

Reviewers: Dr. Edward Topp (Agriculture and Agri-Food, London, Canada) Dr. Kathrin Fenner (Eawag, Dübendorf, Switzerland)

Examinants: Dr. Stéphane Pesce (IRSTEA, Lyon, France) Dr. Isabelle Batisson (CNRS, Clermont-Ferrand, France)

Jury president: Prof. Jean Guzzo (University of Burgundy, Dijon, France)

‘We cannot solve our problems with the same thinking we used when we created them’

Albert Einstein

Contents

Thesis outline 10

Chapter 1 General introduction 16

Chapter 2 Towards a better policy for the European Union 52

Chapter 3 Pesticide dissipation in agricultural soil 82

Chapter 3.1 Dissipation and adsorption of isoproturon, 86 tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Chapter 3.2 Identification and characterization of 118 tebuconazole transformation products in soil by combining suspect screening and molecular typology

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 158

Chapter 4.1 Lab-to-field evaluation of the ecotoxicological 162 impact of three pesticides on the soil bacterial diversity by Illumina nexy-generation sequencing

Chapter 4.2 Impact of repeated treatments of the pesticides 204 chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Chapter 5 Possibilities and limitations of using genes as bioindicators 248 for potential in soil – the case of pdm- encoded isoproturon degradation

Chapter 6 General discussion 274

Summary 306

Acknowledgements 310

List of abbreviations 314

List of authors 320

List of publications 324

List of tables 330

List of figures 336

Thesis outline

Preview of this thesis

This thesis is divided into six chapters (Thesis outline Figure). Chapter 1 is a general introduction containing background information on the topics that chapters 2 to 5 are based upon. It introduces the pesticide benefit to humanity regarding its important role in modern agriculture, and its downside regarding adverse effects in the environment and for human health. It indicates that the environmental risk assessment procedure of pesticides needs improvements that may be achieved by reforms of the current pesticide policy, and states that especially the impact of pesticides on soil microorganisms important for ecosystem functioning is not sufficiently evaluated during the pesticide authorization process. Moreover, this chapter provides an overview of currently available methods able to evaluate potential pesticide impacts on soil microorganisms, including (i) quantitative and qualitative approaches in analytical chemistry to estimate the exposure of an ecosystem to pesticides and their transformation products (TPs), and (ii) International Standard Organization (ISO) approaches and ‘under-development techniques’ such as Illumina next-generation sequencing to study the response of soil microorganisms to pesticide exposure. Furthermore, this chapter presents the interest of using pesticide degradation-encoding genes as bioindicators sensitive to pesticide exposure. It explains the contribution of this thesis to a European Marie Curie project (‘Love-to-Hate’) and describes the main characteristics of the three chosen model pesticides chlorpyrifos (CHL), isoproturon (IPU) and tebuconazole (TCZ). Finally, the chapter provides the aim of this thesis and the research questions that were aimed to answer by this research. Chapter 2 focuses on pesticide environmental risk assessment. It aims to support the ongoing debate of the society about weaknesses in the current pesticide legislation in Europe. The pesticide authorization process is thoroughly explained and weak points in the pesticide environmental risk assessment are identified. Furthermore, the chapter discusses possibilities for improving the robustness and reliability of the pre- and post- registration regulatory framework of pesticides. Chapter 3 investigates the dissipation of CHL, IPU or TCZ in an agricultural soil (agricultural field in North Italy) and is subdivided into two parts. Chapter 3.1 aims to estimate the level and duration of the exposure of soil microorganisms to the three studied pesticides (applied at three different concentrations). Their dissipation and the

Thesis outline 11

formation and dissipation of their main TPs in soil are quantified, as evaluated in a microcosm and a field study. Moreover, the sorption of these molecules to soil is estimated from a batch study. As reference standards of potentially interesting TCZ TPs were not available, their formation, dissipation and soil sorption were not evaluated in this chapter. This lack induced the additional study presented in Chapter 3.2 describing a newly combined approach of (i) suspect screening for the identification of known and unknown (suspected) molecules without the imperative need of reference standards and (ii) molecular typology for the characterization of identified molecules. The development of this approach enabled the comprehensive identification and characterization of TCZ TPs in soil during the field study mentioned before. After defining the pesticide exposure scenario via the research on the environmental dissipation of pesticides, the aim of Chapter 4 is the evaluation of the ecotoxicological impact of pesticides on soil microorganisms. In Chapter 4.1, the impact of a single application of the three pesticides on soil bacteria (lower tier pesticide exposure scenario) is evaluated in the microcosm and field studies described before. To achieve this aim, soil DNA was extracted and bacterial DNA (the hypervariable V3-V4 region of the 16S rDNA) was amplified and then analyzed by Illumina next-generation sequencing. In Chapter 4.2, the impact of the repeated application of CHL and TCZ (higher tier pesticide exposure scenario) on their mineralization pattern in soil, on soil buffering functions for other molecules, and on soil bacteria (the latter evaluated by Illumina next-generation sequencing) was assessed. In Chapter 5, the possibility of using pesticide degradation-encoding genes as bioindicators for the pesticide degradation potential of a soil was tested by quantifying the abundance of pdmA and pdmB genes (involved in IPU degradation) as bioindicators for IPU degradation in two Central Eastern French soils in a microcosm and a field study. Chapter 6 is a general discussion critically analyzing the findings of this thesis. It describes the feasible changes of the current pesticide policy that would be necessary to make the pesticide authorization process scientifically more reliable. Moreover, forces and weaknesses of the proposed changes are discussed. Furthermore, this chapter provides a critical review on the approaches and methods used in this thesis. The originality as well as weak points of both the experimental set-ups and applied methods are elaborated and improvement suggestions are presented. Moreover, the findings for each studied pesticide are separately discussed and compared to the knowledge found

Thesis outline 12

in the literature. It is indicated where further research may be necessary. In addition, suggestions for the establishment of a comprehensive and responsive database consisting of research findings obtained by a broad combination of (alternative) techniques for a more reliable knowledge on the environmental behavior of pesticides are provided.

Thesis outline Figure. Graphic presentation of the thesis outline.

Thesis outline 13

Chapter 1

General introduction

Veronika Storck

Environmental pollution by anthropogenic compounds

The ambition of mankind to improve life quality has great benefits for humanity but has also strongly affected the environment for centuries. According to the Nobel Prize chemist Paul Crutzen, our planet has entered the Anthropocene, a new epoch that is dominated by intense human activities causing global environmental changes (Crutzen, 2002). Nowadays, an incredibly large variety of compounds of anthropogenic origin is present in the environment all over the world and threatens environmental quality and human health (Hutzinger, 2013; Lewis and Maslin, 2015). Among these compounds, pesticides are one of the few chemicals that are intentionally released at large scales into the environment (e.g. into agro-ecosystems to protect crops). As a result of their widespread use, pesticide residues are major environmental contaminants contributing to environmental pollution (Rathore and Nollet, 2012). This paradox has its origin in weighing up pesticide benefits against pesticide hazards.

Pesticides

Today, around 2.4 million tons of pesticides (active substances) are released into the environment worldwide each year (U.S. EPA, 2011), whereof 80 % are used in agriculture (Enserink et al., 2013). Agricultural ecosystems are particularly important for humanity, because they provide us with one of our basic needs: food. Since the early 1940s, pesticides emerged as a powerful tool to secure food yield and quality by protecting crops from various pests, and since then have been extensively used all over the planet (Rathore and Nollet, 2012; Singh, 2012). They can be classified according to their target organisms with three majorly used groups: used against insects, used for growth control of various weeds, and effective to control fungal diseases (Gilden et al., 2010), which account for 30 % (insecticides), 40 % (herbicides) and 10 % (fungicides) of the global pesticide market (Enserink et al., 2013). Other possibilities to classify pesticides are based on their mode of action (Mallory- Smith and Retzinger, 2003; Maltby et al., 2009; Sparks and Nauen, 2015) or their molecular characteristics (Servien et al., 2014). More than 500 different plant protection products (PPPs, containing the active substance(s) and other chemicals) are used worldwide in agriculture (Arias-Estevez et al., 2008). Figure 1.1 shows a world map

Chapter 1 General introduction 18 illustrating the important pesticide use all over the globe between 2005 and 2009 (Enserink et al., 2013).

Figure 1.1. Worldwide pesticide use (amount of pesticide in kg per ha per year) between 2005 and 2009 (Enserink et al., 2013).

Unfortunately, only a tiny part of sprayed pesticides ends up in its target organism, while the rest stays on crops and reaches the agricultural soil, from where it can contaminate surrounding environments such as rivers or groundwater, can enter the food chain and harm non-target organisms (Jeyaratnam, 1990; Malaja et al., 2013; Zhang et al., 2011). Thus, the behavior of pesticides in soil usually determines the extent of pollution (Barriuso et al., 1996). The reduction of their environmental impact requires understanding of the processes to which they are exposed. Depending on their characteristics, pesticides can be rapidly degraded (Yu et al., 2003) or be quite persistent in the environment (Edwards and Adams, 1970). They can sorb to agricultural soil (Baskaran et al., 2003) or be volatilized to the atmosphere or transported to water resources by run-off or leaching (Lindahl and Bockstaller, 2012) (Figure 1.2). Moreover, pesticides tend to turn into transformation products (TPs) by abiotic or biotic transformation in the environment, which can imply greater hazards to non-target organisms than the parent molecule (Fenner et al., 2013). Soil microorganisms play an important role in the dissipation processes as they can contribute to the environmental biodegradation of pesticides, which can constitute a

Chapter 1 General introduction 19 nutrient and energy source to them (Copley, 2009). Extensive knowledge about these processes is crucial to predict potential risks for the environment and needs to be comprehensively involved in the environmental risk assessment of pesticides.

Figure 1.2. Main abiotic and biotic processes involved in pesticide dissipation in the environment (modified from Barriuso et al., 1996).

Environmental risk assessment of pesticides

In the European Union (EU), a statutory pesticide authorization process is the prerequisite for the market permission of a pesticide. After the invention of a new pesticide, its behavior in the environment and its potential hazards for non-target

Chapter 1 General introduction 20 organisms are investigated during an environmental risk assessment procedure at EU and national level prior to its authorization. Only active substances that are registered on the EU’s list of approved active substances and subsequently authorized as PPPs by each EU Member State are released into the environment. Despite these efforts to minimize environmental risks, the safe use of pesticides appears to be one of the biggest challenges of sustainable agriculture, because pesticide residues remain one of the most widespread pollutant groups worldwide (Rathore and Nollet, 2012). The pesticide authorization process is heavily criticized by the society for being relatively inefficient and for compromising a non-precautionary principle (O’Reilly, 2016). In this context, improvements of the current pesticide policy may be a good leverage, and are elaborated in an opinion paper (Chapter 2). Especially the impact of pesticides on soil microorganisms is not sufficiently evaluated during the environmental risk assessment of pesticides. This may be due to the fact that the EU soil protection framework that was proposed to the European Commission (EC) more than 10 years ago (Van-Camp et al., 2004) is still not implemented. Soil is a limited resource that provides a unique habitat for a large variety of organisms including soil microorganisms. One gram of soil can contain one billion of bacteria and one million of fungi that support a range of processes with impacts at the global scale (Blum, 2006; Bondeau et al., 2007; Graham et al., 2016). They contribute to ecosystem functions such as food supply, water quality, carbon cycling, climate regulation, and pollutant degradation (Haygarth and Ritz, 2009; Hillel, 2009), whose annual economic value is estimated to be around 1.3 trillion Euros (Pimentel et al., 1997). The European soil biodiversity expert group identified intensive agricultural use as one of the major pressures on soil organisms (Gardi et al., 2013; Jeffery et al., 2010). Although the preservation of agricultural soil quality was identified as one of the main objectives of the 21st century (Jones et al., 2009; Lal, 2008; Lichtfouse et al., 2009), the soil protection framework is still not implemented as according directive. Till today, the impact of pesticides on soil microorganisms is only slightly reviewed in the respective risk assessment documents, solely based on carbon and nitrogen mineralization tests in soil. These tests are not good indicators for the microbial toxicity of pesticides because these processes can be maintained regardless of significant toxic effects, as tolerant microbes can increase their population at the expense of sensitive microbes (Van Beelen and Doelman, 1997).

Chapter 1 General introduction 21

Recently, Martin-Laurent et al. (2013) proposed to revise the regulatory framework regarding the assessment of the toxicity of pesticides on soil microorganisms. Moreover, the European Food Safety Authority (EFSA) proposed new specific protection goals covering soil ecosystem services potentially affected by the use of pesticides (EFSA, 2010), and identified microorganisms as one of seven key drivers of soil ecosystem services (such as nutrient cycling, water purification, soil remediation, waste treatment, soil formation and retention, pest and disease regulation, genetic resources, education and inspiration, and food) to be protected. EFSA proposed specific protection goals to exclude (i) unacceptable effects on functions of microbial communities and (ii) a decrease of biodiversity. The conservation of microbial diversity in soil is important, as high diversity is thought to provide an ecological insurance with high tolerance to a range of environmental stress. Indeed, it was found that the functional operating range (FOR, i.e. the range of environmental conditions under which a community or ecosystem is able to maintain its functions) of microbial communities decreases when microbial diversity decreases (Hallin et al., 2012). This implies that a high soil microbial diversity ensures a more robust maintenance of ecosystem functions at extreme environmental conditions, while a lower diversity causes more limited ecosystem functions under extreme environmental conditions (Figure 1.3).

Figure 1.3. Presentation of the importance of soil microbial diversity for the maintenance of ecosystem functions under changing environmental conditions.

Chapter 1 General introduction 22

Furthermore, after exposure to a stress, microbial communities have the capability to recover by resilience. The concept of ecological recovery is as an important issue in environmental science (Montoya et al., 2012) (Figure 1.4). This concept implies that an ecological metric (e.g. population size) may decrease in response to the exposure to a chemical pollutant (such as pesticide residues) above environmental thresholds and thus may not act within its normal operating range (NOR) anymore. After the decline of the exposure to a given stress (e.g. exposure), the ecological metric may return to its NOR (ecological recovery).

Figure 1.4. Presentation of the ecological recovery of an ecological metric (e.g. population size) after exposure to a pollutant (e.g. pesticide residues) above the effect threshold (modified from EFSA, 2016a).

EFSA has internalized this concept and recently suggested to consider the temporal and spatial ecological recovery of non-target organisms exposed to EU- regulated stressors in environmental risk assessment frameworks (EFSA, 2016a). However, these two new concepts (i.e. protection of soil ecosystem services and ecological recovery) have only been proposed by EFSA in two scientific opinions that have not yet been fully implemented in the pesticide regulatory frameworks. In this context, as underlined in a recent scientific opinion of EFSA addressing the state of the science on pesticide risk assessment on soil organisms (EFSA, 2016b), the

Chapter 1 General introduction 23 development and implementation of comprehensive and robust methods for the evaluation of potential pesticide effects on soil microorganisms could contribute to the improvement of pesticide environmental risk assessments.

Available approaches for assessing pesticide impacts on soil microorganisms

Assessment of the exposure scenario of soil microorganisms to pesticides

The first step in the assessment of ecotoxicological impacts of pesticides on soil microorganisms is a profound knowledge about their dissipation to enable the estimation of the level and the duration of the exposure of the soil microorganisms to pesticides. In this context, the main abiotic and biotic processes responsible for pesticide dissipation have to be studied, such as (i) sorption of pesticides to soil components, (ii) abiotic degradation or transformation and biotic (metabolic or co-metabolic) degradation or transformation, and (iii) transfer to different parts of the environment by various means (run-off, leaching, volatilization). It is recommended by the Organisation for Economic Co-operation and Development (OECD) to estimate the sorption behavior of a pesticide to soil by application of the standard batch equilibrium method (OECD-Guideline 106, 2000; Chapter 3.1). However, although it estimates the soil sorption affinity of pesticides, this technique has some drawbacks because it does not take into account the mobility of a pesticide. From this point of view, sorption studies carried out in undisturbed soil columns are more realistic (Dagès et al., 2015), notably because they take into account the rapid pesticide transport through soil macropores (Pot et al., 2005). To study the degradation of pesticides and their known TPs in soil, targeted analytical approaches able to quantify a known molecule can be used. One approach is radiorespirometry to monitor the evolution of 14CO2 from a 14C-labelled pesticide. This approach has the advantage of enabling the establishment of a mass balance of 14C- residues in soil (14CO2, 14C-extractable and 14C-bound pesticide residues) and is usually used in the research for pesticide risk assessment studies. Other approaches such as high performance liquid chromatography with photodiode array detection (HPLC-PDA) (Chapter 3.1) or liquid chromatography tandem mass spectrometry (LC-MS/MS) are

Chapter 1 General introduction 24 used depending on the molecule’s features to quantify the disappearance of a given pesticide and the formation of TPs. However, from the regulatory point of view, only a limited number of molecules are defined as targets for analytical monitoring methods, so-defined ‘ecotoxicologically relevant’ pesticide TPs or TPs ‘relevant for groundwater contamination’ (EU-Regulation 1107/2009/EC; EU-Regulation 283/2013/EC; OECD- Guideline ENV/JM/MONO (2007)17, 2007). Moreover, the quantification of TPs is only possible by using synthesized reference standards. For these reasons, most of the available studies focus on ‘relevant’ or already known and synthezised TPs and do not provide any information on other TPs possibly found in the environment. The analytical techniques able to offer more comprehensive insights into the survey of pesticide TPs are high resolution mass spectrometry (MS) such as quadrupole time-of-flight MS (QTOF-MS) and Fourier transform MS (FT-MS), and nuclear magnetic resonance (NMR) spectroscopy (Ibanez et al., 2005), able to detect unknown molecules (for which reference standards are not available). These approaches are often used in forensics (Reitzel et al., 2012) or medicine (Zeng et al., 2013), where QTOF-MS is used for screening for new unknown molecules and NMR spectroscopy is used for the structural elucidation of a molecule. Although NMR spectroscopy is a powerful tool for the elucidation of molecular structures, it is less appropriate for the analysis of pesticide TPs in environmental samples due to its rather poor sensitivity that is generally three orders of magnitude lower than MS (Ibanez et al., 2005). QTOF-MS was shown to readily detect pesticide TPs by suspect screening of water samples with some limitations in the elucidation of their molecular structure (Sancho et al., 2006). Given the high potential of this last approach to search for suspected molecules in environmental samples, it was tested on the detection of pesticide TPs in soil (Chapter 3.2). By the application of these different approaches, the environmental fate of pesticides can be estimated and the scenario of exposure of soil microorganisms to pesticide residues can be determined.

Assessment of the ecotoxicological impact of pesticides on soil microorganisms

Having estimated the scenario of exposure of soil microorganisms to pesticides, the second step consists in estimating their ecotoxicological effects. Soil microorganisms such as bacteria may sensitively react (or not) to pesticide exposure, either by direct or by indirect effects (Imfeld and Vuilleumier, 2012). Microbial ecotoxicology, a

Chapter 1 General introduction 25 multidisciplinary science interfacing microbial toxicology, microbial ecology and chemistry/physics investigates the potential effects caused by pollutant exposure (such as pesticide residues) on microorganisms (such as soil bacteria) in an ecosystem (such as agricultural soil) (Ghiglione et al., 2016) (Figure 1.5).

Figure 1.5. Microbial ecotoxicology, a multidisciplinary science combining microbial toxicology, microbial ecology and chemistry/physics to understand and describe the complex interactions between pollutants, microorganisms and the ecosystem (modified from Ghiglione et al., 2016).

The evaluation of pesticide impacts on soil microorganisms supporting ecosystem functions remains challenging because pesticide-induced processes in soil are highly complex, and experimental designs (microcosm vs. field study, varying applied pesticide doses etc.) and applied methods variate between different laboratories (Philippot et al., 2012) with the result that research findings often differ from one study to another. To counteract this bias, soil microbiologists and microbial ecologists have developed an impressive toolbox to study variations in the abundance, activity and

Chapter 1 General introduction 26 diversity of soil microbial communities in response to the exposure to various stressors (Martin-Laurent et al., 2013) (Figure 1.6).

Figure 1.6. Conceptual presentation of the ecotoxicological impact caused by pesticides on the abundance, activity and diversity of the soil microbial community and possible consequences on supported ecosystem functions.

Microbial ecology methods that are able to analyze those three components of a microbial community are under constant development. The Soil Biological Quality working group of the International Standard Organization (ISO/TC 190/SC4/WG4) with one of the main goals to counteract the loss of soil biodiversity (Nortcliff, 2002) defines internationally standardized (ISO) approaches with the potential to estimate the quality of soils by assessing the ecotoxicological impact of pollutants (such as pesticides) on soil microorganisms (Philippot et al., 2012). Currently available ISO methods assessing the abundance, the activity and/or the diversity of soil microorganisms are listed in Table 1.1.

Chapter 1 General introduction 27

Table 1.1. Available ISO methods in soil microbiology (modified from Martin-Laurent, 2013 and Philippot et al., 2012).

ISO target year method reference Sampling – part 6: Guidance on the collection, handling, and abundance, activity and ISO 10381 2009 storage of soil under aerobic conditions for the assessment of diversity microbiological processes, biomass, and diversity in the laboratory ISO 11063 2012 Method to directly extract DNA from soil samples Determination of soil microbial biomass - part 1: Substrate- ISO 14240-1 1997 induced respiration method Determination of soil microbial biomass - part 2: - ISO 14240-2 1997 abundance extraction method ISO 10832 2009 Effects of pollutants on mycorrhizal fungi-germination test Estimation of abundance of selected microbial gene sequences by ISO 17601 2016 quantitative PCR from DNA directly extracted from soil Determination of abundance and activity of soil microflora using abundance and activity ISO 17155 2002 respiration curves Laboratory incubation systems for measuring the mineralization of ISO 14239 1997 organic chemicals in soil under aerobic conditions Laboratory incubation systems for measuring the mineralization of ISO 15473 1997 organic chemicals in soil under anaerobic conditions Determination of nitrogen mineralization and nitrification in soils ISO 14238 1997 and the influence of chemicals on these processes Guidance on laboratory testing for biodegradation of organic ISO 15473 2002 chemicals in soil under anaerobic conditions ISO 16072 2002 Laboratory methods for determination of microbial soil respiration Determination of dehydrogenase activity in soils - part 1: Method ISO 23753-1 2005 using triphenyltetrazolium chloride (TTC) Determination of dehydrogenase activity in soils - part 2: Method ISO 23753-2 2005 activity using iodotetrazolium chloride (INT) Measurement of enzyme activity patterns in soil samples using ISO 22939 2010 fluorogenic substrates in micro-well plates Measurement of enzyme activity patterns in soil samples using ISO 20130 under development colorimetric substrates in micro-well plates Determination of potential nitrification and inhibition of ISO 15685 2012 nitrification - rapid test by ammonium oxidation Contact test for solid samples using the dehydrogenase activity of ISO 18187 2016 Arthrobacter globiformis Simple laboratory assessments for characterizing the ISO 20131-1 under development denitrification in soil - part 1: Soil denitrifying enzymes activities Simple laboratory assessments for characterizing the ISO 20131-2 under development denitrification in soil - part 2: Assessment of the capacity of soils to

reduce N2O Determination of soil microbial diversity - part 1: Method by PFLA ISO 29843-1 2010 analysis and PLEL analysis diversity Determination of soil microbial diversity - part 2: Method by PLFA ISO 29843-2 2011 analysis using the ‘simple PLFA extraction method’

The distribution of the number of standardized methods assessing the activity, the abundance and/or the diversity of soil microorganisms is not entirely equal, with seven methods targeting abundance, 16 methods targeting activity, and four methods targeting diversity. Another non-ISO method assessing the microbial activity is stable isotope probing (SIP), a technique allowing the characterization or identification of microbial populations actively involved in specific metabolic processes. Regarding approaches to assess the diversity of soil microbial communities, the ISO 29843-1 and -2 open the path for other standards such as approaches based on taxonomic and

Chapter 1 General introduction 28 functional microarrays (Philippot et al., 2012). The ISO 11063 for soil DNA extraction provided a solid basis for the development of ISO 17601 for the quantification of microbial groups by quantitative polymerase chain reaction (qPCR). In addition, ISO 11063 offers the basement for employing new high-throughput sequencing technologies such as Illumina next-generation sequencing of rDNA amplicons to study the diversity of soil microorganisms (Chapter 4.1; Chapter 4.2). In the past, other techniques based on the amplification of rDNA have been applied to assess the soil microbial diversity, such as clone libraries (cloning of PCR products and then sequencing of individual gene fragments) or fingerprinting approaches including denaturing and temperate gradient gel electrophoresis (DGGE and TGGE) (intermediate resolution techniques able to separate PCR product mixtures similar in length but different in base pair composition) and ribosomal intergenic spacer analysis (RISA) and amplified ribosomal DNA restriction analysis (ARDRA). Although they have been intensively applied for two decades, these approaches were largely abandoned on the strength of next-generation sequencing methods.

Use of bioindicators to assess the exposure and potential degradation of pesticides

In response to repeated pesticide exposure, a fraction of the microbial community can adapt to this pressure and acquire the ability to use pesticides as an additional nutrient and/or energy source by the expression of catabolic pesticide degradation-encoding genes, leading to accelerated pesticide degradation (Topp, 2003). It was recently proposed that the abundance of pesticide-degrading genes can be used as a proxy of the pesticide-degrading capability of soil microbial communities (Monard et al., 2013). Indeed, several studies carried out on a range of pesticides including atrazine, 2,4-dichlorophenoxyacetic acid (2,4-D) and diuron showed a positive correlation between the abundance of pesticide-degrading genes (atzA, tfdA and puhA, respectively) and the occurring pesticide degradation (Baelum et al., 2008; Martin- Laurent et al., 2003; Pesce et al., 2013). Monitoring of the pesticide-degrading genetic potential by targeting catabolic genes can not only be used as bioindicator for pesticide degradation but also as bioindicator for pesticide exposure in soil (Monard et al., 2013). This approach is based on direct soil DNA extraction (ISO 11063, 2012) and gene

Chapter 1 General introduction 29 detection by qPCR (ISO 17601, 2016) or functional gene microarrays (Fenner et al., 2013) and relies on the knowledge of the pesticide degradation pathway and in particular of gene sequences coding for the enzymes involved in it. Unfortunately, only a few pesticide degradation pathways have been discovered in sufficient details, wherefore large-scale applications of monitoring approaches based on gene- bioindicators are limited. It is noteworthy that a deep understanding of degradation-encoding genes and notably their clear attribution to the biodegradation process is a prerequisite for reliable conclusions about pesticide exposure or degradation (Bouskill et al., 2007; DeBryn et al., 2007; Kazy et al., 2010; Ogunseitan et al., 2000; Peng et al., 2010). Several drawbacks can make it difficult to use gene-bioindicators for the prediction of pesticide degradation in soil. Indeed, two genes of 99 % similarity can catalyze two different reactions (e.g. the genes atzA and triA that are 99 % similar, but atzA catalyzes the transformation of atrazine while triA calatyzes the transformation of melamine) and may hardly be distinguished by qPCR, therefore potentially leading to an overestimation of pesticide degradation. Similarly, an overestimation of the pesticide degradation can occur if a targeted degrading enzyme is member of a large protein superfamily with various broad-spectrum specificity (Fenner et al., 2013). Moreover, genes with completely different sequences may encode for different enzymes catalyzing the same step in a pesticide degradation pathway, wherefore the quantification of only one of them (as the other may be unknown) can lead to an underestimation of the actual degradation. This was the case in a study by Monard et al. (2010) where the expression of an atrazine- degrading atzA gene in soil was not always detected when atrazine biodegradation occurred, which most likely resulted from the co-existence of trzN, atzB and atzC among the microbial community (Devers et al., 2007). In addition, abiotic pesticide degradation steps further contribute to the underestimation of the actual pesticide degradation or exposure, as it may be the case regarding the abiotic hydrolysis of the chlorpyrifos to its main degradation product 3,5,6-trichloro-2-pyridinol (Racke et al., 1996). On the top of that, the detection and quantification of pesticide-degrading genes in soil DNA targets the pesticide-degrading genetic potential and does not necessarily guarantee its expression as it may not only be present in active bacteria (Pietramellara et al., 2009; Prosser, 2002; Rocca et al., 2015; Wagner 1994). The quantification of mRNA transcribed from the catabolic gene in response to the presence of a pesticide may be a good option to specifically target active microorganisms involved in the

Chapter 1 General introduction 30 degradation process (Kong and Nakatsu, 2010). Indeed, gene expression of a Dehalococcoides sp. strain CBDB1 was successfully used as a functional indicator for chlorobenzoate dehalogenation in contaminated soils (Wagner et al., 2009). However, mRNAs are difficult to retrieve from complex environmental samples such as soil. Their ‘life time’ is very short making it difficult to interprete results at the level of microbial communities. For these reasons, the use of DNA remains the better alternative (Monard et al., 2013). Despite these difficulties, the possibility to use the quantification of catabolic genes as a proxy of the pesticide degradation ability of soil microbial communities was tested by monitoring isoproturon (IPU) mineralization and pdmA and pdmB genes (involved in IPU degradation) in a microcosm and a field study (Chapter 5). Usage of catabolic gene-bioindicators remains difficult to achieve but ideally it is a good completion to other techniques investigating the environmental fate of pesticides, such as measuring the mineralization of a pesticide (Piutti et al., 2002), determining pesticide concentrations in environmental samples (OECD-Guideline ENV/JM/MONO (2007)17, 2007), or detecting new or unknown pesticide TPs (Chapter 3.2). Furthermore, microbial gene sequences coding for pollutant-degrading enzymes can also be used to design probes spotted to DNA microarrays, such as GeoChip 5.0, which contains around 170000 distinct probes covering around 400000 coding sequences (including atz genes involved in atrazine degradation) from around 1500 functional gene families involved in various microbial functions (Gao et al., 2014), to screen soil DNA or RNA for the presence (DNA) or expression (RNA) of microbial functional genes of interest. All of these techniques are under constant development and have their advantages and disadvantages. Their combination can help to obtain a more comprehensive understanding of the environmental fate and ecotoxicological effects of pesticides on soil microorganisms and supported ecosystem services in agro- ecosystems.

This thesis in the context of a European Marie Curie project

This thesis was part of the European Marie Curie project ‘Love-to-Hate’ (FP7- PEOPLE-2012-IAPP) funded by the EC (Grant Agreement number 324349) to shed light into the complex interactions between pesticides and soil microorganisms in order to

Chapter 1 General introduction 31 test innovative methods to assess the ecotoxicological impact of pesticides on soil microorganisms. Within this project, pesticide effects on soil microorganisms are assessed from the yin and the yang point of view1. The yang side signifies the potential stimulation that pesticides may cause on microorganisms (leading to their adaptation), while the yin side represents their potential inhibition (leading to toxic effects) (Karpouzas et al., 2016) (Figure 1.7).

Figure 1.7. Illustration of the main research aim of the European Marie Curie ‘Love-to-Hate’ project: Establishment and development of innovative methods to better assess the ecotoxicological impact of pesticides on soil microorganisms.

The aim of the project was to establish standardized and novel methods for the assessment of the ecotoxicological impact of pesticides on soil microorganisms. With this purpose, the dissipation of three model pesticides in soil and their impact on soil microorganisms involved in ecosystem functions were investigated in a tiered lab-to- field approach. The three chosen pesticides were the organophosphate insecticide chlorpyrifos (CHL), the phenylurea isoproturon (IPU) and the triazole tebuconazole (TCZ) (Figures 1.8, 1.9, 1.10). These three pesticides are on the European market for quite some time (51 years (CHL), 41 years (IPU) and 28 years (TCZ)) and often detected in the environment worldwide (Börner et al., 2009; Hayes and Laws, 1991; Oerke et al., 1994). Their environmental risk assessment needs to be

1 In Chinese philosophy, the terms yin and yang signify two opposite yet related forces.

Chapter 1 General introduction 32 revised with up-to-date approaches to draw further conclusions on their potential ecotoxicological impact on soil microorganisms.

The studied pesticides

The organophosphate insecticide CHL (Figure 1.8) is one of the most extensively used moderately toxic insecticides with a broad spectrum of activity (Joseph and Zarate, 2015; Xu et al, 2008). CHL has a high sorption affinity to soil and its degradation proceeds mainly via abiotic or biotic hydrolysis with DT50 values (time in which 50 % of the applied quantity have depleted) ranging from 10 to 120 days in soil (Racke, 1993). CHL is known to generate (among others) the two TPs 3,5,6-trichloro-2-pyridinol (TCP) and CHL-oxon (Figure 1.8) that have the potential to impact non-target organisms. For instance, the hydrolytic TP TCP was suggested to have antimicrobial properties preventing the proliferation of CHL-degrading microorganisms (Racke, 1993). The desulfuration product CHL-oxon was found to be more toxic than its parent molecule (Duirk and Collette, 2006). The ecotoxicological impact of CHL on soil microorganisms is not sufficiently studied in the current environmental risk assessment. Indeed, the non- toxicity of CHL on soil microorganisms reported in the review report of the EC solely relies on carbon and nitrogen mineralization tests (EC, 2005). The five available risk assessment documents established by the EFSA do not contain any information about potential effects of CHL on soil microorganisms (EFSA, 2011a; EFSA, 2011b; EFSA, 2012; EFSA, 2014a; EFSA, 2015a). Many other studies concluded that CHL had effects on soil microorganisms, evaluated throughout changes in microbial biomass, respiration, microbial or enzyme activity or microbial diversity (Chu et al., 2008; Dutta et al., 2010; Eisenhauer et al., 2009; Fang et al., 2009; Gilani et al., 2010; Gupta et al., 2013; Jastrzebska, 2011; Pozo et al., 1995; Srinivasulu and Rangaswamy, 2013; Wang et al., 2010). Interestingly, in several studies, CHL was found to exert short-term inhibitory effects on soil microorganisms in several studies that did not persist (Fang et al., 2009; Pozo et al., 1995), suggesting a possible ecological recovery of soil microbial communities. In addition, various bacterial and fungal soil populations able to degrade CHL have been isolated from soils in response to repeated exposure to CHL (Table 1.2). Moreover, CHL pathways of biodegradation were described for several microorganisms, and several genes encoding for CHL-degrading enzymes have been identified, such as

Chapter 1 General introduction 33 the mdp gene (methylparathion hydrolase gene) (Yang et al., 2006; Li et al., 2007) and the opdA and opdE genes (organophosphate degradation genes) (Yang et al., 2003; Chino-Flores et al., 2012) (Table 1.2). The phenylurea herbicide IPU (Figure 1.9) is used for the control of annual grasses and broad-leaved weeds in weed and barley crops (Collings et al., 2013). As a result of its widespread use, IPU is frequently detected in surface and groundwater resources (Shark and Zullei-Seibert, 1995; Skark et al., 2004). IPU shows variable persistence in soil with DT50 values ranging from 3 to 200 days (Alletto et al., 2006). IPU degradation and its main TPs mono-desmethyl-IPU (MD-IPU), di-desmethyl-IPU (DD- IPU), 4-isopropyl-aniline (4-IA) and several hydroxylated compounds are pretty well known (Figure 1.9). Some TPs, in particular 4-IA, have been found to exert higher toxicities than their parent molecule (Alletto et al., 2006; Hussain et al., 2015; Tixier et al., 2002). The risks of IPU on soil microorganisms were interpreted as low in the risk assessment by the EFSA, as effects on carbon and nitrogen mineralization in soil were < 25 % (EFSA 2015b). Other studies concluded small to moderate (sometimes temporary) toxic effects of IPU on soil microorganisms, mostly based on decreasing microbial activity and biomass (Harden et al., 1993b; Kuriyal and Pandey, 1994; Perrin- Ganier et al., 2001; Schuster and Schröder, 1990; Tag-El-Din, 1982). It is noteworthy, that various studies observed an initial increase of soil respiration after IPU treatment (Harden et al., 1993; Perrin-Garnier et al., 2001; Schuster and Schröder, 1990). This was hypothesized to be due to a toxic effect of IPU on sensitive soil microorganisms that may have provided an additional easily degradable carbon source for the respiration of microorganisms resistant to IPU. As a result of repeated exposure to IPU, soil microorganisms were found to perform enhanced biodegradation of IPU. An important number of IPU-degrading microorganisms have been isolated and characterized (see Hussain et al., 2015 for a review). Interestingly, most of the known IPU-mineralizing bacterial isolates belong to the Sphingomonadaceae family (Bending et al., 2003; Hussain et al., 2011; Hussain et al., 2009; Sorensen et al., 2001; Sorensen et al., 2002; Sun et al., 2009). An overview of IPU-degrading bacterial and fungal strains is provided in Table 1.2. Recently, the pdmA and pdmB genes (N-demethylase) coding for two subunits of the enzyme catalyzing the first step of the IPU mineralization pathway were identified (Gu et al., 2013). In addition, the puhA and puhB genes initially found to be involved in diuron degradation in Arthrobacter globiformis D47 (Turnbull et al., 2001) and in

Chapter 1 General introduction 34

Mycobacterium bisbanense JK1 (Khurana et al., 2009) were as well shown to degrade IPU (Table 1.2). The triazole fungicide TCZ (Figure 1.10) is used to control the development of a range of fungal plant pathogens in various crops (D’Angelo et al., 2014; Keinath, 2015). It is rather persistent in soil with DT50 values ranging from 49 to 610 days (EFSA, 2014b; Strickland et al., 2014), and can reach surface and groundwater resources (Herrero- Hernandez et al., 2013; Sancez-Gonzales et al., 2013). The dissipation of TCZ in soil has been previously studied (Alvarez-Martin et al., 2016; Herrero-Hernandez et al., 2013; Potter et al., 2005; Strickland et al., 2004) but little is known regarding the existence, abundance and toxicity of potential TPs (Figure 1.10) in the environment (Obanda and Shupe, 2009; Potter et al., 2005; Strickland et al., 2004; White et al., 2010). The potential formation of various triazole TPs is particularly relevant from the ecotoxicological point of view, because they may be recalcitrant to biodegradation and they possibly interact with the hormone regulation network of non-target organisms by inhibition of the cytochrome P450-dependent conversion of lanosterol to ergosterol (Rieke et al., 2014; Shalini et al., 2011). As evaluated by the EFSA during the environmental risk assessment, TCZ was found to impact < 25 % of carbon and nitrogen mineralization in soil, which was interpreted as a low risk for soil microorganisms (EFSA, 2008; EFSA, 2014b; EFSA 2015b). Numerous other studies showed that TCZ has an impact on enzyme activities, microbial biomass and others, with strongest effects observed during the first month after exposure to this fungicide (Ferreira et al., 2009; Munoz-Leoz et al., 2011). Following repeated exposure to TCZ, a few studies reported the isolation of bacterial and fungal strains able to transform TCZ (Obanda and Shupe, 2009; Sehnem et al., 2010; Wallance and Dickinson, 2006) (Table 1.2). In contrast to the other two pesticides considered in this study, no microorganisms able to mineralize TCZ and no genes involved in TCZ degradation have yet been characterized. This brief literature review on these three pesticides highlights that new findings on potential ecotoxicological effects of the pesticides on non-target organisms are ongoing although the pesticides are on the market for decades. Further ecotoxicological studies carried out by applying a range of methods allowing the estimation of ecotoxicological effects of the pesticides on soil microorganisms may provide more clarity.

Chapter 1 General introduction 35

Figure 1.8. Molecular structures of CHL and its known environmental TPs. References for each molecule are given in the reference section.

Chapter 1 General introduction 36

Figure 1.9. Molecular structures of IPU and its known environmental TPs. References for each molecule are given in the reference section.

Chapter 1 General introduction 37

Chapter 1 General introduction 38

Figure 1.10. Molecular structures of TCZ and its known environmental TPs. References for each molecule are given in the reference section.

Chapter 1 General introduction 39

Table 1.2. List of bacterial and fungal strains and genes coding for degradation enzymes that are involved in the biodegradation of CHL, IPU or TCZ (modified from Chishti et al., 2013; Hussain et al., 2015).

bacteria/fungi/genes involved in biodegradation references Alcaligenes faecalis Yang et al., 2005 Bacillus cereus Liu et al., 2012 Bacillus licheniformis ZHU-1 Zhu et al., 2010 Bacillus pumilus sp. C2A Anwar et al., 2009 Enterobacter sp. Singh et al., 2003 Flavobacterium sp. TCC27551 Mallick et al., 1999 Klebsiella sp. Ghanem et al., 2007 bacterial strains Paracoccus sp. TRP Xu et al., 2008 Pseudomonas aeruginosa Lakshmi et al., 2008 Pseudomonas nitroreducens sp. PS-2 Korade and Fulckar, 2009 Pseudomonas stutzeri sp. B-CP5 Awad et al., 2011 Serratia sp. Xu et al., 2007 Sphingomonas sp. Li et al., 2007 CHL Stenotrophomonas sp. Yang et al., 2006 Synechocystis sp. PUPCCC 64 (cyanobacterium) Singh et al., 2011 Acremonium sp. GFRC-1 Kulshrestha and Kumari, 2011 Aspergillus terreus Omar, 1998 Cladosporium cladosporioides sp. Hu-01 Gao et al., 2012 Penicillium brevicompactum Omar, 1998 fungal strains Phanerochaete chrysosporium Bumpus et al., 1993 Trichoderma harzianum Omar, 1998 Trichosporon sp. Xu et al., 2007 Verticillium sp. DSP Fang et al., 2008 mpd (methylparathion hydrolase gene) Yang et al., 2006; Li et al., 2007 genes opdA (organophosphate degradation gene) Yang et al., 2003 opdE (organophosphate degradation gene) Chino-Flores et al., 2012 Arthrobacter globiformis D47 Cullington and Walker, 1999 Arthrobacter sp. N2 Tixier et al., 2002 Methylopila sp. TES El-Sebai et al., 2004 Mycobacterium brisbanense JK1 Khurana et al., 2009 Pseudomonas aeruginosa JS-11 Dwivedi et al., 2011 bacterial strains Sphingobium sp. YBL1 Sun et al., 2009 Sphingobium sp. YBL2 Sun et al., 2009 Sphingobium sp. YBL3 Sun et al., 2009 Sphingomonas sp. F35 Bending et al., 2003 Sphingomonas sp. SH Hussain et al., 2011 Sphingomonas sp. SRS2 Sorensen et al., 2001 IPU Ascomycetes Ronhede et al., 2005 Basidiomycetes Ronhede et al., 2005 Bjerkandera adusta Khadrani et al., 1999 fungal strains Mortierella sp. Gr4 Badawi et al., 2009 Phanerochaete chrysosporium Castillo et al., 2001 Rhizoctonia sp. Steiman et al., 1994 Zygomycetes Ronhede et al., 2005 pdmA (N-demethylase gene) Gu et al., 2013 pdmB (N-demethylase gene) Gu et al., 2013 genes puhA (phenylurea hydrolase gene) Khurana et al., 2009 puhB (phenylurea hydrolase gene) Khurana et al., 2009 Pseudomonas fluorescens Obanda and Shupe, 2009 Rastolia bacteria Wallance and Dickinson, 2006 bacterial strains Enterobacter sakazakii Sehnem et al., 2010 Serratia sp. Sehnem et al., 2010 TCZ Trichoderma harzianum Obanda and Shupe, 2009 Chaetomium globosum Obanda and Shupe, 2009 fungal strains Phanerochaete chrysosporium Obanda and Shupe, 2009 Meruliporia incrassata Obanda and Shupe, 2009

Chapter 1 General introduction 40

Aim of this thesis and research questions

The major purpose of this thesis was to support pesticide environmental risk assessment by (i) the application and development of methods of analytical chemistry to study the environmental fate of pesticides in soil, (ii) the application and development of novel methods of molecular biology and microbiology to assess the ecotoxicological impact of pesticides on soil microorganisms, and (iii) the application of catabolic gene- bioindicators to predict the ability of soil microbial communities to degrade a pesticide. This study focuses on the three chosen pesticides and their TPs. Lower tier pesticide exposure scenarios were carried out in a microcosm and a field study in agricultural soil of North Italy (45°05'20.8"N 9°45'59.4"E, Google Maps) at three different levels (1x, 2x or 10x agronomical doses in the microcosm study, and 1x, 2x or 5x doses in the field study). The three pesticides and their main TPs were quantified in dissipation studies to estimate the level and duration of exposure to soil microorganisms. For each pesticide and its main TPs, soil sorption affinities were estimated in a batch study. In addition, known and unknown (suspected) TPs of TCZ were searched and classified in soil extracts by an approach developed during my PhD combining suspect screening and in silico molecular typology. The ecotoxicological impact of each of the three pesticides (of the lower tier exposure scenario) on the soil bacterial diversity was estimated by a metagenomics approach (Illumina next- generation sequencing of 16S rDNA amplicons generated from soil DNA extracts). A higher tier pesticide exposure scenario was realized in microcosms repeatedly treated with CHL or TCZ (at 10x agronomical doses). Their impact on the soil microbial activity was estimated via the mineralization of CHL and TCZ and various other organic compounds (as a proxy for soil buffering functions). The impact on the soil bacterial diversity was estimated via Illumina next-generation sequencing. In addition, the interest of using the quantification of a pesticide-degrading genetic potential as an indicator for pesticide exposure and impact was assessed by monitoring the abundance of pdm gene sequences and the mineralization of IPU in soils sampled from two agricultural fields in Central Eastern France, regularly exposed to IPU and other pesticides.

Chapter 1 General introduction 41

By the work described in this thesis, the following research questions were aimed to answer:  How is the pesticide authorization process established, and what are the possible ways of improvement for a better environmental risk assessment of pesticides? (Chapter 2)  To what extent do the three studied pesticides CHL, IPU, and TCZ and their main TPs TCP (CHL), and MD-IPU and DD-IPU (IPU) dissipate in soil (degradation and sorption) in a lab-to-field experiment? (Chapter 3.1)  Can suspect screening be used to search for known and new TCZ TPs in soil of a field study? What is the interest of an in silico classification approach to deduce environmental parameters of new TCZ TPs? (Chapter 3.2)  To what extent does each of the three pesticides impact soil bacterial diversity in a lab-to-field experiment, evaluated by Illumina next-generation sequencing of 16S rDNA amplicons? (Chapter 4.1)  To what extent does repeated exposure to CHL and TCZ induce shifts in soil bacterial activity and diversity, evaluated via soil buffering functions and Illumina next- generation sequencing? (Chapter 4.2)  How can the quantification of IPU-degrading genetic potential be applied as a bioindicator for the estimation of the IPU biodegradation potential of a soil? (Chapter 5)  To which extent were the tested and developed techniques of this thesis suitable to improve the pesticide environmental risk assessment for soil microorganisms? (Chapter 6)

Chapter 1 General introduction 42

References

References of Figure 1.8:

chlorpyrifos Sardar and Kole, 2005; Singh and Walker, 2006

chlorpyrifos-oxon Sasikala et al., 2012; Tiwari and Guha, 2014

3,5,6-trichloropyridyl-2-dihydrogenphosphothioate Kulshrestha and Kumari, 2011

3,5,6-trichloropyridinol (TCP) Briceno et al., 2012; Chen et al., 2012 ; Cycon et al., 2013; Dubey et al., 2012; Karanasios et al., 2013; Lu et al., 2013; Sardar and Kole, 2005; Sasikala et al., 2012; Singh et al., 2006; Tiwari and Guha, 2014; Xu et al., 2008; Yu et al., 2006

3,5,6-trichloro-2-methoxypyridine (TMP) Sardar and Kole, 2005; Tiwari and Guha, 2014

3,5,6-trichloro-2-methylpyridine Yu, 2006

3,5-dichloro-2,6-dihydroxypyridine; Feng et al., 1998 2,5-dichloro-3,6-dihydroxypyridine; 2,3-dichloro-5,6-dihydroxypyridine

2-chloro-6-hydroxypyridine Yu et al., 2006

3-chloro-2,5-dihydro-2-pyridinol Feng et al., 1998; Singh and Walker, 2006

3,6-dihydroxypyridine-2,5-dione Li et al., 2010

2,5,6-trihydroxypyridine; Singh and Walker, 2006 2,3-dihydroxypyridine; 2,5-dihydroxypyridine

2-pyridinol Lu et al., 2013; Singh and Walker, 2006

maleamic acid Singh and Walker, 2006

maleamide semialdehyde Feng et al., 1998; Singh and Walker, 2006

fumaric acid; Singh and Walker, 2006 pyruvic acid

diethylhydrogenphosphothioate (DETP) Chen et al., 2012; Dubey et al., 2012 ; Sardar and Kole, 2005; Sasikala et al., 2012; Singh and Walker, 2006; Tiwari and Guha, 2014

ethyldihydrogenphosphothioate Singh and Walker, 2006

diethylhydrogenphosphate (DEP) Tiwari and Guha, 2014

References of Figure 1.9:

isoproturon (IPU) Badawi et al., 2009; Hussain et al., 2009

1-hydroxy-IPU (1-OH-IPU) Badawi et al., 2009; Hangler et al., 2007; Ronhede et al., 2005

2-hydroxy-IPU (2-OH-IPU) Hangler et al., 2007; Lehr et al., 1996; Mudd et al., 1983; Ronhede et al., 2005; Schuelein et al., 1996

mono-desmethyl-IPU (MD-IPU) Alletto et al., 2006; Badawi et al., 2009; El Khattabi et al., 2004; Gu et al., 2013; Hangler et al., 2007; Hussain et al., 2011; Hussain et al., 2009; Lehr et al., 1996; Mamy et al., 2011; Mudd et al., 1983; Perrin-Ganier et al., 2001; Roberts et al., 1998; Ronhede et al., 2005; Schuelein et al., 1996; Sorensen and Aamand, 2001; Sorensen et al., 2001; Zhang et al., 2012

1-hydroxy-MD-IPU (1-OH-MD-IPU) Badawi et al., 2009; Ronhede et al., 2005; Schuelein et al., 1996

2-hydroxy-MD-IPU (2-OH-MD-IPU) El Khattabi et al., 2004; Lehr et al., 1996; Mudd et al., 1983; Pieuchot et al., 1996; Schuelein et al., 1996

di-desmethyl-IPU (DD-IPU) Alletto et al., 2006; Badawi et al., 2009; Gu et al., 2013; Hussain et al., 2009; Lehr et al., 1996; Mudd et al., 1983; Roberts et al., 1998; Sorensen and Aamand, 2001; Sorensen et al., 2001; Zhang et al., 2012

1-hydroxy-DD-IPU (1-OH-DD-IPU) Badawi et al., 2009

2-hydroxy-DD-IPU (2-OH-DD-IPU) Lehr et al., 1996; Mudd et al., 1983

4-isopropyl-aniline (4-IA) Gu et al., 2013; Hussain et al., 2009 ; Mudd et al., 1983; Sorensen and Aamand, 2001; Sorensen et al., 2001; Zhang et al., 2012

2-hydroxy-4-IA (2-OH-4-IA) Mudd et al., 1983

4,4-diisopropyl-azobenzene Perrin-Ganier et al., 2001; Pieuchot et al., 1996

4,5-bis-(4-isopropylphenylamino)-3,5-cyclohexadiene- Scheunert and Reuter, 2000 1,2-dione

4-isopropylcatechol Zhang et al., 2012

catechol Sun, 2009

4-isopropyl-cis,cis-muconic acid; Zhang et al., 2012 4-isopropyl-2-hydroxymuconic semialdehyde

Chapter 1 General introduction 43

cis,cis-muconic acid Sun et al., 2009; Zhang et al., 2012

muconolacetone; Hussain, 2010 β-ketoadipate-enol-lactone; β-ketoadipate; β-ketoadipyl-CoA; succinyl-CoA; acetyl-CoA

References of Figure 1.10:

tebuconazole (TCZ) EFSA, 2014b

5-(4-chlorophenyl)-2,2-dimethyl-3-(1,2,4-triazol-1- Obanda and Shupe, 2009 ylmethyl)pentane-1,3-diol

1-(4-chlorophenyl)-4,4-dimethyl-3-(1,2,4-triazol-1- FAO, 1994 ylmethyl)pentane-2,3-diol

5-chloro-2-[3-hydroxy-4,4-dimethyl-3-(1,2,4-triazol-1- FAO, 1994 ylmethyl)pentyl]phenol

1-(4-chlorophenyl)-4,4-dimethyl-3-(1,2,4-triazol-1- FAO, 1994 ylmethyl)pentane-1,3,5-triol

1-(4-chlorophenyl)-3-hydroxy-4,4-dimethyl-3-(1,2,4- Calza et al., 2002; Herrero-Hernandez et al., 2011; Potter et al., 2005; triazol-1-ylmethyl)pentan-1-one Strickland et al., 2004

5-(4-chlorophenyl)-3-hydroxy-2,2-dimethyl-3-(1,2,4- Obanda and Shupe, 2009 triazol-1-ylmethyl)pentanoic acid

5-(4-chlorophenyl)-3-hydroxy-2,2-dimethyl-5-oxo-3- FAO, 1994 (1,2,4-triazol-1-ylmethyl)pentanoic acid

[5-(4-chlorophenyl)-3-hydroxy-2,2-dimethyl-3-(1,2,4- Obanda and Shupe, 2009 triazol-1-ylmethyl)pentyl]acetate

[5-(4-chlorophenyl)-3-hydroxy-2,2-dimethyl-3-(1,2,4- EFSA, 2008 triazol-1-ylmethyl)pentyl] hydrogen sulfate

2-chloro-5-[4-hydroxy-5,5-dimethyl-3-(1,2,4-triazol-1- EFSA, 2008 ylmethyl)hexyl]phenol

1-(4-chlorophenyl)-4-methyl-3-(1,2,4-triazol-1- FAO, 1994 ylmethyl)pentan-3-ol

4-(4-chlorophenyl)-2-methyl-1-(1,2,4-triazol-1- FAO, 1994 yl)butan-2-ol

4-(4-chlorophenyl)-1-(1,2,4-triazol-1-yl)butan-2-ol FAO, 1994

4-(4-chlorophenyl)-1-(1,2,4-triazol-1-yl)butan-2-one FAO, 1994

(NE)-N-[2-[2-(4-chlorophenyl)ethyl]-2-hydroxy-3,3- Obanda and Shupe, 2009 dimethyl-butylidene]formamidine

1-(4-chlorophenyl)-4,4-dimethyl-3-[(E)- Obanda and Shupe, 2009 (methylenehydrazono)methyl]pentan-3-ol

4-chlorobenzoic acid FAO, 1994

5-tert-butyl-5-(1,2,4-triazol-1- Potter et al., 2005; White et al., 2010 ylmethyl)tetrahydrofuran-2-one

4-hydroxy-5,5-dimethyl-4-(1,2,4-triazol-1- Strickland et al., 2004 ylmethyl)hexanoic acid

3-hydroxy-4,4-dimethyl-3-(1,2,4-triazol-1- Potter et al., 2005 ylmethyl)pentanoic acid

3,3-dimethyl-1-(1,2,4-triazol-1-yl)butan-2-one Strickland et al., 2004

2,2-dimethyl-1-(1,2,4-triazol-1-yl)propan-1-one Potter et al., 2005

2-hydroxy-3-(1,2,4-triazol-1-yl)propanoic acid EFSA, 2008

2-(1,2,4-triazol-1-yl)acetic acid EFSA, 2008

2-(1,2,4-triazol-1-yl)acetaldehyde Schermerhorn et al., 2005

2-(1,2,4-triazol-1-yl)ethane-1,1-diol EFSA, 2008; Schermerhorn et al., 2005

butyl 2-(1,2,4-triazol-1-yl)acetate U.S. EPA, 1990

butyl 2-hydroxy-3-(1,2,4-triazol-1-yl)propanoate U.S. EPA, 1990

2-amino-3-(1,2,4-triazol-1-yl)propanoic acid EFSA, 2008; Schermerhorn et al., 2005

2-imino-3-(1,2,4-triazol-1-yl)propanoic acid Schermerhorn et al., 2005

2-(1,2,4-triazol-1-yl)ethanimine Schermerhorn et al., 2005

1H-1,2,4-triazol-5-ol FAO, 1994

1H-1,2,4-triazole EFSA, 2008; FAO, 1994; Schermerhorn et al., 2005; U.S. EPA, 1990

Chapter 1 General introduction 44

Akallal, R., Debieu, D., Lanen, C., Daboussi, M.-J., Fritz, R., Malosse, C., Bach, J. and Leroux, P. (1998). Inheritance and mechanisms of resistance to tebuconazole, a sterol C14-demethylation inhibitor, in Nectria haematococca. Pesticide Biochemistry and Physiology 60 (3), 147-166. Alletto, L., Coquet, Y., Benoit, P. and Bergheaud, V. (2006). Effects of temperature and water content on degradation of isoproturon in three soil profiles. Chemosphere 64, 1053-1061. Álvarez-Martín, A., Sánchez-Martín, M.J., Pose-Juan, E. and Rodríguez-Cruz, M.S. (2016). Effect of different rates of spent mushroom substrate on the dissipation and bioavailability of cymoxanil and tebuconazole in an agricultural soil. Science of the Total Environment 550, 495-503. Anwar, S., Liaquat, F., Khan, Q.M., Khalid, Z.M. and Iqbal, S. (2009). Biodegradation of chlorpyrifos and its hydrolysis product 3,5,6- trichloro-2-pyridinol by Bacillus pumilus strain C2A1. Journal of Hazardous Materials 168, 400-405. Arias-Estévez, M., López-Periago, E., Martínez-Carballo, E., Simal-Gándara, J., Mejuto, J.-C. and García-Río, L. (2008). The mobility and degradation of pesticides in soils and the pollution of groundwater resources. Agriculture, Ecosystems and Environment 123, 247-260. Awad, N.S., Sabit, H.H., Abo-Aba, S.E.M. and Bayoumi, R.A. (2011). Isolation, characterization and fingerprinting of some chlorpyrifos- degrading bacterial strains isolated from Egyptian pesticides-polluted soils. African Journal of Microbiology Research 5, 2855-2862. Badawi, N., Ronhede, S., Olsson, S., Kragelund, B. B., Johnsen, A. H., Jacobsen, O. S. and Aamand, J. (2009). Metabolites of the phenylurea herbicides chlorotoluron, diuron, isoproturon and linuron by the soil fungus Mortierella sp. Environmental Pollution 157, 2806-2812. Baelum, J., Nicolaisen, M.H., Holben, W.E., Strobel, B.W., Sorensen, J. and Jacobsen, C.S. (2008). Direct analysis of tfdA gene expression by indigenous bacteria in phenoxy acid amended agricultural soil. The ISME Journal 2, 677-687. Barriuso, E., Calvet, R., Schiavon, M. and Soulas, G. (1996). Les pesticides et les polluants organiques des sols. Transformation and dissipation. Forum ‘Le sol, un patrimoine menacé?’, Paris. Baskaran, S., Kookana, R.S. and Naidu, R. (2003). Contrasting behaviour of chlorpyrifos and its primary metabolite, TCP (3,5,6- trichloro-2-pyridinol), with depth in soil profiles. Australian Journal of Soil Research 41 (4), 749-760. Bending, G.D., Lincoln, S.D., Sorensen, S.R., Morgan, J.A.W., Aamand, J., and Walker, A. (2003). In-field spatial variability in the degradation of the phenyl-urea herbicide isoproturon is the result of interactions between degradative Sphingomonas spp. and soil pH. Applied and Environmental Microbiology 69, 827-834. Blum, W.E.H. (2006). Soil resources - the basis of human society and the environment. Bodenkultur 57, 197-202. Bondeau, A., Smith, C.M., Zaehle, S., Schapfhoff, S., Lucht, W., Cramer, W., Gerten, D., Lotze-Campen, H., Mueller, C., Reichstein, K. and Smith, B. (2007). Modelling the role of agriculture for the 20th century global terrestrial carbon balance. Global Change Biology 13, 679-706. Börner, H., Schlüter, K. and Aumann, J. (2009). Pflanzenkrankheiten und Pflanzenschutz. Springer, Berlin Heidelberg, ISBN 978-3- 540-49068-5. Bouskill, N.J., Barnhart, E.P., Galloway, T.S., Handy, R.D. and Ford, T.E. (2007). Quantification of changing Pseudomonas aeruginosa soda, htpX and mt gene abundance in response to trace metal toxicity: A potential in situ biomarker of environmental health. FEMS Microbiology Ecology 60, 276-286. Briceno, G., Fuentes, M. S., Palma, G., Jorquera, M. A. Amoroso and Diez, M. C. (2012). Chlorpyrifos biodegradation and 3,5,6-trichloro- 2-pyridinol production by actinobacteria isolated from soil. International Biodeterioration and Biodegradation 73, 1-7. Bumpus, J.A., Kakkar, S.N. and Coleman, R.D. (1993). Fungal degradation of organophosphorus insecticides. Journal of Applied Biochemistry and Biotechnology 39-40, 715-726. Calza, P., Baudino, S., Aigotti, R., Baiocchi, C., Branca, P. and Pelizzetti, E. (2002). High performance liquid chromatographic/tandem mass spectrometric identification of the photo transformation products of tebuconazole on titanium dioxide. Journal of Mass Spectrometry 37, 566-576. Castillo, M.D.P., Wiren-Lehr, S.V., Scheunert, I., and Torstensson, L. (2001). Degradation of isoproturon by the white rot fungus Phanerochaete chrysosporium. Biology and Fertility of Soils 33, 521-528. Chen, S., Liu, C., Peng, C., Liu, H., Hu, M. and Zhong, G. (2012). Biodegradation of chlorpyrifos and its hydrolysis product 2,5,6- trichloro-2-pyridinol by a new fungal strain Cladosporium cladosporioides Hu-01. Public Library of Science ONE 7 (10), 47205. Chino-Flores, C., Dantan-Gonzalez, E., Vazquez-Ramos, A., Tinoco-Valencia, R., Diaz-Mendez, R., Sanchez-Salinas, E., Castrejon- Godinez, M.L., Ramos-Quintana, F. and Ortiz-Hernandez, M.L. (2012). Isolation of the opdE gene that encodes for a new hydrolase of Enterobacter sp. capable of degrading organophosphorus pesticides. Biodegradation 23, 387-397. Chishti, Z., Hussain, S., Arshad, K.R., Khalid, A. and Arshad, M. (2013). Microbial degradation of chlorpyrifos in liquid media and soil. Journal of Environmental Management 114, 372-380. Chu, X., Fang, H., Pan, X., Wang, X., Shan, M., Feng, B. and Yu, Y. (2008). Degradation of chlorpyrifos alone and in combination with chlorothalonil and their effects on soil microbial populations. Journal of Environmental Sciences 20, 464-469. Collings, L.V., Blair, A.M., Gay, A.P., Dyer, C.J. and Mackay, N. (2003). The effect of weather factors on the performance of herbicides to control Alopecurus myosuroides in winter wheat. Journal of Weed Research 43, 146-153. Copley, S.D. (2009). Evolution of efficient pathways for degradation of anthropogenic chemicals. Nature Chemical Biology 5, 559-566. Crutzen, P.J. (2002). Geology of mankind. Nature 415, 23. Cullington, J.E., and Walker, A. (1999). Rapid biodegradation of diuron and other phenylurea herbicides by a soil bacterium. Soil Biology and Biochemistry 31, 677-686. Cycon, M., Zmijowska, A., Wojcik, M. and Piotrowska-Seget, Z. (2013). Biodegradation and bioremediation potential of diazinon- degrading Serratia marcescens to remove other organophosphorus pesticides from soils. Journal of Environmental Management 117, 7-16. Dagès, C., Samouëlian, S., Negro, S., Storck, V., Huttel, O. and Voltz, M. (2015). Seepage patterns of diuron in a ditch bed during a sequence of flood events. Science of the Total Environment 537, 120-128. D'Angelo, D.L., Bradley, C.A., Ames, K.A., Willeyard, K.T., Madden, L.V. and Paul, P.A. (2014). Efficacy of fungicide applications during and after anthesis against fusarium head blight and deoxynivalenol in soft red winter wheat. Plant Disease Journal 98, 1387-1397. DeBryn, J.M., Chewning, C.S. and Sayler, G.S. (2007). Comparative quantitative prevalence of Mycobacteria and functionally abundant nidA, nahAc, and nagAc dioxygenase genes in coal tar contaminated sediments. Environmental Science and Technology 41, 5426-5432. Devers, M., Azhari, N.E., Kolic, N.-U. and Martin-Laurent, F. (2007). Detection and organization of atrazine-degrading genetic potential of seventeen bacterial isolates belonging to divergent taxa indicate a recent common origin of their catabolic functions. FEMS Microbiology Letters 273, 78-86.

Chapter 1 General introduction 45

Dubey, K. K. and Fulekar, M. H. (2012). Chlorpyrifos bioremediation in Pennisetum rhizosphere by a novel potential degrader Stenotrophomonas maltophilia MHF ENV20. World Journal of Microbiology and Biotechnology 28, 1715-1725. Duirk, S.E. and Collette, T.W. (2006). Degradation of chlorpyrifos in aqueous chlorine solutions: pathways, kinetics, and modeling. Environmental Science and Technology 40 (2), 546-551. Dutta, M., Sardar, D., Pal, R. and Kole, R.K. (2010). Effect of chlorpyrifos on microbial biomass and activities in tropical clay loam soil. Environmental Monitoring and Assessment 160, 385-391. Dwivedi, S., Singh, B.R., Al-Khedhairy, A.A., and Musarrat, J. (2011). Biodegradation of isoproturon using a novel Pseudomonas aeruginosa strain JS-11 as a multi-functional bioinoculant of environmental significance. Journal of Hazardous Materials 185, 938-944. EC (2005). Review report for the active substance chlorpyrifos. Finalised in the Standing Committee on the Food Chain and Animal Health at its meeting on 3 June 2005 in view of the inclusion of chlorpyrifos in Annex I of the Directive 91/414/EEC. SANCO/3059/99 - rev. 1.5. Edwards, C.A. and Adams, R.S. (1970). Persistent pesticides in the environment. Critical Reviews in Environmental Control 1, 1-4. EFSA (2008). Conclusion on the peer review of tebuconazole. EFSA Scientific Report 176, 1-109. EFSA (2010). Scientific opinion on the development of specific protection goal options for environmental risk assessment of pesticides, in particular in relation to the revision of the guidance documents on aquatic and terrestrial ecotoxicology (SANCO/3268/2001 and SANCO/10329/2002), EFSA Panel on Plant Protection Products and their Residues (PPR). EFSA Journal 2010; 8 (10). 1821. EFSA (2011a). Conclusion on the peer review of the pesticide risk assessment of the active substance chlorpyrifos. EFSA Journal 9 (1), 1961. doi:10.2903/j.efsa.2011.1961. EFSA (2011b). Reasoned opinion of EFSA: Modification of the existing MRLs for chlorpyrifos-methyl in various crops. EFSA Journal 9 (6), 2219. doi:10.2903/j.efsa.2011.2219. EFSA (2012). Modification of the existing MRLs for chlorpyrifos in various crops and in products of animal origin. EFSA Journal 10 (1), 2510. doi:10.2903/j.efsa.2012.2510. EFSA (2014a). Conclusion on the peer review of the pesticide human health risk assessment of the active substance chlorpyrifos. EFSA Journal 12 (4), 3640. doi:10.2903/j.efsa.2014.3640. EFSA (2014b). Conclusion on the peer review of the pesticide risk assessment of the active substance tebuconazole. EFSA Journal 12, 3485. EFSA (2015a). Refined risk assessment regarding certain maximum residue levels (MRLs) of concern for the active substance chlorpyrifos. EFSA Journal 13 (6), 4142. doi:10.2903/j.efsa.2015.4142. EFSA (2015b). Conclusion on the peer review of the pesticide risk assessment of the active substance isoproturon. EFSA Journal 13 (8), 4206. doi:10.2903/j.efsa.2015.4206. EFSA (2016a). Recovery in environmental risk assessments at EFSA, EFSA Scientific Committee. EFSA Journal 2016; 14 (2), 4313. EFSA (2016b). Scientific opinion addressing the state oft he science on risk assessment of plant protection products for in-soil organisms. Draft scientific opinion for public consultation. Eisenhauer, N., Klier, M., Partsch, S., Sabais, A.C.W., Scherber, C., Weisser, W.W. and Scheu, S. (2009). No interactive effects of pesticides and plant diversity on soil microbial biomass and respiration. Applied Soil Ecology 42, 31-36. El Khattabi, K., Bouhaouss, A., Perrin-Ganier, C. and Schiavon, M. (2004). Fate of isoproturon in two Moroccan soils. Agronomie 24, 177-183. El-Sebai, T., Lagacherie, B., Soulas, G., and Martin-Laurent, F. (2004). Isolation and characterisation of an isoproturon-mineralising Methylopila sp TES from French agricultural soil. FEMS Microbiology Letters 239, 103-110. Enserink, M., Hines, P.J., Vignieri, S.N., Wigginton, N.S. and Yeston, J.S. (2013). The Pesticide Planet. Science 341, 730-731. EU-Regulation 1107/2009/EC (2009). Regulation (EC) no 1107/2009 of the European parliament and of the council of 21 October 2009 concerning the placing of plant protection products on the market and repealing council directives 79/117/EEC and 91/414/EEC. Official Journal of the European Union 309, 1. EU-Regulation 283/2013/EC of 1 March 2013 setting out the data requirements for active substances, in accordance with Regulation (EC No 1107/2009 of the European Parliament and the Council concerning the placing of plant protection products on the market and repealing Council Directives 79/117/EEC and 91/414/EEC. Fang, H., Xiang, Y.Q., Hao, Y.J., Chu, X.Q., Pan, X.D., Yu, J.Q. and Yu, Y.L. (2008). Fungal degradation of chlorpyrifos by Verticillium sp. DSP in pure cultures and its use in bioremediation of contaminated soil and pakchoi. International Biodeterioration and Biodegradation 61, 294-303. Fang, H., Yu, Y., Chu, X., Wang, X., Yang, X. and Yu, J. (2009). Degradation of chlorpyrifos in laboratory soil and its impact on soil microbial functional diversity. Journal of Environmental Sciences 21, 380-386. FAO (1994). Pesticide Evaluation Tebuconazole 188. Food and Agriculture Organization of the United Nations Report of the Joint FAO/WHO Meeting on Pesticide Residues (JMPR). FAO, Rome. Feng, Y., Minard, R. D. and Bollag, J. M. (1998). Photolytic and microbial degradation of 3, 5, 6-trichloro-2-pyridinol. Environmental Toxicology and Chemistry 17, 814-819. Fenner, K., Canonica, S., Wackett, L.P. and Elsner, M. (2013). Evaluating pesticide degradation in the environment: Blind spots and emerging opportunities. Science 341, 752-758. Ferreira, E.P.B., Dusi, A.N., Costa, J.R., Xavier, G.R. and Rumjanek, N.G. (2009). Assessing insecticide and fungicide effects on the culturable soil bacterial community by analyses of variance of their DGGE fingerprinting data. European Journal of Soil Biology 45, 466-472. Gao, Y., Chen, S., Hu, M., Hu, Q., Luo, J. and Li, Y. (2012). Purification and characterization of a novel chlorpyrifos hydrolase from Cladosporium cladosporioides Hu-01. PLOS ONE 7 (6). Gao, Y., Wang, S., Xu, D., Yu, H., Wu, L., Lin, Q., Hu, Y., Li, X., He, Z., Deng, Y., Zhou, J. and Yang, Y. (2014). GeoChip as a metagenomics tool to analyze the microbial gene diversity along an elevation gradient. Genomics Data 2, 132-134. Gardi, C., Jeffery, S. and Saltelli, A. (2013). An estimate of potential threats levels to soil biodiversity in EU. Global Change Biology 19, 1538-1548. Ghanem, I., Orfi, M. and Shamma, M. (2007). Biodegradation of chlorpyrifos by Klebsiella sp. isolated from an activated sludge sample of waste water treatment plant in Damascus. Folia Microbiologica 52, 423-427. Ghiglione, J.-F., Martin-Laurent, F. and Pesce, S. (2016). Microbial ecotoxicology: an emerging discipline facing contemporary environmental threats. Environmental Science and Pollution Research 23, 3981-3983. Gilani, S.T.S., Ageen, M., Shah, H. and Raza, S. (2010). Chlorpyrifos degradation in soil and its effect on soil microorganisms. The Journal of Animal and Plant Sciences 20 (2), 99-102. Gilden, R.C., Huffling, K. and Sattler, B. (2010). Pesticides and health risks. Journal of Obstetric, Gynecologic, and Neonatal Nursing 39 (1), 103-110.

Chapter 1 General introduction 46

Graham, E.B., Knelman, J.E., Schindlbacher, A., Siciliano, S., Breulmann, M., Yannarell, A., Beman, J.M., Abell, G., Philippot, L., Prosser, J., Foulquier, A., Yuste, J.C., Glanville, H.C., Jones, D.L., Angel, R., Salminen, J., Newton, R.J., Bürgmann, H., Ingram, L.J., Hamer, U., Siljanen, H.M.P., Peltoniemi, K., Potthast, K., Bañeras, L., Hartmann, M., Banerjee, S., Yu, R.-Q., Nogaro, G., Richter, A., Koranda, M., Castle, S.C., Goberna, M., Song, B., Chatterjee, A., Nunes, O.C., Lopes, A.R., Cao, Y., Kaisermann, A., Hallin, S., Strickland, M.S., Garcia-Pausas, J., Barba, J., Kang, H., Isobe, K., Papaspyrou, S., Pastorelli, R., Lagomarsino, A., Lindström, E.S., Basiliko, N. and Nemergut, D.R. (2016). Microbes as engines of ecosystem function: When does community structure enhance predictions of ecosystem processes? Frontiers in Microbiology 7 (214). Gu, T., Zhou, C., Sorensen, S. R., Zhang, J., He, J., Yu, P., Yan, X. and Li, S. (2013). The novel bacterial N-demethylase pdmAB is responsible for the initial step of N,N-dimethyl-substituted phenylurea herbicide degradation. Applied and Environmental Microbiology 79 (24), 7846-7856. Gupta, S., Gupta, R. and Sharma, S. (2013). Impact of chemical- and bio-pesticides on bacterial diversity in rhizosphere of Vigna radiata. Ecotoxicology 22, 1479-1489. Hallin, S., Welsh, A., Stenström, J., Hallet, S., Enwall, K., Bru, D. and Philippot, L. (2012). Soil functional operating range linked to microbial biodiversity and community composition using denitrifiers as model guild. PLOS ONE 7 (12), 51962. Hangler, M., Jensen, B., Ronhede, S. and Sorensen, S. R. (2007). Inducible hydroxylation and demethylation of the herbicide isoproturon by Cunninghamella elegans. FEMS Microbiology Letters 268, 254-260. Hayes, W.J. and Laws, E.R. (1991). Handbook of pesticide toxicology, Vol. 3, Classes of pesticides. Academic Press Inc., ISBN 0-12- 334162-0. Haygarth, P. and Ritz, K. (2009). The future of soils and land use in the UK: Soil systems for the provision of land-based ecosystem services. Land Use Policy 26, 187-197. Herrero-Hernández, E., Andrades, M.S., Alvarez-Martina, A., Pose-Juana, E., Rodriguez-Cruz, M.S. and Sánchez-Martín, M.J. (2013). Occurrence of pesticides and some of their degradation products in waters in a Spanish wine region. Journal of Hydrology 486, 234-245. Herrero-Hernandez, E., Andrades, M.S., Marin-Benito, J.M., Sanchez-Martin, M.J. and Rodriguez-Cruz, M.S. (2011). Field-scale dissipation of tebuconazole in a vineyard soil amended with spent mushroom substrate and its potential environmental impact. Ecotoxicology and Environmental Safety 74, 1480-1488. Hillel, D. (2009). The mission of soil science in a changing world. Journal of Plant Nutrition and Soil Science 172, 5-9. Hussain, S. (2010). Characterization of the isoproturon degrading community: From the field to the genes. Dissertation at National Institute for Agricultural Research (INRA), Dijon, France. tel-00674042, version 1 - 27 Feb 2012. Hussain, S., Arshad, M., Springael, D., Sorensen, S.R., Bending, G.D., Devers-Lamrani, M., Maqbool, Z. and Martin-Laurent, F. (2015). Abiotic and biotic processes governing the fate of phenylurea herbicides in soils: a review. Critical Reviews in Environmental Science and Technology 45, 1947-1998. Hussain, S., Devers-Lamrani, M., El Azhari, N. and Martin-Laurent, F. (2011). Isolation and characterization of an isoproturon mineralizing Sphingomonas sp. strain SH from a French agricultural soil. Biodegradation 22, 637-650. Hussain, S., Sorensen, S. R., Devers-Lamrani, M., El-Sebai, T. and Martin-Laurent, F. (2009). Characterization of an isoproturon mineralizing bacterial culture enriched from a French agricultural soil. Chemosphere 77, 1052-1059. Hutzinger, O. (2013). The handbook of environmental chemistry – anthropogenic compounds, 3rd edition, Springer, Berlin Heidelberg. Ibanez, M., Sancho, J.V., Pozo, O.J., Niessen, W. and Hernandez, F. (2005). Use of quadrupole time-of-flight mass spectrometry in the elucidation of unknown compounds present in environmental water. Rapid Communications in Mass Spectrometry 19, 169-178. Imfeld, G. and Vuillemier, S. (2012). Measuring the effects of pesticides on bacterial communities in soil: A critical review. European Journal of Soil Biology 49, 22-30. Jastrzebska, E. (2011). The effect of chlorpyrifos and teflubenzuron on the enzymatic activity of soil. Polish Journal of Environmental Studies 20 (4), 903-910. Jeffery, S., Gardi, C., Jones, A., Montanarella, L., Marmo, L., Miko, L., Ritz, K., Peres, G., Römbke, J. and van der Putten, W.H. (2010). European Atlas of Soil Biodiversity. European Commission, Publications Office of the European Union, Luxembourg, ISBN 978-92-79-15806-3. Jeyaratnam, J. (1990). Acute : A major global health problem. World Health Statistics Quarterly 43 (3), 139-144. Jones, A., Stolbovoy, V., Rusco, E., Gentile, A.-R., Gardi, C., Marechal, B. and Montanarella, L. (2009). Climate change in Europe. 2. Impact on soil. A review. Agronomy for Sustainable Development 29, 423-432. Joseph, S.V. and Zarate, J. (2015). Comparing efficiency of insecticides against cabbage maggot (Diptera: Anthomyiidae) in the laboratory. Crop Protection 77, 148-156. Karanasios, E. C., Tsiropoulos, N. G. and Karpouzas, D. G. (2013). Quantitative and qualitative differences in the metabolism of pesticides in biobed substrates and soil. Chemosphere 93, 20-28. Karpouzas, D.G., Tsiamis, G., Trevisan, M., Ferrari, F., Malandain, C., Sibourg, O. and Martin-Laurent, F. (2016). "LOVE TO HATE" pesticides : felicity or curse for the soil microbial community ? An FP7 IAPP Marie Curie project aiming to establish tools for the assessment of the mechanisms controlling the interactions of pesticides with soil microorganisms. Environmental Science and Pollution Research 23 (18), 18947-18951. Kazy, W.D., Monier, A.L. and Alvarez, P.J.J. (2010). Assessing the correlation between anaerobic toluene degradation activity and bssA concentrations in hydrocarbon-contaminated aquifer material. Biodegradation 21, 793-800. Keinath, A.P. (2015). Baseline sensitivity of Didymella bryoniae to cyprodinil and fludioxonil and field efficacy of these fungicides against isolates resistant to pyraclostrobin and boscalid. Plant Disease Journal 99, 815-822. Khadrani, A., Seigle-Murandi, F., Steiman, R., and Vroumsia, T. (1999). Degradation of three phenylurea herbicides (chlortoluron, isoproturon and diuron) by micromycetes isolated from soil. Chemosphere 38, 3041-3050. Khurana, J.L., Jackson, C.J., Scott, C., Pandey, G., Horne, I., Russell, R.J., Herlt, A., Easton, C.J. and Oakeshott, J.G. (2009). Characterization of the phenylurea hydrolases A and B: founding members of a novel amidohydrolase subgroup. Biochemistry Journal 418, 431-441. Kong, W.D. and Nakatsu, C.H. (2010). Optimization of RNA extraction for PCR quantification of aromatic compound degradation genes. Applied and Environmental Microbiology 76, 1282-1284. Korade, D.L. and Fulekar, M.H. (2009). Rhizosphere remediation of chlorpyrifos in mycorrhizospheric soil using ryegrass. Journal of Hazardous Materials 172, 1344-1350. Kulshrestha, G. and Kumari, A. (2011). Fungal degradation of chlorpyrifos by Acremonium sp. strain (GFRC-1) isolated from a laboratory-enriched red agricultural soil. Biology and Fertility of Soils 47, 219-225. Kuriyal, S.K. and Pandey, B.N. (1994). Studies of the effects of isoproturon on soil microflora. Journal of the Indian Botanical Society 73, 47-49.

Chapter 1 General introduction 47

Lakshmi, C.V., Kumar, M. and Khanna, S. (2008). Biotransformation of chlorpyrifos and bioremediation of contaminated soil. International Biodeterioration and Biodegradation 62, 204-209. Lal, R. (2008). Soils and sustainable agriculture. A review. Agronomy for Sustainable Development 28, 57-64. Lehr, S., Glaessgen, W.E., Sandermann, H. Jr., Beese, F. and Scheunert, I. (1996). Metabolism of isoproturon in soils originating from different agricultural management systems and in cultures of isolated soil bacteria. International Journal of Environmental Analytical Chemistry 65, 231-243. Lewis, S.L. and Maslin, M.A. (2015). Defining the Anthropocene. Nature 519, 171-180. Li, J., Liu, J., Shen, W., Zhao, X., Hou, Y., Cao, H. and Cui, Z. (2010). Isolation and characterization of 3,5,6-trichloro-2-pyridinol- degrading Ralstonia sp. strain T6. Bioresource Technology 101, 7479-7483. Li, X., He, J. and Li, S. (2007). Isolation of a chlorpyrifos-degrading bacterium, Sphingomonas sp. strain Dsp-2, and cloning of the mpd gene. Research in Microbiology 158, 143-149. Lichtfouse, E., Navarette, M., Debaeke, P., Souchère, V. and Alberola, C. (2009). Sustainable Agriculture, Springer, 1st edition, 645 p., ISBN: 978-90-481-2665-1. Lindahl, A.M.L and Bockstaller, C. (2012). An indicator of pesticide leaching risk to groundwater. Ecological Indicators 23, 95-108. Liu, Z., Chen, X., Shi, Y. and Su, Z. (2012). Bacterial degradation of chlorpyrifos by Bacillus cereus. Advanced Materials Research 356- 360, 676-680. Lu, P., Li, Q., Liu, H., Feng, Z., Yan, X., Hong, Q. and Li, S. (2013). Biodegradation of chlorpyrifos and 3,5,6-trichloro-2-pyridinol by Cupriavidus sp. DT-1. Bioresource Technology 127, 337-342. Malaja, E., von der Ohea, P.C., Groted, M., Kühnee, R., Mondyf, C.P., Usseglio-Polaterag, P., Bracka, W. and Schäfer, R.B. (2013). Organic chemicals jeopardize the health of freshwater ecosystems on the continental scale. Proceedings of the National Academy of Sciences 111 (26), 9549-9554. Mallick, B.K., Banerji, A., Shakli, N.A. and Sethunathan, N.N. (1999). Bacterial degradation of chlorpyrifos in pure culture and in soil. Bulletin of Environmental Contamination and Toxicology 62, 48-55. Mallory-Smith, C. and Retzinger, J. (2003). Revised classification of herbicides by site of action for weed resistance management strategies. Weed Technology 17, 605-619. Maltby, L., Brock, T.C.M. and Van den Brink, P. (2009). Fungicide risk assessment for aquatic ecosystems: Importance of interspecific variation, toxic mode of action, and exposure regime. Environmental Science and Technology 43, 7556-7563. Mamy, L., Vrignaud, P., Cheviron, N., Perreau, F., Belkacem, M., Brault, A., Breuil, S., Delarue, G., Petraud, J.-P., Touton, I., Mougin, C. and Chaplain, V. (2011). No evidence for effect of soil compaction on the degradation and impact of isoproturon. Environmental Chemistry Letters 9, 145-150. Martin-Laurent, F., Kandeler, E., Pertic, I., Djuric, S. and Karpouzas, D.G. (2013). ECOFUN-MICROBIODIV: an FP7 European project for developing and evaluating innovative tools for assessing the impact of pesticides on soil functional microbial diversity- towards new pesticide registration regulation? Environmental Science and Pollution Research 20, 1203-1205. Martin-Laurent, F., Piutti, S., Hallet, S., Wagschal, I., Philippot, L., Catroux, G. and Soulas, G. (2003). Monitoring of atrazine treatment on soil bacterial, fungal and atrazine-degrading communities by quantitative competitive PCR. Pest Management Science 59, 259-268. Monard, C., Martin-Laurent, F., Devers-Lamrani, M., Lima, O., Vandenkoornhuyse, P. and Binet, F. (2010). Atz gene expressions during atrazine degradation in the soil drilosphere. Molecular ecology 19, 749-759. Monard, C., Martin-Laurent, F., Lima, O., Devers-Lamrani, M. and Binet, F. (2013). Estimating the biodegradation of pesticide in soils by monitoring pesticide-degrading gene expression. Biodegradation 24, 203-213. Mudd, P.J., Hance, R.J. and Wright, S.J.L. (1983). The persistence and metabolism of isoproturon in soil. Weed Research 23, 239-246. Munoz-Leoz, B., Ruiz-Romera, E., Antigëdad, I. and Garbisu, C. (2011). Tebuconazole application decreases soil microbial biomass and activity. Soil Biology and Biochemistry 43, 2146-2183. Nortcliff, S. (2002). Standardisation of soil quality attributes. Agriculture, Ecosystems and Environment 88, 161-168. Obanda, D.N. and Shupe, T.F. (2009). Biotransformation of tebuconazole by microorganisms: Evidence of a common mechanism. Wood Fiber Science 41 (2), 157-167. OECD-Guideline 106 (2000). OECD Guidelines for the testing of chemicals. Adsorption – desorption using a batch equilibrium method. Doi 10.1787/9789264069602-en. OECD-Guideline ENV/JM/MONO (2007)17 (2007). Guidance document on pesticide residue analytical methods. Series on pesticides no 39. Series on testing and assessment no 72, p. 34. Oerke, E.-C., Dehne, H.W., Schönebeck, F. and Weber, A. (1994). Crop production and crop protection. Estimated losses in major food and cash crops. Elsevier Science B.V., ISBN 0-444-82095-7. Ogunseitan, O.A., Yang, S. and Ericson, J. (2000). Microbial deltaaminolevulinate dehydratase as a biosensor of lead bioavailability in contaminated environments. Soil Biology and Biochemistry 32, 1899-1906. Omar, S.A. (1998). Availability of phosphorus and sulfur of insecticide origin by fungi. Biodegradation 9, 327-336. O'Reilly, E. (2016). Decision in case 12/2013/MDC on the practices of the European Commission regarding the authorisation and placing on the market of plant protection products (pesticides). Case: 12/2013/MDC. Peng, J.J., Cai, C., Qiao, M., Li, H. and Zhu, Y.G. (2010). Dynamic changes in functional gene copy numbers and microbial communities during degradation of pyrene in soils. Environmental Pollution 158, 2872-2879. Perrin-Ganier, C., Schiavon, F., Morel, J.-L. and Schiavon, M. (2001). Effect of sludge-amendment or nutrient addition on the biodegradation of the herbicide isoproturon in soil. Chemosphere 44, 887-892. Pesce, S., Bequet, J., Rouard, N., Devers-Lamrani, M. and Martin-Laurent, F. (2013). Response of a diuron-degrading community to diuron exposure assessed by real-time quantitative PCR monitoring of phenylurea hydrolase A and B encoding genes. Applied Microbiology and Biotechnology 97 (4), 1661-1668. Philippot, L., Ritz, K., Pandard, P., Hallin, S. and Martin-Laurent, F. (2012). Standardisation of methods in soil microbiology: progress and challenges. FEMS Microbiology Ecology 82, 1-10. Pietramellara, G., Ascher, J., Borgogni, F., Ceccherini, M., Guerri, G. and Nannipieri, P. (2009). Extracellular DNA in soil and sediment: fate and ecological relevance. Biology and Fertility of soils 45, 219-235. Pieuchot, M., Perrin-Ganier, C., Portal, J.-M. and Schiavon, M. (1996). Study on the mineralization and degradation of isoproturon in three soils. Chemosphere 33 (3), 467-478. Pimentel, D., Wilson, C., McCullum, C., Huang, R., Dwen, P., Flack, J., Tran, Q., Saltman, T. and Cliff, B. (1997). Economic and environmental benefits of biodiversity. Bioscience 47, 747-758. Piutti, S., Hallet, S., Rousseaux, S., Philippot, L., Soulas, G. and Martin-Laurent, F. (2002). Accelerated mineralisation of atrazine in maize rhizosphere soil. Biology and Fertility of Soils 36, 434-441. Pot, V., Simůnek, J., Benoit, P., Coquet, Y., Yra, A. and Martínez-Cordón, M.-J. (2005). Impact of rainfall intensity on the transport of two herbicides in undisturbed grassed filter strip soil cores. Journal of Contaminant Hydrology 81 (1-4), 63-88.

Chapter 1 General introduction 48

Potter, T.L., Strickland, T.C., Joo, H. and Culbreath, A.K. (2005). Accelerated soil dissipation of tebuconazole following multiple applications to peanut. Journal of Environmental Quality 34, 1205-1213. Pozo, C., Martinez-Toledo, V., Salmeron, V., Rodelas, B. and Gonzalez-Lopez, J. (1995). Effect of chlorpyrifos on soil microbial activity. Environmental Toxicology and Chemistry 14 (2), 187-192. Prosser, J.I. (2002). Molecular and functional diversity in soil micro-organisms. Plant and Soil 244, 9-17. Racke, K.D. (1993). Environmental fate of chlorpyrifos: review. Environmental Contamination and Toxicology 131, 1-151. Racke, K.D., Steele, K.P., Yoder, R.N., Dick, W.A. and Avidov, E. (1996). Factors affecting the hydrolytic degradation of chlorpyrifos in soil. Journal of Agricultural and Food Chemistry 44 (6), 1582-1592. Rathore, H.S. and Nollet, L.M.L. (2012). Pesticides – evaluation of environmental pollution. CRC Press Taylor & Francis Group. Reitzel, L.A., Dalsgaard, P.W., Muller, I.B. and Cornett, C. (2012). Identification of ten new designer drugs by GC-MS, UPLC-QTOF-MS, and NMR as part of a police investigation of a Danish internet company. Drug Testing and Analysis 4 (5), 342-354. Rieke, S., Koehn, S., Hirsch-Ernst, K., Pfeil, R., Kneuer, C. and Marx-Stoelting, P. (2014). Combination effects of (tri)azole fungicides on hormone production and xenobiotic metabolism in a human placental cell line. International Journal of Environmental Research and Public Health 11, 9660-9679. Roberts, S. J., Walker, A., Cox, L. and Weich, S. J. (1998). Isolation of isoproturon-degrading bacteria from treated soil via three different routes. Journal of Applied Microbiology 85, 309-316. Rocca, J.D., Hall, E.K., Lennon, J.T., Evans, S.E., Waldrop, M.P., Cotner, J.B., Nemergut, D.R., Graham, E.B. and Wallenstein, M.D. (2015). Relationships between protein-encoding gene abundance and corresponding process are commonly assumed yet rarely observed. The ISME Journal 9, 1693-1699. Ronhede, S., Jensen, B., Rosendahl, S., Kragelund, B. B., Juhler, R. K. and Aamand, J. (2005). Hydroxylation of the herbicide isoproturon by fungi isolated from agricultural soil. Applied and Environmental Microbiology 71 (12), 7927-7932. Sancez-Gonzalez, S., Pose-Juan, E., Herrero-Hernandez, E., Alvarez-Martin, A., Sancez-Martin, M.J. and Rodriguez-Cruz, S. (2013). Pesticide residues in groundwaters and soils of agricultural areas in the Águeda River Basin from Spain and Portugal. International Journal of Environmental Analytical Chemistry 93, 1585-1601. Sancho, J.V., Pozo, O.J., Ibanez, M. and Hernandez, F. (2006). Potential of liquid chromatography/time-of-flight mass spectrometry for the determination of pesticides and transformation products in water. Analytical and Bioanalytical Chemistry 386, 987- 997. Sardar, D. and Kole, R. K. (2005). Metabolism of chlorpyrifos in relation to its effect on the availability of some plant nutrients in soil. Chemosphere 61, 1273-1280. Sasikala, C., Jiwal, S., Rout, P. and Ramya, M. (2012). Biodegradation of chlorpyrifos by bacterial consortium isolated from agriculture soil. World Journal of Microbiology and Biotechnology 28, 1301-1308. Schermerhorn, P.G., Golden, P.E., Krynitsky, A.J. and Leimkuehler, W.M. (2005). Determination of 22 triazole compounds including parent fungicides and metabolites in apples, peaches, flour and water by liquid chromatography/tandem mass spectrometry. Journal of the AOAC International 88 (5), 1491-1502. Scheunert, I. and Reuter, S. (2000). Formation and release of residues of the 14C-labelled herbicide isoproturon and its metabolites bound in model polymers and in soil. Environmental Pollution 108, 61-68. Schuelein, J., Glaessgen, W.E., Hertkorn, N., Schroeder, P., Sandermann, H. Jr. and Kettrup, A. (1996). Detection and identification of the herbicide isoproturon and its metabolites in field samples after a heavy rainfall event. International Journal of Environmental Analytical Chemistry 65, 193-202. Schuster, E. and Schröder, D. (1990). Side-effects of sequentially-applied pesticides on non-target soil microorganisms: Field experiments. Soil Biology and Biochemistry 22 (3), 367-373. Sehnem, N.T., Souza-Cruz, P., Do Carmo R. Peralba, M. and Zachia Ayub, M.A. (2010). Biodegradation of tebuconazole by bacteria isolated from contaminated soils. Journal of Environmental Science and Health Part B 45, 67-72. Servien, R., Mamy, L., Li, Z., Rossard, V., Latrille, E., Bessac, F., Patureau, D. and Benoit, P. (2014). TyPol – A new methodology for organic compounds clustering based on their molecular characteristics and environmental behavior. Chemosphere 111, 613-622. Shalini, K., Kumar, N., Drabu, S. and Kumar Sharma, P. (2011). Advances in synthetic approach to and antifungal activity of triazoles. Beilstein Journal of Organic Chemistry 7, 668-677. Singh, B. K. and Walker, A. (2006). Microbial degradation of organophosphorus compounds. FEMS Microbiology Reviews 30, 428- 471. Singh, B.K., Walker, A., Morgan, J.A. and Wright, D.J. (2003). Effects of soil pH on the biodegradation of chlorpyrifos and isolation of a chlorpyrifos-degrading bacterium. Applied and Environmental Microbiology 69, 5198-5206. Singh, D.K. (2012). Pesticide chemistry and toxicology. Bentham Science, Volume 1, ISBN 978-1-60805-531-9. Singh, D.P., Khattar, J.I.S., Nadda, J., Singh, Y., Garg, A., Kaur, N. and Gulati, A. (2011). Chlorpyrifos degradation by the cyanobacterium Synechocystis sp. strain PUPCCC 64. Environmental Science and Pollution Research 18, 1351-1359. Skark, C. and Zullei-Seibert, N. (1995). The occurence of pesticides in groundwater – results of case studies. International Journal of Environmental Analytical Chemistry 58, 387-396. Skark, C., Zullei-Seibert, N., Willme, U., Gatzemann, U. and Schlett, C. (2004). Contribution of non-agricultural pesticides to pesticide load in surface water. Pest Management Science 60, 525-530. Sorensen, S. R, Ronen, Z. and Aamand, J. (2001). Isolation from agricultural soil and characterization of a Sphingomonas sp. able to mineralize the phenylurea herbicide isoproturon. Applied and Environmental Microbiology, 5403-5409. Sorensen, S. R. and Aamand, J. (2001). Biodegradation of the phenylurea herbicide isoproturon and its metabolites in agricultural soils. Biodegradation 12, 69-77. Sorensen, S.R., Ronen, Z. and Aamand, J. (2002). Growth in co-culture stimulates metabolism of the phenylurea herbicide isoproturon by Sphingomonas sp strain SRS2. Applied and Environmental Microbiology 68, 3478-3485. Sparks, T.C. and Nauen, R. (2015). IRAC: Mode of action classification and insecticide resistance management. Pesticide Biochemistry and Physiology 121, 122-128. Srinivasulu, M. and Rangaswamy, V. (2013). Influence of insecticides alone and in combination with fungicides on enzyme activities in soils. International Journal of Environmental Science and Technology 10, 341-350. Steiman, R., Benoit-Guyod, J.-L., Seigle-Murandi, F., Sage, L., and Toe, A. (1994). Biodegradation of pentachlorophenol by micromycetes II. Yeast, Ascomycetes and Basidiomycetes, Environmental Toxicology and Water Quality 9, 1-6. Strickland, T.C., Potter, T.L. and Hyun, J. (2004). Tebuconazole dissipation and metabolism in Tifton loamy sand during laboratory incubation. Pest Management Science 60, 703-709. Sun, J., Huang, X., Chen, Q., Liang, B., Qiu, J., Ali, S. W. and Li, S. (2009). Isolation and characterization of three Sphingobium sp. strains capable of degrading isoproturon and cloning of the catechol 1,2-dioxygenase gene from these strains. World Journal of Microbiology and Biotechnology 25, 259-268.

Chapter 1 General introduction 49

Tag-El-Din, A. (1982). Herbicides and respiration flush of soil biomass. Medical Faculty Landbouww University Gent 47, 191-199. Tiwari, M. K. and Guha, S. (2014). Kinetics of biotransformation of chlorpyrifos in aqueous and soil slurry environments. Water Research 51, 73-85. Tixier, C., Sancelme, M., Aït-Aïssa, S., Widehem, P., Bonnemoy, F., Cuer, A., Truffaut, N. and Veschambre, H. (2002). Biotransformation of phenylurea herbicides by a soil bacterial strain, Arthrobacter sp. N2: structure, ecotoxicity and fate of diuron metabolite with soil fungi. Chemosphere 46, 519-526. Topp, E. (2003). Bacteria in agricultural soils: diversity, role and future perspectives. Canadian Journal of Soil Science 83, 303-309. Turnbull, G.A., Ousley, M., Walker, A., Shaw, E., and Morgan, J.A.W. (2001). Degradation of substituted phenylurea herbicides by Arthrobacter globiformis strain D47 and characterization of a plasmid-associated hydrolase gene, puhA. Applied and Environmental Microbiology 67, 2270-2275. U.S. EPA (1990). Tebuconazole Memorandum. U.S. Environmental Protection Agency (U.S. EPA), Washington, DC. U.S. EPA (2011). Office of Chemical Safety and Pollution Prevention (7503P) EPA 733-R-11-001. Pesticides industry sales and usage - 2006 and 2007 market estimates. Van Beelen, P. and Doelman, P. (1997). Significance and application of microbial toxicity tests in assessing ecotoxicological risks of contaminants in soil and sediment. Chemosphere 34, 455-499. Van-Camp, L., Bujarrabal, B., Gentile, A.-R., Jones, R.J.A., Montanarella, L., Olazabal, C. and Selvaradjou, S.-K. (2004). Reports of the technical working groups established under the thematic strategy for soil protection. EUR 21319 EN/4, 872 pp. Office for Official Publications of the European Communities, Luxembourg. Wagner, A., Adrian, L., Kleinsteuber, S., Andreesen, J.R. and Lechner, U. (2009). Transcription analysis of genes encoding homologues of reductive dehalogenases in “Dehalococcoides“ sp. stain CBDB1 by using terminal restriction fragment length polymorphism and quantitative PCR. Applied and Environmental Microbiology 75, 1876-1884. Wagner, R. (1994). The regulation of ribosomal RNA synthesis and bacterial cell growth. Archives of Microbiology 161, 100-109. Wallance, D.F. and Dickinson, D.J. (2004). Biotransformation of organic – Longevity versus environmental acceptance. COST Secretariat, Brusseld, Belgium. Wang, F., Yao, J., Chen, H., Chen, K., Trebse, P. and Zaray, G. (2010). Comparative toxicity of chlorpyrifos and its oxon derivates to soil microbial activity by combined methods. Chemosphere 78, 319-326. White, P.M., Potter, T.L. and Culbreath, A.K. (2010). Fungicide dissipation and impact on metolachlor aerobic soil degradation and soil microbial dynamics. Science of the Total Environment 408, 1393-1402. Xu, G., Zheng, W., Li, Y., Wang, S., Zhang, J. and Yan, Y. (2008). Biodegradation of chlorpyrifos and 3,5,6-trichloro-2-pyridinol by a newly isolated Paracoccus sp. strain TRP. International Biodeterioration and Biodegradation 62, 51-56. Xu, G.M., Li, Y.Y., Zheng, W., Peng, X., Li, W. and Yan, Y.C. (2007). Mineralization of chlorpyrifos by co-culture of Serratia and Trichosporon sp. Biotechnology Letters 29, 1469-1473. Yang, C., Liu, N., Guo, X.M. and Qiao, C.L. (2006). Cloning of mpd gene from a chlorpyrifos degrading bacterium and use of this strain in bioremediation of contaminated. FEMS Microbiology Letters 265, 118-125. Yang, H., Carr, P.D., McLoughlin, S.Y., Liu, L.W., Horne, I., Qui, X., Jeffries, C.M., Russell, R.J., Oakeshott, J.G. and Ollis, D.L. (2003). Evolution of an organophosphate degrading enzyme: a comparison of natural and directed evolution. Protein Engineering 16, 135-145. Yang, L., Zhao, Y., Zhang, B., Yang, C. and Zhang, X. (2005). Isolation and characterization of a chlorpyrifos and 3,5,6-trichloro-2- pyridinol degrading bacterium. FEMS Microbiology Letters 251, 67-73. Yu, Y. L., Fang, H., Wang, X., Wu, X. M., Shan, M. and Yu, J. Q. (2006). Characterization of a fungal strain capable of degrading chlorpyrifos and its use in detoxification of the insecticide on vegetables. Biodegradation 17, 487-494. Yu, Y.L., Chen, Y.X., Luo, Y.M., Pan, X.D., He, Y.F. and Wong, M.H. (2003). Rapid degradation of butachlor in wheat rhizosphere soil. Chemosphere 50 (6), 771-774. Zeng, W., Han, H., Tao, Y., Yang, L., Wang, Z. and Chen, K. (2013). Identification of bioactive metabolites of gentiopicroside by UPLC/Q- TOF MS and NMR. Biomedical Chromatography 27, 1129-1136. Zhang, J., Hong, Q., Qinfen, L., Li, C., Cao, L., Sun, J., Yan, X. and Li, S. (2012). Characterization of isoproturon degradation pathway in Sphingobium sp. YBL2. International Biodeterioration and Biodegradation 70, 8-13. Zhang, W., Jiang, F. and Ou, J. (2011). Global pesticide consumption and pollution: with China as a focus. Proceedings of the International Academy of Ecology and Environmental Sciences 1 (2), 125-144. Zhu, J., Zhao, Y. and Qiu, J. (2010). Isolation and application of a chlorpyrifos-degrading Bacillus licheniformis ZHU-1. African Journal of Microbiology Research 4, 2410-2413.

Chapter 1 General introduction 50

Chapter 2

Towards a better pesticide policy for the European Union

Veronika Storck Dimitrios G. Karpouzas Fabrice Martin-Laurent

Published in Science of the Total Environment, http://dx.doi.org/10.1016/j.scitotenv.2016.09.167 (2016).

Foreword

The opinion paper of Chapter 2 presents ways how the environmental risk assessment of the pesticide authorization process in the EU could be improved. I wrote it after I had faced several difficulties to get access to detailed information on the environmental risk assessment data provided by the pesticide companies within the authorization process of pesticides. Indeed, during my search on how study results were obtained (regarding certain details), I came across non-transparency impeding the setting-up of my own research. During my search on the existence and the ecotoxicological impact of transformation products (TPs) of each of the three pesticides considered in my PhD, I realized and then questioned that not all known pesticide TPs are included in environmental risk assessment studies. Moreover, understanding that the methods used for evaluating the ecotoxicological impact of pesticides on soil microorganisms are not sufficient to study potential effects, made me assume that methods evaluating impacts in other areas out of my specific expertise may as well exhibit weaknesses. These discoveries motivated me to go deeper into understanding the functioning of the whole pesticide authorization procedure. Based on this critical analysis, I started to develop ideas on how the process could be enhanced for a more sustainable use of pesticides.

Chapter 2 Towards a better pesticide policy for the European Union 54

Abstract

This opinion article aims to foster the debate about pesticide legislation in the European Union (EU). Numerous formerly authorized and widely used pesticides are now banned in the EU because unexpected and unacceptable risks emerged after their initial introduction to the market. Throughout this time lapse, environmental quality and human health have been threatened by the use of these compounds. These hazards could have been prevented by a more responsive pesticide regulatory framework. This article provides detailed insights into the pros and cons of pesticides, and points out weaknesses of the current pesticide environmental risk assessment procedures. Possibilities for improving the robustness and reliability of the pesticide regulatory framework are discussed.

Figure 2.1. Graphical abstract of Chapter 2.

Chapter 2 Towards a better pesticide policy for the European Union 56

An emerging need for changes in pesticide policy

According to the Nobel Prize chemist Paul Crutzen, our planet has entered the Anthropocene, a new epoch dominated by intense human activities causing global environmental changes (Crutzen, 2002). Although scientists still debate about recognizing the Anthropocene (Zalasiewicz et al., 2008), they agree that the global dispersion of man-made chemicals is one of the key drivers of pollution and threatens environmental quality and human health (Hutzinger, 2013; Lewis and Maslin, 2015). Among man-made chemicals, pesticides2 are one of the few chemical groups that are intentionally released into the environment (mainly to protect crops from pest infestations). Approximately 2.4 million tons of active substances are released into the environment worldwide each year (U.S. EPA, 2011), and Europe is the world's largest pesticide consumer (Enserink et al., 2013). Today, the pesticide market is dominated by a limited number of large agrochemical companies (Bayer CropScience/Monsanto, BASF, Syngenta/ChemChina, Dow AgroSciences, and DuPont) which cover approximately 80 % of the pesticide market worldwide (Rathore and Nollet, 2012). Three of these large companies (Bayer CropScience/Monsanto, BASF, and Syngenta/ChemChina) are based in Europe and thus contribute to the development of the European economy. Although authorization is only granted if unacceptable effects of an active substance on the environment are excluded (according to article 4 of EU-Regulation 1107/2009/EC), the application of PPPs containing this active substance at large scales can have undesirable consequences for the environment that have to be reported by the holder of the authorization to the Rapporteur Member State (RMS) (according to article 56 of 1107/2009/EC). Indeed, pesticide may persist in soils for years, and can gradually disperse through different routes over long distances, leading to contamination of non- directly exposed environments such as water resources, the deep sea, or mountain peak regions (Barbash, 2003; Looser et al., 2000; Smalling et al., 2013). As a consequence, our planet is permeated by these compounds, with unknown consequences for environmental quality and human health (Barbash, 2003).

2 The word ‘pesticide’ implies both, the ‘active substance’ (i.e. the molecule with pesticidal effects) and the ‘plant protection product’ (PPP, i.e. the commercial formulation containing the active substance and other chemicals). When distinction is necessary, the according designation ‘active substance’ or ‘PPP’ is used.

Chapter 2 Towards a better pesticide policy for the European Union 58

For instance, the International Agency for Research on Cancer (IARC) recently reported that glyphosate-based PPPs (initially known under the trademark Roundup® and intensively used for decades (Bundesinstitut für Risikobewertung BfR, 2015)) were genotoxic (caused DNA mutations) and probably carcinogenic (involved in causing cancer) to humans (IARC, 2015). A global debate ensued about the unacceptable weaknesses of pesticide policy. During the ongoing peer review of glyphosate, the European Commission (EC) gave a mandate to the European Food Safety Authority (EFSA) to consider the findings by the IARC reporting the potential carcinogenicity of glyphosate PPPs (EFSA, 2015a). Based on this re-evaluation, EFSA concluded that the active substance glyphosate (used in non-selective herbicides for weed control) was unlikely to be genotoxic or to pose a carcinogenic threat to humans. Consequently, it was not proposed for classification as carcinogenic under EU regulation. However, EFSA admitted that this distinction between active substance and PPPs mainly explains the contradictory conclusions of IARC and EFSA. For this reason, it was proposed that the toxicity of these PPPs should be re-assessed at Member State level before granting authorization at national level (EFSA, 2015a). Very recently, the joint FAO/WHO meeting on pesticide residues concluded that glyphosate and glyphosate-based PPPs were not associated with genotoxic effects in the majority of studies conducted on mammals (Summary report from the May 2016 Joint FAO/WHO Meeting on Pesticide Residues, 2016). In this context, we cannot but notice that the big debate around glyphosate approval review is a symptom of weaknesses in the current pesticide policy. In a recent review paper about the hidden and external costs of pesticide use connected to the current debate about pesticide benefits and adverse consequences, the authors concluded that the costs of pesticide use (such as human health costs) may have outreached its benefits and should be included for a more accurate evaluation of pesticide use and for regulatory purposes (Bourguet and Guillemaud, 2016). Furthermore, organizations such as the (PAN), with the specific remit to minimize negative effects of pesticides and replace the use of harmful pesticides by ecologically sound alternatives, have expressed their concern about the current pesticide policy (PAN, 2011). As a consequence, a European ombudsman recently accused the Brussels commission to be too lenient in the authorization process of pesticides and requested its profound reform (O’Reilly, 2016). In this no-win

Chapter 2 Towards a better pesticide policy for the European Union 59

situation, it has become evident that the European pesticide policy needs to be substantially improved to minimize risks for environmental quality and human health.

What are the pros and cons of pesticides?

From an idealistic perspective, the easiest way to reduce or eliminate risks associated with pesticide application is to stop using them. But this is easier said than done because their application brings substantial direct benefits to humans by protecting crops from weeds, insects, and other pests. In a hypothetical scenario without any pesticide use in the USA, the expected crop yield losses were estimated to be 32 % for corn, 57 % for rice, and 24 % for wheat (Knutson, 1999). Furthermore, crop prices were estimated to increase by 38 % for corn, 83 % for rice, and 6 % for wheat in such a scenario. In reality, however, the pesticide use increased approximately 50-fold in the 20th century to feed the rapidly growing human population (Rathore and Nollet, 2012). This trend may even intensify in the next 50 years as the world population is expected to increase by almost 25 % according to United Nation statistics (UN, 2015). Less than 13 % of unused and fertile land areas are thought to remain on earth (Leisinger, 2008), with an ensuing need to further intensify agricultural production, where pesticides can be seen as a key player against food shortage. Despite the benefits of pesticides for humans, their residues pose serious threats to the environmental quality and human health. As a consequence, several persistent and hazardous pesticides massively used in the past have been banned. To get a clear view of the pesticide paradox, one must take a closer look at the downside of pesticides - their environmental contamination and their potential toxicity to non-target organisms. Regarding their environmental contamination, only a tiny part of the sprayed amount of pesticides reaches the target organism (weeds, insects, fungi, bacteria) (Zhang et al., 2011) while the rest ends up in the soil or on the crops from where the pesticides can diffuse to other environmental sectors or enter the food chain. For instance, EFSA recently published a report on pesticide residues detected in food samples collected from the 27 EU Member States, plus Iceland and Norway: 55 % of the samples were free of quantifiable target analytes - but one should keep in mind that only known compounds previously classified as ‘relevant’ were monitored. Potentially

Chapter 2 Towards a better pesticide policy for the European Union 60

harmful compounds, as yet unknown or not yet classified as ‘relevant’, were not included in this study. In the future, new methods in analytical chemistry could enable more comprehensive investigations (Storck et al., 2016). An amount of 42 % of the samples contained pesticide residues below the established threshold values, 3 % exceeded those threshold values, while 27 % of the samples were contaminated with more than one active substance (multiple residues) (EFSA, 2015b). The percentages of positive pesticide detections are usually much higher when only conventionally produced fruit and vegetables are studied (The Chemical and Veterinary Investigation Office Stuttgart, 2013a,b). Therefore, these highly contaminated crops may have been overshadowed by other crops in the recent EFSA report that tested all kinds of food. An example for unintended environmental contamination with pesticides is a study of the EU Member States that monitored pesticide residues in surface and ground water systems within the EU-Water Framework Directive 2000/60/EC: Around half of the European surface water systems are contaminated by pesticide residue levels that may pose a risk to non-target organisms (Malaja et al., 2013). Regarding some examples of the potential toxicity of pesticides to non-target organisms, a number of new-generation insecticides (three neonicotinoids and fipronil) was recently shown to have undesirable effects on honey bees among other insects (European Academies Science Advisory Council EASAC, 2015). This had previously been indicated by EFSA (EFSA, 2012a; EFSA, 2013a). There is growing evidence that the weakening or death of bees is caused by the combined effects of a number of stressors, among which neonicotinoids have raised concern in the last years (Fairbrother et al., 2014). Recently, EFSA published a new guidance document proposing new specific protection goals to conduct the risk assessment of PPPs on bees (EFSA, 2013b). In addition, as mandated by the EC (M-2015-0246), in the context of the revision of the approval of the active substances clothiandin, imidacloprid and thimethoxam applied as seed treatments and granules, EFSA launched a call for new scientific information to assess their risk to bees. The apiculture sector is particularly important for the EU because bees are key players in plant pollination, an essential ecosystem service, responsible for about 85 % of agricultural crop yields (German Beekeeping Organization, 2015). That is why the EU restricted the use of these insecticides and defined steps to improve the procedures for assessing the risks of active substances to bees. The French National Assembly very recently (March 2016) voted for an

Chapter 2 Towards a better pesticide policy for the European Union 61

amendment banning neonicotinoids to preserve pollinators. In addition to the toxic effects of pesticides to the environment, human health can also be impaired. For instance, this was the case in the French West Indies where the extensive use and environmental accumulation of chlordecone was associated with increased frequency of prostate cancer (Multigner et al., 2010) and infantile delays in cognitive, visual, and motor development (Dallaire et al., 2012) in the local population exposed to chlordecone-based PPPs. Although agroecology offers new paths for the development of ecologically sustainable food systems (Gliessman, 2014), it is hard to imagine the worldwide implementation of agriculture without pesticide use. To feed the growing world population, a number of innovations needs to be implemented to intensify agricultural production and simultaneously ensure environmental and human health protection. The safe use of pesticides appears as one of the biggest challenges of agricultural intensification. Within this context, improving pesticide policy may be a good leverage.

How does a new pesticide reach the market?

For the sake of constructive criticism and to have a chance of improving pesticide policy, a good understanding of the various steps of the pesticide authorization process is required. In the EU, only active substances registered on the EU’s list of approved active substances and subsequently authorized as PPPs by each EU Member State are released into the environment (Figure 2.2). The authorization is granted only if proposed uses are not expected (or known) to have harmful effects on environmental, animal, or human health. The procedures and requirements concerning the placement of PPPs on the market are specified in the EU-Regulation 1107/2009/EC, which repeals Council Directives 79/117/EEC and 91/414/EEC. After the submission of a dossier by a pesticide company, the first step of the authorization process is the production of an initial draft assessment report (DAR) by an EU-designated RMS. The RMS can modify and amend, and concludes on the risk assessment provided by the pesticide company, evaluating in the light of uniform principles of EU-Regulation 546/2011/EC. The DAR concludes on the different information provided by the pesticide company including (i) the identity of the active substance and its biological efficacy, (ii) its toxicology and

Chapter 2 Towards a better pesticide policy for the European Union 62

metabolism in mammals, (iii) its metabolic pathway and its residues in plants, crops and livestock and the risk for consumers, (iv) its environmental fate and behavior in soil, water and air, (v) its ecotoxicological impact on several non-target soil and aquatic organisms, including soil microorganisms and (vi) any relevant information found in the literature, if available (EU-Regulation 283/2013/EC; EU-Regulation 284/2013/EC). The data presented in the DAR results from studies carried out by the pesticide producer or assigned subcontractor companies. The second step is the peer review of the RMS’s dossier, including the risk assessment, which is coordinated by the Pesticides Unit of EFSA. In the third step, the Pesticides Unit of EFSA produces a conclusion report. Risk managers of the Standing Committee on Plants, Animals, Food and Feed (SCOPAFF) analyze the EFSA’s conclusions, and decide whether or not to register the compound on the EU’s list of approved active substances (step 4). The first approval of an active substance must not exceed 10 years (15 years for low-risk substances), but review of the approval can be requested by the EC at any time if risk concerns emerge (according to article 21 of 1107/2009). The review of an approval should not exceed 15 years. In the same manner as for the initial approval, the review of the approval starts with the designation of an RMS by the EC in charge of providing a renewed assessment report (RAR) taking into consideration data resulting from post-authorization monitoring within the EU-Air Framework Directive 2008/50/EC and the EU-Water Framework Directive 2000/60/EC and literature search. The RAR is then subjected to the same procedure as the DAR (EU-Regulation 283/2013/EC; EU-Regulation 284/2013/EC). Finally, risk managers of each EU Member State separately take the final decision at the national level of granting or not granting authorization to use an EU-authorized active substance in a PPP without EFSA being involved (step 5). For example, in Germany the DAR is reviewed by the Federal Office of Consumer Protection and Food Safety (BVL) in collaboration with the Julius Kühn Institute (JKI), the Federal Environment Agency (UBA) and the Federal Institute for Risk Assessment (BfR), who decide on authorization of a PPP at a national scale (Federal Institute for Risk Assessment, 2015). EU-Regulation 1107/2009/EC introduced for the first time the division of Europe into three geographical zones according to pedo-climatic criteria (i.e. Scandinavian, temperate, and Mediterranean zones proposed in Annex I) to harmonize and speed up the regulatory process in different Member States. Within the same zone, the peer review procedure on PPPs is similar to that of active substances. First, the RMS evaluates the registration

Chapter 2 Towards a better pesticide policy for the European Union 63

report of the PPP, then the other Member States can comment on the first evaluation. Examples for national institutions in charge of evaluating registration reports of PPPs are the National Agency of Sanitary Security of Alimentation, Environment and Work (ANSES) in France or the National Institute of Investigation, Agrarian Technology and Alimentation (INIA) in Spain. The final decision is then taken by national governments based on the proposal of risk managers. National authorities may also impose risk mitigation measures and monitoring after authorization of a PPP.

Figure 2.2. Schematic presentation of the stepwise process of pesticide authorization in the EU.

Chapter 2 Towards a better pesticide policy for the European Union 64

Are pesticide impacts on the environment and human health excluded?

Despite the substantial efforts of the relevant authorities to regulate the pesticide market, pesticides are still one of the most widespread pollutant groups worldwide (Hutzinger, 2013; Lewis and Maslin, 2015). Pesticide policy is heavily criticized for being relatively inefficient and for compromising the evolution of a non-precautionary principle (EEA, 2013). For example, a pesticide ban does not necessarily mean that the pesticide is totally off the market. Exceptional authorizations for special uses of banned pesticides have increased five-fold within four years (from 2006 to 2010) according to the PAN (2011). With 74 banned pesticides getting exceptional authorization for special uses, France was nominated the ‘European backdoor champion’ in 2010 - followed by Greece (54), Portugal (31), and Germany (24). The most frequent argument of the authorities to explain this questionable procedure is the lack of equally efficient pesticides for the particular use. To stop this increasing trend, each special use has to be strictly considered on its merits. The decision processes for the establishment of the threshold values for pesticide residues in food are also subject to criticism. As an example, the initial maximum residue level (MRL) of glyphosate in lentils was set by the Pesticide Unit of EFSA at the limit of analytical quantification and had no reference to toxicology. In 2011, following the frequent detection of glyphosate exceeding the MRL, it was increased 100-fold as the toxicological reference value was higher and therefore glyphosate residues were considered not to pose a public health concern (EFSA, 2012b). Instead of increasing the MRL, a better option may be solving the problem at its source by improving agronomical practices. All over Europe, the level of pesticide contamination has not substantially diminished despite the efforts of different Member States to reach the objectives of relevant directives like the EU-Water Framework Directive 2000/60/EC and of the EU- Directive 2009/128/EC for sustainable use of pesticides. The Federation for Environmental Studies and Nature Conservation ‘BUND’ (2011) has pointed out the lack of a specific regulation regarding the use of pesticides in drinking water protection areas important for the production of clean drinking water, and criticized the authorization to

Chapter 2 Towards a better pesticide policy for the European Union 65

simultaneously use two or more pesticides without knowing their combined effects on non-target organisms. Furthermore, the BUND asked for clarifications in the regulatory documents since phrases commonly used in such documents like ‘good practice’ or ‘imminent danger’ are too broad and have to be clearly defined to enable a closer insight into agricultural practices through regular controls and repressive actions. Among other voices criticizing the current pesticide policy, the PAN (2011) wishes a constitutional change of the fundaments of pesticide risk assessment, notably because of the involvement of pesticide manufacturers in the first step of the decision- making process (Figure 2.2), when they can be seen as both judge and party. Although the multi-actor decision chain of pesticide authorization was originally designed to guarantee consensus about pesticide authorizations or bans, the procedure presently arouses suspicion because it is perceived as a potential source of conflict of interest. Thus, pesticide manufacturers can appear suspect to the society who presumes that they produce biased risk assessment results and influence authorization decisions, although they actually operate within legal possibilities and duties. Changes in the fundamentals of the pesticide authorization processes appear as a possible way to protect pesticide manufacturers from suspicion by the society. In addition, the popularization of scientific knowledge could improve the acceptability of the pesticide authorization process by the society. In addition, specific environmental protection goals targeted by pesticide environmental risk assessment are currently debated. Possible harmful effects on soil biodiversity, which supports numerous ecosystem services such as food production or climate regulation, appear as a particularly relevant proposal. To tackle this issue, EFSA proposed new specific protection goals covering ecosystem services potentially affected by the use of pesticides (EFSA, 2010). It also suggested the development of a new environmental risk assessment framework to assess the temporal and spatial ecological recovery of non-target organisms exposed to regulated stress factors including pesticides (EFSA, 2016). However, these new concepts have not yet been fully implemented in a new regulatory framework, except for aquatic organisms in edge-of- field surface waters carried out according to the guidance on aquatic risk assessment (EFSA, 2013c). On top of that, the post-authorization evaluation of pesticide impacts is weak. This is particularly the case for the description of pesticide degradation processes in

Chapter 2 Towards a better pesticide policy for the European Union 66

soils: The lack of progress could be attributed to the absence of a soil protection framework, although such a framework has already been proposed to the EC (Van-Camp et al., 2004). Although, the current regulation estimates possible risks based on the calculation of predicted environmental concentrations (PECs) and toxicity endpoints (according to FOCUS scenarios on water resources), it is still difficult to accurately predict the fate and the impact of pesticides in the environment on the long term (Fenner et al., 2013). The data required to conduct environmental risk assessment do not take into account all pesticide transformation processes or the environmental parameters that influence pesticide fate, as it is very difficult to perform risk assessment studies in all kinds of environments. For instance, pesticide degradation estimated according to the OECD-Guideline 307 (2002) impose testing in three soils which do not necessarily originate from the three different pedo-climatic zones (EU-Regulation 1107/2009/EC) for authorization, and may not be enough to predict pesticide transformation under a range of various situations found across Europe, particularly in the context of climate change. Although the precautionary principle should be applied according to Article 8 of EU-Regulation 1107/2009/EC, several organizations such as the BUND and the PAN wonder whether it is done, as pesticides remain on the market for too long after new risks have emerged because the post-authorization decision process is slow (BUND, 2011; PAN, 2011). This point of view is shared by the European Environment Agency (EEA) that, on the basis of historical case studies, underlines that early warnings were ignored or side-lined until damage to health and the environment was inevitable (EEA, 2013). Moreover, these organizations criticize a lack of transparency in the decision- making processes and results because it impedes public research and causes suspicion. Although, EFSA contributes to the transparency of pesticide authorization by releasing conclusions on peer review of risk assessment for each approved active substance, these documents are barely understandable for non-experts and often not sufficient for researchers who would need more details. This leads to delayed emergence of risks, and meanwhile the environment and our health may be endangered.

Chapter 2 Towards a better pesticide policy for the European Union 67

Confessions of risks after years of pesticide use are not uncommon

Pesticides typically undergo different life stages from their ‘birth’ (creation) to their ‘death’ (ban) (Figure 2.3). After its approval, an active substance enters the market in an authorized PPP, where it is meant to become a success story and thus allows the manufacturer to make a profit invested in the development of new products among others. Review of the active substance approval is undertaken not more than 10 years after the market introduction (or 15 years after the latest risk assessment) or can be requested by the EC at any time by considering information from post-authorization monitoring. Meanwhile, the PPP is sold and applied to control pests in crops worldwide. As a result, pesticides can accumulate in the soil, diffuse to other environments, and come in contact with non-target organisms. Once in the environment, most active substances are transformed into a huge number of transformation products through various abiotic and biotic processes (just like throwing a Lego® brick construction to the ground where it disaggregates into pieces). The properties of the transformation products can greatly differ from those of the original active substance (Sinclair et al., 2010; Tixier et al., 2000). Transformation products are often more soluble than the parent compound, so that they more easily migrate to water bodies (Fenner et al., 2013). Although the ecotoxicological impact of most transformation products is usually lower than that of the parent compound (Boxall et al., 2004), some of them may be more toxic (Fenner et al., 2013; Rochkind et al., 1986). At the time of the initial authorization, information on all possible transformation products is not known. Environmental risk assessment only considers transformation products classified as ‘relevant’ (i.e. those that represent at least 10 % of the initial amount of the applied active substance) (EU- Regulation 283/2013/EC; Fenner et al., 2013), whereas other transformation products that are not classified as ‘relevant’ (based on their detection at low concentrations in the tested soil and aquatic environments) can be mentioned but do not need to be considered in risk assessment. The prerequisites for the inclusion of transformation products into the ‘relevant list’ should be redefined and permanently renewed not only based on quantified amounts (as it is now), but also based on their inherent toxicity on non-target organisms. As a result of today’s definition of ‘relevant’ transformation

Chapter 2 Towards a better pesticide policy for the European Union 68

products, harmful molecules can slip through the precautionary principle and public research is impeded if the existence of newly emerged transformation products is confidential. If unacceptable risks become evident through scientific - often public - research after years of pesticide usage, its authorization can be reviewed by EFSA mandated by the EC on the request of a Member State at any time. Usage restrictions can be imposed by the authorities, such as the inscription on the list of candidates for substitution (CFS, according to Annex II, Paragraph 4 of EU-Regulation 1107/2009/EC), and/or the pesticide can be banned. This life cycle is not uncommon, as illustrated by 77 inscriptions on the list of CFS (EC, 2015) and 48 bans of formerly approved active substances in the EU (PAN, 2015). Past experience showed that ecotoxicologically relevant transformation products typically emerged only 20 to 30 years after the market introduction of certain pesticides (Fenner et al., 2013). Examples are the late report of chloridazon (market introduction in 1964) transformation products in surface and ground water (Buttiglieri et al., 2009) and the delayed monitoring of tolylfluanid (market introduction in 1971) transformation products like N-nitroso-dimethylamine which is carcinogenic (Schmidt and Brauch, 2008). This pesticide lifespan, together with the continuous introduction of new products and the bans of old ones, have created a pesticide market which is under constant evolution.

Figure 2.3. Presentation of a typical pesticide life cycle.

Chapter 2 Towards a better pesticide policy for the European Union 69

Improved pesticide policy is necessary and feasible

Pre-authorization changes

Pre-authorization changes may include (i) the reduction of the time lag between the market introduction of a new pesticide and the awareness of risks, (ii) the commitment of one authority regulating both active substances and PPPs, (iii) the assignment of environmental risk assessment studies to anonymous accredited subcontractors, (iv) redefining which transformation products are considered ‘relevant’, and (v) a limitation of the rapidly evolving pesticide market. Due to the continuous and rapid changes of the pesticide market, manufacturers are one step ahead of public bodies, which stay behind because they are hindered by the current pesticide policy and shortage of financial resources. It is not unusual to have a 20 to 30 years time lag between the development of a new pesticide and the awareness demonstration of its harmfulness by academic studies. This is a long period of time during which the environment and human health are exposed to potential risks associated to these pesticides. To reduce this time lag, EFSA could release in real time information about the introduction of a new active substance in the registration process (and not at the time of its market introduction). This would enable a more contemporaneous initiation of private and public investigations. This way, it would be easier for public research to keep up with private research and be taken into account as early as the first authorization procedure, which would thus be potentially enriched by supplemental information. Another fundamental issue for improving the risk assessment of the whole pesticide authorization process is the fact that decisions are not all taken by the same authority. Active substances are approved at the EU level, while PPPs are authorized at the national level of each EU Member State (Figure 2.1). This leads to responsibility issues, as described earlier for glyphosate: It was stated unlikely to be carcinogenic by EFSA (EFSA, 2015a), but its PPPs were considered as potentially genotoxic by the IARC (IARC, 2015). As an active substance is always applied in a formulation, splitting their risk assessment is not logical and leads to a poorly coordinated and weighed-down authorization process. This kind of issues would be solved if EFSA not only managed the risk assessment and approval process of active substances but also the authorization of

Chapter 2 Towards a better pesticide policy for the European Union 70

their related PPPs. This goal may be difficult to achieve because each Member State relies on deputed institutions to fulfil the authorization process of PPPs on the national level, but EFSA could at least be the responsible authority for the collection and publication of conducted studies. To improve the trustworthiness of the pesticide authorization process, instead of relying on data from the industry, EFSA could anonymously choose accredited subcontractors, not contractually related to the industry, who would perform the risk assessment investigations, on which the DAR (done by the RMS) is based. This would (i) protect the industry from suspicion by the public since it would not be directly involved in the production and delivery of data used for preparing the DAR, (ii) avoid possible conflicts of interest between the industry and subcontractors (since EFSA would assign the investigations to the relevant subcontractor), and (iii) bring more credibility to the overall registration process which would be fully performed under the responsibility of EC authorities (from data production to data evaluation). Compliance with the precautionary principle could be enhanced by improving other steps of the process such as the definition of the relevance of pesticide transformation products. At present, environmental risk assessment only considers active substances and transformation products classified as ‘relevant’ (EU-Regulation 283/2013/EC). Each active substance and its relevant transformation products are examined for their effects on non-target organisms and are included in post- authorization monitoring programs. Transformation products that are not classified as relevant are largely ignored. The criteria for classifying transformation products as relevant should not be solely based on their amounts in the environment, but also on their inherent toxicity. Environmental characteristics of pesticide transformation products including degradation half-lives in different environmental compartments, bioaccumulation factors, mobility in the environment and their inherent toxicity to various organisms could provide a reliable estimate of the potential of a compound to cause undesirable toxic effects to receiving ecosystems. Sustainable risk reduction of transformation products will only be achieved by using such environmental parameters as selection criteria. In silico classification methods such as TyPol (Servien et al., 2014) can be used to prioritize transformation products to be incorporated into the environmental risk assessment (Storck et al., 2016).

Chapter 2 Towards a better pesticide policy for the European Union 71

Alternative ways to reduce the risks associated with the rapidly evolving market of new pesticides could be to only allow the market introduction of new active substances with improved efficiency and better environmental profile than existing ones. The highly discussed ‘cocktail effects’ (i.e. combined effects of several pesticides on non-target organisms) (BUND, 2011) would probably not be solved this way, as several 100 of pesticides are still authorized and simultaneously applied. However, (i) a reduction in the number of exceptional authorizations for special uses of banned pesticides would decrease the chemical pressure exerted on the environment, (ii) the rapidly changing pesticide market would be stabilized and thus public research would be given time to provide additional information for post-authorization environmental risk assessment, and (iii) the list of molecules present in the environment and to be studied and monitored for post-authorization risk assessment would be shortened in the long term.

Post-authorization changes

Post-authorization changes may include (i) the implementation of the soil protection EU-Soil Framework Directive (Van-Camp et al., 2004) to preserve soil ecosystem services, (ii) the transparency of studies influencing pesticide risk assessment, (iii) a standardization of the mode of data collection for monitoring studies, (iv) the application of an action plan to speed up the post-authorization response to newly identified risks, (v) the limitation of the number of exceptional authorizations for special uses of banned pesticides, and (vi) a legal ban for sales and exports of EU-banned pesticides to other continents. To achieve a more sustainable post-authorization management, the proposed EU- Soil Framework Directive COM(2006)232 (Van-Camp et al., 2004) should be implemented. This directive was initially created to protect soil resources. The introduction of the EU-Water Framework Directive 2000/60/EC in 2000, which introduced fixed active substance thresholds in water resources, forced pesticide manufacturers to develop new classes of active substances that are effective at low doses, have a high affinity for the target organism, and show reduced dispersion in the environment (Cessna et al., 2006). In a similar way, implementing the soil directive could put pressure on pesticide manufacturers and politicians to introduce chemicals

Chapter 2 Towards a better pesticide policy for the European Union 72

and practices that will minimize undesirable impacts on environmental quality and human health. The lack of transparency of the authorization and post-authorization processes is another source of serious criticism. Although EFSA publishes conclusions of the risk assessment for transparency, study details often remain out of reach, hindering follow- up research on the same compound by public bodies. One could recommend that all the documents reporting new post-authorization findings (such as new transformation products) should be immediately made available to the public to stimulate parallel research from other bodies and therefore facilitate faster decisions for a more responsive post-authorization risk reduction. The studies or documents upon which the manufacturer’s risk assessment is based should also be included, as potential recommendations are hardly understandable otherwise. Transparency should be enforced by law - similarly to the law in European Member States that grants each citizen the right to know which chemicals are found at which concentrations in natural water bodies. Furthermore, the mode of data collection for monitoring studies of pesticide residues should be standardized to make studies more representative. In addition, data regarding the levels of active substances in the environment (e.g. recreation lakes) should be made available to the public and managed by an EU authority in charge of collecting, standardizing and publishing the data in a fully accessible format, in the same manner as for pesticide residue monitoring in food. Another important aspect is the implementation of an action plan to speed up the post-authorization response to newly identified risks. As an unfortunate matter of fact, science is not omniscient. Even for experts, it remains difficult to predict and anticipate potential risks that can be initially overlooked and only emerge after decades of pesticide use. Therefore, decision makers should (i) not rely on previous risk assessment documents implemented 10 to 20 years ago but employ a new risk assessment based on newly produced datasets and recent studies which guarantee an accurate risk assessment, (ii) take quick decisions on the temporal ban or use restriction of an active substance after its frequent or continuous detection in water resources at levels exceeding the threshold levels or when new potentially toxic transformation products have been detected.

Chapter 2 Towards a better pesticide policy for the European Union 73

Furthermore, measures should be enforced to limit the number of exceptional authorizations for special uses of banned pesticides to an absolute minimum. The reason why some pesticides are now banned at the EU level is that their negative effects on environmental quality and human health have been clearly proved by scientific research. At a national level, improvements have to be made to sustainably protect the environment by thoroughly applying the sustainable use directive (EU-Directive 2009/128/EC). For instance, specific rules for the use of pesticides in drinking water protection areas should be implemented in Europe. Furthermore, pesticide-free areas should be declared as refuge areas for insects (e.g. the honey bee) or other organisms at the national level to counteract the loss of biodiversity. Reduction of the respective national pesticide uses should be enforced by law. To that end, the so-called ‘recommended doses’ (given on the packaging label by the pesticide manufacturer) could be converted into maximum legal doses. Moreover, farmers need to be further made aware that they should not use higher doses than the recommended ones. For this purpose, the various initiatives launched in response to the EU-Directive 2009/128/EC for the ‘sustainable use of pesticides’, such as ‘TOPPS’ (Training the Operators to prevent Pollution from Point Sources, EU), ‘NROSO’ (National Register of Sprayer Operators, UK) and ‘OpenTea’ (Open Training and Education Association, Italy) should be coordinated and spread all over Europe to promote a safer use of pesticides. Last but not least, European pesticide legislation could become an ethical example for the rest of the world by legally banning sales and exports of EU-banned pesticides used for crop protection (if not indispensable to counteract a certain pest) to countries of other continents that may have less strict pesticide legislation. From the ethical point of view, the use of banned insecticides for disease vector control can be acceptable (e.g. DDT approved by the World Health Organization (WHO) for prevention of malaria). This question could be brought to the United Nations within the framework of a global initiative to regulate the world market of man-made chemicals, including pesticides, taking into consideration risks and benefits of pesticide use.

Chapter 2 Towards a better pesticide policy for the European Union 74

Epilogue

After this constructive criticism, facts currently remain as follows: policy makers are going to keep being confronted with questions about how to weigh the benefits of pesticides against their deleterious effects on environmental quality and human health. Possibilities for improvements do exist to achieve a safer use of pesticides defining ‘a new deal’ between farmers, industry and society. In last year's movement against a powerful European chemical company that tests EU-banned pesticides in Hawaii, the Hawaiian politician Gary Hooser said: ‘The only thing that can help us is regulating political decisions. And that can only be enforced by strong public awareness.’ (Tageswoche Basel, 2015).

Acknowledgements

The PhD studies of Veronika Storck performed at the Doctoral School of Environment and Health (ED E2S) at the University of Burgundy were funded by the French Ministry of Education and Research (grant number 2013-31). This work was done within the framework of the FP7-PEOPLE-2012-IAPP (Industry-Academia Partnerships and Pathways) Marie-Curie project ‘Love-to-Hate’ funded by the European Commission (grant number 389 324349). We thank Annie Buchwalter for English editing. We are grateful to the editor and reviewers for helpful comments.

Chapter 2 Towards a better pesticide policy for the European Union 75

References

Barbash, J.E. (2003). The geochemistry of pesticides. Treatise on Geochemistry 9, DOI: 10.1016/B0-08-043751-6/09056-3. Bourguet, D and Guillemaud, T. (2016). The hidden and external costs of pesticide use. Springer International Publishing Switzerland, E. Lichtfouse (ed.), Sustainable Agriculture Reviews 19. Boxall, A.B.A., Sinclair, C.J., Fenner, K., Kolpin, D. and Maund, S.J. (2004). When synthetic chemicals degrade in the environment. Environmental Science and Technology 38 (19), 368A-375A. BUND (2011). Stellungnahme zum Gesetz zur Neuordnung des Pflanzenschutzrechtes. Az. 512-32001/9. Bundesinstitut für Risikobewertung BfR (2015). Löst Glyphosat Krebs aus? Mitteilung 007/2015 des BfR vom 23. März 2015. Buttiglieri, G., Peschka, M., Frömel, T., Müller, J., Malpei, F., Seel, P. and Knepper, T.P. (2009). Environmental occurrence and degradation of the herbicide n-chloridazon. Water Research 43 (11), 2865-2873. Cessna, A.J., Donald, D.B., Bailey, J., Waiser, M. and Headley, J.V. (2006). Persistence of the sulfonylurea herbicides thifensulfuron- methyl, ethametsulfuron-methyl, and metsulfuron-methyl in farm dugouts (ponds). Journal of Environmental Quality 35, 2395-2401. Crutzen, P.J. (2002). Geology of mankind. Nature 415, 23. Dallaire, R., Muckle, G., Rouget, F., Kadhel, P., Bataille, H., Guldner, L., Seurin, S., Chajes, V., Monfort, C., Boucher, O., Thome, J.P., Jacobson, S.W., Multigner, L. and Cordier, S. (2012). Cognitive, visual, and motor development of 7-month-old Guadeloupean infants exposed to chlordecone. Environmental Research 118, 79-85. EC (2015). List of candidates for substitution. http://ec.europa.eu/food/plant/pesticides/approval_active_substances/index_en.htm. EEA (2013). Report No 1/2013. Late lessons from early warnings: Science, precaution, innovation - summary. European Environment Agency 2013 - 44 pp. - 21 x 29.7 cm, ISBN 978-92-9213-349-8, doi:10.2800/70069. EFSA (2010). Scientific opinion on the development of specific protection goal options for environmental risk assessment of pesticides, in particular in relation to the revision of the guidance documents on aquatic and terrestrial ecotoxicology (SANCO/3268/2001 and SANCO/10329/2002), EFSA Panel on Plant Protection Products and their Residues (PPR). EFSA Journal 2010, 8 (10): 1821. EFSA (2012a). Scientific Opinion on the science behind the development of a risk assessment of Plant Protection Products on bees (Apis mellifera, Bombus spp. and solitary bees). EFSA Journal 2012, 10 (5): 2668. EFSA (2012b). Modification of the existing MRL for glyphosate in lentils. EFSA Journal 2012, 10 (1): 2550. EFSA (2013a). Conclusion on the peer review of the pesticide risk assessment for bees for the active substance imidacloprid. EFSA Journal 2013, 11 (1): 3068. EFSA (2013b). EFSA guidance document on the risk assessment of plant protection products on bees (Apis mellifera, Bombus spp. and solitary bees). EFSA Journal 2013, 11 (7): 3295. EFSA (2013c). Guidance on tiered risk assessment for plant protection products for aquatic organisms in edge-of-field surface waters. EFSA Journal 2013, 11 (7): 3290. EFSA (2015a). Conclusion on the peer review of the pesticide risk assessment of the active substance glyphosate. EFSA Journal 2015, 13 (11): 4302. EFSA (2015b). The 2013 European Union report on pesticide residues iESA, 2013n food. EFSA Journal 2015, 13 (3): 4038. EFSA (2016). Recovery in environmental risk assessments at EFSA, EFSA Scientific Committee. EFSA Journal 2016, 14 (2): 4313. Enserink, M., Hines, P.J., Vignieri, S.N., Wigginton, N.S. and Yeston, J.S. (2013). The pesticide planet. Science 341, 730-731. EU-Air Framework Directive 2008/50/EC. Directive 2008/50/EC of the European Parliament and of the Council on ambient air quality and cleaner air for Europe. EU-Directive 2009/128/EC of the European Parliament and of the council of 21 October 2009 establishing a framework for Community action to achieve the sustainable use of pesticides. EU-Regulation 1107/2009/EC of the European Parliament and the council of 21 October 2009 concerning the placing of plant protection products on the market and repealing Council Directives 79/117/EEC and 91/414/EEC. EU-Regulation 283/2013/EC of 1 March 2013 setting out the data requirements for active substances, in accordance with Regulation EC No 1107/2009 of the European Parliament and of the Council concerning the placing of plant protection products on the market. EU-Regulation 284/2013/EC of 1 March 2013 setting out the data requirements for plant protection products, in accordance with Regulation (EC) No 1107/2009 of the European Parliament and of the Council concerning the placing of plant protection products on the market. EU-Regulation 546/2011/EC of 10 June 2011 implementing Regulation (EC) No 1107/2009 of the European Parliament and of the Council as regards uniform principles for evaluation and authorisation of plant protection products. European Academies Science Advisory Council EASAC (2015). Ecosystem services, agriculture and neonicotinoids. http://www.easac.eu/fileadmin/Reports/Easac_15_ES_web_complete_01.pdf. EU-Water Framework Directive 2000/60/EC. Directive 2000/60/EC of the European Parliament and of the Council establishing a framework for community action in the field of water policy. Fairbrother, A., Purdy, J., Anderson, T. and Fell, R. (2014). Risks of neonicotinoid insecticides to honeybees. Environmental Toxicology and Chemistry 33, 719-713. Federal Institute for Risk Assessment (2015). Gesetzliche Verfahren im Bereich Pflanzenschutz. Hintergrundinformation Nr. 029/2015 des BfR vom 14. September 2015. Fenner, K., Canonica, S., Wackett, L.P. and Elsner, M. (2013). Evaluating pesticide degradation in the environment: Blind spots and emerging opportunities. Science 341, 752-758. German Beekeeping Organization (2015). Bienen als Bestäuber. http://www.deutscherimkerbund.de/162- Bienen_und_Bestaeubungsleistung. Gliessman, S.R. (2014). Agroecology: The ecology of sustainable food systems. 3rd Edition, CRC Press. Hutzinger, O. (2013). The handbook of environmental chemistry – anthropogenic compounds, 3rd, ed., Springer, Berlin Heidelberg. IARC (2015). IARC Monographs on the evaluation of carcinogenic risks to humans. Some organophosphate insecticides and herbicides: Diazinon, Glyphosate, Malathion, Parathion, and Tetrachlorvinphos. Glyphosate. IARC monograph 112. Knutson, R.D. (1999). Economic impacts of reduced pesticide use in the United States: Measurement of costs and benefits. AFPC Policy Issues Paper 99-2.

Chapter 2 Towards a better pesticide policy for the European Union 76

Leisinger, K.M. (2008). Weltbevölkerungswachstum und Vernichtung fruchtbarer Böden. Das Online-Handbuch Demografie des Berlin-Instituts, 1-7. Lewis, S.L. and Maslin, M.A. (2015). Defining the Anthropocene. Nature, 519, 171-180. Looser, R., Froescheis, O., Cailliet, G.M., Jarman, W.M. and Ballschmiter, K. (2000). The deep-sea as a final global sink of semivolatile persistent organic pollutants? Part II: Organochlorine pesticides in surface and deep-sea dwelling fish of the north and south Atlantic and the Monterey Bay Canyon (California). Chemosphere, 40 (6), 661-670. Malaja, E., von der Ohea, P.C., Groted, M., Kühnee, R., Mondyf, C.P., Usseglio-Polaterag, P., Bracka, W. and Schäfer, R.B. (2013). Organic chemicals jeopardize the health of freshwater ecosystems on the continental scale. Proceedings of the National Academy of Sciences 111 (26), 9549-9554. Multigner, L., Ndong, J.R., Giusti, A., Romana, M., Delacroix-Maillard, H., Cordier, S., Jegou, B., Thome, J.P. and Blanchet, P. (2010). Chlordecone exposure and risk of prostate cancer. Journal of Clinical Oncology 28 (21), 3457-3462. OECD-Guideline 307 (2002). OECD guideline for the testing of chemicals. Aerobic and anaerobic transformation in soil. O'Reilly, E. (2016). Decision in case 12/2013/MDC on the practices of the European Commission regarding the authorisation and placing on the market of plant protection products (pesticides). Case: 12/2013/MDC. http://www.ombudsman.europa.eu/en/cases/decision.faces/en/64069/html.bookmark. PAN (2011). Meet (chemical) agriculture. The world of backdoors, derogations, sneaky pathways and loopholes. Part 1: the 120-days derogation. http://www.pan-europe.info/old/Resources/Reports. PAN (2015). Consolidated list of banned pesticide: pesticide action network releases list of highly hazardous pesticides banned in countries around the world. http://pan-international.org/release/consolidated-list-of-banned-pesticides-pesticide- action-network-releases-list-of-highly-hazardous-pesticides-banned-in-countries-around-the-world/. Rathore, H.S. and Nollet, L.M.L. (2012). Pesticides – evaluation of environmental pollution. CRC Press Taylor & Francis Group. Rochkind, M.L., Blackburn, J.W. and Sayler, G.S. (1986). Decomposition of chlorinated aromatic compounds. US Environmental Protection Egency, Cincinatti, 115-128. Schmidt, C.K. and Brauch, H.-J. (2008). N,N-Dimethylsulfamide as precursor for N-nitrosodimethylamine (NDMA) formation upon ozonation and its fate during drinking water treatment. Environmental Science and Technology 42, 6340-6346. Servien, R., Mamy, L., Li, Z., Rossard, V., Latrille, E., Bessac, F., Patureau, D. and Benoit, P. (2014). TyPol - a new methodology for organic compounds clustering based on their molecular characteristics and environmental behavior. Chemosphere 111, 613-622. Sinclair, C., Van Beinum, W., Adams, C., Bevan, R., Levy, L., Parsons, S., Goslan, E. and Baumann, G. (2010). A desk study on pesticide metabolites, degradation and reaction products to inform the inspectorate’s position on monitoring requirements. Final report for drinking water inspectorate. The Food and Environment Research Agency. Smalling, K.L., Fellers, G.M., Kleeman, P.M. and Kuivila, K.M. (2013). Accumulation of pesticides in pacific chorus frogs (Pseudacris regilla) from California's Sierra Nevada Mountains, USA. Environmental Toxicology and Chemistry, 32 (9), 2026-2034. Storck, V., Lucini, L., Mamy, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., Karas, P.A., Servien, R., Karpouzas, D.G., Trevisan, M., Benoit, P. and Martin-Laurent, F. (2016). Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology. Environmental Pollution 208, 537-545. Summary report from the May 2016 Joint FAO/WHO Meeting on Pesticide Residues, 2016. Tageswoche Basel (2015). Anti-Expo-Kongress - Syngenta-Kritiker Gary Hooser: „Man log uns offen ins Gesicht”. The Chemical and Veterinary Investigation Office Stuttgart (2013a). Pflanzenschutzmittelrückstände in Frischobst 2012. The Chemical and Veterinary Investigation Office Stuttgart (2013b). Pflanzenschutzmittelrückstände in Frischgemüse 2012 – Zusammenfassung der Rückstandsbefunde in Erzeugnissen aus konventionellem Anbau. Tixier, C., Bogaerts, P., Sancelme, M., Bonnemoy, F., Twagilimana, L., Cuer, A., Bohatier, J. and Veschambre, H. (2000). Fungal biodegradation of a phenylurea herbicide, diuron: structure and toxicity of metabolites. Pest Management Science 56, 455- 462. U.S. EPA (2011). Office of Chemical Safety and Pollution Prevention (7503P) EPA 733-R-11-001. Pesticides industry sales and usage - 2006 and 2007 market estimates. UN (2015). World Population Prospects: The 2015 Revision, Key Findings and Advance Tables. UN Department of Economic and Social Affairs, Population Division Working Paper No. ESA/P/WP.241. Van-Camp, L., Bujarrabal, B., Gentile, A.-R., Jones, R.J.A., Montanarella, L., Olazabal, C. and Selvaradjou, S.-K. (2004). Reports of the technical working groups established under the thematic strategy for soil protection. EUR 21319 EN/4, 872 pp. Office for Official Publications of the European Communities, Luxembourg. Zalasiewicz, J., Williams, M., Smith, A., Barry, T.L., Coe, A.L., Brown, P.R., Brenchley, P., Cantrill, D., Gale, A., Gibbard, P., Gregory, F.J., Hounslow, M.W., Kerr, A.C., Pearson, P., Knox, R., Powell, J., Waters, C., Marshall, J., Oates, M., Rawson, P. and Stone, P. (2008). Are we now living in the Anthropocene? Geological Society of America (GSA) Today 18 (2), 4-8. Zhang, W., Jiang, F. and Ou, J. (2011). Global pesticide consumption and pollution: with China as a focus. Proceedings of the International Academy of Ecology and Environmental Sciences 1(2), 125-144.

Chapter 2 Towards a better pesticide policy for the European Union 77

Conclusion

Chapter 2 provides an introduction into the pesticide paradox, as it describes the difficulty of responsible authorities to content in equal parts the high demand of food and the concerns about negative pesticide effects in the environment. Moreover, the chapter proposes a simplified and intelligible view of the myriad steps involved in the pesticide authorization process. The profound understanding of this process is a real challenge (for a non-expert), as publicly available information is hardly deducible and needs to be complemented by many other sources of information. This was also perceptible from the reviewers’ comments on this opinion paper, as they provided divergent views on the details of this process. Once, I had managed to unlock the nebulosity of the pesticide authorization process, I was able to define its weaknesses potentially leading to pesticide-related environmental issues. Moreover, I suggested ways of improvements of the overall process. To increase the potential that this opinion paper may contribute to the discussion on a better pesticide policy in the EU (or at least that it may contribute to a broader public knowledge about hidden pesticide effects in the environment), one perspective is the vulgarization of this opinion paper in outreach activities of the European Marie Curie project ‘Love-to-hate’. As a first step, some contents will be presented at the next scientific EFSA meeting by Dr. Fabrice Martin-Laurent (‘Environmental risk assessment of pesticides: 25 years of scientific advancements since the adoption of Directive 91/414/EEC’, 15th-16th November 2016, Parma, Italy). Other planned outreach activities are a publication of some parts of this opinion paper in daily newspapers of France, Germany and Greece, and a presentation at the 7th Society of Environmental Toxicology and Chemistry (SETAC) World Congress/SETAC North America 37th Annual Meeting (6th-10th November 2016, Orlando, USA).

Chapter 2 Towards a better pesticide policy for the European Union 80

Chapter 3

Pesticide dissipation in agricultural soil

Veronika Storck Luigi Lucini Laure Mamy Federico Ferrari Evangelia S. Papadopoulou Sofia Nikolaki Panagiotis A. Karas Remi Servien George Tsiamis Dimitrios G. Karpouzas Marco Trevisan Pierre Benoit Fabrice Martin-Laurent

Foreword

Knowledge on the environmental fate of pesticides is essential in order to characterize the scenario of exposure of organisms to pesticide residues. Indeed, studying the exposure scenario of pesticides in soil is a prerequisite to assess the ecotoxicological effects of pesticides on soil microorganisms. Therefore, the main processes involved in pesticide dissipation (i.e. degradation, sorption and transformation to TPs) were studied (Chapter 3). Chapter 3.1 elaborates the fate of each of the three studied pesticides CHL, IPU or TCZ (applied at three different concentrations) and their main TPs in a lab-to-field experiment. Furthermore, their soil sorption affinities were measured in a batch study. The physicochemical and environmental properties of pesticide TPs (formed by abiotic and biotic processes in the environment) can greatly differ from those of the original compound. Although most TPs have a lower ecotoxicity than the parent molecule, some of them are more soluble, so that they possess higher tendencies to migrate to water bodies and to be more bioavailable, resulting in potentially higher toxicities. Knowledge of the existence of TPs relevant for the environment is one of the biggest challenges to fully characterize the scenario of exposure of organisms to pesticide residues. With the aim to face this challenge, I investigated the interest of combining suspect screening to in silico molecular typology to track formed TCZ TPs in soil (Chapter 3.2). Past experiences have shown that TPs of relevant ecotoxicity typically emerged only 20 to 30 years after the market introduction of certain pesticides. Thus, the development of comprehensive suspect screening methods to search for both known and yet unknown TPs may be a step towards a broader knowledge of the existence of pesticide TPs. Moreover, even if the existence of TPs is already known, EU- coordinated environmental risk assessment only considers TPs classified as ‘relevant’ (i.e. representing at least 10 % of the initially applied amount of an active substance). We proposed to use in silico molecular typology to categorize TPs according to their physicochemical properties. From this classification, their main environmental properties and consequently their potential environmental risk can be deduced towards a more precautionary classification of a TP to be ‘relevant’.

Chapter 3 Pesticide dissipation in agricultural soil 84

Chapter 3.1

Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Evangelia S. Papadopoulou Panagiotis A. Karas Sofia Nikolaki Veronika Storck Federico Ferrari Marco Trevisan George Tsiamis Fabrice Martin-Laurent Dimitrios G. Karpouzas

Published in Science of the Total Environment 569-570, 86-96 (2016).

Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Abstract

Assessment of dissipation constitutes an integral part of pesticides risk assessment since it provides an estimate of the level and the duration of exposure of the terrestrial ecosystem to pesticides. Within the frame of an overall assessment of the soil microbial toxicity of pesticides, we investigated the dissipation of a range of dose rates of three model pesticides, isoproturon (IPU), tebuconazole (TCZ), and chlorpyrifos (CHL), and the formation and dissipation of their main transformation products following a tiered lab-to-field approach. The adsorption of pesticides and their transformation products was also determined. IPU was the least persistent pesticide showing a dose-dependent increase in its persistence in both laboratory and field studies. CHL dissipation showed a dose-dependent increase under laboratory conditions and an exact opposite trend in the field. TCZ was the most persistent pesticide under lab conditions showing a dose-dependent decrease in its dissipation, whereas in the field TCZ exhibited a biphasic dissipation pattern with extrapolated DT90s ranging from 198 to 603 days in the x1 and x2 dose rates, respectively. IPU was demethylated to mono- (MD-IPU) and di-desmethyl- isoproturon (DD-IPU) which dissipated following a similar pattern with the parent compound. CHL was hydrolyzed to 3,5,6-trichloro-2-pyridinol (TCP) which dissipated showing a reverse dose-dependent pattern compared to CHL. Pesticides adsorption affinity increased in the order IPU

Figure 3.1.1. Graphical abstract of Chapter 3.1.

Chapter 3 Pesticide dissipation in agricultural soil 88

Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Introduction

Pesticides still constitute an integral part of modern agriculture despite the environmental and health risks associated with their use (Muñoz-Leoz et al., 2011). Upon their application, either intentionally or unintentionally, they reach the soil environment which acts as a sink for their further distribution to other environmental compartments depending on their soil dissipation rates (Arias-Estévez et al., 2008). Thus, good knowledge of the dissipation and metabolism of pesticides in soil is an essential part of risk assessment since it determines the scenario of exposure of the soil ecosystem. The level and the duration of exposure has to be known in order to assess the toxicity of pesticides to non-target soil organisms including microorganisms, which have been identified by the European Food Safety Agency as one of the specific protection goals (EFSA, 2010). Recently, Martin-Laurent et al. (2013), proposed a revision of the regulatory framework regarding the assessment of the soil microbial toxicity of pesticides. This involves the implementation of a tiered lab-to-field experimental approach where the dissipation and transformation of the studied pesticides (measure of exposure) coupled with standardized advanced biochemical and molecular methods (measure of toxicity) could provide a more robust estimation of the soil microbial toxicity of pesticides (Karpouzas et al., 2014). Within this framework, we studied the dissipation and the transformation of a range of dose rates of three model pesticides, isoproturon (IPU), tebuconazole (TCZ), and chlorpyrifos (CHL) which were chosen based on their widespread use in Europe. IPU [3-(4-isopro- pylphenyl)-1,1-dimethylurea] is a phenylurea herbicide used for the control of annual grasses and broad-leaved weeds in spring and winter cereals

(Collings et al., 2003). It shows variable persistence in soil with DT50 values varying from 3 to 200 days (Alletto et al., 2006). Its extensive use has resulted in its common detection in surface and groundwater resources (Skark and Zullei-Seibert, 1995; Skarket al., 2004). Mono-desmethyl-isoproturon [3-(4-isopropylphenyl)-1-methylurea (MD- IPU), di-desmethyl-isoproturon [3-(4-phenyl)-urea] (DD-IPU), 4-isopropyl-aniline (4-IA) and several hydroxylated compounds (Lehr et al., 1996; EFSA, 2015) have been identified as the most common transformation products of IPU in soil. Some of these have been found to exert higher toxicity than the parent compound (Hussain et al., 2015;

Chapter 3 Pesticide dissipation in agricultural soil 90 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Alletto et al., 2006; Tixier et al., 2002). However, little is known regarding their dissipation kinetics and adsorption in soil. TCZ [(RS)-1-p-chlorophenyl-4,4-dimethyl-3-(1H-1,2,4-triazol-1-yl methyl) pentan-3-ol] is a systemic triazole fungicide which is used for the control of a range of plant fungal pathogens in different crops including winter cereals (D’Angelo et al., 2014;

Keinath, 2015). It is rather persistent in the soil environment with DT50s ranging from 49 to 610 days (Strickland et al., 2004; EFSA, 2014). Monitoring studies in agricultural areas where TCZ is used have verified its frequent presence in surface and groundwater systems (Herrero-Hernandez et al., 2013; Sanchez-Gonzalez et al., 2013). Although its dissipation and transformation in soil has been studied previously (Álvarez-Martín et al., 2016; Herrero-Hernandez et al., 2013; Potter et al., 2005; Strickland et al., 2004), little is known regarding the fate of its transformation products in the environment (Storck et al., 2016). CHL [O,O-diethyl O-3,5,6-trichloro-2-pyridinyl phosphorothioate] is one of the most extensively used organophosphate insecticides with a broad spectrum of activity (Joseph and Zarate, 2015). Its degradation in soil proceeds mainly via hydrolysis, abiotic and biotic, with DT50 values ranging from 10 to 120 days (Racke, 1993). Hydrolysis of CHL leads to the formation of 3,5,6-trichloro-2-pyridinol (TCP) which is known to have adverse effects on soil microbial activity and on the degradation of the parent compound (Racke et al., 1990). However, limited knowledge is available regarding the dissipation and adsorption of TCP in agricultural soils where CHL has been applied. The main objectives of the present study were (i) to determine the dissipation of the three model-pesticides applied to soil at different dose rates under laboratory (x1, x2 and x10 the recommended dose rate) and field conditions (x1, x2 and x5 the recommended dose rate), (ii) to follow the dynamics of formation and dissipation of the main transformation products of the model pesticides and (iii) to determine the soil adsorption affinity of the model pesticides and their main transformation products. This will provide a thorough view of the soil persistence of the studied pesticides and their transformation products which defines the scenario of exposure of non-target soil organisms to be taken into account to estimate potential toxicity risks in follow up studies.

Chapter 3 Pesticide dissipation in agricultural soil 91 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Material and methods

Pesticides

Analytical standards of IPU (99.9 %), MD-IPU (99.5 %), DD-IPU (99.0 %), TCZ (98.8 %), CHL (97 %), and TCP (99.0 %) were purchased by Dr. Ehrenstorfer, (Germany). The analytical standard of 4-IA (99.0 %) was purchased by Sigma-Aldrich (Germany). Commercial formulations of IPU (QUINTIL® 500SC), TCZ (FOLICUR 25SE) and CHL (CARPOSAN 48EC), were provided by Phytorus, Bayer CropScience and ISAGRO, respectively.

Microcosm experiment

The used soil was collected in July 2013 from an agricultural field situated in North Italy (area of Mortizza, 45°05'20.8"N 9°45'59.4"E (Google Maps), Supplementary Figure 3.1.1), which was also used for the execution of the field experiment described below. The physicochemical characteristics of the soil are shown in Supplementary Table 3.1.1. The field site did not have a recent history of treatment with IPU, TCZ, and CHL. Topsoil samples (0-10 cm depth) were collected from the field site following the W non-systematic pattern of sampling, according to the ISO 10381-1 and -2 guidelines (2002), and mixed thoroughly to provide a single bulk soil sample. The soil was then partially air-dried, sieved to pass through a 2 mm mesh sieve and divided into ten subsamples (6 kg each). For each pesticide, three subsamples were treated with appropriate amounts of aqueous solutions of IPU, TCZ and CHL (prepared from their commercial formulations) aiming an application of x1, x2 and x10 the recommended dose rates (Supplementary Table 3.1.2). The final soil subsample received the same amount of water without pesticide to serve as non-treated control. After pesticide application, the soils were left to equilibrate for one hour and water was added to adjust moisture to 40 % of the water holding capacity. Soil samples were separated into 150 g subsamples which were placed in aerated plastic bags and incubated in the dark at 20 °C. Immediately after pesticide application and 3, 7, 14, 21, 35, 56, 70, 100 and 125 days after pesticide application, triplicate samples from each treatment were removed from the incubator and stored at -20 °C for pesticide analysis.

Chapter 3 Pesticide dissipation in agricultural soil 92 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Field experiment

A field experiment was conducted on the same field site as described above (Supplementary Figure 3.1.1). The field site had been cultivated with winter cereals for the last five years. A randomized complete block design was established with four replicate plots (4 m × 15 m) for each combination of pesticide x dose rate. A 2 m wide buffer zone between plots was maintained to minimize possible cross-contamination between treatments. On the 7thNovember 2013, the field was seeded with a mixture of cereals (60% Hordeum vulgare L., 25% Triticum spp., and 15% Triticosecale sp.) and a weather station was installed in the field site to collect daily weather data (rain precipitation (mm), mean daily air and soil temperature (°C), solar radiation (W/m2), relative humidity and wind speed (m/sec)). On the 12th November 2013, the pesticides IPU, TCZ and CHL were applied to the established plots at three dose rates (x1, x2 and x5 the recommended dose). The highest dose rate was selected in compliance with the maximum pesticide application rate allowed to be used for experimental purposes in Italy. Pesticides were applied with a backpack sprayer at a spaying rate of 250 L/ha (exact dose rates are given in Supplementary Table 3.1.2). Four plots were not treated with pesticides to serve as untreated controls. Immediately after treatment and 3, 7, 14, 21, 35, 56, 70, 100 and 125 days after treatment, soil samples were collected from each plot to assess pesticide dissipation. Nine random samples collected from the top 10 cm of each plot were homogenized providing a composite sample per plot. Each sampling point was marked with a wooden stick to avoid sampling from the same point in the following sampling campaigns. All samples were stored at -20 °C until further analysis.

Soil adsorption of pesticides and their transformation products

The adsorption of the tested substances was determined using the standard batch equilibrium method according to the OECD-Guideline 106 (2000). Preliminary kinetic studies were employed to determine the most appropriate soil:solution ratios and equilibration times for all chemicals. Thus, the most appropriate soil:solution ratios to achieve 20 to 80 % adsorption of the studied chemicals were 1:10 for IPU, MD-IPU, and DD-IPU, 1:50 for 4-IA, 1:25 for TCZ, 1:200 for CHL, and 1:5 for TCP. Equilibrium was reached within 24 hours for all the tested substances except for CHL and TCP, for which

Chapter 3 Pesticide dissipation in agricultural soil 93 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions equilibrium was reached at 12 hours. Stock solutions of each substance in acetone (1 g/L) or methanol (4 g/L) (only for CHL due to its low water solubility) were prepared using analytical standards. Appropriate amounts of the stock solutions were dissolved in

0.01 M CaCl2 leading to the preparation of six solutions with pesticide concentrations ranging from 5 to 50 mg/L for IPU, 2 to 20 mg/L for TCZ, and 3 to 40 mg/L for CHL. Regarding pesticide transformation products, a series of solutions with concentrations ranging from 1 to 10 mg/L were used. The range of concentrations used for each compound in the adsorption study was selected assuming (i) application of the recommended dose rate for each compound (Supplementary Table 3.1.2), (ii) soil bulk density of 1.3 g/mL and (iii) soil moisture content of 20 %. Regarding transformation products, the concentrations used were based on their amounts formed in the laboratory and field dissipation studies. In all cases, the content of the organic solvent in the final solution phase did not exceed 0.1 %. Triplicate soil samples (1 to 10 g) were mixed with 50 or 200 mL of each of the above solutions in screw-cupped vials and they were shaken overnight on an orbital shaker (200 rpm) at room temperature. When equilibrium was reached, samples were centrifuged at 2500 rpm for five minutes and the supernatant was collected, extracted and analyzed by HPLC with a photodiode array (HPLC-PDA) detector as described below. The adsorption of the model pesticides was also determined using commercial formulations instead of analytical standards, following the same procedure with the only difference that pesticides were directly dissolved in 0.01 M CaCl2 without the need for organic solvent addition.

Pesticide residue analysis

Extraction of pesticides and their transformation products from soil All studied compounds, except 4-IA, were extracted from soil with the same procedure. Thus, sealed glass bottles (250 mL) containing 40 g of soil, 2 mL of 2 M ammonium acetate and 50 mL of acetone were placed on a bed-shaker operated at 210 rev/min for 30 min. The extract was passed through a glass-fiber filter (Whatman GF/F) on a Buchner funnel under vacuum. The soil was re-extracted with 50 mL of acetone and the two extracts were pooled and mixed with 200 mL of a 4 % solution of anhydrous sodium sulfate and 100 mL dichloromethane. The mixture was shaken for 10 minutes at 190 rev/min and then allowed the phases to separate: an aqueous upper

Chapter 3 Pesticide dissipation in agricultural soil 94 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions phase and a dichloromethane lower phase. The organic phase was collected while the aqueous phase was re-extracted twice with 50 mL of dichloromethane. The organic phase from all extraction steps was pooled and passed through a glass column containing 1 g of glass wool and 40 g of anhydrous sodium sulfate. The filtrates were collected in a round bottom flask and evaporated to dryness on a rotary evaporator at 40 °C. Pesticide residues were re-dissolved in 10 mL dichloromethane and brought to complete dryness under a stream of nitrogen gas. The final pesticide residues were re- dissolved in 1 mL of methanol: H2O + 0.01 % H3PO4 (70:30, v/v) and analyzed by HPLC- PDA. 4-IA was extracted from soil samples according to the protocol proposed by Fenoll et al. (2012) with slight modifications. Briefly, 10 g of soil were extracted with 10 mL of acetonitrile:water (1:1, v/v) by 30 min sonication followed by a salting-out step of 2 g NaCl. The tube was then vortexed for one minute and centrifuged for 10 minutes at 4500 rpm. An aliquot of the supernatant was analyzed by HPLC-PDA.

HPLC-PDA analyses Analyses were performed in an ΗPLC 1100HP system, equipped with a UV/VIS PDA detector. A Gemini C18 (4 x 2.0mm ID) (SecurityGuard Cartridges) pre-column, connected to a Gemini 3 μm-C18 column 110A (150 mm x 2mm i.d.) (Phenomenex, Cheshire, UK) was used for separation of the studied pesticides. The injection volume was 20 μL, the flow rate of the mobile phase was 0.3 mL/min and the column temperature was set at 25 °C. Separation of compounds was achieved following a gradient elution program of a mobile phase composed of (A) acetonitrile and (B) H2O acidified with ortho-phosphoric acid (0.01% by volume). At the time of injection, the solvent composition was 30 % A and 70 % B. This was maintained for 5 min increased linearly to 80 % A and 20 % B until 15 min where it was maintained until 32 min. The mobile phase was then reverted to its initial composition (30 % A and 70 % B) until 35 min, followed by a further stabilization period of 3 min. Detection was achieved at 240 nm for IPU, MD-IPU, and DD-IPU, at 220 nm for TCZ, and at 230 nm for CHL and TCP. Pesticides were quantified by the external standard method using calibration curves obtained by the injection of matrix-matched standard solutions. Matrix-matched standard solutions were prepared by diluting the methanolic working solutions of each of the studied compounds in blank soil extract.

Chapter 3 Pesticide dissipation in agricultural soil 95 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Analysis of 4-IA was achieved under isocratic elution conditions with a mobile phase composed of a 30:70 (v:v) mixture of (A) acetonitrile + methanol (1:1 by volume) acidified with acetic acid (0.01 vol%) and (B) 5 mM KH2PO4 acidified with acetic acid (0.01 vol%) (Jühler et al., 2001). Total analysis time was 30 min. The injection volume was 20 μL, the flow rate was 0.3 mL/min and the column temperature was set at 45 °C. Detection of 4-IA was achieved at 240 nm.

Analytical method validation Analysis of soil samples fortified at three levels (0.05, 1 and 10 mg/kg for pesticides, and 0.1, 0.5, 2.5 mg/kg for their transformation products) was employed to assess the efficiency of the extraction methods described above. Triplicate samples for each compound and concentration level were processed. The mean percentage recoveries for IPU, TCZ, and CHL were 86.7 %, 86.4 %, and 88.1 %, respectively (CV ≤ 15.3 %), while the recoveries for MD-IPU, DD-IPU, 4-IA, and TCP were 77.6 %, 70.1 %, 101.9 %, and 82.2 %, respectively (CV ≤ 15.3 %). The limit of detection was 0.5 μg/kg for IPU, TCZ, and DD-IPU, 0.75 μg/kg for MD-IPU, CHL, and TCP, and 10 μg/kg for 4-IA. The limit of quantification was 1.25 μg/kg for IPU, TCZ, and DD-IPU; 2.5 μg/kg for MD-IPU, CHL, and TCP, and 25 μg/kg for 4-IA.

Calculation of dissipation kinetic parameters

The four kinetic models proposed by the FOCUS working group on pesticide degradation kinetics (FOCUS, 2006) were used to calculate pesticide dissipation kinetic parameters: the single first order kinetic model (SFO), and the biphasic models hockey stick (HS), first order multi-compartment model (FOMC) and double first order in parallel model (DFOP). The goodness of fit was assessed using the χ2 test as well as visual inspection and the distribution of the residuals. In general, the biphasic kinetic models were used only in cases where the linear model (SFO) failed to acceptably describe (χ2 > 15 %) pesticides dissipation. Regarding the dissipation kinetics of the transformation products, the guidelines of the FOCUS working group on pesticides degradation kinetics were followed (FOCUS, 2006). Briefly the DT50s of the transformation products produced either directly from the parent compound (i.e. MD- IPU, TCP) or through the transformation of a preceding metabolite (DD-IPU), were

Chapter 3 Pesticide dissipation in agricultural soil 96 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions calculated using the following assumptions: (i) there was a flow of the parent compound to a sink (unidentified metabolites or bound residues), (ii) the dissipation of the parent compound and the transformation products followed first-order kinetics, and (iii) the initial concentration for all compounds was the one measured at time 0. The endpoints needed to be calculated were the initial amount (Pini), and the transformation rate constant of the parent compound (kP), the formation fraction (ff) and transformation rate constant for each transformation product (kM). Finally, the DT50 values for the transformation products were calculated by the following equation:

DT50 (transformation product) = ln2 / kM. Parameters of the kinetic models and their standard errors were obtained by least square non-linear regression analysis using the statistical program R and the mkin package (version 09.40, 2015).

Results

Laboratory study

Dissipation and transformation of IPU The dissipation patterns of IPU in all three dose rates were well described (χ2 < 15 %) by the SFO kinetic model (Figure 3.1.2a, Supplementary Table 3.1.3) with estimated DT50 values increasing with the dose rate: 16.5, 18.2 and 25.7 days for the x1, x2 and x10 dose rates, respectively (Table 3.1.1). IPU dissipation proceeded via sequential demethylation to MD-IPU (Figure 3.1.3a), which constituted the major transformation product of IPU, and to DD-IPU, which was formed in lower amounts (Figure 3.1.3b). No residues of 4-IA were detected throughout the study. The formation of the two transformation products of IPU peaked between 14 to 21 days for all the different doses applied to the soil and dissipated thereafter. Their dissipation followed

SFO kinetics with estimated DT50 values of 13.0, 14.5 and 15.0 days for MD-IPU, and 13.8, 12.0, and 22.5 days for DD-IPU in the samples treated with the x1, x2 and x10 dose rates, respectively (Table 3.1.1).

Chapter 3 Pesticide dissipation in agricultural soil 97 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Dissipation of TCZ TCZ dissipation showed a biphasic pattern (Figure 3.1.2b) and it was best described by the FOMC model (Supplementary Table 3.1.3). The calculated DT50 values of TCZ showed an increasing trend with increasing dose rates: 60.7, 74.8 and 97.1 days in the soils treated with x1, x2 and x10 dose rates, respectively (Table 3.1.1).

Dissipation and transformation of CHL The dissipation of CHL showed a biphasic pattern (Figure 3.1.2c). The FOMC model provided the best fit to the dissipation data (Supplementary Table 3.1.3) with

DT50 values showing a decreasing trend with increasing dose rates; 52.6, 43.9 and 28.6 days in the x1, x2 and x10 dose rates (Table 3.1.1), respectively. CHL was transformed via hydrolysis to TCP whose formation reached a maximum between 56 to 100 days after application and dissipated thereafter (Figure 3.1.3c). Its dissipation was best described by the SFO model with DT50 values of 2.9, 2.2, and 17.7 days for the x1, x2 and x10 dose rates, respectively (Table 3.1.1).

Figure 3.1.2. The laboratory dissipation of (a) IPU, (b) TCZ and (c) CHL in soil samples treated with x1, x2, or x10 the recommended dose of each of these compounds. Each value is the mean of three replicates ± standard deviation.

Chapter 3 Pesticide dissipation in agricultural soil 98 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Table 3.1.1. The dissipation parameters of IPU, TCZ, CHL and their main transformation products in soil in the laboratory experiment. Dissipation kinetic parameters were calculated with the single first order (SFO) kinetic model or with the biphasic first order multi compartment (FMOC) model.

Figure 3.1.3. The formation and dissipation patterns of the transformation products of IPU ((a) MD-IPU and (b) DD-IPU) and of CHL((c) TCP) in the soil samples treated with x1, x2, or x10 the recommended dose of the pesticide in the laboratory. Each value is the mean of three replicates ± the standard deviation.

Chapter 3 Pesticide dissipation in agricultural soil 99 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Field study

Weather data were collected during the field study (Supplementary Figure 3.1.2). Eight intense precipitation events with daily mean precipitation > 10 mm and seven with mean daily precipitation > 5 mm occurred during the field study. The most important precipitation events occurred in the following periods: 3-11, 43-53, and 73- 107 days after treatment. Temperature levels in soil and air ranged from -1 to 11 °C. The lowest temperatures were observed from 15 to 36 days (mean daily temperatures < 4 °C), whereas an increase in temperatures was observed from 94 days onwards (mean daily temperature > 6 °C). Solar radiation showed large fluctuations with no consistent pattern. Relative humidity and wind speed ranged from 60 to 100 % and 0.1 to over 2.5 m/sec, respectively.

Dissipation and transformation of IPU In contrast to the laboratory microcosm experiment, IPU dissipation in the field showed a biphasic pattern characterized by an initial rapid dissipation phase lasting for two weeks and followed by a slow dissipation phase until the end of the experiment (Figure 3.1.4a). The HS model showed the best fit to the dissipation data

(Supplementary Table 3.1.4) with calculated DT50s of 7.4, 10.4 and 12.8 days for the x1, x2, and x5 dose rates, respectively (Table 3.1.2). Similarly to the laboratory experiment, MD-IPU was the major transformation product while low amounts of DD-IPU were also detected (Figure 3.1.5a and 3.1.4b). MD-IPU concentrations peaked between 14 to 35 days compared to DD-IPU whose concentrations in soil peaked at 35 days. The transformation products of IPU did not persist and dissipated with SFO-obtained DT50s of 9.1, 27.1 and 39.4 days for MD-IPU, and 6.9, 14.1, and 18.9 days for DD-IPU in the plots treated with the x1, x2 and x5 dose rates, respectively.

Dissipation of TCZ The field dissipation of TCZ showed a clear biphasic pattern with a drastic decline in its concentrations during the first 3 days, when the first precipitation event of 6.4 mm occurred, followed by a slow dissipation phase until the end of the experiment (Figure 3.1.4b). This dissipation pattern of TCZ was best described by the HS model

(Supplementary Table 3.1.4) with DT50 values of 1.5, 2.3 and 2.5 days for the x1, x2 and

Chapter 3 Pesticide dissipation in agricultural soil 100 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions x5 dose rates, respectively (Table 3.1.2). Extrapolated DT90 values of 198, 603.4, 186.7 days for the x1, x2 and x5 dose rates, respectively, showed a much longer persistence of TCZ (Table 3.1.2).

Dissipation and transformation of CHL As observed for the two other pesticides, the dissipation of CHL in the field experiment was biphasic and it was best described by the HS model (Supplementary

Table 3.1.4) with DT50s of 6.7, 33.9 and 119.6 days in the plots treated with the x1, x2 and x5 dose rates, respectively (Figure 3.1.4c, Table 3.1.2). In accordance with the results of the laboratory microcosm experiment, TCP was the only detected transformation product of CHL (Figure 3.1.5c). The dissipation of TCP for all dose rates was best described by the SFO model with estimated DT50 values of 52.8, 49.1, and 17.4 days for the x1, x2 and x5 dose rates, respectively (Table 3.1.2).

Figure 3.1.4. The field dissipation of (a) IPU, (b) TCZ and (c) CHL in soil samples treated with x1, x2, or x5 the recommended dose of each of these compounds. Each value is the mean of three replicates ± standard deviation.

Chapter 3 Pesticide dissipation in agricultural soil 101 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Table 3.1.2. The dissipation parameters of IPU, TCZ, CHL and their main transformation products in soil in the field experiment. Dissipation kinetic parameters were calculated with the single first order (SFO) kinetic model or with the biphasic hockey stick (HS) model.

Figure 3.1.5. The formation and dissipation patterns of the transformation products of IPU ((a) MD-IPU and (b) DD-IPU) and of CHL((c) TCP) in the soil samples treated with x1, x2, or x5 the recommended dose of the pesticide in the field. Each value is the mean of three replicates ± the standard deviation.

Chapter 3 Pesticide dissipation in agricultural soil 102 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Soil adsorption of pesticides and their transformation products

The adsorption of pesticides and their transformation products were fitted to the Freundlich equation (R2 > 0.96) which was used for the calculation of the adsorption coefficients Kf and Kfoc (Figure 3.1.6, Table 3.1.3). IPU, MD-IPU and DD-IPU showed a weak adsorption affinity in the tested soil with Kf values of 4.0, 4.43 and 4.17 mg1-N LN kg-1, respectively. On the contrary, 4-IA showed a higher adsorption affinity with a Kf value of 34.9 mg1-N LN kg-1. IPU adsorption affinity increased when its commercial formulation was used (Kf = 4.99 mg1-N LN kg-1). TCZ was moderately adsorbed to soil components with Kf values of 21.9 and 30.1 mg1-N LN kg-1, for the active substance and the commercial formulation, respectively. CHL showed the highest adsorption affinity with Kf values of 195.2 and 211.4 mg1-N LN kg-1, for the active substance and the commercial formulation, respectively. In contrast, the hydrolysis product of CHL, TCP, showed the weakest adsorption affinity among all the tested substances (Kf = 1.55 mg1-N LN kg-1). According to the classification system of isotherms introduced by Giles et al. (1960), MD-IPU, DD-IPU, and TCP gave a C-type isotherm compared to the rest of the compounds studied which gave L-shaped isotherms.

Figure 3.1.6. Adsorption isotherms of (a) IPU, (b) TCZ and (c) CHL, prepared with the use of the active substance or the commercial formulation, and (d) their main transformation products MD-IPU, DD-IPU, 4- IA and TCP.

Chapter 3 Pesticide dissipation in agricultural soil 103 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Table 3.1.3. The adsorption coefficients of IPU, TCZ, CHL, determined either with the use of the active substance or the commercial formulation, and of their main transformation products MD-IPU, DD-IPU, 4- IA, and TCP.

Discussion

Estimation of the dissipation of pesticides in soil constitutes an integral part of pesticide environmental risk assessment required for the authorization and placement on the market of plant protection products. We investigated the dissipation and the transformation of three model pesticides following a lab-to-field experimental approach as a proxy of the scenario of soil exposure to pesticides and their main transformation products.

Dissipation and transformation of IPU

IPU showed the lowest persistence from the pesticides tested with DT50s which were within the range reported in the literature (6.5 to 40 days) (Walker et al., 2001;

Chapter 3 Pesticide dissipation in agricultural soil 104 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Rodríguez-Cruz et al., 2006). A trend towards reduced DT50 values of IPU in the field compared to the corresponding DT50 values in the laboratory was observed. This is in line with previous studies for other pesticides (Laabs et al., 2000; Dolaptsoglou et al., 2009) and it is not surprising considering that in laboratory studies the contribution of other dissipation processes like volatilization, leaching or photolysis are impeded (volatilization) or eliminated (leaching and photolysis) compared to field studies where these processes may significantly contribute to pesticide loss (EEC, 2000). IPU dissipation in soil proceeded via sequential demethylation to MD-IPU which was the main transformation product, and to DD-IPU which was a minor transformation product. This is in agreement with previous studies which showed that early stages of IPU transformation by soil microorganisms proceeds via successive demethylations (Sorensen et al., 2001), followed by cleavage of the urea side chain resulting in the transitory accumulation of 4-IA (Hussain et al.,2009) which was not detected in our study. 4-IA has been previously detected in liquid cultures of microorganisms degrading IPU (Johannesen et al., 2003; Sorensen and Aamand, 2001) but it has been rarely detected in soil studies (Muddet al., 1983). Our adsorption studies confirmed that 4-IA was strongly adsorbed onto soil colloids. Thus, it is probable that the transformation of IPU in our study led to the formation of 4-IA which became strongly bound to soil colloids and it was not available for extraction. The demethylated transformation products of IPU were transitory and did not accumulate in soil. Their dissipation rates in both laboratory and field studies showed a decreasing trend with increasing dose rates in agreement with the dissipation behavior of the parent compound. No previous studies have investigated the dissipation kinetics of MD-IPU and DD-IPU and their DT50s at both laboratory and field scale. The only data available are from the registration documents of IPU which reported laboratory DT50 values of 10.5-100.5 d for MD-IPU and 38.4-78.1 d for DD-IPU (EFSA, 2015), which is in agreement with our findings.

Dissipation of TCZ

TCZ showed a moderate to high persistence in the laboratory study. Its DT50 values (60.7-97.1 d) were within the range reported by previous laboratory studies (Strickland et al., 2004; Potter et al., 2005; Li et al., 2015). However, they were

Chapter 3 Pesticide dissipation in agricultural soil 105 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions significantly lower than the DT50s reported in the registration documents where 14C- labelled TCZ was used (DT50 > one year) (EFSA, 2014). DT50 values measured under laboratory conditions indicated a trend for increasing persistence with increasing dose rate, in accordance with previous laboratory studies (Muñoz-Leoz et al., 2011; Wang et al., 2016).

TCZ showed a limited persistence in the field with DT50 values of 1.5-2.5 days for the different dose rates. Previous field studies have also showed a lower persistence of TCZ in the field compared to the laboratory (Herrero-Hernández et al., 2011; Wang et al.,

2015). Despite that, the DT50s obtained for TCZ in the field study are at the lower part of the range of DT50 values reported in the literature which vary from 5.8-6.5 d (temperature 15-30 °C) (Wang et al., 2015) to 91.6 days (average temperature 10.5 °C) (EFSA, 2014). The reduced field persistence of TCZ observed in our study was the result of a biphasic dissipation pattern which was composed of a very rapid dissipation phase within the first three days (70 % dissipation) followed by a slow dissipation phase thereafter where limited loss of TCZ occurred. A more realistic measure of the soil exposure to TCZ would be given by the DT90 values which ranged from 198-603.4 days in agreement with the DT90-values reported for the field dissipation of TCZ (115- 304 days) (EFSA, 2014). The rapid decline of TCZ observed in the field study during the first 3 days after application could not be attributed to a rapid degradation of TCZ as it is indicated by the moderate to high persistence of the compound in the laboratory where degradation, biotic or abiotic, constitutes the main dissipation process. Rapid formation of bound residues or movement of the pesticide below the top 10 cm of the soil have been proposed by Herrero-Hernandez et al. (2011) as possible reasons to explain the initial rapid field dissipation of TCZ. The rapid formation of bound residues of TCZ is not fully supported by the moderate soil adsorption affinity of TCZ and previous regulatory studies which reported the formation of only 19.5 % of soil bound residues after 30 d (EFSA, 2014). Significant losses of TCZ due to photolysis or volatilization are not expected considering its soil photostability and low volatility (EFSA, 2014). A precipitation event occurred at day 3 (Supplementary Figure 3.1.2), right before the collection of the soil samples, might have facilitated the vertical leaching of a large fraction of TCZ residues below the top 10 cm which was the soil layer sampled. Regarding the second slow dissipation phase of TCZ, it could be attributed to a strongly

Chapter 3 Pesticide dissipation in agricultural soil 106 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions adsorbed fraction of the pesticide that could be less accessible to dissipation processes (Muñoz-Leoz et al., 2011; Herrero-Hernández et al., 2011). Previous studies have identified four main transformation products of TCZ: a lactone, a pentanoic acid, a triazolyl pinacoline, and a 5-keto derivative (Strickland et al., 2004; Potter et al., 2005). In contrast, regulatory documents suggest that 1,2,4-triazole is the most relevant metabolite of TCZ in soil, whereas the derivatives reported by Strickland et al. (2004) were only detected at trace amounts (EFSA, 2014). In the present study, no transformation products of TCZ were determined due to the lack of relevant analytical standards. However, a follow-up study combining suspect screening time-of-flight mass spectrometry with in silico molecular typology reported the presence of 22 empirical and 12 yet unknown transformation products of TCZ in the field experiment soil samples treated with x5 dose rate (Storck et al., 2016). Among them, some of the transformation products reported by Strickland et al. (2004) were found.

Dissipation and transformation of CHL

CHL showed a moderate persistence with DT50s within the range reported in the literature: 10 to 120 days in laboratory studies (Racke, 1993; Papadopoulou et al., 2016) and 0.6 to 121 days in field studies (Jin and Webster, 1997; Laabs et al., 2002). In the laboratory study, the persistence of CHL decreased with increasing dose rates (DT50s 28.6 to 52.6 days in the x10 and x1 dose rate, respectively) which is not in accordance with the general trend for increasing CHL persistence at increasing application rates

(John and Saike, 2015) and to the opposite trend observed in the field experiment (DT50s 6.7 to 119.6 days in the x1 and x5 dose rate, respectively). The higher laboratory dissipation of CHL at increasing dose rates might be attributed to its high adsorption affinity (verified in our study) which at low dose rates in a static soil laboratory incubation system could be mostly adsorbed resulting in low bioavailability and longer persistence compared to the higher dose rates which might have saturated the soil adsorption sites and the remaining fraction of CHL is found dissolved in the soil solution phase where it was degraded by biotic and abiotic mechanisms. If we compare the persistence of CHL in the laboratory and in the field study at the two common dose rates

Chapter 3 Pesticide dissipation in agricultural soil 107 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions x1 and x2, a faster field dissipation of CHL was evident, which is in agreement with the generally more rapid field dissipation of the other pesticides tested. In both laboratory and field experiments, the dissipation of CHL proceeded via hydrolysis to TCP. Several previous studies have reported the vulnerability of CHL to hydrolysis in alkaline soils which is controlled by both abiotic and biotic processes (Racke et al., 1996; Singh et al., 2003). The accumulation of TCP upon hydrolysis of CHL in agricultural soils has been linked to the resistance of the parent compound to the phenomenon of enhanced biodegradation due to the antimicrobial characteristics of TCP (Racke, 1990). However, no accumulation of TCP was observed in our study and the TCP amounts formed were dissipated showing a contrasting dose-dependent pattern compared to the parent compound. In particular, the persistence of TCP in soil increased with increasing dose rates in the laboratory experiment, whereas the opposite trend was observed in the field experiment. Little is known regarding the dissipation rates of TCP and its fate upon CHL dissipation has been scarcely explored. In agreement with our findings, Baskaran et al. (2003) reported laboratory DT50 values of 42-49 days in topsoil and observed a contrasting dissipation behavior of TCP compared to its parent compound.

Adsorption of pesticides and their transformation products in soil

Based on the dissipation patterns of the pesticides and their transformation products in the laboratory and the field study, we determined the soil adsorption of the studied compounds and of their main transformation products for which little is known. The three pesticides showed different soil adsorption affinity which increased in the following order IPU

IPU (log Kow 2.5). The Kf values obtained were within the range reported in the literature for IPU (0.67-5 mg1-N LN kg-1) (Benoit et al., 1998; EFSA, 2015), TCZ (30-80 mg1-N LN kg-1) (Čadková et al., 2013a; 2013b) and CHL (116-271 mg1-N LN kg-1) (Gebremariam et al., 2013). Based on the shape of the adsorption isotherms of the compounds studied we could obtain information on the adsorption mechanism. Thus the C-type isotherms of MD-IPU, DD-IPU, and TCP suggest a constant partition of these compounds between

Chapter 3 Pesticide dissipation in agricultural soil 108 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions solution and substrate (Giles et al., 1960). On the other hand, the L-shaped isotherms observed for all the other compounds suggest a high dependence of adsorption on the initial solution concentration of these compounds with higher adsorption observed at lower solute concentration (Teng and Hsieh, 1998). In most cases adsorption studies are performed with the active substances although pesticides are applied in the field as commercial formulations which might exhibit a different adsorption behavior. All pesticides studied showed a higher soil adsorption affinity in the form of commercial formulations compared to the pure active substance. This is not surprising considering the presence of organic additives in the commercial formulations including dispersing or wetting agents and other surfactants that have the ability to decrease surface tension and facilitate the adsorption of pesticides onto soil mineral surfaces (Čadková et al., 2012). Previous studies have clearly demonstrated a positive effect of the commercial formulation on the adsorption of TCZ (Čadková et al., 2013a; 2013b) while no such studies are available regarding the other two pesticides.

IPU transformation products MD-IPU (Kf = 4.43 mg1-N LN kg-1) and DD-IPU (Kf =

4.17 mg1-N LN kg-1) showed an adsorption behavior similar to the parent compound (Kf =

4.00 mg1-N LN kg-1). On the contrary, 4-IA (Kf = 34.9 mg1-N LN kg-1) showed a substantially higher adsorption affinity than IPU, MD-IPU, and DD-IPU. Little is known about the adsorption behavior of IPU transformation products. Registration documents of IPU reported Kf values of 0.57-4.41 mg1-N LN kg-1 for MD-IPU (EFSA, 2015), while studies by

Johannesen et al. (2003) showed Kf values of 21.2 mg1-N LN kg-1 for 4-IA. In contrast, no adsorption measurements are available for DD-IPU. Our adsorption data suggest that the demethylated transformation products of IPU are expected to be moderately mobile in soil similarly to the parent compound. In contrast to the high adsorption affinity of CHL, its hydrolysis product TCP exhibited low adsorption affinity (Kf = 1.55 mg1-N LN kg-1). This is in line with previous studies by Baskaran et al. (2003) who reported that TCP was 100 times less adsorbed than the parent compound. Our findings suggest that the accumulation of TCP in the soil environment could induce a reciprocal risk for the contamination of groundwater resources and should deserve greater attention in future environmental fate studies and pesticide risk assessment.

Chapter 3 Pesticide dissipation in agricultural soil 109 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Conclusions

In the present study, we evaluated the soil dissipation and metabolism of three model pesticides, IPU, TCZ and CHL, following a tiered lab-to-field experimental approach. Overall, IPU was the least persistent chemical, showed an increasing persistence at increasing dose rates and was demethylated to MD-IPU and DD-IPU. The latter showed similar dissipation as the parent compound. TCZ was the most persistent chemical in the laboratory compared to the field where it dissipated more rapidly following a biphasic pattern. Finally, CHL showed a contrasting dose-dependent behavior in the laboratory and in the field study and it was hydrolyzed to TCP which was further dissipated. Soil adsorption affinity increased in the order IPU

Acknowledgements

This study was conducted within the frame of the FP7-PEOPLE-2012-IAPP (Industry-Academia Partnerships and Pathways) Marie-Curie project ‘Love-to-Hate: Pesticides: Felicity or curse for the soil microbial community?’ which was funded by the European Commission (grant number 324349). The PhD studies of Veronika Storck performed at the Doctoral School of Environment and Health (ED E2S) at the University of Burgundy (grant number 2013-31) was funded by the French Ministry of Education and Research (MESR).

Chapter 3 Pesticide dissipation in agricultural soil 110 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions Supplementary material

Supplementary Figure 3.1.1. Location of the North Italian agricultural field.

Supplementary Table 3.1.1. The physicochemical characteristics of the soil used in the study.

Chapter 3 Pesticide dissipation in agricultural soil 111 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Supplementary Table 3.1.2. Dose levels (mg/kg soil dwt) of the model pesticides used in the laboratory and in the field study. The expected pesticide concentrations in soil were calculated based on the assumption of a soil bulk density of 1.3 kg/L, diffusion of pesticide residues to a soil depth of 5 cm for IPU and TCZ, and to 10 cm for CHL (soil incorporated).

Supplementary Table 3.1.3. Soil dissipation kinetics of IPU, TCZ, and CHL in the laboratory experiment.

Chapter 3 Pesticide dissipation in agricultural soil 112 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Supplementary Figure 3.1.2. Meteorological data recorded during the field experiment: (a) Mean daily precipitation [mm], (b) mean daily air and soil temperature [°C], (c) mean daily solar radiation [W/m2], (d) mean daily air relative humidity [%] and (e) mean daily wind speed [m/s].

Chapter 3 Pesticide dissipation in agricultural soil 113 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Supplementary Table 3.1.4. Soil dissipation kinetics of IPU, TCZ, and CHL in the field experiment.

Chapter 3 Pesticide dissipation in agricultural soil 114 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions References

Alletto, L., Coquet, Y., Benoit, P. and Bergheaud, V. (2006). Effects of temperature and water content on degradation of isoproturon in three soil profiles. Chemosphere 64, 1053-1061. Álvarez-Martín, A., Sánchez-Martín, M.J., Pose-Juan, E. and Rodríguez-Cruz, M.S. (2016). Effect of different rates of spent mushroom substrate on the dissipation and bioavailability of cymoxanil and tebuconazole in an agricultural soil. Science of the Total Environment 550, 495-503. Andreasen and Andersen (1930). Etude de l’influence de la dilution sur les résultats de l’analyse granulometrique par sedimentation. Kolloid 50, 217. Arias-Estévez, M., López-Periago, E., Martínez-Carballo, E., Simal-Gándara, J., Mejuto, J.-C. and García-Río, L. (2008). The mobility and degradation of pesticides in soils and the pollution of groundwater resources. Agriculture, Ecosystems and Environment. 123, 247-260. Baskaran, S., Kookana, R.S. and Naidu R. (2003). Contrasting behaviour of chlorpyrifos and its primary metabolite, TCP (3,5,6- trichloro-2-pyridinol), with depth in soil profiles. Australian Journal of Soil Science 41, 749-760. Benoit, P., Barriuso, E., Vidon, P. and Real, B. (1998). Isoproturon sorption and degradation in a soil from grassed buffer strip. Journal of Environmental Quality 28, 121-129. Bremner, J.M. and Mulvaney, C.S. (1982). In: Methods of soil analysis, Part 2, Chemical and microbiological properties, Monograph 9.2, pp. 595-624. A.L. Page, R.H. Miller (Eds.) Čadková, E., Komárek, M., Kaliszová, R., Koudelková, V., Dvořák, J. and Vaněk, A. (2012). Sorption of tebuconazole onto selected soil minerals and humic acids. Journal of Environmental Science and Health, Part B 47, 336-342. Čadková, E., Komárek, M., Kaliszová, R., Száková, J., Vaněk, A., Bordas, F. and Bollinger, J.-C. (2013a). The influence of copper on tebuconazole sorption onto soils, humic substances, and ferrihydrite. Environmental Science and Pollution Research 20, 4205-4215. Čadková, E., Komárek, M., Kaliszová, R., Vaněk, A. and Balíková, M. (2013b). Tebuconazole sorption in contrasting soil types. Soil and Sediment Contamination 22, 404-414. Collings, L.V., Blair, A.M., Gay, A.P., Dyer, C.J. and Mackay, N. (2003). The effect of weather factors on the performance of herbicides to control Alopecurus myosuroides in winter wheat. Weed Research 43, 146-153. D’Angelo, D.L., Bradley, C.A., Ames, K.A., Willeyard, K.T., Madden, L.V. and Paul, P.A. (2014). Efficacy of fungicide applications during and after anthesis against fusarium head blight and deoxynivalenol in soft red winter wheat. Plant Disease Journal 98, 1387-1397. Dolaptsoglou, C., Karpouzas, D., Menkissoglu-Spiroudi, U., Eleftherohorinos, I. and Voudrias, E.A. (2009). Influence of different organic amendments on the leaching and dissipation of terbuthylazine in a column and a field study. Journal of Environmental Quality 38, 782-791. EEC (2000). Guidance Document No 9188/VI/97-ver. 8, Persistence in Soil, Commission of the European Communities, Directorate General for Agriculture, GG VI B II-1, Brussels, Belgium, Issued on July 2000. EFSA (2010). Scientific Opinion on the development of specific protection goal options for environmental risk assessment of pesticides, in particular in relation to the revision of the guidance documents on aquatic and terrestrial ecotoxicology (SANCO/3268/2001 and SANCO/10329/2002). EFSA Panel on Plant Protection Products and their Residues. EFSA Journal 8, 1821-1876. EFSA (2014). Conclusion on the peer review of the pesticide risk assessment of the active substance tebuconazole. EFSA Journal 12, 3485. EFSA (2015). Conclusion on the peer review of the pesticide risk assessment of the active substance isoproturon. EFSA Journal 13, 4206. Egner H., Riehm H. and Domingo W.R. (1960). Untersuchungen über die chemische Bodenanalyse als Grundlage für die Beurteilung des Nährstoffzustandes der Böden. II. Chemische Extraktionsmetoden zur Phosphor- und Kaliumbestimmung. K. Lantbr. Hogsk. Annlr. 26, 199-21. Fenoll, J., Hellín, P., Martínez, C.M., Flores, P. and Navarro, S. (2012). High performance liquid chromatography-tandem mass spectrometry method for quantifying phenylurea herbicides and their main metabolites in amended and unamended soils. Journal of Chromatography A, 1257, 81-88. FOCUS (2006). Guidance document on estimating persistence and degradation kinetics from environmental fate studies on pesticides in EU registration. Report of the FOCUS Work Group on Degradation Kinetics, EC Document Reference Sanco/10058/2005 version 2.0, 2006, 434pp. Gebremariam, S.Y., Beutel, M.W., Yonge, D.R., Flury, M. and Harsh, J.B. (2012). Adsorption and desorption of chlorpyrifos to soils and sediments. Reviews of Environmental Contamination and Toxicology 215, 123-175. Giles, C.H., Macewans, T.H., Nakhwa, N. and Smith, D. (1960). Studies in adsorption. Part XI. A system of classification of solution adsorption isotherms, and its use in diagnosis of adsorption mechanisms and in measurement of specific surface areas of solids. Journal of the Chemical Society 786, 3973-3993. Herrero-Hernández, E. Andrades, M.S., Alvarez-Martina, A., Pose-Juana, E., Rodriguez-Cruz, M.S and Sánchez-Martín, M.J. (2013). Occurrence of pesticides and some of their degradation products in waters in a Spanish wine region. Journal of Hydrology 486, 234-245. Herrero-Hernández, E., Andrades, M.S., Marín-Benito, J.M., Sánchez-Martín, M.J. and Rodríguez-Cruz, M.S. (2011). Field-scale dissipation of tebuconazole in a vineyard soil amended with spent mushroom substrate and its potential environmental impact. Ecotoxicology and Environmental Safety 74, 1480-1488. Hussain, S., Arshad, M., Springael, D., Sørensen, S. R., Bending, G. D., Devers-Lamrani, M., Maqbool, Z., and Martin-Laurent, F. (2015). Abiotic and biotic processes governing the fate of phenylurea herbicides in soils: A review. Critical Reviews in Environmental Science and Technology 45, 1947-1998. Hussain, S., Sorensen, S.R., Devers-Lamrani, M., El-Sebai, T. and Martin-Laurent, F. (2009). Characterization of an isoproturon mineralizing bacterial culture enriched from a French agricultural soil. Chemosphere 77, 1052-1059. ISO 10381-1 and -2. (2002). Soil quality - Sampling - Part 1: Guidance on the design of sampling programmes. Part 2: Guidance on sampling techniques. Jin H. and Webster G.R.B. (1997). Dissipation of chlorpyrifos, oxon and 3,5,6-tricholor-2-pyridinol in litter and elm forest soil. Journal of Environmental Science and Health Part B, 32, 879-900.

Chapter 3 Pesticide dissipation in agricultural soil 115 Chapter 3.1 Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions

Johannesen, H., Sørensen, S.R. and Aamand, J. (2003). Mineralization of soil-aged isoproturon and isoproturon metabolites by Sphingomonas sp. strain SRS2. Journal of Environmental Quality 32, 1250-1257. John, E. M. and Shaike, J.M. (2015). Chlorpyrifos: pollution and remediation. Environmental Chemistry Letters 13, 269-291. Joseph, S.V. and Zarate, J. (2015). Comparing efficacy of insecticides against cabbage maggot (Diptera: Anthomyiidae) in the laboratory. Crop Protection 77, 148-156. Juhler, R. K., Sørensen, S. R. and Larsen, L. (2001). Analysing transformation products of herbicide residues in environmental samples. Water Research 35, 1371-1378. Karpouzas, D.G., Kandeler, E., Bru, D., Friedel, I., Auer, Y., Kramer, S., Vasileiadis, S., Petric, I., Udikovic-Kolic, N., Djuric, S. and Martin- Laurent, F. (2014). A tiered assessment approach based on standardized methods to estimate the impact of nicosulfuron on the abundance and function of the soil microbial community. Soil Biology and Biochemistry 75, 282-291. Keinath, A.P. (2015). Baseline sensitivity of Didymella bryoniae to cyprodinil and fludioxonil and field efficacy of these fungicides against isolates resistant to pyraclostrobin and boscalid. Plant Disease 99, 815-822. Laabs, V., Amelung, W., Fent, G., Zech, W. and Kubiak, R. (2002). Fate of 14C-labeled soybean and corn pesticides in tropical soils of Brazil under laboratory conditions. Journal of Agricultural and Food Chemistry 50, 4619-4627. Laabs, V., Amelung, W., Pinto, A., Altstaedt, A. and Zech, W. (2000). Leaching and degradation of corn and soybean pesticides in anoxisol of the Brazilian Cerrados. Chemosphere 41, 1441-1449. Lehr, S., Scheunert, I., and Beese, F. (1996). Mineralization of free and cell-wall-bound isoproturon in soils in relation to soil microbial from environmental pollutants. Soil Biology and Biochemistry 28, 1-8. Li, Y., Dong, F., Liu, X., Xu, J., Han, Y. and Zheng, Y. (2015). Enantioselectivity in tebuconazole and myclobutanil non-target toxicity and degradation in soils. Chemosphere 122, 145-153. Martin-Laurent, F., Kandeler, E., Pertic, I., Djuric, S. and Karpouzas, D.G. (2013). ECOFUN-MICROBIODIV: an FP7 European project for developing and evaluating innovative tools for assessing the impact of pesticides on soil functional microbial diversity- towards new pesticide registration regulation? Environmental Science and Pollution Research 20, 1203-1205. Mele, P. M. and Carter, M. R. (1995). Estimation of microbial biomass by ninhydrin-reactive N using liquid chloroform. Canadian Journal of Soil Science 76, 37-40. Mudd, P.J., Hance, R.J. and Wright, S.J.L. (1983). The persistence and metabolism of isoproturon in soil. Weed Research 23, 239-347. Muñoz-Leoz, B., Ruiz-Romera, E., Antigüedad, I. and Garbisu, C. (2011). Tebuconazole application decreases soil microbial biomass and activity. Soil Biology and Biochemistry 43, 2176-2183. OECD-Guideline 106 (2000). OECD Guidelines for the testing of chemicals. Adsorption – desorption Using a Batch Equilibrium Method. Papadopoulou, E.S., Lagos, S., Spentza, F., Vidiadakis, E., Karas, P.A., Klitsinaris, T. and Karpouzas, D.G. (2016). The dissipation of fipronil, chlorpyrifos, fosthiazate and ethoprophos in soils from potato monoculture areas: first evidence for the enhanced biodegradation of fosthiazate. Pest Management Science 72, 1040-1050. Potter, T.L., Strickland, T.C., Joo, H. and Culbreath, A.K. (2005). Accelerated soil dissipation of tebuconazole following multiple applications to peanut. Journal of Environmental Quality 34, 1205-1213. Racke, K.D. (1993). Environmental fate of chlorpyrifos. Reviews of Environmental Contamination and Toxicology 131, 1-150. Racke, K.D., Laskowski, D.A. and Schultz, M.R. (1990). Resistance of chlorpyrifos to enhanced biodegradation in soil. Journal of Agricultural and Food Chemistry 38, 1430-1436. Racke, K.D., Steele, K.P., Yoder, R.N., Dick W.A. and Avidov, E. (1996). Factors effecting the hydrolytic degradation of chlorpyrifos in soil. Journal of Agricultural and Food Chemistry 44, 1582-1592. Rodríguez-Cruz, M.S., Jones, J.E. and Bending, G.D. (2006). Field-scale study of the variability in pesticide biodegradation with soil depth and its relationship with soil characteristics. Soil Biology and Biochemistry 38, 2910-2918. Sancez-Gonzalez, S., Pose-Juan, E., Herrero-Hernandez, E., Alvarez-Martin, A., Sancez-Martin, M.J. and Rodriguez-Cruz, S. (2013). Pesticide residues in groundwaters and soils of agricultural areas in the Águeda River Basin from Spain and Portugal. International Journal of Environmental Analytical Chemistry 93, 1585-1601. Singh, B.K., Walker, A., Alun, J., Morgan, W. and Wright, D.J. (2003). Effects of soil pH on the biodegradation of chlorpyrifos and isolation of a chlorpyrifos-degrading bacterium. Applied and Environmental Microbiology 69, 5198-5206. Skark, C. and Zullei-Seibert, N. (1995). The occurrence of pesticides in groundwater – Results of case studies. International Journal of Environmental Analytical Chemistry 58, 387-396. Skark, C., Zullei-Seibert, N., Willme, U., Gatzemann, U. and Schlett, C. (2004). Contribution of non-agricultural pesticides to pesticide load in surface water. Pest Management Science 60, 525-530. Sørensen, S.R. and Aamand, J. (2001). Biodegradation of the phenylurea herbicide isoproturon and its metabolites in agricultural soils. Biodegradation 12, 69-77. Sørensen, S.R., Ronen, Z. and Aamand, J. (2001). Isolation from agricultural soil and characterization of a Sphingomonas sp. able to mineralize the phenylurea herbicide isoproturon. Applied and Environmental Microbiology 67, 5403-5409. Storck, V., Lucini, L., Mamy, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., Karas, P.A., Servien, R., Karpouzas, D.G., Trevisan, M., Benoit, P. and Martin-Laurent, F. (2016). Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology. Environmental Pollution 208, 537-545. Strickland, T.C., Potter, T.L. and Joo, H. (2004). Tebuconazole dissipation and metabolism in Tifton loamy sand during laboratory incubation. Pest Management Science 60, 703-709. Teng, H. and Chien-To, H. (1998). Influence of surface characteristics on liquid phase adsorption of phenol by activated carbon prepared from Bituminons coal. Industrial and Engineering Chemistry Research 37, 3168-3624. Tixier, C., Sancelme, M., Aït-Aïssa, S., Widehem, P., Bonnemoy, F., Cuer, Truffaut, N. and Veschambre, H. (2002). Biotransformation of phenylurea herbicides by a soil bacterial strain, Arthrobacter sp. N2: structure, ecotoxicity and fate of diuron metabolite with soil fungi. Chemosphere 46, 519-526. Walker, A., Jurado-Exposito, M., Bending, G.D. and Smith, V.J.R. (2001). Spatial variability in the degradation rate of isoproturon in soil. Environmental Pollution 111, 407-415. Walkley, A. and Black, C.A. (1934). An examination of the Djetgarref method for determining organic matter and proposed modification of the chromic acid titration method. Soil Science 5, 401-407. Wang, C., Wang, F., Zhang, Q. and Liang, W. (2016). Individual and combined effects of tebuconazole and carbendazim on soil microbial activity. European Journal of Soil Biology 72, 6-13. Wang, Y., Wang, C., Gao, J., Liu, C., Cui, L. and Li, A. (2015). Dissipation, residues, and safety evaluation of trifloxystrobin and tebuconazole on ginseng and soil. Environmental Monitoring and Assessment 187, 344-356.

Chapter 3 Pesticide dissipation in agricultural soil 116

Chapter 3.2

Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Veronika Storck Luigi Lucini Laure Mamy Federico Ferrari Evangelia S. Papadopoulou Sofia Nikolaki Panagiotis A. Karas Remi Servien Dimitrios G. Karpouzas Marco Trevisan Pierre Benoit Fabrice Martin-Laurent

Published in Environmental Pollution 208, 537-545 (2016).

Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Abstract

Pesticides generate transformation products (TPs) when they are released into the environment. These TPs may be of ecotoxicological importance. Past studies have demonstrated how difficult it is to predict the occurrence of pesticide TPs and their environmental risk. The monitoring approaches mostly used in current regulatory frameworks target only known ecotoxicologically relevant TPs. Here, we present a novel combined approach which identifies and categorizes known and unknown pesticide TPs in soil by combining suspect screening time- of-flight mass spectrometry with in silico molecular typology. We used an empirical and theoretical pesticide TP library for compound identification by both non-target and target time- of-flight (tandem) mass spectrometry, followed by structural proposition through a molecular structure correlation program. In silico molecular typology was then used to group TPs according to common molecular descriptors and to indirectly elucidate their environmental parameters by analogy to known pesticide compounds with similar molecular descriptors. This approach was evaluated via the identification of TPs of the triazole fungicide tebuconazole occurring in soil during a field dissipation study. Overall, 22 empirical and 12 yet unknown TPs were detected, and categorized into three groups with defined environmental properties. This approach combining suspect screening time-of-flight mass spectrometry with molecular typology could be extended to other organic pollutants and used to rationalize the choice of TPs to be investigated towards a more comprehensive environmental risk assessment scheme.

Figure 3.2.1. Graphical abstract of Chapter 3.2.

Chapter 3 Pesticide dissipation in agricultural soil 120

Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology Introduction

Pesticides are used in agriculture to ensure high crop yields and quality. This large group of xenobiotic compounds and their transformation products (TPs) have globally been identified as major contaminants of natural water resources (Hutzinger, 2013; Lewis and Maslin, 2015). The United States Environmental Protection Agency (U.S. EPA) estimates that 2.4 billion kilograms of pesticides are released per year. They support important economic sectors of agriculture, but also significantly contribute to environmental pollution (U.S. EPA, 2011). Another aspect that makes pesticides a chemical group of particular interest is their rapidly evolving market which is continuously enriched with new molecules. In order to reach the market, new pesticide molecules have to undergo an environmental risk assessment (ERA) procedure based on a large set of data submitted by the pesticide companies and evaluated by the official authorities (e.g. the Plant Protection Products and their Residues panel (Regulation EC 1107/2009, 2009) of the European Food Safety Authority (EFSA) for Europe). This procedure is periodically repeated to re-evaluate the currently registered pesticide molecules. Once released into the agroecosystem, pesticides and their TPs can diffuse away from their target point and contaminate surrounding water resources (U.S. EPA, 2015), with reciprocal impact on non-target organisms supporting key ecosystemic services (Bozdogan, 2014). The properties of TPs such as bioavailability, mobility and ecotoxicity can substantially differ from the parent compound (Sinclair et al., 2010; Tixier et al., 2000), and although their biological efficacy is usually lower (Boxall et al., 2004), they could be highly ecotoxicologically relevant (Fenner et al., 2013). In Europe, a posteriori monitoring of the contamination of water and air resources by pesticides is imposed by the corresponding directives (EU-Air Framework Directive 2008/50/EC; EU-Water Framework Directive 2000/60/EC). By contrast, no monitoring of soil contamination by pesticides is required at the EU level. This ensues from the lack of a corresponding directive, although the need for some kind of regulation was proposed to the European Commission (Van-Camp et al., 2004). Thus, if an unacceptable risk for the environment becomes evident several years after the introduction of a pesticide to the market, its use becomes restricted, and if monitoring studies reveal that no alleviation of the risk is achieved, the pesticide is removed from the market. Past experience has shown that the knowledge of the existence of ecotoxicologically relevant TPs typically

Chapter 3 Pesticide dissipation in agricultural soil 122 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology emerges only 20 to 30 years after a given pesticide has first been marketed (Fenner et al., 2013). Currently used monitoring approaches for pinpointing ecotoxicologically relevant pesticide TPs are often restricted to water samples. They do not prevent the contamination of natural resources, and impose mitigation actions only after contamination has been reported. Novel approaches able to identify potential TPs and in silico assessment of their ecotoxicity and environmental contamination could offer a solution to this issue. The knowledge gap about the existence or properties of pesticide TPs derived from abiotic and biotic processes in the environment may be the reason why relevant TPs have been overlooked for so long. Conventional analytical methods focus on the monitoring of the so-defined ‘ecotoxicologically relevant’ pesticide TPs or TPs ‘relevant for groundwater contamination’ (OECD-Guideline ENV/JM/MONO(2007)17, 2007; EU- Regulation 1107/2009). Consequently, most studies focus on a limited number of molecules identified as targets through monitoring approaches and calibrated against reference standards. Therefore, they often provide only a partial view of the complex transformation patterns of pesticides in the environment. Thus, there is a need to develop an innovative approach by combining powerful analytical methods also detecting unknown TPs, with categorization methods based on the chemical features of TPs. The analytical techniques suitable to detect unknown and suspected molecules in the absence of reference standards are high resolution mass spectrometry (MS), namely quadrupole time-of-flight MS (QTOF-MS) and Fourier transform MS (FT-MS), as well as nuclear magnetic resonance (NMR) spectroscopy (Ibanez et al., 2005). These techniques are often combined in forensics (Reizel et al., 2012) or medicine (Zeng et al., 2013), where QTOF-MS is used for screening for new unknown molecules and then supplemented by NMR spectra for the structural elucidation of a compound. NMR spectroscopy is an effective method for confirming molecular structures, but it is less appropriate for analyzing TPs from environmental samples because of its rather poor sensitivity, generally three orders of magnitude lower than MS (Ibanez et al., 2005). Categorization of unknown and suspected molecules can be achieved by in silico molecular typology that clusters potential TPs according to their properties (Servien et al., 2014).

Chapter 3 Pesticide dissipation in agricultural soil 123 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology In this study, we present a new approach for the analysis of agricultural soil samples treated with a pesticide by combining high-resolution QTOF-MS with in silico molecular typology. Our approach provides the profiles and the molecular structures of unknown and suspected pesticide TPs in soil, using a compound library containing empirical (already known) and theoretical (yet unknown) TPs. It consists of three main steps: (i) construction of an in-house pesticide-specific TP library based on a literature survey and listing of suspect TPs, (ii) QTOF-MS analyses of soil extracts by non-target MS to screen TPs against this library, followed by target tandem MS using an ion inclusion list to propose their structure via accurate mass daughter ions and a molecular structure correlation (MSC) program, and (iii) categorization of detected TPs based on in silico analysis using molecular typology. To demonstrate the potential of this novel approach, the transformation of the triazole fungicide tebuconazole (TCZ) was investigated in a field dissipation study. TCZ was chosen as a model compound because, although it has been on the market for quite some time (27 years) (Börner et al., 2009), there are still significant knowledge gaps regarding its environmental fate, its transformation in soil and the ecotoxicological impact of potential TPs (EFSA, 2014). The potential formation of various triazole TPs is particularly relevant, because they are recalcitrant to biodegradation, and they possibly interact with the hormone regulation network of non-target organisms by inhibiting the cytochrome P450-dependent conversion of lanosterol to ergosterol (Rieke et al., 2014; Shalini et al., 2011). The applied approach is of interest for improving future environmental fate studies to become more comprehensive, thus strengthening ERA.

Material and methods

Library setup of TCZ TPs

An in-house TCZ TP library was constructed based on a thorough literature review including research papers (cited in the results section entitled ‘Empirical TCZ TPs’) and data sets from the pesticide company on the transformation pattern of TCZ within various matrices. Literature search was performed via ‘Web of Science’ and ‘Google’ using search terms such as ‘tebuconazole AND metabolite’, ‘tebuconazole AND

Chapter 3 Pesticide dissipation in agricultural soil 124 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology transformation product’, ‘tebuconazole AND degradation’. The resulting library was composed of 47 empirical (already detected in at least one study) and 29 theoretical TPs. Theoretical TPs were created based on expert knowledge taking into account common reaction processes in organic chemistry and biochemistry likely leading to probable molecules. The TCZ TP library contains the name, molecular formula, molecular structure and mass of each molecule (Supplementary Table 3.2.1). Initial pretests of chemical analysis were carried out on a subset of soil samples randomly chosen from the field study to search for TCZ TPs. Retention times of TPs proposed via MSC were then added to the library as an additional parameter for identification purposes in MS screening.

TCZ field dissipation study

The field dissipation study was conducted on an agricultural field in North Italy (45°05'22.1"N 9°45'58.6"E) in fall/winter 2013/14. The soil was characterized as loamy sand (4.2 % clay, 13.5 % silt, 82.3 % sand) with a pH of 7.3 and an organic matter content of 4.5 %. The field had not been treated with pesticides for more than five years, and was cropped with winter wheat. A commercial formulation of TCZ (Folicur® SE, Bayer Garden) was applied on 6 m2 plots at a dose of 2.5 kg TCZ per hectare (5 times the recommended dose) (Bayer CropScience, 2011). Soil samples were taken 0, 3, 7, 21, 35, 56, 70, 105, and 125 days after pesticide application, following procedures described in the ISO standard for collection and handling of soil samples (ISO 10381-6, 2009). For each plot and sampling date, a composite soil sample was prepared from 10 sub-samples collected with a 5 cm diameter iron cylinder at a depth of 5 cm. Soil samples were manually homogenized, sieved (5 mm), and kept at -20°C until further processing. The field dissipation study was part of a larger study investigating field dissipation of three model pesticides including TCZ at various application doses (four plots were used for this study: two non-treated control plots and two TCZ plots (5 times the recommended dose)). TCZ field dissipation was monitored in soil extracts via high performance liquid chromatography coupled to a photodiode array detector (HPLC-PDA), and followed a biphasic pattern with a decline time 90 % (DT90) of 186.6 days, as calculated with the hockey-stick model (Papadopoulou et al., personal communication).

Chapter 3 Pesticide dissipation in agricultural soil 125 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology TCZ and TP extraction from soil

TCZ and its related TPs were extracted with acetone from 40 g of soil, as explained in Supplementary Data 3.2.1. The final residues were re-dissolved in 1 mL of 70 % methanol and 30 % ultrapure water containing 0.01 % of phosphoric acid, and used for chemical analysis. Extraction efficiency for TCZ was 86.36 % ± 0.45 %.

UHPLC-ESI-QTOF-MS analysis

In order to investigate TCZ TPs, samples were first analyzed by non-target UHPLC-ESI-QTOF mass spectrometry. TCZ TPs were primarily identified by screening the MS-only raw data against the TP library, using accurate mass, isotope spacing and isotope ratio for identification. A difference in accurate mass of 5 ppm, together with an isotopic pattern score higher than 85 % were used as cutoff values. Compounds that passed both thresholds, that had plausible chromatogram peak features, and that were not detected in the controls (non-treated soil samples), were chosen for further analysis by target UHPLC-ESI-QTOF tandem MS. For that purpose, the three most abundant precursor ions (generated during MS-only) and the retention time of each MS-only TP were exported to be included into a target list for tandem MS. Details are given in Supplementary Data 3.2.2.

Molecular structure correlation

Elucidation of the molecular structure of the compounds detected by tandem MS was obtained through an MSC program (Agilent Mass Hunter Molecular Structure Correlator B.05.00). The program correlated accurate mass daughter ions for a compound of interest with the molecular structures available in the in-house custom library by using a ‘systematic bond-breaking’ approach, as described by Hill and Mortishire-Smith (2005). Briefly, an overall correlation score was generated by taking into account the individual scores from each daughter ion, the mass accuracy of these fragments, and the overall percentage of ion intensities being plausibly explained with substructures.

Chapter 3 Pesticide dissipation in agricultural soil 126 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology Molecular typology for categorizing TCZ TPs

TCZ and its TPs were categorized using the molecular typology approach ‘TyPol’ (Typology of Pollutants) proposed by Servien et al. (2014), which clusters molecules based on the use of molecular descriptors and of environmental parameters. The environmental parameters of TCZ TPs are unknown, so clustering was only done using a set of 40 molecular descriptors. After clustering, TyPol was used to estimate the environmental parameters (vapor pressure, dissipation half-life, sorption coefficient and bioconcentration factor) of TCZ TP clusters by analogy to reference compounds whose environmental parameters are known. Details about the procedure are given in Supplementary Data 3.2.3.

Results and discussion

Empirical TCZ TPs

The literature search revealed that studies reporting TCZ dissipation often ignored the occurrence of potential TPs and mainly followed targeted analytical approaches (Sehnem et al., 2010; Woo et al., 2010). It is noteworthy that non-target approaches resulted in the detection of previously unknown TPs in studies looking at the degradation of TCZ in bacterial or fungal cultures (Obanda and Shupe, 2009), during photolysis (Calza et al., 2002), in field studies (Herrero-Hernandez et al., 2011; Potter et al., 2005), in microcosm studies (Strickland et al., 2004; White et al., 2010) and in fruit (Schermerhorn et al., 2005). Additional information was found in regulatory documents mainly citing unpublished or confidential studies reporting TPs in soil and crops, partially based on MS and NMR spectroscopy (EFSA, 2008; FAO, 1994; U.S. EPA, 1990). Altogether, the literature search led to the construction of a library containing 47 empirical and 29 theoretical TCZ TPs (Supplementary Table 3.2.1), which was used for TCZ TP suspect screening in soil samples from the field dissipation study described above.

Chapter 3 Pesticide dissipation in agricultural soil 127 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology Detection of TCZ TPs

The QTOF-MS analyses of soil extracts gained from the field dissipation study led to the detection of the parent compound TCZ, as well as 22 empirical and 12 yet unknown TPs, which were in the library. Among the 34 detected potential TPs, some of them had exactly the same mass and retention times. 12 TPs passed all identification steps, from non-target MS-only identification to target tandem MS and structure correlation. Three TPs passed only the first two steps (selected by tandem MS but not further identified through structural correlation), and 19 TPs were only detected by MS identification with plausible chromatogram features, but were not further selected by tandem MS due to method restrictions, or selection criteria, or molecular characteristics (e.g. low mass or high volatility). Table 3.2.1 summarizes the characteristics of all our TCZ TPs. It reports the compound name, molecular structure, mass, reference if empirical, and retention time of each TP. Moreover, the time points when the TCZ TPs were detected are provided (expressed in days after the TCZ treatment was applied), together with information regarding the confidence of identification. TPs are coded according to their level of identification confidence (a: detected by MS-only; a,b: detected by MS and selected by tandem MS; a,b,c: detected by MS, selected by tandem MS and identified by MSC). According to Schymanski et al. (2014), level ‘a’ (MS-only detection) could be assigned to ‘level 4 - identification confidence by unequivocal molecular formula’; level ‘a,b’ (selection by tandem MS) to ‘level 3 - tentative candidate - identification confidence by structure, substituent and class’; level ‘a,b,c’ (structure proposition by MSC) to ‘level 2,b - probable structure - identification confidence by diagnostic evidence’. The molecular formulas of the compounds that were only detected by non-target MS but were not further selected by target tandem MS are rather unlikely to be false positives, as they (i) were not detected in the controls, (ii) have plausible chromatogram features and isotopic patterns, and (iii) do not differ in retention time between replicates and different time points. However, their molecular structure remains tentative because those compounds were not selected by tandem MS, and molecular structure elucidation by MSC is only possible based on the detected product ions produced during tandem MS. The non-selection of those compounds by target tandem MS could be attributed either to peak saturation at high concentrations (as occurred with the parent molecule TCZ when injected at high volumes), high sensitivity

Chapter 3 Pesticide dissipation in agricultural soil 128 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology of compounds present at low concentrations, too low masses of the product ions, or high volatility of the precursor. Indeed, electron spray ionization (ESI) conditions and the chosen collision energies have to be efficient enough to produce precursor and product ions while avoiding fragmenting the precursor into too small product ions, as the minimum m/z value was chosen to be 60 due to noise exclusion. Regarding low m/z values, the putative low-molecular-weight TPs undetected by the non-target MS-only screening - e.g. the well-known triazole fungicide TP 1-H-1,2,4-triazole (TP 63) (EFSA, 2008; FAO, 1994; Schermerhorn et al., 2005; U.S. EPA, 1990) - may be considered as false negatives: they may not have been efficiently ionized by ESI due to their volatility at the high temperature of the ion source. Nevertheless, a compound detected by MS- only and selected by target tandem MS, but whose molecular structure could not be proposed by MSC, might be correct in its molecular formula but its molecular structure might not be in the TP library. Additionally, the poor correlation for low-molecular- weight TPs might be affected by the formation of product ions, which were below the m/z threshold and thus were missing for MSC. Strictly speaking, all the molecules detected by our approach still remain inferred until confirmed by the retention times and fragmentation patterns of pure reference standards that are not presently available on the market in most cases.

Chapter 3 Pesticide dissipation in agricultural soil 129 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Table 3.2.1. TCZ TPs detected in this study by suspect screening analysis. Reference to the degree of confidence on identification (a: detected by MS; b: selected by target tandem MS; c: proposed molecular structure by MSC). The name, structure, mass and reference if empirical of each TP are provided. The time points when a TP was detected and the cluster into which it was categorized are given. Blue boxes group TPs with the same mass.

Chapter 3 Pesticide dissipation in agricultural soil 130 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Chapter 3 Pesticide dissipation in agricultural soil 131 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Chapter 3 Pesticide dissipation in agricultural soil 132 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology Despite this weakness, the benefits of suspect screening coupled to MSC for elucidating molecular structures of TPs represent a powerful alternative when reference standards are not available. These benefits can be illustrated by the example of TPs 13 and 15, which are constitutional isomers. As they have the same molecular formula and mass, they were not discriminated by MS-only and tandem MS (Supplementary Figure 3.2.1), even using MS instruments with a higher resolution power. MSC correlated the product ions to the respective precursors for elucidation of their molecular structure (Supplementary Figure 3.2.2), and hence discriminated between these TPs without requiring any additional separation step or the use of reference standards. Furthermore, a significantly higher number of TPs can be considered in suspect screening compared to targeted approaches. This provides a more comprehensive picture of the transformation of one or more pesticides into TPs in the environment. The pesticide dissipation study carried out at field scale was suitable for the purposes of our study. Our TCZ concentration was 5 times the recommended dose (Bayer CropScience, 2011), and nine time points within 125 days after treatment provided for a comprehensive investigation of TP formation and decay patterns. Yet, detailed data on the temporal patterns were not established due to the absence of reference standards. As TCZ TPs were extracted from the top 5 cm layer of soil, only those TPs present in that soil layer were studied. TP properties can strongly deviate from those of the parent molecule (Giacomazzi and Cochet, 2004) and might be vulnerable to leaching, runoff or volatilization, although the parent molecule is considered as relatively immobile or non-volatile (EFSA, 2008; Horn et al., 2003). Thus, TPs that potentially dissipated from the top soil layer, that were not extractable via our extraction method or that were not listed in the TP library slipped through the net of this study. Moreover, TCZ and several of its TPs have two enantiomers or even possess more than one chiral center. This may be important for stereoselective ecotoxicity and dissipation (Wang et al., 2012), but was not considered in this study. To complete our analytical approach and provide comprehensive information on the actual environmental fate of TPs, molecular typology can be applied to theoretically estimate the potential properties of TPs. This technique can be used to estimate the properties of TPs from any man-made substances present in the environment.

Chapter 3 Pesticide dissipation in agricultural soil 133 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology Categorization of TCZ TPs

TCZ and its 34 TPs were categorized into three different clusters using TyPol (Table 3.2.1; Figure 3.2.2) according to their molecular descriptors including molecular weight, number of atoms, dipole moment, polarizability or total energy (Supplementary Table 3.2.2). TCZ and 23 of its TPs were categorized into cluster 1. They exhibited the highest simple and valence molecular connectivity indices (MCIs), energy of the highest occupied molecular orbital (EHOMO), molecular weight (MW), surface (S), polarizability

(), number of atoms, and the lowest total energy (Etot). Cluster 2 was composed of seven TPs that were characterized by the highest dipole moment (µ), energy of the lowest unoccupied molecular orbital (ELUMO) and the lowest number of chlorine atoms. The last four TPs were grouped in cluster 3. They were characterized by the lowest

MCIs, EHOMO, MW, S and , but by the highest Etot. This means that TPs classified into the same cluster as their parent molecule TCZ (including newly detected and yet unknown TPs) may have a similar behavior as TCZ in the environment. TPs clustered away from TCZ may have different environmental fates and ecotoxicological impacts than TCZ. This can be a true issue if their formation is ignored and thus only TCZ and known TPs are quantified and monitored in the environment. To evaluate these hypotheses, TyPol was applied for a second time to categorize 116 various well-known pesticides and 19 TPs available from the TyPol database, with an analysis of both molecular descriptors and their known environmental parameters (half-life in soil (DT50), sorption coefficient (KOC), vapor pressure (Pvap), and bioconcentration factor (BCF)). The compounds formed six clusters, and median values of the molecular descriptors for each cluster were obtained (an example is given for KOC in Supplementary Table 3.2.3, Supplementary Figure 3.2.3 and Supplementary Table 3.2.4). In order to obtain a preliminary idea of the potential environmental behavior of TCZ TPs, their molecular descriptors were compared to the median values of the six clusters formed by clustering the 116 pesticides and 19 TPs. When the molecular descriptors of the well-known compounds of one cluster were highly similar to those of a TCZ TP cluster, it was assumed that environmental parameters could be in a similar range. The TPs in the same cluster as their parent compound TCZ (cluster 1) overlapped with a cluster of pesticides characterized by (i) low Pvap (median Pvap = 0.026 mPa), indicating low volatility from soil and plants and thus a low risk for short or long range

Chapter 3 Pesticide dissipation in agricultural soil 134 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology transfer through the atmosphere (FOCUS, 2008), (ii) low mobility through soil (median

KOC = 825 L/kg) suggesting a low risk for groundwater contamination (McCall et al.,

2014), (iii) low persistence in soil (median DT50 = 37 days), and (iv) low potential for bioaccumulation (median BCF = 78) (Supplementary Table 3.2.2). In the same manner, the TCZ TPs grouped in cluster 2 had molecular descriptor values similar to pesticides and TPs exhibiting (i) higher soil mobility than TCZ (median KOC = 217 L/kg), hence a moderate potential risk for groundwater contamination (McCall et al., 1980), (ii) a higher potential for volatilization losses (median Pvap = 0.098 mPa) than TCZ, (iii) lower persistence in soil (median DT50 = 20 days), and (iv) a lower potential for bioaccumulation (median BCF = 44). Finally, the TCZ TPs in cluster 3 were the most mobile ones (median KOC = 123 L/kg), suggesting a potentially higher risk for groundwater contamination. Furthermore, cluster 3 TPs exhibited the highest risk for volatilization losses (median Pvap = 1.600 mPa), suggesting a significant risk for long or short range atmospheric transfer. Their persistence in soil was assumed to be similar to that of cluster 2 (DT50 = 20 days), and lower than that of TCZ (cluster 1), and their potential for bioaccumulation was estimated to be lower (BCF = 1) as compared to cluster 1 (BCF = 78) and cluster 2 (BCF = 44) (Supplementary Table 3.2.2). These results are in agreement with previous observations showing that when EHOMO, MCIs, and MW decrease, sorption of organic compounds tends to decrease, and when , µ and MW increase, Pvap tends to decrease (Mamy et al., 2015). Overall, the in silico approach using TyPol suggests that 23 of the 34 TCZ TPs targeted in this study can be expected to have a similar environmental behavior as the parent compound. On the other hand, the remaining 11 TPs may be more mobile and volatile than TCZ, suggesting a higher risk for groundwater and air contamination. However, their lower persistence and lower bioaccumulation potential than TCZ may counterbalance this risk. This combined method provides a first assessment of the environmental fate of theoretical TPs and allows a first identification of those of potential concern that need to be investigated in a refined risk assessment approach. If we consider TCZ as a case study, certain TPs from each cluster could be selected based on their molecular features (e.g. selecting between some constitutional isomers, keto- enol tautomers or others) and further considered for a refined ERA.

Chapter 3 Pesticide dissipation in agricultural soil 135 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Figure 3.2.2. Categorization of TCZ and its 34 TPs into three clusters according to their molecular descriptors on the two main components of the PLS regression (PLS 1 and PLS 2), and their estimated environmental parameters. Clusters are represented by different color codes. Each TP has its respective number. The inserted graph is an enlargement of a part of cluster 1.

To our knowledge, our approach is the first to combine (i) comprehensive screening of environmental samples (particularly soils) for suspected pesticide TPs, based on a library of empirical and theoretical TPs, (ii) detection of TPs without using reference standards, (iii) tentative proposal for their molecular structures without using NMR spectroscopy, and then (iv) categorization of TPs according to their physicochemical properties to assess their environmental behavior. The approach described in this study could be applied for refining pesticide risk assessment. It would enable regulators and agrochemical industries to identify potential pesticide TPs with relevant properties for further environmental and ecotoxicological assessment. Once the selected TPs are synthesized, it will be possible to confirm and quantify them, and thereby their fate. This way, environmental concentrations and the toxicity to

Chapter 3 Pesticide dissipation in agricultural soil 136 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology ecosystems and human health will be assessed, providing a refined and comprehensive risk assessment of a given pesticide.

Acknowledgements

The Ph.D. studies of Veronika Storck performed at the Doctoral School of Environment and Health (ED E2S) at the University of Burgundy were funded by the French Ministry of Education and Research (MESR). This work was done within the framework of the FP7-PEOPLE-2012-IAPP (Industry-Academia Partnerships and Pathways) Marie-Curie project ‘Love-to-Hate’ funded by the European Commission (grant number 324349). We thank Annie Buchwalter for English editing. We are grateful to Reviewers and Editor for helpful comments.

Chapter 3 Pesticide dissipation in agricultural soil 137 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology Supplementary material

Supplementary Table 3.2.1. TCZ TP library containing TCZ and 76 TPs (47 empirical TPs and 29 theoretical TPs). TPs with the same mass are indicated by blue boxes.

Chapter 3 Pesticide dissipation in agricultural soil 138 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Chapter 3 Pesticide dissipation in agricultural soil 139 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Chapter 3 Pesticide dissipation in agricultural soil 140 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Chapter 3 Pesticide dissipation in agricultural soil 141 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Chapter 3 Pesticide dissipation in agricultural soil 142 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Supplementary Data 3.2.1. TCZ and TP extraction from soil samples. Soil samples (40 g) were extracted twice with 50 mL of acetone and 2 mL of 2 M ammonium acetate for 30 minutes under shaking. The liquid phase was decanted through a glass fiber filter (Whatman GF/F) on a Buchner funnel stand under vacuum into a cone-shaped separator funnel containing 200 mL of 2 M sodium sulfate and 100 mL of dichloromethane. The separator funnel was shaken for 10 min. The dichloromethane phase was removed and the aqueous phase was extracted two more times with 50 mL of dichloromethane. The dichloromethane phases from the repeated extraction cycles were then combined and loaded onto a glass column packed with 1 g of

Chapter 3 Pesticide dissipation in agricultural soil 143 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology glass wool and 40 g of anhydrous sodium sulfate. The eluent was subsequently collected into an evaporation balloon and evaporated to complete dryness on a rotary evaporator under vacuum at 40 °C. The residues of TCZ were re-suspended in 10 mL of dichloromethane and re-evaporated to complete dryness under a stream of ultrapure nitrogen gas. TCZ residues were finally dissolved in 1 mL of 70 % methanol and 30 % ultrapure water containing 0.01 % of phosphoric acid and were used for chromatographic analysis as described below.

Supplementary Data 3.2.2. UHPLC-ESI-QTOF-tandem MS analysis. Samples were first analyzed by non-target MS-only. The screening was carried out on a hybrid quadrupole-time-of-flight mass spectrometer coupled to a UHPLC chromatographic system (UHPLC/Q-TOF). The instrument was run in the positive scan mode and operated to acquire MS-only spectra. A 1290 UHPLC chromatographic system, equipped with a binary pump and a Dual Electrospray Jet Stream ionization system, and coupled to a G6550 mass spectrometer (Agilent technologies Santa Clara, CA, USA) was used. UHPLC separation was performed at a flow rate of 0.3 mL/min at 35 °C, using ultrapure water and methanol as solvents. The solvent program consisted of 50 % of water and 50 % of methanol for 8 min, then a first gradient was applied to increase methanol to 70 % in 2 min, and a second one to reach 95 % of methanol in 3 min. The injection volume was 7 µL, and reverse phase separation was carried out on an Agilent Zorbax Eclipse-plus column (75 x 2.1 mm i.d., 1.8 μm). ESI was operated at 14 L/min gas flow, 250 °C gas temperature, 11 L/min sheath gas flow, 300 °C sheath gas temperature, 45 psig nebulizing, 300 V nozzle voltage, 3500 V capillary voltage, and 200 V fragmentor. MS was performed with an m/z range of 60 to 1000, with a scan rate of 1 spectrum/sec and a relative threshold of 0.01 %. Raw data gained from the time-of-flight analyzer were processed using MassHunter Qualitative Analysis B.06 software program (Agilent Technologies). A first identification of TCZ TPs was achieved by screening the raw chromatogram data against the TP library (Agilent MassHunter PCDL Manager B.04 software program), using a find-by-formula algorithm. Identification was done taking into account monoisotopic accurate mass (with a threshold of 5 ppm from the nominal), isotope spacing and ratio to gain an overall identification score. Compounds with scores higher than 85 % and plausible chromatogram shapes were chosen for further analysis by target UHPLC-ESI-QTOF-tandem MS. Detected TP precursor ions were included into

Chapter 3 Pesticide dissipation in agricultural soil 144 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology the list for targeted tandem MS, which was performed at low, middle and high collision energies (10, 20 and 40 V, respectively). Chromatography conditions, ESI and MS parameters were the same as for MS analysis, except that acquisition was done in MS/MS mode.

Supplementary Data 3.2.3. Typology for categorizing TCZ TPs. The organic compound clustering methodology ‘TyPol’ (Typology of Pollutants) (Servien et al., 2014) was used to categorize TPs. As no environmental parameters such as sorption coefficient, or degradation half-life were available for most of the compounds identified in our study, the typology was initially solely based on structural molecular descriptors such as the number of atoms in the molecule, molecular surface, dipole moment, or energy of the orbitals. For TCZ and each TP, 40 molecular descriptors were calculated using several software programs (ChemOffice Ultra 12.0, Dragon 5.5) and added to the TyPol database. TyPol clustering is based on partial least square regression (PLS) analysis. PLS regression is used to find the multidimensional directions of the maximum observable variables space and presents individuals as plots with two axis components. After PLS analysis, a hierarchical clustering algorithm was used to categorize TCZ TPs by grouping similar compounds into one cluster. The final number of clusters was chosen by comparing heights in the dendrogram, a statistical map resuming Ward clustering. The number of clusters was selected by plotting the heights of the dendrogram’s node and looking for a break. The results showed that the best choice (i.e. creating a number of clusters that minimized intra-variability and maximized inter- variability) was to classify the compounds into three clusters. The molecular descriptors of TCZ TPs were then compared to those of 116 pesticides and 19 TPs for which environmental parameters such as sorption coefficient or degradation half-life are already known. These comparisons allowed us to propose a hypothetical environmental behavior for TCZ TPs according to their position in the three clusters.

Chapter 3 Pesticide dissipation in agricultural soil 145 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

(a)

(b)

(c)

Supplementary Figure 3.2.1. Chromatogram examples of TPs with the molecular formula C16H20ClN3O2. (a) MS chromatogram, total ion current (TIC), (b) MS chromatogram, extracted ion current (EIC), (c) tandem MS chromatogram.

Chapter 3 Pesticide dissipation in agricultural soil 146 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

precursor: precursor:

TP 13 TP 15 product ions: product ions:

Supplementary Figure 3.2.2. MSC of the constitutional isomers TP 13 and 15.

Chapter 3 Pesticide dissipation in agricultural soil 147 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Supplementary Table 3.2.2. TyPol clustering results for TCZ and its TPs: Median values of molecular descriptors and median values of environmental parameters estimated from compounds with similar molecular descriptors and known environmental parameters.

Chapter 3 Pesticide dissipation in agricultural soil 148 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Supplementary Table 3.2.3. TyPol clustering of the 116 pesticides and 19 TPs into six clusters

(Pesticide-clusters 1 to 6) according to their molecular descriptors and KOC: median values of molecular descriptors and median values of KOC. The list of compounds in each cluster is indicated in Table 3.2.4. The comparison of these median molecular descriptor values to those of TCZ and its TPs (Supplementary Table 3.2.2) showed that cluster 1 can be assimilated to Pesticide-cluster 4, cluster 2 to Pesticide-cluster 5, and cluster 3 to Pesticide-cluster 1.

Chapter 3 Pesticide dissipation in agricultural soil 149 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Chapter 3 Pesticide dissipation in agricultural soil 150 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Supplementary Figure 3.2.3. TyPol clustering of the 116 pesticides and 19 TPs into six clusters on the two main components of the PLS (PLS1 and PLS2): Example of clustering according to their molecular descriptors and KOC values.

Chapter 3 Pesticide dissipation in agricultural soil 151 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Supplementary Table 3.2.4. TyPol clustering of the 116 pesticides and 19 TPs into six clusters

(pesticide-clusters 1 to 6): example of clustering according to their molecular descriptors and KOC values (see Supplementary Table 3.2.3 and Supplementary Figure 3.2.3). TPs are indicated in italics.

Chapter 3 Pesticide dissipation in agricultural soil 152 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology References

Bayer CropScience, 2011. Folicur® EW250 - directions for use. http://www.bayercropscience.ie/manual/Folicur_250_EW.pdf. Börner, H., Schlüter, K. and Aumann, J. (2009). Pflanzenkrankheiten und Pflanzenschutz. Springer, Berlin Heidelberg. Boxall, A.B.A., Sinclair, C.J., Fenner, K., Kolpin, D. and Maund, S.J. (2004). Peer reviewed: When synthetic chemicals degrade in the environment. Environmental Science and Technology 38 (19), 368-375. Bozdogan, A.M. (2014). Assessment of total risk on non-target organisms in fungicide application for agricultural sustainability. Sustainability 6, 1046-1058. Calza, P., Baudino, S., Aigotti, R., Baiocchi, C., Branca, P. and Pelizzetti, E. (2002). High-performance liquid chromatographic/tandem mass spectrometric identification of the phototransformation products of tebuconazole on titanium dioxide. Journal of Mass Spectrometry 37, 566-576. EFSA (2008). Conclusion on the peer review of tebuconazole. EFSA Scientific Report 176, 1-109. EFSA (2014). Conclusion on the peer review of the pesticide risk assessment of the active substance tebuconazole. EFSA Journal 12 (1): 3485, 98pp. doi:10.2903/j.efsa.2014.3485. EU-Air Framework Directive 2008/50/EC. Directive 2008/50/EC of the European Parliament and of the Council on ambient air quality and cleaner air for Europe. EU-Regulation 1107/2009/EC (2009). Regulation (EC) no 1107/2009 of the European parliament and of the council of 21 October 2009 concerning the placing of plant protection products on the market and repealing council directives 79/117/EEC and 91/414/EEC. Official Journal of the European Union 309, 1. EU-Water Framework Directive 2000/60/EC. Directive 2000/60/EC of the European Parliament and of the Council establishing a framework for community action in the field of water policy. FAO (1994). Pesticide evaluation tebuconazole 188. Food and Agriculture Organization of the United Nations Report of the Joint FAO/WHO Meeting on Pesticide Residues (JMPR). FAO, Rome. Fenner, K., Canonica, S., Wackett, L.P. and Elsner, M. (2013). Evaluating pesticide degradation in the environment: Blind spots and emerging opportunities. Science 341, 752-758. FOCUS (2008). Pesticides in air: Considerations for exposure assessment. Report of the FOCUS Working Group on Pesticides in Air, EC Document Reference SANCO/10553/2006 Rev 2 June 2008. 327 pp. Giacomazzi, S. and Cochet, N. (2004). Environmental impact of diuron transformation: a review. Chemosphere 56, 1021-1032. Herrero-Hernandez, E., Andrades, M. S., Marin-Benito, J. M., Sanchez-Martin, M. J. and Rodriguez-Cruz, M.S. (2011). Field-scale dissipation of tebuconazole in a vineyard soil amended with spent mushroom substrate and its potential environmental impact. Ecotoxicology and Environmental Safety 74, 1480-1488. Hill, A. W. and Mortishire-Smith R. J. (2005). Automated assignment of high-resolution collisionally activated dissociation mass spectra using a systematic bond disconnection approach. Rapid Communications in Mass Spectrometry 19, 3111-3118. Horn, W., Jann, O. and Wilke, O. (2003). Suitability of small environmental chambers to test the emission of biocides from treated materials into the air. Atmospheric Environment 37, 5477-5483. Hutzinger, O. (2013). The handbook of environmental chemistry – anthropogenic compounds. 3rd ed., Springer, Berlin Heidelberg. Ibanez, M., Sancho, J.V., Pozo, O.J., Niessen, W. and Hernandez, F. (2005). Use of quadrupole time-of-flight mass spectrometry in the elucidation of unknown compounds present in environmental water. Rapid Communications in Mass Spectrometry 19, 169-178. ISO 10381-6 (2009). Soil quality - Sampling - Part 6: Guidance on the collection, handling and storage of soil under aerobic conditions for the assessment of microbiological processes, biomass and diversity in the laboratory. International Organization for Standardization (ISO). Lewis, S.L. and Maslin, M.A. (2015). Defining the Anthropocene. Nature 519, 171-180. Mamy, L., Patureau, D., Barriuso, E., Bedos, C., Bessac, F., Louchart, X., Martin-Laurent, F., Miège, C. and Benoit, P. (2015). Prediction of the fate of organic compounds in the environment from their molecular properties: A review. Critical Reviews in Environmental Science and Technology 45, 1277-1377. McCall, P.J., Swann, R.L., Laskowski, D.A., Unger, S.M., Vrona, S.A. and Dishburger, H.J. (1980). Estimation of chemical mobility in soil from liquid chromatographic retention times. Bulletin of Environmental Contamination and Toxicology 24, 190-195. Obanda, D.N. and Shupe, T.F. (2009). Biotransformation of tebuconazole by microorganisms: Evidence of a common mechanism. Wood and Fiber Science 41 (2), 157-167. OECD-Guideline ENV/JM/MONO (2007)17 (2007). Guidance document on pesticide residue analytical methods. Series on pesticides no 39. Series on testing and assessment no 72, p. 34.Potter, T.L., Strickland, T.C., Joo, H. and Culbreath, A.K. (2005). Accelerated soil dissipation of tebuconazole following multiple applications to peanut. Journal of Environmental Quality 34, 1205-1213. Reitzel, L.A., Dalsgaard, P.W., Muller, I.B. and Cornett, C. (2012). Identification of ten new designer drugs by GC-MS, UPLC-QTOF-MS, and NMR as part of a police investigation of a Danish internet company. Drug Testing and Analysis 4 (5), 342-254. Rieke, S., Koehn, S., Hirsch-Ernst, K., Pfeil, R., Kneuer, C. and Marx-Stoelting, P. (2014). Combination effects of (tri)azole fungicides on hormone production and xenobiotic metabolism in a human placental cell line. International Journal of Environmental Research and Public Health 11, 9660-9679. Schermerhorn, P.G., Golden, P.E., Krynitsky, A.J. and Leimkuehler, W.M. (2005). Determination of 22 triazole copounds including parent fungicides and metabolites in apples, peaches, flour and water by liquid chromatography/tandem mass spectrometry. Journal of AOAC International 88 (5), 1491-1502. Schymanski, E.L., Jeon, J., Gulde, R., Fenner, K., Ruff, M., Singer, H.P. and Hollender, J. (2014). Identifying small molecules via high resolution mass spectrometry: Communicating confidence. Environmental Science and Technology 48, 2097-2098. Sehnem, N., Souza-Cruz, P., Peralba, M. and Zachia Ayub, M. (2010). Biodegradation of tebuconazole by bacteria isolated from contaminated soils. Journal of Environmental Science and Health Part B 45, 67-72. Servien, R., Mamy, L., Li, Z., Rossard, V., Latrille, E., Bessac, F., Patureau, D. and Benoit, P. (2014). TyPol – A new methodology for organic compounds clustering based on their molecular characteristics and environmental behavior. Chemosphere 111, 613-622. Shalini, K., Kumar, N., Drabu, S. and Kumar Sharma, P. (2011). Advances in synthetic approach to and antifungal activity of triazoles. Beilstein Journal of Organic Chemistry 7, 668-677.

Chapter 3 Pesticide dissipation in agricultural soil 153 Chapter 3.2 Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology

Sinclair, C., Van Beinum, W., Adams, C., Bevan, R., Levy, L., Parsons, S., Goslan, E. and Baumann, G. (2010). A desk study on pesticide metabolites, degradation and reaction products to inform the inspectorate’s position on monitoring requirements. Final report for drinking water inspectorate. The Food and Environment Research Agency (Fera), Fera Project: S3VB, 478 pp. Strickland, T.C., Potter, T.L. and Hyun, J. (2004). Tebuconazole dissipation and metabolism in Tifton loamy sand during laboratory incubation. Pest Management Science 60, 703-709. Tixier, C., Bogaerts, P., Sancelme, M., Bonnemoy, F., Twagilimana, L., Cuer, A., Bohatier, J. and Veschambre, H. (2000). Fungal biodegradation of a phenylurea herbicide, diuron: structure and toxicity of metabolites. Pest Management Science 56, 455- 462. U.S. EPA (1990). Tebuconazole Memorandum. U.S. Environmental Protection Agency (U.S. EPA), Washington, DC. U.S. EPA (2011). Pesticides industry sales and usage - 2006 and 2007 market estimates. U.S. Environmental Protection Agency (U.S. EPA), Washington, DC. U.S. EPA (2015). Technical overview of ecological risk assessment. Analysis phase: exposure characterization - fate and transport of pesticides. U.S. Environmental Protection Agency (U.S. EPA), Washington, DC. Van-Camp, L., Bujarrabal, B., Gentile, A.-R., Jones, R.J.A., Montanarella, L., Olazabal, C. and Selvaradjou, S.-K. (2004). Reports of the Technical Working Groups Established under the Thematic Strategy for Soil Protection. EUR 21319 EN/4, 872 pp. Office for Official Publications of the European Communities, Luxembourg. Wang, X., Wang, X., Zhang, H., Wu, C., Wang, X., Xu, H., Wang, X. and Li, Z. (2012). Enantioselective degradation of tebuconazole in cabbage, cucumber and soils. Chirality 24, 104-111. White, P.M., Potter, T.L. and Culbreath, A.K. (2010). Fungicide dissipation and impact on metolachlor aerobic soil degradation and soil microbial dynamics. Science of the Total Environment 408, 1393-1402. Woo, C., Daniels, B., Stirling, R. and Morris, P. (2010). Tebuconazole and propiconazole tolerance and possible degradation by Basidiomycetes: A wood-based bioassay. International Biodeterioration and Biodegradation 64, 403-408. Zeng, W., Han, H., Tao, Y., Yang, L., Wang, Z. and Chen, K. (2013). Identification of bio-active metabolites of gentiopicroside by UPLC/Q-TOF MS and NMR. Biomedical Chromatography 27, 1129-1136.

Chapter 3 Pesticide dissipation in agricultural soil 154

Conclusion

Chapter 3 provides a set of data describing the main processes involved in the fate of the three studied pesticides in soil of a North-Italian agricultural field. It provides a profound evaluation of the exposure scenario of soil microorganisms to residues of the three tested pesticides. As these three pesticides are still in use in Europe, the results may enrich their future post-authorization risk assessments that are due in 2017 (IPU), 2020 (CHL) and 2023 (TCZ). The approach presented in Chapter 3.1 was able to quantify the fate of pesticides and their main known TPs (of which reference standards were available) in dissipation and sorption studies. This constitutes an important dataset to define the exposure scenario of soil microorganisms to pesticide residues. In addition, the application of a new approach combining suspect screening to in silico molecular typology was found to be suitable to detect and categorize known as well as unknown TCZ TPs in soil (Chapter 3.2). One major force of this approach is that reference standards are not compulsory necessary (albeit preferable, as they are needed for the identity confirmation and quantification of detected transformation products), and thus a comprehensive inventory of existent TPs in a complex environmental matrix, such as agricultural soil, is enabled. In addition, in silico molecular typology allowed to categorize TPs according to their physicochemical properties and to estimate their environmental properties, which are key elements to be considered to deduce their ecotoxicity. This information may be used to choose potentially relevant TPs among the numerous TCZ TPs, which may be synthesized to assess their environmental fate and their ecotoxicological impact on soil microorganisms in order to verify the hypotheses resulting from the application of this combined approach.

Chapter 3 Pesticide dissipation in agricultural soil 156

Chapter 4

Ecotoxicological impact of pesticides on soil microorganisms

Veronika Storck Chiara Perruchon Nadine Rouard Giorgia Pertile Jérémie Béguet Céline Baguelin Federico Ferrari Olivier Sibourg Evangelia S. Papadopoulou Cedric Malandain Sofia Nikolaki George Tsiamis Panagiotis A. Karas Dimitrios G. Karpouzas David Bru Marco Trevisan Lucio Botteri Marion Devers-Lamrani Aymé Spor Fabrice Martin-Laurent

Foreword

Being involved in complex processes with impacts at the global scale, microorganisms are key players in a range of soil ecosystem services such as primary production, climate regulation and water purification. Agricultural practices, including the application of pesticides, have been identified as a major threat to soil ecosystem functioning. Indeed, a range of studies have shown that exposure of non-target organisms such as soil microorganisms to pesticides can cause ecotoxicological effects, with unknown consequences for soil functioning on the long term. Knowledge of ecotoxicological effects of pesticides on soil microorganisms is thus crucial to approach more sustainable agricultural practices. At EU-level, the current evaluation of ecotoxicological pesticide effects on soil microorganisms solely relies on global carbon and nitrogen mineralization tests that are are weak indicators for the microbial toxicity of a pesticide (as potential effects tend to remain undetected). Therefore, there is a need to develop and implement innovative methods to evaluate the ecotoxicological risk of pesticides on microorganisms supporting soil ecosystem functions. With the research of Chapter 4, I aimed to test up-to-date metagenomic methods and alternative radiorespirometric techniques to monitor the ecotoxicological impact of each of the three model pesticides on the diversity and activity of soil microorganisms. In Chapter 4.1, the ecotoxicological effects of CHL, IPU or TCZ (applied once at doses simulating a lower tier exposure scenario) on the diversity of the soil bacterial community were assessed by Illumina next-generation sequencing of 16S rDNA amplicons produced from DNA extracts of soil samples of the lab-to-field experiment (described in Chapter 3). In Chapter 4.2, the potential effect of repeated pesticide treatments (CHL or TCZ at 10x agronomical doses simulating a higher tier exposure scenario) on soil microorganisms was investigated. I studied the evolution of (i) the ability of the soil microbial community to mineralize the applied pesticide, (ii) the capacity of soil microbial communities to mineralize various other organic compounds as a proxy of soil buffering functions, and (iii) the diversity of the soil bacterial community by Illumina next-generation sequencing of 16S rDNA amplicons.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 160

Chapter 4.1

Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Veronika Storck Olivier Sibourg Sofia Nikolaki Cedric Malandain Chiara Perruchon Marco Trevisan Giorgia Pertile Federico Ferrari Céline Baguelin Dimitrios G. Karpouzas Panagiotis A. Karas George Tsiamis Aymé Spor Fabrice Martin-Laurent

Manuscript in preparation for publication.

Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing Abstract

Pesticides are intentionally applied to agricultural fields for crop protection. They can harm non-target organisms such as soil microorganisms involved in important ecosystem functions with impacts at the global scale. During the pesticide authorization process, the ecotoxicological impact of pesticides on soil microorganisms is only estimated using carbon and nitrogen mineralization tests, despite the availability of more extensive approaches analyzing the abundance, activity or diversity of soil microorganisms. In this study, we evaluate the possible interest of using 16S rDNA next-generation sequencing to estimate the impact of either the organophosphate insecticide chlorpyrifos (CHL), the phenyl-urea herbicide isoproturon (IPU) or the triazole fungicide tebuconazole (TCZ) on the diversity and composition of the soil bacterial community. To our knowledge, it is the first time that this metagenomic approach is applied to assess the impact of the three pesticides in a lab-to-field experimental design. The bacterial diversity and composition were found to vary over time and (as expected), this trend was more marked in the microcosm than in the field study. Only slight but significant transient effects of CHL or TCZ were observed in the microcosm and the field study, respectively. IPU was not found to significantly modify the soil bacterial diversity or composition. Our results are in accordance with EFSA conclusions indicating that these three pesticides may have a low risk towards soil microorganisms.

Figure 4.1.1. Graphical abstract of Chapter 4.1.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 164

Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing Introduction

Soil is a limited and largely non-renewable resource providing a unique and complex habitat for a wide range of microorganisms supporting soil ecosystem functions contributing to complex processes with impacts at the global scale (Blum, 2006, Bondeau et al., 2007, Graham et al., 2016). For instance, soil microorganisms contribute to food supply, water quality, carbon cycling, and climate regulation (Haygarth and Ritz, 2009; Hillel, 2009). The annual economic value of soil-based ecosystem functions is estimated to be around 1.3 trillion Euros (Pimentel et al., 1997). In this context, the preservation of soil quality by sustainable agriculture was identified to be one of the greatest objectives of the 21st century (Jones et al., 2009; Lal, 2008; Lichtfouse et al., 2009). To achieve this objective, the main pressures on soil organisms were weighted by the European soil biodiversity expert group, ranking human intensive exploitation on place one (Gardi et al., 2013; Jeffery et al., 2010), to which pesticide effects may contribute. Pesticides are intentionally applied in conventional agriculture to protect crops from various pests. They can persist in soil from where they can be dispersed by different means leading to the contamination of surrounding environments (Barbash, 2003; Looser et al., 2000; Smalling et al., 2013) where they can harm non-target organisms (Pimentel, 1995; Zhang et al., 2011). The evaluation of ecotoxicological effects of pesticides on soil microorganisms remains difficult, as research results and their complex interpretation often differ from one study to another. Effects of the organophosphate insecticide chlorpyrifos (CHL), the phenyl-urea herbicide isoproturon (IPU) and the triazole fungicide tebuconazole (TCZ) are particularly interesting to study as these pesticides are on the market for 51 (CHL), 41 (IPU) and 28 (TCZ) years and thus they are widely present in the environment worldwide (Börner et al., 2009; Hayes and Laws, 1991; Oerke et al., 1994). However, in respective risk assessment documents, their impacts on soil microorganisms are only reviewed based on the sole estimation of their impact on carbon and nitrogen mineralization by soil microorganisms. For CHL, the review report of the European Commission concludes that CHL has no significant effects on carbon and nitrogen mineralization in soil (EC, 2005). The effects of IPU and TCZ on carbon and nitrogen mineralization in soil were < 25 % for both pesticides and were interpreted as a low risk for soil microorganisms (EFSA, 2008; EFSA, 2014b; EFSA, 2015a). However, carbon and

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 166 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing nitrogen mineralization tests are global indicators which have been criticized for their use to estimate the microbial toxicity of pesticides because these processes can be maintained regardless of significant toxic effects, as resistant microbes can increase their population at the expense of sensitive microbes, due to the redundancy of microorganisms contributing to carbon and nitrogen mineralization functions within the soil microbial community (Van Beelen and Doelman, 1997). Moreover, another source of criticism relies on the fact that pesticide risk assessment for approval at EU level is based on the sole active substance, and not on the formulated plant protection products (PPPs) (containing the active substance and other chemicals), which are broadly used in agriculture and can be more toxic than the active substance alone (EFSA, 2015b). Several studies have investigated the impact of CHL, IPU and TCZ on soil microorganisms more comprehensively throughout the analysis of the microbial biomass, diversity or activity in addition to carbon and nitrogen mineralization tests. Studies on CHL effects were mostly performed in soil microcosms (Chu et al., 2008; Dutta et al., 2010; Fang et al., 2009; Gilani et al., 2010; Greaves and Shales, 1991; Jastrzebska, 2011; Pozo et al., 1995; Singh et al., 2002; Srinivasulu and Rangaswamy, 2013; Wang et al., 2010) and more rarely in field studies (Eisenhauer et al., 2009; Gupta et al., 2013). Most of these studies used CHL in a formulated PPP (Chu et al., 2008; Eisenhauer et al., 2009; Gilani et al., 2010; Gupta et al., 2013; Jastrzebska, 2011; Pozo et al., 1995; Singh et al., 2002; Srinivasulu and Rangaswamy, 2013) rather than as sole active substance (Dutta et al., 2010; Fang et al., 2009; Wang et al., 2010). One study concluded small, insignificant, or no CHL effects on soil microorganisms (Singh et al., 2002) confirming the conclusions of the risk assessment of regulatory authorities. Other studies reported effects of CHL on microbial biomass, respiration, enzyme activities, microbial diversity or activity in soil (Chu et al., 2008; Dutta et al., 2010; Eisenhauer et al., 2009; Fang et al., 2009; Gilani et al., 2010; Gupta et al., 2013; Jastrzebska, 2011; Pozo et al., 1995; Srinivasulu and Rangaswamy, 2013; Wang et al., 2010). It is noteworthy that several studies showed CHL exposure to cause short-term inhibitory effects within the first month after treatment, which were recovered in subsequent months, suggesting the potential recovery of soil microbial communities (Fang et al., 2009; Pozo et al., 1995). Similarly, most of the studies evaluating the effects of IPU on soil microorganisms were performed in microcosms (Harden et al., 1993; Kuriyal and Pandey, 1994; Perrin- Ganier et al, 2001; Tag-El-Din, 1982) and only a few in field studies (Schuster and

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 167 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing Schröder, 1990). These studies were mainly conducted using formulated PPPs (Harden et al., 1993; Kuriyal and Pandey, 1994; Schuster and Schröder, 1990; Tag-El-Din, 1982) and scarcely with the sole active substance (Perrin-Ganier et al., 2001). The studies concluded small to moderate (sometimes temporary) toxic effects of IPU to soil microbial biomass and activity. Interestingly, in response to IPU exposure, an initial increase in soil respiration was observed (Harden et al., 1993; Perrin-Ganier et al., 2001; Schuster and Schröder, 1990). It was hypothesized that this may be due to a toxic effect of IPU on sensitive soil microorganisms, which may have provided an additional easily degradable carbon source for the respiration of microorganisms resistant to IPU. A number of studies evaluating the ecotoxicological impact of TCZ on soil microorganisms were performed in soil microcosms using the formulated PPP (Anuradha et al., 2016; Bending et al., 2007; Cycon et al., 2006) or the active substance (Ferreira et al., 2009; Munoz-Leoz et al., 2011; Wang et al., 2016). All studies concluded TCZ effects on soil microorganisms, observed by different means including the analysis of soil enzyme activities, microbial biomass, carbon and nitrogen mineralization, and others. In the same manner as for CHL and IPU, strongest TCZ effects were observed during the first month after treatment (Ferreira et al., 2009; Munoz-Leoz et al., 2011). This recovery of soil microbial communities may be facilitated by the decrease of pesticide exposure due to its dissipation by abiotic and biotic processes (EFSA, 2016). The aim of this study was to test the interest of 16S rDNA amplicon next- generation sequencing to assess the ecotoxicological impact of CHL, IPU or TCZ on the soil microbial diversity in a lab-to field approach with tiered pesticide exposure scenarios. For this purpose, we established a microcosm and a field experiment consisting in applying each of the three formulated pesticides at different doses (0x, 1x, 2x or 10x (microcosm study) and 0x, 1x, 2x or 5x (field study) the recommended dose). The ecotoxicological impact of each pesticide on the diversity of the soil bacterial community was assessed during three months of pesticide exposure by next-generation sequencing of 16S rDNA amplicons (generated from soil DNA extracts). The dissipation of CHL, IPU or TCZ and their main transformation products during these pesticide exposure scenarios had previously been determined by Papadopoulou et al. (2016). To our best knowledge, this is the first time that the ecotoxicological impact of CHL, IPU and TCZ on the soil bacterial diversity is estimated by this approach.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 168 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing Material and methods

Microcosm and field study

The studied site was an agricultural field situated in North Italy (45°05'20.8"N 9°45'59.4"E, Google Maps), an area intensively exploited by agriculture where winter cereal is a representative crop. The soil was characterized as loamy sand (4.2 % clay, 13.5 % silt, 82.2 % sand) with an organic carbon content of 1.5 %, a microbial biomass of 160.2 mg C/kg soil dwt and a pH of 7.5. The field had not been treated with the studied pesticides for more than five years. For the microcosm study, topsoil samples (10 cm depth) were taken following procedures described by the International Standardization Organization (ISO) for collection and handling of soil samples (ISO 10381-6, 2009) in July 2013. The soil was manually homogenized, partially air-dried, sieved (2 mm) and stored at 4 °C for approximately one week before the final application of the pesticides. For each pesticide and dose, three replicates were treated with an aqueous solution of CHL, IPU or TCZ (prepared from their commercial formulations Carposan® (CHL), Quintil® (IPU) and Folicur® (TCZ)) at 1x, 2x or 10x the recommended dose (2.0 (CHL), 1.9 (IPU) and 0.6 (TCZ) mg/kg soil dwt). Three control sub-samples were treated with water. The water content of all subsamples was adjusted to 40 % of the water holding capacity (WHC = 46.8 wt%), which was kept constant throughout the whole study. 150 g sub- samples were placed in aerated plastic bags and incubated in the dark at 20 °C. Immediately (0 days) and 7, 42, 56, and 100 days after pesticide application, soil was sampled and stored at -80 °C until DNA extraction. In the field study, three replicates of 6 m2 field plots were treated with 1x, 2x or 5x the recommended dose of each pesticide in November 2013. Three replicate control plots were treated with water. Soil samples were taken at 0, 14, 35, 70 and 105 days after treatment, following the same procedure as for the microcosm study (ISO 10381-6, 2009). Sampled soil was manually homogenized, sieved (2 mm) and kept at -80 °C until DNA extraction.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 169 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing DNA extraction

DNA was extracted and isolated from 300 mg soil samples following the ISO 11063 procedure (2012), derived from the method described by Martin-Laurent et al. (2001). The procedure involved three principal steps: (i) microbial cell lysis by physical (bead beating) and chemical (sodium dodecyl sulfate) actions, (ii) deproteination by precipitation and (iii) DNA precipitation, washing and purification.

Illumina next-generation sequencing

150 microcosm and 150 field study DNA samples (5 time points x (3 controls + 3 pesticides x 3 doses x 3 replicates) = 150) were analyzed using an amplicon Illumina next-generation sequencing approach. In a first-step PCR, fusion primers U341F (5’- CCTACGGGRSGCAGCAG-3’) and 805R (5’-GTGCCAGCMGCCGCGGTAA-3’) (Klindworth et al., 2012) were used to amplify the hypervariable V3-V4 region of the bacterial 16S rDNA (around 460 bp). Primer sequences contained corresponding overhang adapters (forward adapter: TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG, reverse adapter: GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG) needed to add multiplexing index- sequences in a second-step PCR. Concerning the microcosm study, each PCR reaction was carried out in a volume of 25 μL containing 1 μL of template DNA, 5 μL Kapa 5x High-Fidelity Hot Start Buffer, 0.3 μL of each primer (25 μM), 0.75 μL dNTPs (10mM), High-Fidelity DNA Kapa polymerase (1 unit/μL) and 17.15 μL water. PCR reactions were performed using a PTC- 200 thermocycler (MJ Research Inc., USA). Cycling conditions were 5 min at 95 °C, 30 cycles at 98 °C for 20 sec, 60 °C for 30 sec and 72 °C for 45 sec, followed by a final extension of 5 min at 72 °C. The size of the PCR products was visually checked with agarose gel electrophoresis using appropriate size markers. The PCR products were precipitated by polyethyleneglycol (20 % PEG, 2.5 M NaCl) (Hartley and Bowen, 2003) and resuspended in 15 μL water. Concentrations of the PCR products were determined with a Nanodrop Quawell UV-VIS spectrophotometer (Q5000). The PCR amplicons were then used as templates in a second-step PCR for further amplification and for addition of the multiplexing index-sequences to the overhang adapters. Amplicons were purified with AMPure XP beads and re-suspended in 30 μL water. Products were quantified using

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 170 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing the Qubit dsDNA High-Sensitivity assay (Life Technologies), and an Agilent Bioanalyzer High-Sensitivity DNA chip (Agilent). Amplicons were pooled at equimolar concentrations and the library was quantified using a Qubit Fluorometer and dsDNA HS Assay kit (Life Technologies, CA, USA). Samples were pooled at equimolar concentrations for size-selection by Pippin-Prep (Sage Science), where a size range of 350 to 450 bp was extracted. Reaction negatives were included in the pool. Size-selected fragments were purified and quantified in the same manner as described above prior to sequencing. Sequencing was performed at IMGM SA on an Illumina MiSeq platform using 2x 300 bp paired-end read chemistry (Illumina). Concerning the field study, first-step PCR amplifications were carried out in 50 μL reaction volumes containing 2.5 U of Accuzyme polymerase (Bioline), 0.2 μM of each primer, 1x AccuBuffer with 2 mM Mg2+, 0.5 mM of each dNTP, and 0.4 μg/μL of bovine serum albumin (BSA, Sigma) (Kreader, 1996). The thermal cycling conditions were 95 °C for 5 min, followed by 35 cycles of 95 °C for 15 sec, 58 °C for 15 sec and 72 °C for 1 min, with a final extention of 72 °C for 5 min. DNA samples were brought to 0.5 ng/µL before adding 4 µL to the PCR reaction. The PCR products were visualized on a 1.5 % agarose gel in order to verify the correct size of amplicons (around 560 bp). Amplicon sizes of randomly selected samples were further analyzed by Bioanalyzer DNA 1000 chip (Agilent Technologies). The amplicons were cleaned-up with a Nucleospin gel and PCR clean-up kit (Macherey-Nagel) following the manufacturer’s instructions. Subsequently, amplicon concentrations were determined with NanoDrop 2000c (Thermo Scientific). For the addition of multiplexing index-sequences to the overhang adapters, a second- step PCR was performed using a 384 libraries Nextera XT index kit v2 (Illumina). Second-step PCR amplifications were carried out in 50 μL reaction volumes containing 1x Titanium Taq polymerase (Clontech), 1x Titanium Taq PCR buffer, 0.2 mM of each dNTP, 5 μL of each Nextera XT index primer using the TrutSeq Index Plate Fixture (Illumina). All cleaned-up amplicons from the first-step PCR were diluted to 10 ng/μL and added to the second-step PCR reaction mix at a final concentration of 1 ng/μL. Thermal cycling conditions were 96 °C for 3 min, followed by 8 cycles of 95 °C for 30 sec, 55 °C for 30 sec and 72 °C for 30 sec, with a final extension of 72 °C for 5 min. Amplicon sizes (around 630 bp) of randomly selected sampled were then analyzed by Bioanalyzer DNA 1000 chip (Agilent Technologies). The libraries were cleaned-up before quantification using AMPure XP beads (Beckman-Coulter Genomics) following the

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 171 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing manufacturer’s instructions. Amplicon concentrations were then fluorometrically measured with Qubit (Invitrogen) and converted to nM, assuming an average library size of 630 bp: Amplicon concentration in ng/µL x (660 g/mol x 630 bp) x 106 = amplicon concentration in nM. Libraries with amplicon concentrations >40 nM were first diluted to 40 nM and then all samples were brought to 4 nM. For pooling of the libraries, 2 µL of each sample were brought together and denatured with 0.2 N NaOH and diluted to 4 pM in hybridization buffer HT1, following the Illumina manufacturer’s instructions. The pooled amplicon library was then combined with a PhiX control library (Illumina) before the Illumina Miseq run (2x 301 bp) was performed.

Illumina sequencing data analysis

The raw data sequences of Illumina next-generation sequencing of the microcosms and the field study were analyzed in a similar manner. Reads were de- multiplexed and converted to FASTQ format. For the microcosm study, Cutadapt 1.2.1 (Martin, 2011) was used to trim Illumina adapter sequences from FASTQ files. Reads were trimmed if 3 bp or more of the 3’ end of a read matched the adapter sequence. Sickle version 1.200 (Joshi and Fass, 2011) was used to trim reads based on quality: Any read with a quality score of < 20, or < 10 bp size after trimming were discarded. Paired- end reads were assembled, trimmed by length and further corrected using PandaSeq (Masella et al., 2012). For the field study, reads were quality-controlled and assembled using PEAR (Zhang and Sun, 2014). Unassembled reads and once-assembled reads outside the expected range were discarded. All subsequent analyses were conducted in QIIME 1.9.0 (Caporaso et al., 2010a). Sequences were clustered into operational taxonomic units (OTUs) using USEARCH (Edgar, 2010) by open-reference OTU picking. Chimeras were detected and omitted using the program UCHIME (Edgar et al., 2011) with the QIIME-compatible version of the SILVA aligned version of the Gold database (Quast et al., 2013) for the microcosm study and with the Greengenes version of the Gold database for the field study. Taxonomy was assigned to representative sequences using the SILVA 111 release database (Quast et al., 2013) for the microcosm study and with the Greengenes 13.8 release for the field study. Representative sequences were aligned using the SILVA 111 core reference alignment using PyNAST (Caporaso et al., 2010b). In

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 172 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing total, 4,414,311 (microcosm study) and 5,159,912 (field study) sequences in 150 samples each were analyzed using QIIME. Several α-diversity indices, as well as indices depicting the population structure, were calculated with the QIIME pipeline (Caporaso et al., 2010a) based on the rarefied OTU table at a depth of 5000 sequences/sample for the microcosm and the field study (observed species, PD whole tree, chao1 and simpson reciprocal). The indices were compared using analyses of variances (ANOVAs) followed by the Tukey HSD test (p < 0.05). Potential phylum composition differences were visualized using a comparative bar chart. Phyla whose total relative abundances in all samples were < 0.05 % were excluded from the analysis. Principal coordinate analysis (PCoA) of weighted unifrac distance matrices was performed and plotted. ANOSIM analyses were performed to identify potentially significant differences at the community level between treatments of the same sampling day. Significantly variating OTUs were identified using the R package ‘pamR’. ANOVA on OTUs were followed by the Bonferroni test grouping treatments per each OTU.

Results

Estimation of the impact of CHL, IPU or TCZ on α-diversity indices of the soil bacterial community

Analysis of the α-diversity indices ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’ indicated that the soil bacterial diversity changed in response to time but not in response to pesticide exposure in both the microcosm and the field experiment (Table 4.1.1 and 4.1.2, Supplementary Figures 4.1.1 to 4.1.6). In the soil microcosm study, diversity indices significantly decreased with time (p < 0.05) (Table 4.1.1, Supplementary Figures 4.1.1 to 4.1.3). Regardless the applied pesticide or dose, no significant differences between control (0x) and pesticide-treated (1x, 2x, 10x) soil microcosms were recorded within equal time points (Table 4.1.1, Supplementary Figures 4.1.1 to 4.1.3), indicating that none of the three pesticides significantly modified the α-diversity of the soil bacterial community.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 173 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing Time effects in the field study were less marked than in the microcosm study (Table 4.1.2, Supplementary Figures 4.1.4 to 4.1.6). Similarly to the microcosm study, no significant differences between control (0x) and pesticide-treated (1x, 2x, 5x) soil microcosms were recorded within the same time point (with one exception: PD whole tree, control versus IPU 2x, 14 days), suggesting that none of the three pesticides significantly modified the α-diversity of the soil bacterial community.

Table 4.1.1. Microcosm study - Bacterial α-diversity indices at various time points (0, 7, 42, 56 and 100 days after treatment) in soil exposed to each of the three studied pesticides (CHL, IPU or TCZ) applied at different doses (0x, 1x, 2x or 10x). Diversity indices of the bacterial community are 'observed species', 'PD whole tree', 'chao1' and 'simpson reciprocal'. For each diversity index, ANOVAs followed by the Tukey HSD test, (p < 0.05) were done by grouping (for each tested pesticide) the five sampling time points and the four pesticide doses. Significant differences are indicated by letters.

observed species PD whole tree 0d 7d 42d 56d 100d 0d 7d 42d 56d 100d 0x 2619a ± 57 2425ab ± 27 2365bc ± 28 2191bc ± 31 2188c ± 43 227a ± 5 211ab ± 2 212abc ± 2 191bc ± 2 198c ± 5 1x 2577a ± 77 2327ab ± 11 2268bc ± 40 2301bc ± 47 2139c ± 21 228a ± 2 214ab ± 8 210abc ± 2 206bc ± 5 188c ± 2 CHL 2x 2581a ± 60 2447ab ± 24 2282bc ± 19 2283bc ± 34 2131c ± 29 223a ± 7 213ab ± 5 202abc ± 2 205bc ± 2 191c ± 5 10x 2539a ± 91 2380ab ± 20 2360bc ± 35 2187bc ± 37 2157c ± 17 224a ± 2 214ab ± 9 204abc ± 5 196bc ± 5 193c ± 2 0x 2619a ± 57 2425ab ± 27 2365bc ± 28 2191bc ± 31 2188c ± 43 227a ± 5 211ab ± 2 212ab ± 2 191c ± 2 198bc ± 5 1x 2663a ± 41 2398ab ± 92 2386bc ± 35 2251bc ± 15 2200c ± 60 230a ± 7 212ab ± 12 216ab ± 4 198c ± 1 195bc ± 6 IPU 2x 2406a ± 241 2322ab ± 67 2272bc ± 30 2237bc ± 24 2166c ± 13 205a ± 23 204ab ± 8 204ab ± 3 192c ± 2 193bc ± 2 10x 2584a ± 115 2448ab ± 23 2353bc ± 25 2184bc ± 35 2148c ± 18 223a ± 9 217ab ± 2 212ab ± 3 191c ± 3 198bc ± 2 0x 2619a ± 57 2425ab ± 27 2365bc ± 28 2191c ± 31 2188c ± 43 227a ± 5 211ab ± 2 212bc ± 2 191c ± 2 198c ± 5 1x 2630a ± 23 2444ab ± 77 2323bc ± 21 2100c ± 26 2134c ± 32 216a ± 11 202ab ± 4 202bc ± 5 185c ± 4 191c ± 3 TCZ 2x 2575a ± 77 2403ab ± 35 2249bc ± 24 2226c ± 25 2172c ± 26 222a ± 6 216ab ± 2 202bc ± 2 191c ± 2 196c ± 1 10x 2624a ± 15 2412ab ± 89 2305bc ± 60 2279c ± 38 2118c ± 50 219a ± 5 210ab ± 4 210bc ± 3 201c ± 5 187c ± 6 chao1 simpson reciprocal 0d 7d 42d 56d 100d 0d 7d 42d 56d 100d 0x 6727a ± 483 6516a ± 173 6310a ± 173 5987a ± 139 6319a ± 416 774a ± 13 421b ± 16 440b ± 14 259b ± 14 299b ± 28 1x 6590a ± 528 5910a ± 189 6206a ± 337 7053a ± 486 5847a ± 144 731a ± 34 430b ± 31 329b ± 23 382b ± 22 351b ± 12 CHL 2x 6241a ± 414 6559a ± 201 5873a ± 188 6793a ± 183 5761a ± 209 758a ± 37 537b ± 12 420b ± 9 427b ± 18 357b ± 16 10x 6401a ± 249 6039a ± 266 6673a ± 273 5838a ± 408 5771a ± 165 589a ± 147 475b ± 30 354b ± 19 367b ± 9 336b ± 11 0x 6727a ± 483 6516a ± 173 6310a ± 173 5987a ± 139 6319a ± 416 774a ± 13 421b ± 16 440b ± 14 259b ± 14 299b ± 28 1x 6909a ± 359 6317a ± 608 6570a ± 355 6213a ± 161 5974a ± 313 817a ± 29 439b ± 16 411b ± 35 349b ± 74 346b ± 35 IPU 2x 5830a ± 1346 5745a ± 513 5703a ± 360 6174a ± 207 5714a ± 147 680a ± 21 423b ± 11 434b ± 13 379b ± 19 380b ± 16 10x 6606a ± 753 5745a ± 513 6458a ± 251 5938a ± 210 6129a ± 197 701a ± 53 457b ± 16 452b ± 13 335b ± 14 337b ± 10 0x 6727a ± 483 6516ab ± 173 6310ab ± 173 5987ab ± 139 6319b ± 416 774a ± 13 421b ± 16 440b ± 14 259b ± 14 299b ± 28 1x 6995a ± 179 6257ab ± 275 6029ab ± 115 5448ab ± 152 6082b ± 154 787a ± 29 540b ± 93 425b ± 29 309b ± 8 345b ± 12 TCZ 2x 6535a ± 610 6317ab ± 234 5681ab ± 269 6289ab ± 198 5898b ± 150 680a ± 93 447b ± 22 377b ± 10 345b ± 14 357b ± 11 10x 6956a ± 176 6221ab ± 610 6302ab ± 386 6815ab ± 330 5721b ± 273 742a ± 29 507b ± 22 407b ± 23 386b ± 15 343b ± 18

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 174 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Table 4.1.2. Field study - Bacterial α-diversity indices at various time points (0, 14, 35, 70 and 105 days after treatment) in soil exposed to each of the three studied pesticides (CHL, IPU or TCZ) applied at different doses (0x, 1x, 2x or 5x). Diversity indices of the bacterial community are 'observed species', 'PD whole tree', 'chao1' and 'simpson reciprocal'. For each diversity index, ANOVAs followed by the Tukey HSD test, (p < 0.05) were done by grouping (for each tested pesticide) the five sampling time points and the four pesticide doses. Significant differences are indicated by letters.

observed species PD whole tree 0d 14d 35d 70d 105d 0d 14d 35d 70d 105d 0x 1823abc ± 82 2224a ± 273 1997abc ± 109 1735abc ± 151 1862abc ± 81 137abc ± 5 180a ± 24 154abc ± 8 133bc ± 10 143abc ± 3 1x 1655bc ± 82 1924abc ± 55 2123ab ± 136 1913abc ± 68 1926abc ± 59 128bc ± 6 146abc ± 2 160ab ± 11 148abc ± 5 148abc ± 5 CHL 2x 1561c ± 95 1875abc ± 69 1866abc ± 91 1789abc ± 38 1851abc ± 20 121c ± 6 144abc ± 4 139abc ± 3 138abc ± 3 140abc ± 1 5x 1872abc ± 21 1983abc ± 93 2067abc ± 27 1662bc ± 83 1711abc ± 61 141abc ± 2 150abc ± 7 155abc ± 3 131bc ± 5 132bc ± 5 0x 1823ab ± 82 2224a ± 273 1997ab ± 109 1735ab ± 151 1862ab ± 81 137ab ± 5 180a ± 24 154ab ± 8 133ab ± 10 143ab ± 3 1x 1564b ± 82 1839ab ± 19 1974ab ± 104 1888ab ± 70 1843ab ± 93 122b ± 5 141ab ± 2 150ab ± 7 148ab ± 4 141ab ± 3 IPU 2x 1916ab ± 11 1755ab ± 136 1785ab ± 137 1794ab ± 104 1907ab ± 11 147ab ± 2 132b ± 9 139ab ± 8 142ab ± 6 144ab ± 1 5x 1890ab ± 117 1797ab ± 110 1828ab ± 84 1935ab ± 75 1890ab ± 67 143ab ± 8 135ab ± 7 142ab ± 4 150ab ± 5 142ab ± 3 0x 1823a ± 82 2224a ± 273 1997a ± 109 1735a ± 151 1862a ± 81 137ab ± 5 180a ± 24 154ab ± 8 133b ± 10 143ab ± 3 1x 1871a ± 71 1931a ± 107 1840a ± 98 1870a ± 41 1894a ± 18 145ab ± 5 144ab ± 7 144ab ± 7 143ab ± 3 142ab ± 1 TCZ 2x 2044a ± 26 1963a ± 41 1935a ± 124 1885a ± 49 1889a ± 52 156ab ± 1 148ab ± 2 151ab ± 9 145ab ± 2 143ab ± 2 5x 1816a ± 9 1869a ± 53 1924a ± 84 1807a ± 67 1780a ± 12 135b ± 2 142ab ± 3 147ab ± 6 141ab ± 3 136b ± 3 chao1 simpson reciprocal 0d 14d 35d 70d 105d 0d 14d 35d 70d 105d 0x 2887ab ± 193 4227a ± 865 3374ab ± 481 2787ab ± 386 2907ab ± 132 687abc ± 52 774a ± 60 731abc ± 44 489c ± 51 668abc ± 49 1x 2475b ± 161 3180ab ± 122 3878ab ± 586 3257ab ± 279 3026ab ± 134 568abc ± 31 680abc ± 28 670abc ± 55 506bc ± 46 754ab ± 45 CHL 2x 2356b ± 143 2941ab ± 196 2912ab ± 257 2837ab ± 86 2939ab ± 62 547abc ± 62 606abc ± 34 611abc ± 23 516abc ± 23 661abc ± 47 5x 3203ab ± 146 3448ab ± 306 3646ab ± 78 2626ab ± 163 2680ab ± 115 542abc ± 47 524abc ± 7 650abc ± 46 549abc ± 52 589abc ± 15 0x 2887ab ± 193 4227a ± 865 3374ab ± 481 2787ab ± 386 2907ab ± 132 687ab ± 52 774a ± 60 731ab ± 44 489ab ± 51 668ab ± 49 1x 2323b ± 174 2976ab ± 76 3405ab ± 382 3056ab ± 228 3033ab ± 99 542ab ± 43 640ab ± 29 644ab ± 50 592ab ± 53 645ab ± 41 IPU 2x 3145ab ± 169 2854ab ± 346 2818ab ± 345 2762ab ± 222 3180ab ± 114 607ab ± 56 529ab ± 50 636ab ± 33 656ab ± 26 615ab ± 17 5x 3088ab ± 291 2934ab ± 266 2924ab ± 214 3175ab ± 172 3170ab ± 216 672ab ± 56 585ab ± 90 630ab ± 40 669ab ± 45 624ab ± 9 0x 2887ab ± 193 4227a ± 865 3374a ± 481 2787a ± 386 2907a ± 132 687a ± 52 774a ± 60 731a ± 44 489a ± 51 668a ± 49 1x 3027ab ± 129 3133a ± 301 3069a ± 333 2979a ± 112 3054a ± 90 700a ± 85 628a ± 60 591a ± 21 621a ± 45 633a ± 35 TCZ 2x 3482ab ± 95 3350a ± 200 3170a ± 382 3074a ± 164 3088a ± 128 722a ± 22 667a ± 32 691a ± 28 697a ± 48 638a ± 37 5x 3000ab ± 36 2975a ± 115 3175a ± 237 2780a ± 190 2802a ± 38 601a ± 63 671a ± 42 619a ± 36 611a ± 17 580a ± 25

Estimation of the impact of CHL, IPU or TCZ on the β-diversity of the soil bacterial community

Examination of bacterial phyla showed that neither the applied pesticide nor the time affected the distribution entirety of the most abundant bacterial phyla in the soil (Figures 4.1.2 and 4.1.3). The six most abundant phyla in the microcosm and the field experiments were Proteobacteria, Acidobacteria, Chloroflexi, Bacteroidetes, Actinobacteria and Gemmatimonadetes that represented around 85 % of all phyla. This configuration was not affected by time and remained constant.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 175 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Figure 4.1.2. Microcosm study - Relative abundance (%) of bacterial phyla identified in the untreated (control) and treated (CHL, IPU or TCZ at 10x dose) soil samples at different sampling time points (0, 7, 42, 56 and 100 days after treatment). Each color represents a phylum. 16S rDNA sequences that could not be assigned to a phylum are grouped as ‘unassigned’.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 176 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Figure 4.1.3. Field study - Relative abundance (%) of bacterial phyla identified in the untreated (control) and treated (CHL, IPU or TCZ at 5x dose) soil samples at different sampling time points (0, 14, 35, 70 and 105 days after treatment). Each color represents a phylum. Bacterial phyla, whose abundance sum of all samples was < 0.05 % are grouped as ‘others’. 16S rDNA sequences that could not be assigned to a phylum are grouped as ‘unassigned’.

A more extensive analysis of the 16S rDNA amplicon sequences was performed by ANOSIMs (Supplementary Table 4.1.1) on PCoAs (Supplementary Figures 4.1.7 to 4.1.14) performed on OTU weighted unifrac distance matrices. It suggested that CHL and TCZ had slight but significant transient impacts on the soil bacterial β-composition, while IPU did not significantly affect the soil bacterial composition. In the microcosm study, significant differences (p < 0.01) were only observed for CHL at 56 days (Supplementary Table 4.1.1, Supplementary Figure 4.1.7). In the field study, only TCZ was found to significantly (p < 0.01) impact the soil bacterial composition at 35 days (Supplementary Table 4.1.1, Supplementary Figure 4.1.13). Comparisons of the untreated samples of all time points of the microcosm and the field study revealed significant time effects for both the microcosm and the field study

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 177 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing (Supplementary Figure 4.1.10 and 4.1.14, Supplementary Table 4.1.1), further confirming the findings of the α-diversity analysis. Analysis of detected OTUs at species level by the R package ‘pamR’ was employed to find the OTUs responsible for the significant differences detected by ANOSIM. In the microcosm study, eight OTUs were responsible for the differences between CHL-treated samples and untreated samples at 56 days: an unknown species of the (i) genus Aeromicrobium, (ii) class Acidobacteria, (iii) genus Flexibacter, (iv) order Chloroacidobacterium, (v) order Sphingobacteriales, and three unknown species of the phylum Chloroflexi. (Supplementary Figure 4.1.15). In the field study, two OTUs were found to be responsible for differences found by ANOSIM for TCZ-treated samples at 35 days: an unknown species of the (i) class Acidobacteria and (ii) the order Acidimicrobiales (Supplementary Figure 4.1.16).

Discussion

The ecotoxicological impact of the three formulated pesticides CHL, IPU or TCZ applied at different doses in a lab-to-field experiment (0x, 1x, 2x or 10x (microcosms), or 0x, 1x, 2x or 5x (field) the agronomical doses) on the soil bacterial diversity was assessed by Illumina next-generation sequencing of 16S rDNA amplicons. To our knowledge, it was the first time that their effects on the soil bacterial diversity were assessed by this approach.

Time effects more distinctive in the microcosm study than in the field study

Diversity indices in control samples (not exposed to pesticide) significantly decreased with time in the microcosm study. This decrease in the bacterial diversity was less marked in the field study. This is in accordance with one of the most known drawbacks of long-term soil microcosm experiments causing changes in microbial abundance, composition, diversity and activity because of the lack of nutrient inputs (Edwards et al., 1997). The ISO recommends not to incubate soil microcosms for more than one month to limit this drawback (ISO 10381-6, 2009). The changes in diversity observed in the controls of the field experiment are not expected to be due to a shortage

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 178 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing in nutrients but most likely to pedoclimatic conditions (i.e. cold periods during winter). Phylum analysis revealed Proteobacteria, Acidobacteria, Chloroflexi, Bacteroidetes, Actinobacteria and Gemmatimonadetes to dominate the total bacterial community in the microcosm and the field study, as also found in a range of other soils (Acosta-Martinez et al., 2008; Janssen, 2006; Merlin et al., 2015; Petric et al, 2011; Romdhane et al., 2016; Wessen et al., 2010). The abundance of these phyla remained relatively constant during the three months of the study in both the microcosm and the field experiments. The PCoA and ANOSIM at species level confirmed the significant time effects in the microcosm and the field study.

Slight but significant transient pesticide effects on the soil bacterial composition

IPU exposure did not modify the bacterial diversity in both microcosms and field experiments. This is in agreement with the conclusions of EFSA (2015b) (who interpreted low impact of IPU on carbon and nitrogen mineralization as low risk for soil microorganisms) and with other studies reporting small to moderate (sometimes temporary) effects of IPU on soil microorganisms (mainly based on decreasing microbial activity and biomass) (Harden et al., 1993; Kuriyal and Pandey, 1994; Perrin-Ganier et al., 2001; Schuster and Schröder, 1990; Tag-El-Din, 1982). Slight but significant differences between control and CHL-treated soils in PCoA and OTU abundances were observed after 56 days of exposure to CHL in the microcosm experiment. Interestingly, in soil treated with 10x the agronomical dose, one could observe the transient accumulation of the main transformation product of CHL (3,5,6- trichloropyridinol, TCP) during CHL dissipation (reaching its maximum at 56 days (1 mg/kg dwt soil), which disappeared thereafter (< 0.1 mg/kg dwt soil after 125 days) (Papadopoulou et al., 2016). In the field experiment, where CHL rapidly dissipated and TCP did not accumulate (< 0.1 mg/kg dwt soil), the bacterial composition was not significantly changed. One could therefore hypothesize that (i) changes in the bacterial composition in response to CHL exposure recorded at 56 days may not only be due to CHL (that had partially dissipated at 40 % of the initially applied dose at that time point) but also due to TCP exposure and that (ii) resilience resulting in the recovery of the bacterial composition may be attributed to the rapid dissipation of TCP and the constant dissipation of CHL. On one hand, the importance of TCP in the ecotoxicological impact of

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 179 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing CHL has to be taken with care because it was not found to accumulate in soils treated with the 1x and 2x agronomical CHL doses (Papadopoulou et al., 2016). On the other hand, its contribution to changes in bacterial composition is supported by a number of studies indicating its toxicity towards (micro)organisms (Caceres et al., 2007; Feng et al., 1997). Although the review report of the European Commission (EC, 2005) and one other study (Singh et al., 2002) conclude CHL not to impact soil microorganisms, our study reports a slight but significant ecotoxicological effect of CHL on the soil bacterial composition in the microcosm experiment where CHL and TCP cannot dissipate away from the tested soil by transport processes (leaching, run-off, volatilization), as it can be the case in field experiments. The effect was transient and the bacterial community recovered as shown by the rapid resilience of the bacterial composition. The biological importance (as compared to other stresses such as extreme weather conditions) and the consequences of these modifications on the functions supported by the bacterial community remain unknown. Our findings are in the line of studies reporting that CHL impacts soil microorganisms in many ways (Chu et al., 2008; Dutta et al., 2010; Eisenhauer et al., 2009; Fang et al., 2009; Gilani et al., 2010; Gupta et al., 2013; Jastrzebska, 2011; Pozo et al., 1995; Srinivasulu and Rangaswamy, 2013; Wang et al., 2010). Slight but significant differences between control and TCZ-treated soils were observed in PCoA and OTU abundance analyses after 35 days of exposure to TCZ in the field experiment. TCZ was rapidly dissipated in the field experiment (60 % dissipation after 35 days) (Papadopoulou et al., 2016). In addition, an important number and variety of TCZ transformation products have been detected but not quantified (due to the absence of appropriate reference standards) in the field soil samples (treated with the 5x agronomical dose) (Storck et al., 2016). As discussed before for CHL, observed changes in the bacterial composition may be attributed to the combined exposure to the parent compound and related transformation products. Keeping in mind that in silico typology led to their classification into different clusters characterized by different estimated environmental parameters (Storck et al., 2016), one could hypothesize that they may impact soil microorganisms to different extents. Especially triazole dead-end transformation products recalcitrant to biodegradation may possibly interact with the hormone regulation network of non-target organisms (Rieke et al., 2014; Shalini et al., 2011). Nevertheless, the bacterial composition was resilient suggesting the recovery of

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 180 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing the bacterial community. Although the evaluations of EFSA (2008; 2014) concluded that TCZ has low risks on soil microorganisms (based on carbon and nitrogen mineralization in soil), our study reports slight but significant and transient effect of TCZ on the bacterial composition in the field experiment. This is in agreement with previous studies reporting TCZ effects on soil microorganisms, evaluated by analysis of soil enzyme activities, microbial biomass, carbon and nitrogen mineralization, and others (Anuradha et al., 2016; Bending et al., 2007; Cycon et al., 2006; Ferreira et al., 2009; Munoz-Leoz et al., 2011; Wang et al., 2016). Our study showed that small and transient changes in the diversity of the bacterial community can be detected for two (CHL and TCZ) of the three tested pesticides by 16S rDNA amplicon next-generation sequencing. Contrariwise to Romdhane et al. (2016) reporting big shifts in the bacterial diversity exposed to the natural triketone herbicide, only slight modifications were observed in our study. This may be due to the fact that, contrary to natural triketone for which the target (4- hydroxyphenylpyruvate dioxygenase, HPPD) is present in non-target organisms including soil microorganisms, the targets of the three studied pesticides are (to our best knowledge) not found in soil bacteria (IPU targets photosystem I, CHL blocks acetylcholine neurotransmission, and TCZ blocks sterol 14α-demethylase and impedes any additional modes of action on sterol biosynthesis). For this reason, no direct effects of the three tested pesticides on soil bacteria were expected. Nonetheless, transformation products of each of the three tested pesticides that are known for their toxicity (namely 4-isopropyl-aniline (4-IA) for IPU, TCP for CHL, and triazole transformation products for TCZ) may directly impact soil microorganisms. In addition, indirect effects of the three pesticides can be expected notably for TCZ because of close interactions between fungal and bacterial species in soil. Despite the new insights offered by next-generation sequencing of 16S rDNA amplicons to assess pesticide effects on the soil bacterial diversity, bioindicators specifically responding to pesticide exposure may represent an alternative. For instance, the adaptation of soil microorganisms to pesticides leading to enhanced pesticide degradation in soil may constitute a bioindicator for exposure (Copley, 2009). Such adaptive responses to cope with the stress imposed by pesticides have already been described for a range of pesticides (Crouzet et al., 2010; Hussain et al., 2011, Weaver et al., 2007). Unfortunately, up to now it remains difficult to measure adaptation by sequencing methods because

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 181 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing (i) the abundance of degraders in soil remains quite low (between 103 and 104 cells per g of soil) (Piutti et al., 2002) and may be unequally distributed at different scales (from a soil aggregate to the field) (Pinheiro et al., 2015) , (ii) the expression of pesticide- degrading genes is transient (Baelum et al., 2008; Martin-Laurent et al., 2003) and (iii) the pesticide-degrading genetic potential may be exchanged by horizontal gene transfer (Devers et al., 2007). In any cases, the interpretation of the biological significance of pesticide ecotoxicological impact remains difficult in the absence of data on the normal operating range for each studied bioindicator (such as bacterial diversity), making it challenging to define specific protection goals to protect ecosystem functions provided by soil microorganisms (Bell et al., 2005).

Acknowledgements

The Ph.D. studies of Veronika Storck performed at the Doctoral School of Environment and Health (ED E2S) at the University of Burgundy were funded by the French Ministry of Education and Research (grant number 2013-31). This work was done within the framework of the FP7-PEOPLE-2012-IAPP (Industry-Academia Partnerships and Pathways) Marie-Curie project ‘Love-to-Hate’ funded by the European Commission (grant number 324349).

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 182 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary material

Supplementary Figure 4.1.1. Microcosm study - Bacterial diversity in soil treated with CHL at various doses (0x, 1x, 2x, 10x) at various time points after treatment (0, 7, 42, 56 and 100 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 183 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.2. Microcosm study - Bacterial diversity in soil treated with IPU at various doses (0x, 1x, 2x, 10x) at various time points after treatment (0, 7, 42, 56 and 100 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 184 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.3. Microcosm study - Bacterial diversity in soil treated with TCZ at various doses (0x, 1x, 2x, 10x) at various time points after treatment (0, 7, 42, 56 and 100 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 185 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.4. Field study - Bacterial diversity in soil treated with CHL at various doses (0x, 1x, 2x, 5x) at various time points after treatment (0, 14, 35, 70 and 105 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 186 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.5. Field study - Bacterial diversity in soil treated with IPU at various doses (0x, 1x, 2x, 5x) at various time points after treatment (0, 14, 35, 70 and 105 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 187 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.6. Field study - Bacterial diversity in soil treated with TCZ at various doses (0x, 1x, 2x, 5x) at various time points after treatment (0, 14, 35, 70 and 105 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 188 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.7. Microcosm study - PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) and CHL-treated (1x, 2x or 10x doses) soil samples at different time points (0, 7, 42, 56 and 100 days). Replicates of the same treatment are represented in same colors.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 189 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.8. Microcosm study - PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) and IPU-treated (1x, 2x or 10x doses) soil samples at different time points (0, 7, 42, 56 and 100 days). Replicates of the same treatment are represented in same colors.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 190 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.9. Microcosm study - PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) and TCZ-treated (1x, 2x or 10x doses) soil samples at different time points (0, 7, 42, 56 and 100 days). Replicates of the same treatment are represented in same colors.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 191 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.10. Microcosm study - PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) soil samples at different time points (0, 7, 42, 56, 100 days). Replicates of the same treatment are represented in same colors.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 192 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.11. Field study - PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) and CHL-treated (1x, 2x or 5x doses) soil samples at different time points (0, 14, 35, 70 and 105 days). Replicates of the same treatment are represented in same colors.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 193 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.12. Field study - PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) and IPU-treated (1x, 2x or 5x doses) soil samples at different time points (0, 14, 35, 70 and 105 days). Replicates of the same treatment are represented in same colors.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 194 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.13. Field study - PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) and TCZ-treated (1x, 2x or 5x doses) soil samples at different time points (0, 14, 35, 70 and 105 days). Replicates of the same treatment are represented in same colors.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 195 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.14. Field study - PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) soil samples at different time points (0, 14, 35, 70 and 105 days). Replicates of the same treatment are represented in same colors.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 196 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Table 4.1.1. Microcosm and field study - ANOSIM of the PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) and treated (CHL, IPU or TCZ at 1x, 2x or 10x (microcosms) or 1x, 2x or 5x (field) doses) soil samples at different time points (0, 7, 42, 56 and 100 days (microcosms) or 0, 14, 35, 70 and 105 days (field)). Significant differences between groups are indicated by p < 0.01.

Supplementary Figure 4.1.15. Microcosm study - Abundance [number/sample] of OTUs responsible for significant differences of ANOSIM, detected by ‘pamR’ in untreated (control) and treated (CHL at 1x, 2x or 10x doses) soil samples at 56 days. ANOVAs were followed by the Bonferroni test per each taxon. Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 197 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Supplementary Figure 4.1.16. Field study - Abundance [sequence/sample] of an OTUs responsible for significant differences of ANOSIM, detected by ‘pamR’ in untreated (control) and treated (TCZ at 1x, 2x or 10x doses) soil samples at 35 days. ANOVAs were followed by the Bonferroni test per each taxon. Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 198 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing References

Martin-Laurent, F., Piutti, S., Hallet, S., Wagschal, I., Philippot, L., Catroux, G. and Soulas, G. (2003). Monitoring of atrazine treatment on soil bacterial, fungal and atrazine-degrading communities by quantitative competitive PCR. Pest Management Science 59, 259-268. Baelum, J., Nicolaisen, M.H., Holben, W.E., Strobel, B.W., Sorensen, J. and Jacobsen, C.S. (2008). Direct analysis of tfdA gene expression by indigenous bacteria in phenoxy acid amended agricultural soil. The ISME Journal 2, 677-687. Devers, M., El Azhari, N., Udikovic-Kolic, N. and Martin-Laurent, F. (2007). Detection and organization of atrazine-degrading genetic potential of seventeen bacterial isolates belonging to divergent taxa indicate a recent common origin of their catabolic functions. FEMS Microbiology Letters 273, 78-86. Pinheiro, M., Garnier, P., Beguet, J., Martin-Laurent, F. and Vieublé Gonod, L. (2015). The millimetre-scale distribution of 2,4-D and ist degraders drives the fate of 2,4-D at the soil core scale. Soil Biology and Biochemistry 88, 90-100. Rieke, S., Koehn, S., Hirsch-Ernst, K., Pfeil, R., Kneuer, C. and Marx-Stoelting, P. (2014). Combination effects of (tri)azole fungicides on hormone production and xenobiotic metabolism in a human placental cell line. International Journal of Environmental Research and Public Health 11, 9660-9679. Shalini, K., Kumar, N., Drabu, S. and Kumar Sharma, P. (2011). Advances in synthetic approach to and antifungal activity of triazoles. Beilstein Journal of Organic Chemistry 7, 668-677. Storck, V., Lucini, L., Mamy, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., Karas, P.A., Servien, R., Karpouzas, D.G., Trevisan, M., Benoit, P. and Martin-Laurent, F. (2016). Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology. Environmental Pollution 208, 537-545. Feng, Y., Racke, K.D. and Bollag, J. (1997). Isolation and characterization of a chlorinated-pyridinol-degrading bacterium Applied and Environmental Microbiology 63 (10), 4096-4098. Caceres, T., He, W., Naidu, R. and Megharaj, M. (2007). Toxicity of chlorpyrifos and TCP alone and in combination to Daphnia carinata: the influence of microbial degradation in natural water. Water Research 41 (19), 4497-4503. Edwards, C.A., Subler, S., Parmelee, R., Knacker, T. and Pokarzhevskij, A.A. (1997). Use of soil microcosms in assessing the effect of pesticides on soil ecosystems. Environmental Behaviour of Crop Protection Chemicals 28 (13), 435-451. ISO 10381-6 (2009). Soil quality - Sampling - Part 6: Guidance on the collection, handling and storage of soil under aerobic conditions for the assessment of microbiological processes, biomass and diversity in the laboratory. Acosta-Martinez, V., Dowd, S., Sun, Y., and Allen, V. (2008). Tag-encoded pyrosequencing analysis of bacterial diversity in a single soil type as affected by management and land use. Soil Biology and Biochemistry 40, 2762-2770. Anuradha, B., Rekhapadmini, A. and Rangaswamy, V. (2016). Influence of tebuconazole and copper hydroxide on phosphatase and urease activities in red sandy loam and black clay soils. 3 Biotech (6), 78. Barbash, J.E. (2003). The geochemistry of pesticides. Treatise on Geochemistry 9, DOI: 10.1016/B0-08-043751-6/09056-3. Bell, T., Newman, J.A., Silverman, B.W., Turner, S.L. and Lilley, A.K. (2005). The contribution of species richness and composition to bacterial services. Nature 436, 1157-1160. Bending, G.D., Rodriguez-Cruz, M.S. and Lincoln, S.D. (2007). Fungicide impacts on microbial communities in soils with contrasting management histories. Chemosphere 69, 82-88. Blum, W.E.H. (2006). Soil resources - the basis of human society and the environment. Bodenkultur 57, 197-202. Bondeau, A., Smith, C.M., Zaehle, S., Schapfhoff, S., Lucht, W., Cramer, W., Gerten, D., Lotze-Campen, H., Mueller, C., Reichstein, K. and Smith, B. (2007). Modelling the role of agriculture for the 20th century global terrestrial carbon balance. Global Change Biology 13, 679-706. Börner, H., Schlüter, K. and Aumann, J. (2009). Pflanzenkrankheiten und Pflanzenschutz. Springer, Berlin Heidelberg, ISBN 978-3- 540-49068-5. Brodie, E.L., Desantis, T.Z., Joyner, D.C., Baek, S.M., Larsen, J.T., Andersen, G.L., Hazen, T.C., Richardson, P.M., Herman, D.J., Tokunaga, T.K., Wan, J.M. and Firestone, M.K. (2006). Application of a high-density oligonucleotide microarray approach to study bacterial population dynamics during uranium reduction and reoxidation. Applied and Environmental Microbiology 72, 6288-6298. Caporaso, J.G., Bittinger, K., Bushman, F.D., DeSantis, T.Z., Andersen, G.L. and Knight, R. (2010b). PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics 26, 266-267. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D. and Costello, E.K. (2010a). QIIME allows analysis of high- throughput community sequencing data. Nature Methods 7, 335-336. Chu, X., Fang, H., Pan, X., Wang, X., Shan, M., Feng, B. and Yu, Y. (2008). Degradation of chlorpyrifos alone and in combination with chlorothalonil and their effects on soil microbial populations. Journal of Environmental Sciences 20, 464-469. Copley, S.D. (2009). Evolution of efficient pathways for degradation of anthropogenic chemicals. Nature Chemical Biology 5, 559-566. Crouzet, O., Batisson, I., Besse-Hoggan, P., Bonnemoy, F., Bardot, C., Poly, F., Bohatier, J. and Mallet, C. (2010). Response of soil microbial communities to the herbicide mesotrione: a dose-effect microcosm approach. Soil Biology and Biochemistry 42, 193-202. Cycon, M., Piotrowska-Seget, Z. and Kaczynska, A. (2006). Microbiological characteristics of a sandy loam soil exposed to tebuconazole and λ-cyhalothrin under laboratory conditions. Ecotoxicology 15, 639-646. Dutta, M., Sardar, D., Pal, R. and Kole, R.K. (2010). Effect of chlorpyrifos on microbial biomass and activities in tropical clay loam soil. Environmental Monitoring and Assessment 160, 385-391. EC (2005). Review report for the active substance chlorpyrifos. Finalised in the Standing Committee on the Food Chain and Animal Health at its meeting on 3 June 2005 in view of the inclusion of chlorpyrifos in Annex I of the Directive 91/414/EEC. SANCO/3059/99 - rev. 1.5. Edgar, R.C. (2010). Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26, 2460-2461. Edgar, R.C., Haas, B.J., Clemente, J.C., Quince, C. and Knight, R. (2011). UCHIME improves sensitivity and speed of chimera detection. Bioinformatics 27, 2194-2200. EFSA (2008). Conclusion on the peer review of tebuconazole. EFSA Scientific Report 176, 1-109. EFSA (2014). Conclusion on the peer review of the pesticide risk assessment of the active substance tebuconazole. EFSA Journal 12 (1), 3485. http://dx.doi.org/10.2903/j.efsa.2014.3485, 98. EFSA (2015a). Conclusion on the peer review of the pesticide risk assessment of the active substance isoproturon. EFSA Journal 13 (8), 4206. doi:10.2903/j.efsa.2015.4206.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 199 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

EFSA (2015b). Conclusion on the peer review of the pesticide risk assessment of the active substance glyphosate. EFSA Journal 13 (11), 4302 [107 pp.] EFSA (2016). Recovery in environmental risk assessments at EFSA, EFSA Scientific Committee. EFSA Journal 2016; 14 (2), 4313. Eisenhauer, N., Klier, M., Partsch, S., Sabais, A.C.W., Scherber, C., Weisser, W.W. and Scheu, S. (2009). No interactive effects of pesticides and plant diversity on soil microbial biomass and respiration. Applied Soil Ecology 42, 31-36. Fang, H., Yu, Y., Chu, X., Wang, X., Yang, X. and Yu, J. (2009). Degradation of chlorpyrifos in laboratory soil and its impact on soil microbial functional diversity. Journal of Environmental Sciences 21, 380-386. Ferreira, E.P.B., Dusi, A.N., Costa, J.R., Xavier, G.R. and Rumjanek, N.G. (2009). Assessing insecticide and fungicide effects on the culturable soil bacterial community by analyses of variance of their DGGE fingerprinting data. European Journal of Soil Biology 45, 466-472. Gardi, C., Jeffery, S. and Saltelli, A. (2013). An estimate of potential threats levels to soil biodiversity in EU. Global Change Biology 19, 1538-1548. Gilani, S.T.S., Ageen, M., Shah, H. and Raza, S. (2010). Chlorpyrifos degradation in soil and its effect on soil microorganisms. The Journal of Animal and Plant Sciences 20 (2), 99-102. Graham, E.B., Knelman, J.E., Schindlbacher, A., Siciliano, S., Breulmann, M., Yannarell, A., Beman, J.M., Abell, G., Philippot, L., Prosser, J., Foulquier, A., Yuste, J.C., Glanville, H.C., Jones, D.L., Angel, R., Salminen, J., Newton, R.J., Bürgmann, H., Ingram, L.J., Hamer, U., Siljanen, H.M.P., Peltoniemi, K., Potthast, K., Bañeras, L., Hartmann, M., Banerjee, S., Yu, R.-Q., Nogaro, G., Richter, A., Koranda, M., Castle, S.C., Goberna, M., Song, B., Chatterjee, A., Nunes, O.C., Lopes, A.R., Cao, Y., Kaisermann, A., Hallin, S., Strickland, M.S., Garcia-Pausas, J., Barba, J., Kang, H., Isobe, K., Papaspyrou, S., Pastorelli, R., Lagomarsino, A., Lindström, E.S., Basiliko, N. and Nemergut, D.R. (2016). Microbes as engines of ecosystem function: When does community structure enhance predictions of ecosystem processes? Frontiers in Microbiology 7 (214), doi: 10.3389/fmicb.2016.00214. Gupta, S., Gupta, R. and Sharma, S. (2013). Impact of chemical- and bio-pesticides on bacterial diversity in rhizosphere of Vigna radiata. Ecotoxicology 22, 1479-1489. Hamady, M., Lozupone, C. and Knight, R. (2010). Fast UniFrac: facilitating high-throughput phylogenetic analyses of microbial communities including analysis of pyrosequencing and PhyloChip data. The ISME Journal 4, 17-27. Harden, T., Joergensen, R.G., Meyer, B. and Wolters, V. (1993). Soil microbial biomass estimated by fumigation-extraction and substrate-induced respiration in two pesticide-treated soils. Soil Biology and Biochemistry 25 (6), 679-683. Hartley J.L. annd Bowen H. (2003). PEG precipitation for selective removal of small DNA fragments. Focus 18, 27. Hayes, W.J. and Laws, E.R. (1991). Handbook of pesticide toxicology, Vol. 3, Classes of pesticides. Academic Press Inc., ISBN 0-12- 334162-0. Haygarth, P. and Ritz, K. (2009). The future of soils and land use in the UK: Soil systems for the provision of land-based ecosystem services. Land Use Policy, 26, 187-197. Hillel, D. (2009). The mission of soil science in a changing world. Journal of Plant Nutrition and Soil Science 172, 5-9. Hussain, S., Devers-Lamrani, M., El Azhari, N. and Martin-Laurent, F. (2011). Isolation and characterization of an isoproturon mineralizing Sphingomonas sp. strain SH from a French agricultural soil. Biodegradation 22, 637-650. Imfeld, G. and Vuillemier, S. (2012). Measuring the effects of pesticides on bacterial communities in soil: A critical review. European Journal of Soil Biology 49, 22-30. ISO 10381-6 (2009). Soil Quality - Sampling - Part 6: Guidance on the collection, handling and storage of soil under aerobic conditions for the assessment of microbiological processes, biomass and diversity in the laboratory. International Organization for Standardization (ISO). ISO 11063 (2012). Soil quality - Method to directly extract DNA from soil samples. Janssen, P. H. (2006). Identifying the dominant soil bacterial taxa in libraries of 16S rRNA and 16S rRNA genes. Applied and Environmental Microbiology 72, 1719-1728. Jastrzebska, E. (2011). The effect of chlorpyrifos and teflubenzuron on the enzymatic activity of soil. Polish Journal of Environmental Studies 20 (4), 903-910. Jeffery, S., Gardi, C., Jones, A., Montanarella, L., Marmo, L., Miko, L., Ritz, K., Peres, G., Römbke, J. and van der Putten, W.H. (2010). European Atlas of Soil Biodiversity. European Commission, Publications Office of the European Union, Luxembourg, ISBN 978-92-79-15806-3. Jones, A., Stolbovoy, V., Rusco, E., Gentile, A.-R., Gardi, C., Marechal, B. and Montanarella, L. (2009). Climate change in Europe. 2. Impact on soil. A review. Agronomy for Sustainable Development 29, 423-432. Joshi, N. and Fass, J. (2011). Sickle: A sliding-window, adaptive, quality-based trimming tool for FastQ files. http://github.com/najoshi/sickle, accessed 1st February, 2016. Klindworth, A., Pruesse, E., Schweer, T., Peplies, J., Quast, C., Horn, M., and Glockner, F.O. (2012). Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Research 41 (1), 1. Kreader, C.A. (1996). Relief of amplification inhibition in PCR with bovine serum albumin or T4 gene 32 protein. Applied and Environmental Microbiology 62 (3), 1102-1106. Kuriyal, S.K. and Pandey, B.N. (1994). Studies of the effects of isoproturon on soil microflora. Journal of the Indian Botanical Society 73, 47-49. Lal, R. (2008). Soils and sustainable agriculture. A review. Agronomy for Sustainable Development 28, 57-64. Lane, D. (1991). 16S/23S rRNA sequencing. Nucleic acid techniques in bacterial systematics, Wiley, 329. Lichtfouse, E., Navarette, M., Debaeke, P., Souchère, V. and Alberola, C. (2009). Sustainable Agriculture, Springer, 1st ed., 645 p., ISBN: 978-90-481-2665-1. Looser, R., Froescheis, O., Cailliet, G.M., Jarman, W.M. and Ballschmiter, K. (2000). The deep-sea as a final global sink of semivolatile persistent organic pollutants? Part II: Organochlorine pesticides in surface and deep-sea dwelling fish of the North and South Atlantic and the Monterey Bay Canyon (California). Chemosphere, 40 (6), 661-70. Martin, M. (2011). Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet Journal 17, 10-12. Martin-Laurent, F., Philippot, L., Hallet, S., Chaussod, R., Germon, J.C., Soulas, G. and Catroux, G. (2001). DNA extraction from soils: old bias for new microbial diversity analysis methods. Applied and Environmental Microbiology 67 (5), 2354-2359. Masella, A.P., Bartram, A.K., Truszkowski, J.M., Brown, D.G. and Neufeld, J.D. (2012). PANDAseq: paired-end assembler for Illumina sequences. BMC Bioinformatics 13, 31. Merlin, C., Devers, M., Béguet, J., Boggio, B., Rouard, N., and Martin-Laurent, F. (2015). Evaluation of the ecotoxicological impact of the organochlorine chlordecone on soil microbial community structure, abundance, and function. Environmental Science and Pollution Research 23, 4185-4198.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 200 Chapter 4.1 Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina next-generation sequencing

Munoz-Leoz, B., Ruiz-Romera, E., Antigëdad, I. and Garbisu, C. (2011). Tebuconazole application decreases soil microbial biomass and activity. Soil Biology and Biochemistry 43, 2146-2183. Oerke, E.-C., Dehne, H.W., Schönebeck, F. and Weber, A. (1994). Crop production and crop protection. Estimated losses in major food and cash crops. Elsevier Science B.V., ISBN 0-444-82095-7. Papadopoulou, E.S., Karas, P., Nikolaki, S., Storck, V., Ferrari, F., Trevisan, M., Tsiamis, G., Martin-Laurent, F. and Karpouzas, D.G. (2016). Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions. Science of the Total Environment 569-570, 86-96. Perrin-Ganier, C., Schiavon, F., Morel, J.-L. and Schiavon, M. (2001). Effect of sludge-amendment or nutrient addition on the biodegradation of the herbicide isoproturon in soil. Chemosphere 44, 887-892. Petric, I., Bru, D., Udikovic-Kolic, N., Hrsak, D., Philippot, L., and Martin-Laurent, F. (2011). Evidence for shifts in the structure and abundance of the microbial community in a long-term PCB-contaminated soil under bioremediation. Journal of Hazardous Materials 195, 254-260. Pimentel, D. (1995). Amounts of pesticides reaching target pests: Environmental impacts and ethics. Journal of Agricultural and Environmental Ethics 8 (1), 17-29. Pimentel, D., Wilson, C., McCullum, C., Huang, R., Dwen, P., Flack, J., Tran, Q., Saltman, T. and Cliff, B. (1997). Economic and environmental benefits of biodiversity. Bioscience, 47, 747-758. Piutti, S., Semon, E., Landry, D., Hartmann, A., Dousset, S., Lichtfouse, E., Topp, E., Soulas, G. and Martin-Laurent, F. (2003). Isolation and characterisation of Nocardioides sp. SP12, an atrazine degrading bacterial strain possessing the gene trzN from bulk- and maize rhizosphere soil. FEMS Microbiology Letters 221, 111-117. Pozo, C., Martinez-Toledo, V., Salmeron, V., Rodelas, B. and Gonzalez-Lopez, J. (1995). Effect of chlorpyrifos on soil microbial activity. Environmental Toxicology and Chemistry 14 (2), 187-192. Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T. and Yarza, P. (2013). The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Research 41, 590-596. Romdhane, S., Devers-Lamrani, M., Barthelmebs, L., Calvayrac, C., Bertrand, C., Cooper, J.-F., Dayan, F.E. and Martin-Laurent, F. (2016). Ecotoxicological impact of the Leptospermone on the microbial community of two arable soils. Frontiers in Microbiology 7, 775. Schuster, E. and Schröder, D. (1990). Side-effects of sequentially-applied pesticides on non-target soil microorganisms: Field experiments. Soil Biology and Biochemistry 22 (3), 367-373. Singh, B.K., Walker, A. and Wright, D. (2002). Degradation of chlorpyrifos, fenamiphos, and chlorothalonil alone and in combination and their effects on soil microbial activity. Environmental Toxicology and Chemistry 21 (12), 2600-2605. Smalling, K.L., Fellers, G.M., Kleeman, P.M. and Kuivila, K.M. (2013). Accumulation of pesticides in pacific chorus frogs (Pseudacris regilla) from California's Sierra Nevada Mountains, USA. Environmental Toxicology and Chemistry, 32 (9), 2026-2034. Srinivasulu, M. and Rangaswamy, V. (2013). Influence of insecticides alone and in combination with fungicides on enzyme activities in soils. International Journal of Environmental Science and Technology 10, 341-350. Tag-El-Din, A. (1982). Herbicides and respiration flush of soil biomass. Medical Faculty Landbouww University Gent 47, 191-199. Tsiamis, G., Katsaveli, K., Ntougias, S., Kyrpides, N., Andersen, G., Piceno, Y. and Bourtzis, K. (2008). Prokaryotic community profiles at different operational stages of a Greek solar saltern. Research in Microbiology 159, 609-627. Van Beelen, P. and Doelman, P. (1997). Significance and application of microbial toxicity tests in assessing ecotoxicological risks of contaminants in soil and sediment. Chemosphere 34, 455-499. Wang, C., Wang, F., Zhang, Q. and Liang, W. (2016). Individual and combined effects of tebuconazole and carbendazim on soil microbial activity. European Journal of Soil Biology 72, 6-13. Wang, F., Yao, J., Chen, H., Chen, K., Trebse, P. and Zaray, G. (2010). Comparative toxicity of chlorpyrifos and its oxon derivates to soil microbial activity by combined methods. Chemosphere 78, 319-326. Weaver, M. A., Krutz, L. J., Zablotowicz, R. M., and Reddy, K. N. (2007). Effects of glyphosate on soil microbial communities and its Mississippi soil. Pest Management Science 63, 388-393. Wessen, E., Hallin, S., and Philippot, L. (2010). Differential responses of bacterial and archaeal groups at high taxonomical ranks to soil management. Soil Biology and Biochemistry 42, 1759-1765. Zhang, W., Jiang, F. and Ou, J. (2011). Global pesticide consumption and pollution: with China as a focus. Proceedings of the International Academy of Ecology and Environmental Sciences 1(2), 125-144. Zhang, Y. and Sun, Y. (2014). A method for de novo nucleic acid diagnostic target discovery. Bioinformatics 30 (22), 3174-3180.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 201

Chapter 4.2

Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Veronika Storck Nadine Rouard Jérémie Béguet David Bru Lucio Botteri Aymé Spor Marion Devers-Lamrani Fabrice Martin-Laurent

Manuscript in preparation for publication.

Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods Abstract

Pesticides can harm non-target organisms such as soil microorganisms involved in important ecosystem functions. The techniques that are used during a pesticide authorization process to evaluate potential risks on soil microorganisms have a weak potential to detect the microbial toxicity of pesticides. In consequence, risk assessment conclusions by responsible authorities can differ from other research studies using more advanced techniques. This is the case for the two pesticides chlorpyrifos (CHL) and tebuconazole (TCZ) whose impact on soil microorganisms was investigated in this study via alternative techniques, such as (i) the evaluation of pesticide degradation in soil, (ii) the assessment of soil buffering functions, and (iii) the analysis of the diversity of soil bacteria by DNA sequencing. Both pesticides were found to impact soil microorganisms in this study. The capability of soil microorganisms to degrade CHL increased with treatment repetitions, while it remained constant (low) for TCZ. Soil buffering functions were slightly (CHL) or not (TCZ) affected by pesticide treatments. The soil bacterial composition was altered by both pesticides. Nitrospira (reducing effects) and Lysobacter (enhancing effects) were the most interesting affected genera. Results of this study have the potential to improve future post-authorization risk assessments of CHL and TCZ.

Figure 4.2.1. Graphical abstract of Chapter 4.2.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 206

Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods Introduction

Pesticides are extensively applied to protect crops from pests in conventional agriculture worldwide (Enserink et al., 2013). Besides their advantages, they can contaminate the environment (Hutzinger, 2013; Lewis and Maslin, 2015) and harm non- target organisms (Jeyaratnam, 1990; Malaja et al., 2013), such as soil microorganisms catalyzing an important number of reactions in the environment that support soil ecosystem functions (Bondeau et al., 2007; EFSA, 2010; Graham et al., 2016). Microbial functions are supported by important processes with impacts at the global scale and contribute to food supply, climate regulation and water quality (Haygarth and Ritz, 2009; Hillel, 2009). In this context, complex environmental risk assessment is performed during a pesticide authorization process to minimize potential risks for the environment (Storck et al., 2016a). Despite the complexity of such risk assessments, the impact of pesticides on soil microorganisms is not sufficiently assessed and solely relies on simple carbon and nitrogen mineralization tests in soil (although techniques to assess microbial biomass, diversity or abundance are available). These mineralization tests are weak indicators for the microbial toxicity of pesticides because carbon and nitrogen mineralization processes can remain unchanged regardless of toxic effects, if resistant microorganisms increase their activity or abundance in expense of sensitive microorganisms (Van Beelen and Doelman, 1997). However, more advanced methods to study the impact of pesticides on soil microorganisms do exist. During the past two decades, microbial ecologists have developed an impressive tool box (including methods of classical microbiology, biochemistry, molecular biology and omics) allowing to study the abundance, diversity and activity of the microbial community in various environmental matrices (Schloter et al., 2003). Among these methods, some have been standardized by the International Standard Organization (ISO) and are now available to private companies (Philippot et al., 2012). Recently, it was proposed to revise the risk assessment of pesticides on soil microorganisms by considering ISO-standardized and under-development microbiology methods (Martin-Laurent et al., 2013; Karpouzas et al., 2016). The effects of pesticides on soil microorganisms can be assessed from the yin and the yang point of view, where the yang side signifies the potential stimulation that pesticides may cause on

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 208 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods microorganisms (leading to their adaptation), while the yin side represents their potential inhibition (leading to toxic effects) (Karpouzas et al., 2016). In response to repeated exposure to pesticides, their biodegradation can be accelerated in soils, as it has previously been observed for various pesticides including isoproturon (El-Sebai et al., 2004), atrazine (Piutti et al., 2002) or 2,4-dichlorophenoxyacetic acid (2,4-D) (Baelum et al., 2008). This phenomenon called enhanced biodegradation relies on the adaptation of soil microorganisms to use a pesticide as additional nutrient and/or energy source, providing degraders with a selective advantage favoring their emergence within the soil microbial community (Copley, 2009). Therefore, the appearance of enhanced biodegradation in a soil can be viewed not only as an indicator of pesticide exposure to microorganisms but also as an indicator of the resilience of a soil (contributing to decrease pesticide persitance in soil and therefore favoring microbial recovery). On the other hand, pesticide exposure can modify the activity of microorganisms with consequences on supported ecosystem services. Among them, water purification largely relies on soil buffering functions, the ability of a soil to capture or neutralize pollutants (such as pesticides) by sorption or by abiotic and biotic degradation, preventing their dispersion to surrounding environments like rivers, lakes or groundwater (Keesstra et al., 2012). Therefore, assessing the ecotoxicological impact of pesticides on soil buffering functions, using as a proxy the capacity of a soil to degrade a range of xenobiotic organic compounds relatively recalcitrant to biodegradation, may be of interest for the evaluation of potential pesticide effects on soil microorganisms. As there are some limitations in measuring microbial functions as proxy of ecosystem services, another option to assess the ecotoxicological impact of pesticides on soil microorganisms is to monitor the microbial diversity in response to pesticide exposure. Although the link between microbial diversity and microbial functions remains a matter of debate among microbial ecologists, there is a general consensus on the fact that microbial diversity can be seen as a kind of ecological insurance preserving microbial functions and ecosystem services. Monitoring of shifts in the microbial diversity by next- generation sequencing of 16S rDNA amplicons of extracted soil DNA can be used as an alternative method to assess the response of soil microorganisms to pesticide exposure (Romdhane et al., 2016). In order to test the interest of these alternative methods for the assessment of pesticide risks on soil microorganisms, we established a soil microcosm experiment

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 209 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods using a worst case pesticide exposure scenario consisting in the repeated application of the organophosphate insecticide chlorpyrifos (CHL) or the triazole fungicide tebuconazole (TCZ) on soil microorganisms. These two pesticides have been chosen because they are on the EU market for 51 (CHL) and 28 (TCZ) years and are present in agricultural environments worldwide. However, their impact on soil microorganisms is only slightly recorded in respective risk assessment documents, solely based on carbon and nitrogen mineralization in soil. CHL was concluded not to have any significant effects on carbon and nitrogen mineralization in soil (EC, 2005). TCZ effects on these processes were found to be < 25 % and interpreted as a low risk for microorganisms (EFSA, 2008; EFSA, 2014). Other studies investigating the impact of CHL or TCZ on soil microorganisms via the analysis of microbial abundance, diversity or activity concluded effects of CHL (Chu et al., 2008; Dutta et al., 2010; Eisenhauer et al., 2009; Fang et al., 2009; Gilani et al., 2010; Gupta et al., 2013; Jastrzebska, 2011; Pozo et al., 1995; Srinivasulu and Rangaswamy, 2013; Wang et al., 2010) or TCZ (Anuradha et al., 2016; Bending et al., 2007; Cycon et al., 2006; Ferreira et al., 2009; Munoz-Leoz et al., 2011; Wang et al., 2016) on soil microorganisms. Aim of this study was to enrich post-authorization environmental risk assessment of CHL and TCZ by innovative research studying the impact of repeated CHL or TCZ treatments on soil microorganisms via the use of alternative methods such as the evaluation of pesticide degradation in soil, the assessment of soil buffering functions, and the analysis of the diversity of soil bacteria by Illumina next-generation sequencing.

Material and methods

Overview of the experimental procedure

The impact of repeated treatments of the pesticides CHL or TCZ on soil microorganisms was evaluated via (i) the mineralzation of each pesticide applied to the soil, (ii) the mineralization of a range of organic compounds (including several pesticides showing different levels of recalcitrance to biodegradation) as proxy of soil buffering functions, and (iii) the diversity of soil bacteria. For this purpose, soil mesocosms (four individual replicates per treatment) were repeatedly treated by CHL

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 210 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods or TCZ once per month (Figure 4.2.2). Soil was sampled once per month to (i) measure the mineralization of the applied pesticides (CHL or TCZ, as proxy of enhanced biodegradation) and of three other compounds (2,4-D, glyphosate and sodium acetate, as proxy of soil buffering functions), and (ii) extract soil DNA used to generate 16S rDNA amplicons to be analysed by next-generation sequencing (as proxy of bacterial diversity).

Figure 4.2.2. Flow chart of the experimental procedure.

Soil sampling and soil properties

The soil used in this study originated from an agricultural field in North Italy (45°05'20.8"N 9°45'59.4"E, Google Maps). The soil was characterized as loamy sand (4.2 % clay, 13.5 % silt, 82.3 % sand) with an organic matter content of 4.5 %, a pH of 7.3 and a water holding capacity of 46.8 %. The field had not been treated with pesticides for more than five years. A composite soil sample of 15 kg was prepared from randomly located sub-samples collected with a 5 cm diameter iron cylinder at a depth of 5 cm in August 2015. The soil was sieved (5 mm) and kept for one week at 20 °C in the dark until the establishment of mesocosms.

Experimental set-up

A quantity of 36 individual mesocosms of 150 g soil dwt was prepared (four replicates per treatment and per sampling date) (Supplementary Table 4.2.1). The mesocosms were kept at 40 % of the water holding capacity and treated by CHL or TCZ (active substances) or not (controls) once per month at a dose that was 10 times higher

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 211 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods than the recommended agronomical dose (2 mg/kg soil dwt (CHL) or 0.6 mg/kg soil dwt (TCZ)). Both active substances were purchased from Sigma-Aldrich (99 % chemical purity). Due to their low water solubility, they were solved in methanol and applied to 7.5 g soil dwt sub-samples of the mesocosms (i.e. 5 % of the mesocosms’ mass). The sub- samples had been dried and ground for a better dispersion of the active substance. After evaporation of the methanol, the dried sub-samples spiked with the respective pesticide were re-added to the corresponding mesocosms. The water content was re-adjusted to 40 % of the water holding capacity. The mesocosms were kept at 20 °C in the dark at constant humidity throughout the whole experiment. Control mesocosms were treated in the same manner with methanol. After 1, 2 and 3 months of incubation, the mesocosms were exploited (divided into sub-samples) for analyses (Supplementary Table 4.2.1). Microcosms of 20 g were established to measure the mineralization of the two applied pesticides (CHL or TCZ) and of the three other compounds (2,4-D, glyphosate and sodium acetate), and sub- samples of 1 g were frozen for DNA extraction.

Mineralization of 14C-labelled compounds in soil microcosms

The ability of the soil microbial community to mineralize CHL, TCZ, 2,4-D, glyphosate or sodium acetate was determined using 14C-labelled compounds as described by El-Sebai et al. (2007). Four replicates of 20 g soil microcosms were treated with CHL or TCZ at 10x recommended agronomical doses following the same procedure as described before (methanol solutions applied on dried and ground 5 % soil mass of the 20 g microcosms), or with 2,4-D, glyphosate or sodium acetate (active substances) at doses of 1.5 mg/kg (2,4-D), 3 mg/kg (glyphosate), or 2.2 mg/kg (sodium acetate) directly applied in water solutions to the 20 g microcosms. The active substances of 2,4- D, glyphosate and sodium acetate were purchased from Sigma Aldrich (99 % chemical purity). Amounts of around 0.068 µCi of 14C-ring-labelled CHL (Izotop, specific activity = 118.108 µCi/mg), 0.077 µCi of 14C-phenyl-ring-labelled TCZ (Izotop, specific activity = 143.459 µCi/mg), 0.068 µCi of 14C-triazole-ring-labelled TCZ (Izotop, specific activity = 127.567 µCi/mg), 0.091 µCi of 14C-ring-labelled 2,4-D (ARC, specific activity = 55000 µCi/mmol), 0.063 µCi of 14C-labelled glyphosate (Monsanto, specific activity = 57000 µCi/mmol) or 0.054 µCi of 14C-labelled sodium acetate (Amersham, specific

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 212 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods activity = 52000 µCi/mmol) were added to the microcosms (Supplementary Figure 4.2.1). All microcosms were kept at 40 % of the water holding capacity in the dark for 150 days in a closed jar containing a 5 mL sodium hydroxide trap (Supplementary

Figure 4.2.2). Generated 14CO2 evolved from the respective 14C-labelled compound was trapped in the sodium hydroxide solution and was quantified by liquid scintillation counting (LS 6500 Multi-Purpose Scintillation Counter, Beckman) using ACSII scintillation fluid (Amersham).

Determination of extractable and bound 14C-labelled residues in soil microcosms

For the quantification of remaining 14C-labelled extractable residues, they were extracted from the soil microcosms as described by El-Sebai et al. (2005) at the end of the incubation (150 days). Briefly, the residues were extracted by 40 mL methanol at 20 °C for 15 h, the soil solutions were centrifuged at 20 °C and 6000 rpm for 10 min, 5 mL of the supernatant were mixed with 10 mL scintillation fluid and analyzed for their radioactivity content by liquid scintillation counting. For the quantification of remaining 14C-labelled bound residues, the soil pellets were dried at 20 °C and ground. An aliquot of 0.5 g was then combusted under O2 flow at 900 °C for 4 min using a Biological Oxidizer OX-500 (EG&G Instruments) and the radioactivity was measured by liquid scintillation counting. For each microcosm, a mass balance was established containing 14CO2, 14C-labelled extractable residues and 14C- labelled bound residues, expressed as fraction of initially applied 14C-labelled compound.

Molecular analysis of soil bacterial diversity

Soil DNA from 1 g soil samples was extracted according to the ISO 11063 method (2012) initially described by Martin-Laurent et al. (2001). The procedure involved three main steps: (i) microbial cell lysis by physical (bead beating) and chemical (sodium dodecyl sulfate at 60 °C for 10 min) actions, (ii) deproteination and (iii) DNA precipitation and washing. DNA concentrations were determined with a Thermo Fisher

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 213 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods StepOnePlus thermocycler (Applied Biosystems) by using fluorescent dye of a Quant-iT™ PicoGreen® dsDNA assay kit (Invitrogen). The hypervariable V3-V4 region of the 16S rDNA (464 base pairs) of the extracted soil DNA was amplified in two runs of a first-step PCR using primers 341f_iII (5’-CCT ACG GGN GGC WGC AG-3’) and 805r (5’-GAC TAC NVG GGT ATC TAA TCC-3’). The primers contained corresponding overhang adapters (forward adapter: AAT GAT ACG GCG ACC ACC GAG ATC TAC AC, reverse adapter: CAA GCA GAA GAC GGC ATA CGA GAT) needed to add multiplexing index-sequences in a second-step PCR. First-step PCRs were carried out in 15 μL reaction volumes containing 4.25 μL sterile water, 7.5 μL 2X Phusion HF mastermix, 0.5 μL T4 gp32 (500 ng/μL), 0.375 μL of each primer (10 μM) and 2 μL of extracted DNA (1 ng/μL). The thermal cycling conditions of the first-step PCRs were 98 °C for 3 min, followed by 25 cycles of 98 °C for 30 s, 55 °C for 30 s and 72 °C for 30 s, with a final extention of 72 °C for 10 min. The second-step PCR was performed using a 384 Nextera XT index kit (Illumina) for the addition of multiplexing index-sequences. It was carried out in 30 μL reaction volumes containing 2.5 μL sterile water, 15 μL 2X Phusion HF mastermix, 0.5 μL T4 gp32 (500 ng/μL), 3 μL of each primer (10 μM) and 6 μL of the step-one PCR product. The thermal cycling conditions of the second-step PCR was 98 °C for 3 min, followed by 8 cycles of 98 °C for 30 s, 55 °C for 30 s and 72 °C for 30 s, with a final extention of 72 °C for 10 min. Second-step PCR products were further prepared (amplicon library purification, PicoGreen® quantification and pooling) and sequenced (Illumina next-generation sequencing) by Microsynth (Switzerland).

Data analysis

Estimation and comparison of mineralization curve parameters Obtained mineralization curves of 14C-ring labelled CHL, 14C-phenyl-ring-labelled TCZ, 14C-ring-labelled 2,4-D, 14C-labelled glyphosate and 14C-labelled sodium acetate under the influence of pesticide treatment (CHL or TCZ) or untreated (control) were modeled with SigmaPlot or R to calculate the time (in days) that was needed to mineralize 5 % (TCZ), 25 % (CHL, 2,4-D, glyphosate) or 50 % (sodium acetate) of initially applied radioactivity to 14CO2 (t5%, t25%, t50%). Curves of 14C-triazole-ring-labelled

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 214 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods TCZ were not modeled as the mineralization for all curves was very low (< 0.6 % after 150 days). CHL curves of soil repeatedly treated with this insecticide were fitted to a hyperbola double rectangular model with four parameters (y=ax/(b+x)+cx/(d+x)). CHL curves of untreated soil were fitted to a sigmoidal Gompertz model with three parameters (y=ae^(-e^(-x(x-x0)/b))). TCZ (phenyl-ring-labelled) curves were fitted to a hyperbola single rectangular II model with three parameters (y=ax/(b+x)+cx). 2,4-D, glyphosate and sodium acetate curves were fitted to a hyperbola double rectangular model with four parameters (y=ax/(b+x)+cx/(d+x)). The variable y signifies the mineralized percentage of initially applied radioactivity in function of the variable x signifying the time in days. Analyses of variances (ANOVA) on (i) the radioactivity that was mineralized after 150 days (14CO2) in % of initially applied radioactivity (A150d) and on (ii) t5%, t25% or t50% were performed on four replicates followed by the Tukey HSD test (p < 0.05) with XLstat.

Analysis of obtained sequences Reads were de-multiplexed, converted to FASTQ and quality-controlled and assembled using PEAR (Zhang and Sun, 2014). Unassembled reads and once-assembled reads outside the expected range were discarded. All subsequent analyses were conducted in QIIME 1.9.0 (Caporaso et al., 2010). Sequences were clustered into operational taxonomic units (OTUs) using USEARCH (Edgar, 2010) by open-reference OTU picking. Chimeras were detected and omitted using the program UCHIME (Edgar et al., 2011) with the QIIME-compatible Greengenes version of the Gold database. Taxonomy was assigned to representative sequences using the Greengenes 13.8 release. In total, 3470820 sequences in 34 samples were analyzed using QIIME (two samples were not successfully sequenced and were discarded before the analysis). Analysis was performed on all samples except for two of the four controls at 3 months, where sequencing was not successful. Several α-diversity indices were calculated with the QIIME pipeline (Caporaso et al., 2010) based on the rarefied OTU table at a depth of 22000 sequences/sample (observed species, PD whole tree, chao1 and simpson reciprocal). The indices were compared using ANOVA followed by the Tukey HSD test (p < 0.05) in XLstat. Potential phylum composition differences were investigated using a comparative bar chart. Phyla whose total relative abundances in all samples were < 0.05 % were excluded from the analysis. A Venn diagram was

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 215 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods established with the R package ‘gplots’ to indicate if time impacted the number of unique and shared genera between the different months (1 month, 2 months, 3 months). Principal coordinates analysis (PCoA) of OTU weighted unifrac distance matrices was performed and plotted with QIIME. Analyses of similarities (ANOSIM) were performed with QIIME to identify significant differences at the community level between treated and untreated samples of the same time point and between untreated samples of all time points. Significantly variating OTUs were identified using the R package ‘pamR’. ANOVA on each OTU were followed by the Tukey HSD test (p < 0.05) on four replicates with XLstat.

Results

Impact of repeated pesticide treatments on their mineralization in soil

Repeated treatments of CHL significantly enhanced the capacity of the soil to mineralize CHL (Figure 4.2.3) indicating the adaptation of CHL-degrading microorganisms. Mineralization curves completely differed in their shapes between untreated and treated samples and did not fit to the same model. Curve shapes of treated samples were exponential (i.e. followed first order kinetics), while curve shapes of untreated samples (controls) were rather linear (i.e. followed zero order kinetics) with an important lag phase. In the control samples, the mineralized radioactivity

(14CO2) after 150 days (A150d) was significantly lower than in CHL-treated samples

(Table 4.2.1). A150d values in the controls varied between 45 ± 7 % (1 month), 53 ± 1 % (2 months) and 44 ± 9 % (3 months) of initially applied radioactivity (14C-ring-labelled CHL). In comparison, treated and repeatedly treated soil samples exhibited significantly higher A150d-values of 86 ± 1 % (1 month), 80 ± 2 % (2 months) and 87 ± 1 % (3 months). Values of the time that was needed to mineralize 25 % of CHL (t25%) showed a similar trend. The 25 % mark was reached much faster in repeatedly treated samples than in untreated samples. T25%-values in treated samples were 9 ± 3 days (1 month), 2 ± 1 days (2 months) and 1 ±0 days (3 months) and significantly varied from those of untreated samples being 91 ± 7 days (1 month), 76 ± 4 days (2 months) and 98 ± 19 days (3 months). Comparison between treated samples showed that the

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 216 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods increase in CHL mineralization was proportional to the number of CHL treatments (Figure 4.2.3). The incubation time had no significant impact on CHL mineralization, as observed by comparison between the untreated samples of all months (Table 4.2.1). Complementary information to the mineralization data was obtained via a mass balance summing up (in % of initially applied 14C-labelled CHL) the mineralized fraction (A150d), the extractable fraction (bioavailable residues) and the non-extractable fraction (bound residues) (Figure 4.2.3). The overall mass balance yield was > 90 % for all samples. Both the extractable and the non-extractable fraction were lower in the repeatedly treated samples of all months (extractable: 2 ± 1 %, non-extractable: 8 ± 2 %) than in the control samples (extractable: 17 ± 7 %, non-extractable: 21 ± 4 %), reflecting the higher mineralization of CHL in treated samples. Almost no extractable CHL residues remained in CHL-treated soil. Contrariwise, repeated TCZ treatments of did not induce any changes in the capacity of the soil to degrade TCZ (Figure 4.2.3). 14C-triazole-ring-labelled TCZ (TCZ-T) was almost not mineralized at all (A150d-values remaining < 0.6 %). A few % of 14C- phenyl-ring-labelled TCZ (TCZ-P) was initially mineralized and then not anymore. Its mineralization remained very low and showed a slight but significant decrease in treated samples in comparison to untreated samples at 2 and 3 months. A150d-values of

TCZ-P remained low varying between 8 ± 0 % and 11 ± 1 % (Table 4.2.1). T5%-values varied between 31 ± 1 days and 69 ± 1 days. Although, the incubation time was found to have a significant effect on TCZ-P at 2 months (Table 4.2.1), no clear trend of time effects was observed. Data obtained from the mass balance complemented the mineralization data. The overall mass balance yield of all samples was > 87 % (Figure 4.2.3). Extracted fractions of all samples treated with TCZ-P made up 52 ± 7 %. Non-extractable fractions made up 32 ± 6 %. Extracted fractions of all samples treated with TCZ-T made up 56 ± 8 %. Non-extractable fractions made up 39 ± 8 %.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 217 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

14 14 14 Figure 4.2.3. Kinetics of CO2 evolution from C-labelled pesticides and mass balance of C-labelled residues (mineralized, extractable and non-extractable fraction) in soil untreated (control) or repeatedly treated with pesticides (CHL or TCZ). Mineralization of 14C-labelled CHL or 14C-labelled TCZ was monitored after 1, 2 and 3 months of incubation (i.e. one month after the 1st, 2nd and 3rd pesticide treatment). Initial radioactivity was applied to soil as 14C-ring-labelled CHL, as 14C-phenyl-ring-labelled TCZ (TCZ-P) or as 14C-triazole-ring-labelled TCZ (TCZ-T). Mass balance fractions represent the

14 radioactivity mineralized during 150 days ( CO2), the radioactivity extracted by methanol (extractable residues), and the radioactivity released during soil combustion (non-extractable residues). Error bars represent standard errors between four replicates.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 218 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

14 14 14 14 Table 4.2.1. Amount of CO2 evolved from the C-labelled pesticides ( C-ring-labelled CHL or C- phenyl-ring-labelled TCZ) after 150 days in % of initially applied radioactivity (A150d), and time needed to

14 mineralize 25 % (CHL) or 5% (TCZ) of the initially applied radioactivity to CO2 in days (t25% or t5%) in soil untreated (control) or repeatedly treated with pesticides (CHL or TCZ). Mineralization was monitored after 1, 2 and 3 months of incubation (i.e. one month after the 1st, 2nd and 3rd pesticide treatment). ANOVA on four replicates were performed followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

Impact of repeated pesticide treatments on soil buffering functions

CHL treatments slightly but significantly enhanced the degradation of 2,4-D and sodium acetate in soil at some time points, but had no impact on glyphosate degradation (Figure 4.2.4). 2,4-D mineralization curves had similar shapes but showed slight differences in A150d-values varying between 49 ± 1 % and 56 ± 2 % and t25%-values varying between 7 ± 0 days and 13 ± 1 days (Table 4.2.2). Significant differences (although slight) were obtained between the untreated and treated samples at 2 months and persisted at 3 months. A150d- and t25%-values at 2 months were 49 ± 1 % and

13 ± 1 days in untreated samples, and 55 ± 1 % and t25% = 9 ± 1 days in treated samples.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 219 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

A150d- and t25%-values at 3 months were 52 ± 1 % and 10 ± 0 days in untreated samples and 56 ± 2 % and 8 ± 1 days in treated samples. Regarding glyphosate mineralization in soil, no treatment effect trend was found (Figure 4.2.4). A150d-values varied between

60 ± 1 % and 64 ± 1 % (Table 4.2.2). T25%-values varied between 6 ± 0 days and 9 ± 1 days. CHL treatments significantly enhanced sodium acetate mineralization in soil at 1 month, where both the untreated (A150d = 70 ± 1 %) and the CHL-treated samples

(A150d = 75 ± 1 %) exhibited A150d-values that were higher than in all other samples

(A150d-values varying between 59 ± 2 % and 63 ± 1 %). Complementary data was obtained by the mass balance (Supplementary Figure 3.2.3). Mass balance yields of all samples were > 88 %. Extracted 2,4-D, glyphosate and sodium acetate (residue) fractions of untreated and CHL-treated samples were < 2 %, indicating that almost all easily available 2,4-D, glyphosate or sodium acetate residues were transformed by soil microorganisms. Non-extractable fractions of 2,4-D residues made up 43 ± 3 %, non- extractable fractions of glyphosate residues made up 29 ± 2 %, and non-extractable fractions of sodium acetate residues made up 29 ± 5 % in all untreated or CHL-treated samples. TCZ treatments had no effect on the soil buffering function for none of the three tested compounds (Figure 4.2.4). A150d-values varied between 49 ± 1 % and 55 ± 2 % for 2,4-D mineralization curves, between 61 ± 0 % and 64 ± 0 % for glyphosate, and between 59 ± 2 % and 70 ± 1 % for sodium acetate (Table 4.2.2). T25%-values (for 2,4-D and glyphosate) or t50%-values (for sodium acetate) varied between 7 ± 0 days and 13 ± 1 days for 2,4-D mineralization curves, between 6 ± 0 days and 9 ± 1 days for glyphosate, and between 3 ± 0 d and 29 ± 11 d for sodium acetate. The mass balance situation for untreated or TCZ-treated soil samples was the same as for untreated or CHL-treated soil samples with less than 2 % of extractable residues of 2,4-D, glyphosate and sodium acetate, and non-extractable fractions making up 44 ± 3 % (2,4-D residues), 30 ± 2 % (glyphosate residues), and 29 ± 5 % (sodium acetate residues) (Supplementary Figure 4.2.3). The incubation time was found to have significant effects on all three compounds at 2 and 3 months, as observed by comparison of A150d- and t25%-values of the untreated control samples of all months (Table 4.2.2). However, no clear trend of time effects was observed.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 220 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

14 14 Figure 4.2.4. Kinetics of CO2 evolution from C-labelled compounds (in % of initially applied radioactivity) in soil microcosms untreated (control) or repeatedly treated with pesticides (CHL or TCZ). Mineralization of 14C-ring-labelled 2,4-D, 14C-labelled glyphosate, or 14C-labelled sodium acetate was monitored after 1, 2 and 3 months of incubation (i.e. one month after the 1st, 2nd and 3rd pesticide treatment). Error bars represent standard errors between four replicates.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 221 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

14 14 14 14 Table 4.2.2. Amount of CO2 evolved from C-labelled compounds ( C-ring-labelled 2,4-D, C-labelled

14 glyphosate or C-labelled sodium acetate) after 150 days in % of initially applied radioactivity (A150d), and time needed to mineralize 25 % (2,4-D and glyphosate) or 50 % (sodium acetate) of the initially applied

14 radioactivity to CO2 in days (t25% or t50%) in soil untreated (control) or repeatedly treated with pesticides (CHL or TCZ). Mineralization was monitored after 1, 2 and 3 months of incubation (i.e. one month after the 1st, 2nd and 3rd pesticide treatment). ANOVA on four replicates were performed followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 222 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods Impact of repeated pesticide treatments on soil bacterial diversity

A first analysis of the bacterial composition in the soil microcoms revealed that it varied over the incubation time, regardless of pesticide treatment. Two of the four diversity indices (‘PD whole tree’ and ‘simpson reciprocal’) significantly increased in the untreated samples during the 3 months incubation (Table 4.2.3, Supplementary Figure 4.2.4). PCoA ordinations of OTU weighted unifrac distance matrices visualized a clear evolution of OTUs over time (Figure 4.2.5) that was identified by ANOSIM to be significant (Supplementary Table 4.2.2). However, the phyla composition did not reveal any noteworthy changes over time (Supplementary Figure 4.2.5). The most abundant phyla in all samples (control, CHL, and TCZ) were Proteobacteria (around 33 % of all phyla), Acidobacteria (around 24 %), Chloroflexi (around 4 %), Gemmatimonadetes (around 5 %), Bacteroidetes (around 4 %), and Actinobacteria (around 4 %). Analysis of shared and unique genera between all soil samples (control, CHL and TCZ) of all months indicated no time effect on the genera distribution in the samples (Supplementary Figure 4.2.6). The number of unique genera varied between 12 and 15, while the number of shared genera was 486. As indicated by these analyses, the incubation time affected the abundances of some species (composition), but the overall diversity of the microflora at higher taxonomic levels remained largely constant. To exclude that time effects disturb the detection of pesticide effects, results of control samples were compared to those of pesticide-treated samples for each sampling time point. Unfortunately, two of the four untreated samples at 3 months were not successfully sequenced and thus were excluded from the analysis, leading to less robust statistics at 3 months. Although CHL had no significant impact on the diversity indices (‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’) (Table 4.2.3, Supplementary Figure 4.2.4), repeated CHL treatments affected the composition of soil bacteria as shown by PCoA ordinations of OTU weighted unifrac matrices (Figure 4.2.5), which was identified as significant (p < 0.1) by ANOSIM (Supplementary Table 4.2.2). The strongest effect of CHLwas observed at 2 months. The significant effect of CHL at 3 months was weaker than that of 2 months, leaving the open question if the presence of four controls (instead of two controls) may have impacted the anlysis. OTUs that were impacted by repeated exposure to CHL were identified using the R package ‘pamR’. A number of 15

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 223 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods OTUs was found to be impacted by repeated exposure to CHL (Table 4.2.4, Supplementary Figure 4.2.7). CHL exposure was found to significantly decrease the abundance of six OTUs, among which one OTU was related to an unknown species of the genus Nitrospira and another one was related to an unknown species of the order Rhizobiales (class: Alphaproteobacteria). Contrariwise, the abundance of nine OTUs was significantly increased in response to CHL exposure, among which one OTU was related to an unknown species of the order Legionellales (class: Gammaproteobacteria) and another one was related to an unknown species of the genus Lysobacter (class: Gammaproteobacteria). Similarly, repeated exposure to TCZ was found to modify the composition of soil bacteria. The observed modifications were slightly more pronounced than those caused by CHL and seemed to increase with the number of TCZ treatments (strongest effects were observed at 3 months). The diversity indices ‘observed species’, ‘PD whole tree’ and ‘chao1’ were not significantly impacted by TCZ treatments. However, after the 3rd treatment (after 3 months of incubation) the ‘simpson reciprocal’ diversity index was significantly lower in soil samples exposed to TCZ than in the control samples (Table 4.2.3, Supplementary Figure 4.2.4). The ‘simpson reciprocal’ diversity index is the only index of the four used ones that considers the eveness in addition to the richness of obtained taxa, which may explain why only this diversity index was affected. In response to repeated exposure to TCZ, the soil bacterial composition at OTU level was modified as visualized by PCoA ordinations of OTU weighted unifrac matrices (Figure 4.2.5). This observation was confirmed by ANOSIM which approved TCZ effects to be significant at all time points (Supplementary Table 4.2.2). The TCZ effects seemed to increase with treatment repetitions (the most visible effect was obtained at 3 months). However, despite the increasing visual effect (Figure 4.2.5), the significance of the effect (obtained by ANOSIM) was lower at 3 months than at 2 months (Supplementary Table 4.2.2), which may be attributed to the lack of two controls at this time point. Supporting the visual assumption that the TCZ effect increased with time points (i.e. that it was higher at 3 months than at 2 months), comparison of the calculated values of the OTU weighted unifrac distance matrices led to the same conclusion. Furthermore, this finding matched with the diversity index results, where the ‘simpson reciprocal’ index was significantly decreased in TCZ-treated samples after the 3rd treatment (i.e. after 3 months of incubation). The ‘pamR’ analysis led to the identification of 23 OTUs significantly

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 224 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods impacted by repeated TCZ exposure (Table 4.2.4, Supplementary Figure 4.2.8). The abundance of 11 OTUs was significantly decreased in response to repeated exposure to TCZ, of which one OTU was related to an unknown species of the genus Nitrospira. In contrast, the abundance of 12 other OTUs was significantly higher in soil repeatedly exposed to TCZ than in control samples, of which one OTU was related to an unknown species of the genus Kaistobacter (order: Sphingomonadales, class: Alphaproteobacteria), one OTU was related to an unknown species of the genus Lysobacter (class: Gammaproteobacteria), and three OTUs were related to unknown species of the family Saprospiraceae (phylum: Bacteroidetes).

Table 4.2.3. Diversity indices of the soil bacterial community repeatedly exposed to CHL or TCZ (at 10x agronomical doses) or untreated (control). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVA of each diversity index were followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 225 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Figure 4.2.5. PCoA ordinations of OTU weighted unifrac distance matrices for soil microcosms repeatedly exposed to CHL or TCZ or untreated (control). Replicates of the same treatment per month (1st, 2nd or 3rd pesticide treatment) are represented in the same colors.

Discussion

In this paper, the impact of repeated pesticide treatments on soil microorganisms was studied via a range of methods that are not yet implemented in the pesticide risk assessment (regarding effects on soil microorganisms) of the authorization process (EC, 2015; EFSA, 2008; EFSA, 2014). The impact of repeated CHL or TCZ treatments was assessed via (i) microbial adaptation to pesticide exposure resulting in enhanced biodegradation, (ii) soil buffering functions, and (iii) the bacterial diversity in soil.

Time effects

Analyses of the results obtained for the untreated control samples all along the experiment revealed that the time had no effect on the capability of soil microorganisms to mineralize CHL or TCZ-T, but affected their capability to mineralize the other tested compounds (TCZ-P, 2,4-D, glyphosate, sodium acetate), without revealing any trend. In

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 226 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods addition, the composition of soil bacteria was shown to be influenced by time, as suggested by increasing diversity indices and significant ANOSIM of the PCoA. The bacterial community was dominated by the phyla Proteobacteria, Acidobacteria, Chloroflexi, Gemmatimonadetes, Bacteroidetes and Actinobacteria that had also been found to be dominant in a range of other soils (Acosta-Martinez et al., 2008; Janssen, 2006; Merlin et al., 2015; Petric et al, 2011; Romdhane et al., 2016; Wessen et al., 2010). The abundances of these dominant phyla were not affected by time.

CHL effects

Repeated exposure to CHL had a clear effect on soil microorganisms. Indeed, the capability of soil microorganisms to mineralize CHL was significantly higher (almost twice higher after 150 days) and faster (25 % were mineralized around 85 days faster) in soil microcosms repeatedly treated with this insecticide as compared to untreated soil microcosms. This suggests that repeated CHL exposure led to the adaptation of the soil microbial community leading to enhanced biodegradation of CHL, because of the emergence of CHL-degrading microorganisms able to use CHL as an additional nutrient and/or energy source providing them with a selective advantage over other microorganisms (Copley, 2009). Adaptation of soil microorganisms resulting in enhanced biodegradation seems to be a common response to pesticide treatments, as it was also already reported for several other pesticides (Crouzet et al., 2010; De Andrea et al., 2003; Hussain et al., 2011; Vischetti et al., 2008; Weaver et al., 2007). Adaptation of microorganisms to pesticides leading to their enhanced biodegradation is an environmentally friendly process that decreases the persistence of pesticide residues in the soil and thereby limits their dispersion and ecotoxicological impact to non-target organisms (Topp, 2003). Mass balance results indicated that the extractable CHL fraction had decreased in soil microcosms repeatedly treated with CHL as compared to respective controls, where the extractable fraction was clearly higher. This further indicates that adaptation of soil microorganisms and the resulting enhanced CHL biodegradation led to the almost entire mineralization of bioavailable CHL. This is in accordance with the general reality that bioavailable compounds are easily metabolized by microorganisms (Krell et al., 2013). However, bound CHL residues that are hardly bioavailable and known to be persistent in soil (Reid et al., 2000) were formed in lower

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 227 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods amounts in repeatedly treated soil microcosms than in respective controls. Around 8 % of each CHL treatment was found to persist and accumulate as bound residues in the adapted soil microcosms, while around 21 % persisted in the non-adapted controls, representing a long-lasting source of contamination (Barraclough et al., 2005). Analysis of the soil buffering functions estimated by using as proxy the mineralization of three organic compounds (2,4-D, glyphosate or sodium acetate) revealed that all tested molecules were readily mineralized by soil microorganisms. Mass balance results showed that extracted fractions of 2,4-D, glyphosate or sodium acetate in all samples (control, CHL and TCZ) were < 2 %, suggesting that the soil microorganisms were able to degrade all the bioavailable fraction of 2,4-D, glyphosate or sodium acetate. This conforms to several other studies describing them as degradable compounds in the environment (Dick and Quinn, 1995; Ka et al., 1994; Rueppel et al., 1977; Sandmann and Loos, 1984). 2,4-D and sodium acetate mineralizations were slightly but significantly increased in response to repeated CHL exposure (for 2,4-D significant after 2 and 3 months of incubation and for sodium acetate significant after 1 month of incubation). This may be due to their degradation by soil microorganisms that were stimulated by repeated exposure to CHL. However, to draw meaningful conclusions about the impact of CHL on soil buffering functions, the mineralizations of a wider range of compounds need to be studied. Glyphosate degradation was not affected by CHL treatments, suggesting either that glyphosate-degrading microorganisms were not affected by CHL, or that the ability of glyphosate degradation was redundant enough among the soil microbial community to be impacted by repeated CHL exposure. In addition, this observation indicates that potential impacts of pesticides on the mineralization of various compounds are compound-specific making difficult their interpretation in terms of consequences on soil buffering functions from a general point of view. In addition, CHL was found to affect the composition of bacteria in soil. Although the effects did not appear in diversity indices or as visual effects in the phyla composition, PCoA ordinations of OTU weighted unifrac distance matrices displayed significant effects of CHL treatments on soil bacteria at OTU level (15 OTUs were found to be affected). The abundance of six OTUs decreased in soil microcosms repeatedly exposed to CHL among which one OTU was related to an unknown species of the genus Nitrospira and and another one was related to an unknown species of the order

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 228 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods Rhizobiales. Species of the genus Nitrospira are important key players in the biogeochemical nitrogen cycle that oxidize nitrite to nitrate (Altmann et al., 2003). This process is especially important in the global N2O greenhouse gas production (Ma et al., 2008; Philippot and Germon, 2005) or in aquatic ecosystems to counteract the accumulation of toxic nitrite concentrations (Hovanec et al., 1998). Species of the order Rhizobiales are likewise important for the environment, as some of them fix nitrogen and are symbiotic with leguminous plant roots making them important for soil fertility in agriculture (Gage, 2004). The presence of anthropogenic chemicals (such as pesticide residues) was found to generally reduce Rhizobium populations in soil (Slattery et al., 2001). Potential indirect effects of CHL on processes involved in the nitrogen cycle thus have to be further investigated. The abundance of nine OTUs significantly increased in response to CHL treatments, among which one OTU was related to an unknown species of the order Legionellales and another OTU was related to an unknown species of the genus Lysobacter. Species of the order Legionellales notably include famous human pathogens such as Legionella pneumophila causing Legionnaires’ disease (Garrity et al., 2001). Thus, a potential impact of CHL on the abundance of human pathogenic species in soil may have an interest to be further studied to prevent consequences for the environment and human health. Species of the genus Lysobacter are known for their production of extracellular enzymes and antibiotics (Ahmed et al., 2003; Kimura et al., 1996; Von Tigerstrom, 1980). These features make Lysobacter an effective competitor against rival microorganisms (Vasilyeva et al., 2008) in the constant microbial war for nutrients in soil (Fontaine et al., 2003). Interestingly, it is possible that Lysobacter was in some ways indirectly strengthened by CHL exposure that potentially weakened a range of other soil bacteria getting weaker rivals for Lysobacter. It is not presumed that CHL- degrading species appeared as OTUs whose abundance significantly increased due to CHL treatments (if CHL was degraded by bacteria and not by fungi). First, the degradation may be a result of a relatively complex functional community of different species, as previously shown for atrazine (Devers et al., 2007). Second, the abundance of pesticide-degraders in soil usually remains quite low (around 103 to 104 cells per g of soil) (Piutti et al., 2003) making them undetectable within the microbial community dominated by populations averaging 108 cells per g of soil. This illustrates that possible changes in the abundance of rare species with big impacts in ecosystem functions (here: degradation of CHL residues) remain difficult to be detected by next-generation

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 229 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods sequencing of 16S rDNA amplicons, suggesting that although powerful, this method is not yet appropriate to study the ecology of rare species (such as pesticide-degrading bacteria) in soil. Moreover, one could hypothesize that detectable changes in the composition of the microbial community may hide modifications of the rare biosphere, thereby conducting to misleadings in trying to link shifts in diversity to consequences on supported ecosystem functions. In this study, the CHL dose repeatedly applied to the soil microcosms was 10 times higher than the recommended agronomical dose, therefore constituting a worst case scenario of exposure, which accords with a range ofother studies that concluded CHL to produce ecotoxicological effects on the soil microbial abundance, diversity or activity (Chu et al., 2008; Dutta et al., 2010; Eisenhauer et al., 2009; Fang et al., 2009; Gilani et al., 2010; Gupta et al., 2013; Jastrzebska, 2011; Pozo et al., 1995; Srinivasulu and Rangaswamy, 2013; Wang et al., 2010).

TCZ effects

Repeated exposure of the soil microbial community to TCZ did not induce enhanced TCZ mineralization. Mineralization of 14C-triazole-ring-labelled TCZ was negligible (< 0.6 %). Mineralization of 14C-phenyl-ring-labelled TCZ was low (around 10 %) in all samples (treated or untreated). This indicated that only the phenyl-ring of TCZ was slightly degraded by soil microorganisms while the triazole-ring remained largely untouched. This is in accordance with results of a study performed with the same soil (Storck et al., 2016b), where 34 TCZ transformation products were detected of which 31 had an intact triazole-ring. Only 3 transformation products had a cleaved, differentiated or degraded triazole-ring while 10 transformation products had a cleaved, differentiated or degraded phenyl-ring. However, detected transformation products were not quantified in that study (due to the lack of available reference standards). Yet, this information would be needed for a more substantial comparison between the findings of this study and the findings of Storck et al. (2016b). This trend is further confirmed by other studies performed in (mostly soil) environments all over the world leading to the detection of around 50 TCZ transformation products in the environment. The phenyl-ring is cleaved, differentiated or degraded in around 30 of them, while only 3 transformation products are known that have a cleaved, differentiated or degraded

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 230 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods triazole-ring (see Storck et al., 2016b for a review). Mass balance results showed that large parts of the bioavailable TCZ residues fraction remained in the soil (extractable fractions made up around 50 %). Around 36 % remained in the soil as bound TCZ residues (non-extractable fraction). Moreover, this study suggested that around 90 % of applied TCZ (of each of the three treatments) may remain in the tested soil as the sum of TCZ and transformation products with intact phenyl-ring, while > 99 % may remain in the soil as sum of TCZ and transformation products with intact triazole ring. Transformation products with intact triazole ring are particularly relevant as they are persistent and possibly interact with the hormone regulation network of non-target organisms by inhibiting the cytochrome P450-dependent conversion of lanosterol to ergosterol (Rieke et al., 2014; Shalini et al., 2011). Their accumulation in exposed soils due to the lack of microbes able to degrade them may thus be a potential threat to non- target organisms. Repeated exposure to TCZ was not found to modify the soil buffering functions assessed by the measurement of the mineralization of 2,4-D, glyphosate and sodium acetate, that were mineralized at 50 to 80 %. Amounts of < 2 % accounted to the bioavailable (extractable) fraction that remained in the soil after 150 days, indicating that easily available fractions were readily degraded. The rest remained as bound residues in the soil (non-extractable fraction). Contrariwise to CHL that induced slight but significant changes of tested soil buffering functions, TCZ did not affect the tested soil buffering functions, therefore indicating that effects on soil buffering functions are pesticide-dependent. Although repeated exposure to TCZ neither induced microbial adaptation leading to its enhanced mineralization, nor affected the tested soil buffering functions, it was found to have clear effects on the composition of soil bacteria. Even though only one of the four diversity indices was significantly impacted by repeated exposure to TCZ treatments (‘simpson reciprocal’ index at 3 months) and no visual effects appeared in the phyla composition, PCoA ordinations of OTU weighted unifrac distance matrices clearly displayed the effects of repeated TCZ treatments at OTU level (23 OTUs were found to be affected). TCZ effects were significant at all sampling time points and even seemed to increase with treatment repetitions (although an increase could not be proven by ANOSIM, probably due to the lack of two controls at 3 months). Interestingly, among the 11 sensitive OTUs whose abundance significantly decreased in response to

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 231 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods TCZ exposure, one OTU was related to an unknown species of the genus Nitrospira. It is noteworthy that the abundance of a similar OTU of this nitrite-oxidizing genus was also found to be decreased by CHL treatments in this study. This suggests that Nitrospira may be sensitive to the presence of various pesticides. If this hypothesis is correct, the common application of pesticide cocktails (mix of several pesticides) in conventional agriculture (Faasen, 1995) may even strengthen the negative effect on Nitrospira important in global nitrogen cycling (Altmann et al., 2003; Ma et al., 2008; Philippot and Germon, 2005). Contrariwise, the abundance of 12 OTUs was increased in response to TCZ exposure including one OTU related to an unknown species of the genus Kaistobacter, one OTU related to an unknown species of the genus Lysobacter, and three OTUs related to unknown species of the family Saprospiraceae. The genus Kaistobacter belongs to the family of Sphingomonadaceae that is known for the ability of some species to degrade some aromatic compounds, including substituted phenyl-urea herbicides (e.g. Spingomonas sp. SRS2 (Sorensen et al., 2001), Sphingobium sp. YBL2 (Gu et al., 2013) or Sphingomonas sp. SH (Hussain et al., 2011)). It is thus not excluded that Kaistobacter may be involved in TCZ degradation, but this hypothesis has to be tested by further research. Compatibly to the enhancing effects of CHL on the abundance of the genus Lysobacter, the abundance of another OTU of this genus was also increased by TCZ treatments. This parallelism reinforces the hypothesis that Lysobacter (that is known for its features to effectively compete against rival microorganisms by using extracellular enzymes and antibiotics (Ahmed et al., 2003; Kimura et al., 1996; Von Tigerstrom, 1980)) may be indirectly strengthened by the application of pesticides such as CHL or TCZ potentially weakening rivals of Lysobacter. Moreover, this hypothesis is further supported, as the abundance of three OTUs related to unknown species of the family Saprospiraceae was increased by TCZ treatments. One species of this family (Saprospira grandis) is known for its ability to capture other microorganisms via the production of slime and to lyse them via the production of exotoxins and enzymes with the purpose of using the released nutrients (Lewin, 1997). In the same manner as for Lysobacter species, species of the family Saprospiraceae may be indirectly strengthened by the presence of microbial stress factors such as pesticides that may weaken potential victims of these predators. Further studies may be of interest to investigate if the pesticide effects on these species have a measurable effect on soil functions or if effects on some species may be buffered by others. Furthermore, species that were affected by

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 232 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods pesticide treatments in this study may be used as indicator species for a screening/monitoring approach testing the ecotoxicity of pollutants such as pesticide residues in environmental soil samples. TCZ was repeatedly applied to soil microcoms at 10 times the recommended agronomical dose in order to simulate a high pesticide exposure scenario and was found to affect soil microorganisms, which is in agreement with the other studies, where TCZ was found to impact soil microorganisms, as evaluated by microbial abundance, diversity or activity analyses (Anuradha et al., 2016; Bending et al., 2007; Cycon et al., 2006; Ferreira et al., 2009; Munoz-Leoz et al., 2011; Wang et al., 2016).

Conclusion

Aim of this study was to pesticide risk assessment with innovative research using alternative techniques to study the impact of pesticides on soil microorganisms. The findings of this study fit into the category of studies concluding pesticide impacts on soil microorganisms, and may be more comprehensive than risk assessment studies, where conclusions are based on broad indicators (carbon and nitrogen mineralization) that are weak indicators for the microbial toxicity of pesticides, as they are too global and thus impacts tend to remain undetected if they are buffered by functional redundancy. The methods tested in this study to investigate pesticide impacts on the diversity and activity of soil microorganisms were found to be useful (within the limits of what is possible today) to detect ecotoxicological impacts of pesticides. An adaptation of degrading microorganisms to repeated treatments seemed to be sensitively detectable via measuring the mineralization of the applied pesticides. Quantification of the mineralization of other organic pollutants to estimate soil buffering functions seemed to be compound-specific and needs further development, notably by testing the mineralization of more compounds and by acquiring more data on different soil types to establish normal operating ranges of soil buffering capacities. In contrast, sequencing methods were concluded to be suitable for the detection of shifts in the bacterial composition in response to repeated pesticide exposure, but one could observe that changes in the rare species, notably those responsible for CHL mineralization, remained undetected. In addition, it remains open to which extent such detectable impacts on soil

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 233 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods microorganisms can be related to potential effects on ecosystem functions. It is difficult to understand the functioning of all these complex processes and interactions, and that is why it is a real challenge to drive steady conclusions about the impact of any pesticide on non-target organisms in the environment. Anyway, results of this study have the potential to enrich future post- authorization risk assessments of CHL (due in 2020) and TCZ (due in 2023) with additional information.

Acknowledgements

The PhD studies of Veronika Storck performed at the Doctoral School of Environment and Health (ED E2S) at the University of Burgundy were funded by the French Ministry of Education and Research (MESR) (grant number 2013-31). This work was done within the framework of the FP7-PEOPLE-2012-IAPP (Industry-Academia Partnerships and Pathways) Marie-Curie project ‘Love-to-Hate’ funded by the European Commission (grant number 324349).

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 234 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods Supplementary material

Supplementary Table 4.2.1. Details of the sampled soil amounts of the established mesocosms per treatment (CHL, TCZ or control) and time (1 month, 2 months, 3 months). TCZ-P abbreviates for 14C- phenyl-ring-labelled TCZ. TCZ-T abbreviates for 14C-triazole-ring-labelled TCZ.

Supplementary Figure 4.2.1. Location of 14C atoms in 14C-ring-labelled CHL, 14C-phenyl-ring-labelled TCZ, 14C-triazole-ring-labelled TCZ, 14C-ring-labelled 2,4-D, 14C-labelled glyphosate and 14C-labelled sodium acetate.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 235 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Supplementary Figure 4.2.2. Setup of mineralization jars (closed system). A 14C-labelled compound is

14 applied to a soil microcosm and generated CO2 is trapped in sodium hydroxide that can be exchanged to quantify the generated radioactivity.

Supplementary Figure 4.2.3. Mass balance in % of initially applied radioactivity. The fractions represent

14 the radioactivity mineralized during 150 d ( CO2), the radioactivity extracted by methanol, and the radioactivity released during soil combustion (non-extractable fraction). Initial radioactivity was applied as 14C-ring-labelled 2,4-D, as 14C-labelled glyphosate or as 14C-labelled sodium acetate to soil untreated (control) or repeatedly treated (once per month) with pesticides (CHL or TCZ). Error bars represent standard errors between four replicates.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 236 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Supplementary Figure 4.2.4. Bacterial diversity (observed species, PD whole tree, chao1, simpson reciprocal) in soil repeatedly treated (once per month) with CHL or TCZ or untreated (control). ANOVA for each diversity index were followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 237 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Supplementary Table 4.2.2. ANOSIM of the PCoA ordinations of OTU weighted unifrac distance matrices for soil repeatedly treated (once per month) with CHL or TCZ or untreated (control). Significant differences between groups are indicated by p < 0.1.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 238 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Supplementary Figure 4.2.5. Relative abundance (%) of prokaryotic phyla in soil repeatedly treated (once per month) with CHL or TCZ or untreated (control). Each color represents a phylum. Phyla whose abundance sum of all samples was < 0.05 % are grouped as ‘others’. Sequences that could not be assigned to a phylum are grouped as ‘unassigned’.

Supplementary Figure 4.2.6. Venn diagrams indicating the number of shared and unique genera over time in soil repeatedly treated with CHL or TCZ or untreated.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 239 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Supplementary Figure 4.2.7. Number of sequences per sample of OTUs altering between soil samples repeatedly treated with CHL (once per month) or untreated (control). Each bar represents four replicates. ANOVA for each taxon were followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 240 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Supplementary Figure 4.2.8. Number of sequences per sample of OTUs altering between soil samples repeatedly treated with TCZ (once per month) or untreated (control). Each bar represents four replicates. ANOVA for each taxon were followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 241 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods References

Acosta-Martinez, V., Dowd, S., Sun, Y., and Allen, V. (2008). Tag-encoded pyrosequencing analysis of bacterial diversity in a single soil type as affected by management and land use. Soil Biology and Biochemistry 40, 2762-2770. Ahmed, K.., Chohnan, S., Ohashi, H., Hirata, T., Masaki, T. and Sakiyama, F. (2003). Purification, bacteriolytic activity, and specificity of β-lytic protease from Lysobacter sp. IB-9374. Journal of Bioscience and Bioengineering 95, 27-34. Altmann, D., Stief, P., Amann, R., De Beer, D. and Schramm, A. (2003). In situ distribution and activity of nitrifying bacteria in freshwater sediment. Environmental Microbiology 5 (9), 798-803. Anuradha, B., Rekhapadmini, A. and Rangaswamy, V. (2016). Influence of tebuconazole and copper hydroxide on phosphatase and urease activities in red sandy loam and black clay soils. 3 Biotech (6), 78. Baelum, J., Nicolaisen, M.H., Holben, W.E., Strobel, B.W., Sorensen, J. and Jacobsen, C.S. (2008). Direct analysis of tfdA gene expression by indigenous bacteria in phenoxy acid amended agricultural soil. The ISME Journal 2, 677-687. Barraclough, D., Kearney, T. and Croxford, A. (2005). Bound residues: environmental solution or future problem? Environmental Pollution 133 (1), 85-90. Bending, G.D., Rodriguez-Cruz, M.S. and Lincoln, S.D. (2007). Fungicide impacts on microbial communities in soils with contrasting management histories. Chemosphere 69, 82-88. Bondeau, A., Smith, C.M., Zaehle, S., Schapfhoff, S., Lucht, W., Cramer, W., Gerten, D., Lotze-Campen, H., Mueller, C., Reichstein, K. and Smith, B. (2007). Modelling the role of agriculture for the 20th century global terrestrial carbon balance. Global Change Biology 13, 679-706. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D. and Costello, E.K. (2010). QIIME allows analysis of high- throughput community sequencing data. Nature Methods 7, 335-336. Chu, X., Fang, H., Pan, X., Wang, X., Shan, M., Feng, B. and Yu, Y. (2008). Degradation of chlorpyrifos alone and in combination with chlorothalonil and their effects on soil microbial populations. Journal of Environmental Sciences 20, 464-469. Copley, S.D. (2009). Evolution of efficient pathways for degradation of anthropogenic chemicals. Nature Chemical Biology 5, 559-566. Crouzet, O., Batisson, I., Besse-Hoggan, P., Bonnemoy, F., Bardot, C., Poly, F., Bohatier, J and Mallet, C. (2010). Response of soil microbial communities to the herbicide mesotrione: a dose-effect microcosmapproach. Soil Biology and Biochemistry 42, 193-202. Cycon, M., Piotrowska-Seget, Z. and Kaczynska, A. (2006). Microbiological characteristics of a sandy loam soil exposed to tebuconazole and λ-cyhalothrin under laboratory conditions. Ecotoxicology 15, 639-646. De Andrea, M.M., Peres, T.B., Luchini, L.C., Bazarin, S., Papini, S., Matallo, M.B. and Tedeschi Savoy, V.L. (2003). Influence of repeated applications of glyphosate on its persistence and soil bioactivity. Pesquisa Agropecuaria Brasileira 38 (11), 1329-1335. Devers, M., El Azhari, N., Udikovic-Kolic, N. and Martin-Laurent, F. (2007). Detection and organization of atrazine-degrading genetic potential of seventeen bacterial isolates belonging to divergent taxa indicate a recent common origin of their catabolic functions. FEMS Microbiology Letters 273, 78-86. Dick, R.E. and Quinn, J.P. (1995). Glyphosate-degrading isolates from environmental samples: occurrence and pathways of degradation. Applied Microbiology and Biotechnology 43 (3), 545-550. Dutta, M., Sardar, D., Pal, R. and Kole, R.K. (2010). Effect of chlorpyrifos on microbial biomass and activities in tropical clay loam soil. Environmental Monitoring and Assessment 160, 385-391. EC (2005). Review report for the active substance chlorpyrifos. Finalised in the Standing Committee on the Food Chain and Animal Health at its meeting on 3 June 2005 in view of the inclusion of chlorpyrifos in Annex I of the Directive 91/414/EEC. SANCO/3059/99 - rev. 1.5. Edgar, R.C. (2010). Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26, 2460-2461. EFSA (2008). Conclusion on the peer review of tebuconazole. EFSA Scientific Report 176, 1-109. EFSA (2010). Scientific opinion on the development of specific protection goal options for environmental risk assessment of pesticides, in particular in relation to the revision of the guidance documents on aquatic and terrestrial ecotoxicology (SANCO/3268/2001 and SANCO/10329/2002), EFSA Panel on Plant Protection Products and their Residues (PPR). EFSA Journal 2010; 8 (10). 1821. EFSA (2014). Conclusion on the peer review of the pesticide risk assessment of the active substance tebuconazole. EFSA Journal 12 (1), 3485. http://dx.doi.org/10.2903/j.efsa.2014.3485, 98. Eisenhauer, N., Klier, M., Partsch, S., Sabais, A.C.W., Scherber, C., Weisser, W.W. and Scheu, S. (2009). No interactive effects of pesticides and plant diversity on soil microbial biomass and respiration. Applied Soil Ecology 42, 31-36. El-Sebai, T., Lagacherie, B., Soulas, G. and Martin-Laurent, F. (2004). Isolation and characterisation of an isoproturon-mineralising Methylopila sp. TES from French agricultural soil. FEMS Microbiology Letters 239, 103-110. El-Sebai, T., Lagacherie, B., Cooper, J.F., Soulas, G. and Martin-Laurent, F. (2005). Enhanced isoproturon mineralization in a clay silt loam agricultural soil. Agronomy for Sustainable Development 25, 271-277. El-Sebai, T., Lagacherie, B., Soulas, G. and Martin-Laurent, F. (2007). Spatial variability of isoproturon mineralizing activity within an agricultural field: geostatistical analysis of simple physicochemical and microbiological soil parameters. Environmental Pollution 145, 680-690. Enserink, M., Hines, P.J., Vignieri, S.N., Wigginton, N.S. and Yeston, J.S. (2013). The pesticide planet. Science 341, 730-731. Faasen, R. (1995). Agricultural pesticide use – a threat to the European environment? European Water Pollution Control 5 (2), 34-40. Fang, H., Yu, Y., Chu, X., Wang, X., Yang, X. and Yu, J. (2009). Degradation of chlorpyrifos in laboratory soil and its impact on soil microbial functional diversity. Journal of Environmental Sciences 21, 380-386. Ferreira, E.P.B., Dusi, A.N., Costa, J.R., Xavier, G.R. and Rumjanek, N.G. (2009). Assessing insecticide and fungicide effects on the culturable soil bacterial community by analyses of variance of their DGGE fingerprinting data. European Journal of Soil Biology 45, 466-472. Fontaine, S., Mariotti, A. and Abbadie, L. (2003). The priming effect of organic matter: a question of microbial competition? Soil Biology and Biochemistry 35 (6), 837-843. Gage, D.J. (2004). Infection and invasion of roots by symbiotic, nitrogen-fixing Rhizobia during nodulation of temperate legumes. Microbiology and Molecular Biology Reviews 68 (2), 280-300. Garrity, G.M., Boone, D.R. and Castenholz, R.W. (2001). Bergey’s Manual of Systematic Bacteriology. 2nd edition, vol. 1, Springer New York.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 242 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Gilani, S.T.S., Ageen, M., Shah, H. and Raza, S. (2010). Chlorpyrifos degradation in soil and its effect on soil microorganisms. The Journal of Animal and Plant Sciences 20 (2), 99-102. Graham, E.B., Knelman, J.E., Schindlbacher, A., Siciliano, S., Breulmann, M., Yannarell, A., Beman, J.M., Abell, G., Philippot, L., Prosser, J., Foulquier, A., Yuste, J.C., Glanville, H.C., Jones, D.L., Angel, R., Salminen, J., Newton, R.J., Bürgmann, H., Ingram, L.J., Hamer, U., Siljanen, H.M.P., Peltoniemi, K., Potthast, K., Bañeras, L., Hartmann, M., Banerjee, S., Yu, R.-Q., Nogaro, G., Richter, A., Koranda, M., Castle, S.C., Goberna, M., Song, B., Chatterjee, A., Nunes, O.C., Lopes, A.R., Cao, Y., Kaisermann, A., Hallin, S., Strickland, M.S., Garcia-Pausas, J., Barba, J., Kang, H., Isobe, K., Papaspyrou, S., Pastorelli, R., Lagomarsino, A., Lindström, E.S., Basiliko, N. and Nemergut, D.R. (2016). Microbes as engines of ecosystem function: When does community structure enhance predictions of ecosystem processes? Frontiers in Microbiology 7 (214). Gu, T., Zhou, C., Sørensen, S.R., Zhang, J., He, J., Yu, P., Yan, X. and Li, S. (2013). The novel bacterial N-demethylase pdmAB is responsible for the initial step of N,N-dimethyl-substituted phenylurea herbicide degradation. Applied and Environmental Microbiology 79 (24), 7846-7856. Gupta, S., Gupta, R. and Sharma, S. (2013). Impact of chemical- and bio-pesticides on bacterial diversity in rhizosphere of Vigna radiata. Ecotoxicology 22, 1479-1489. Haygarth, P. and Ritz, K. (2009). The future of soils and land use in the UK: Soil systems for the provision of land-based ecosystem services. Land Use Policy 26, 187-197. Hillel, D. (2009). The mission of soil science in a changing world. Journal of Plant Nutrition and Soil Science 172, 5-9. Hovanec, T.A., Taylor, L.T., Blakis, A. and Delong, E.F. (1998). Nitrospira-like bacteria associated with nitrite oxidation in freshwater aquaria. Applied and Environmental Microbiology 64 (1), 258-264. Hussain, S., Devers-Lamrani, M., El Azhari, N. and Martin-Laurent, F. (2011). Isolation and characterization of an isoproturon mineralizing Sphingomonas sp. strain SH from a French agricultural soil. Biodegradation 22, 637-650. Hutzinger, O. (2013). The handbook of environmental chemistry - anthropogenic compounds, 3rd edition, Springer, Berlin Heidelberg. Janssen, P. H. (2006). Identifying the dominant soil bacterial taxa in libraries of 16S rRNA and 16S rRNA genes. Applied and Environmental Microbiology 72, 1719-1728. Jastrzebska, E. (2011). The effect of chlorpyrifos and teflubenzuron on the enzymatic activity of soil. Polish Journal of Environmental Studies 20 (4), 903-910. Jeyaratnam, J. (1990). Acute pesticide poisoning: A major global health problem. World Health Statistics Quarterly 43 (3), 139-144. Ka, J.O., Holben, W.E. and Tiedje, J.M. (1994). Genetic and phenotypic diversity of 2,4-dichlorophenoxyacetic acid (2,4-D)-degrading bacteria isolated from 2,4-D-treated field soils. Applied and Environmental Microbiology 60 (4), 1106-1115. Karpouzas, D.G., Tsiamis, G., Trevisan, M., Ferrari, F., Malandain, C., Sibourg, O. and Martin-Laurent, F. (2016). "LOVE TO HATE" pesticides : felicity or curse for the soil microbial community ? An FP7 IAPP Marie Curie project aiming to establish tools for the assessment of the mechanisms controlling the interactions of pesticides with soil microorganisms. Environmental Science and Pollution Research 23 (18), 18947-18951. Keesstra, S.D., Geissen, V., Mosse, K., Piiranen, S., Scudiero, E., Leistra, M. and van Schaik, L. (2012). Soil as a filter for groundwater quality. Current Opinion on Environmental Sustainability 4 (5), 507-516. Kimura, H., Izawa, M. and Sumino, Y. (1996). Molecular analysis of the gene cluster involved in cephalosporin biosynthesis from Lysobacter lactamgenus YK90. Applied Microbiology and Biotechnology 44, 589-596. Krell, T., Lacal, J., Reyes-Darias, J.A., Jumenez-Sanchez, C., Sungthong, R. and Ortega-Calvo, J.J. (2013). Bioavailability of pollutants and chemotaxis. Current Opinion in Biotechnology 24 (451-456). Lewin, R.A. (1997). Saprospira grandis: A flexibacterium that can catch bacterial prey by ‘ixotrophy’. Microbial Ecology 34 (3), 232- 236. Lewis, S.L. and Maslin, M.A. (2015). Defining the Anthropocene. Nature 519, 171-180. Ma, W.K., Bedard-Haughn, A., Siciliano, S.D. and Farrell, R.E. (2008). Relationship between nitrifier and denitrifier community composition and abundance in predicting nitrous oxide emissions from ephemeral wetland soils. Soil Biology and Biochemistry 40 (5), 1114-1123. Malaja, E., von der Ohea, P.C., Groted, M., Kühnee, R., Mondyf, C.P., Usseglio-Polaterag, P., Bracka, W. and Schäfer, R.B. (2013). Organic chemicals jeopardize the health of freshwater ecosystems on the continental scale. Proceedings of the National Academy of Sciences 111 (26), 9549-9554. Martin-Laurent, F., Kandeler, E., Pertic, I., Djuric, S. and Karpouzas, D.G. (2013). ECOFUN-MICROBIODIV: an FP7 European project for developing and evaluating innovative tools for assessing the impact of pesticides on soil functional microbial diversity- towards new pesticide registration regulation? Environmental Science and Pollution Research 20, 1203-1205. Merlin, C., Devers, M., Béguet, J., Boggio, B., Rouard, N., and Martin-Laurent, F. (2015). Evaluation of the ecotoxicological impact of the organochlorine chlordecone on soil microbial community structure, abundance, and function. Environmental Science and Pollution Research 23, 4185-4198. Munoz-Leoz, B., Ruiz-Romera, E., Antigëdad, I. and Garbisu, C. (2011). Tebuconazole application decreases soil microbial biomass and activity. Soil Biology and Biochemistry 43, 2146-2183. Petric, I., Bru, D., Udikovic-Kolic, N., Hrsak, D., Philippot, L., and Martin-Laurent, F. (2011). Evidence for shifts in the structure and abundance of the microbial community in a long-term PCB-contaminated soil under bioremediation. Journal of Hazardous Materials 195, 254-260. Philippot, L. and Germon, J.C. (2005). Contribution of bacteria to initial input and cycling of nitrogen in soils. Microorganisms in Soils: Roles in Genesis and Functions 3, 159-176. Philippot, L., Ritz, K., Pandard, P., Hallin, S. and Martin-Laurent, F. (2012). Standardisation of methods in soil microbiology: progress and challenges. FEMS Microbiology Ecology 82, 1-10. Piutti, S., Hallet, S., Rousseaux, S., Philippot, L., Soulas, G. and Martin-Laurent, F. (2002). Accelerated mineralisation of atrazine in maize rhizosphere soil. Biology and Fertility of Soils 36, 434-441. Piutti, S., Semon, E., Landry, D., Hartmann, A., Dousset, S., Lichtfouse, E., Topp, E., Soulas, G. and Martin-Laurent, F. (2003). Isolation and characterisation of Nocardioides sp. SP12, an atrazine degrading bacterial strain possessing the gene trzN from bulk- and maize rhizosphere soil. FEMS Microbiology Letters 221, 111-117. Pozo, C., Martinez-Toledo, V., Salmeron, V., Rodelas, B. and Gonzalez-Lopez, J. (1995). Effect of chlorpyrifos on soil microbial activity. Environmental Toxicology and Chemistry 14 (2), 187-192. Reid, B.J., Jones, K.C. and Semple, K.T. (2000). Bioavailability of persistent organic pollutants in soils and sediments - a perspective on mechanisms, consequences and assessment. Environmental Pollution 108 (1), 102-112.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 243 Chapter 4.2 Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods

Rieke, S., Koehn, S., Hirsch-Ernst, K., Pfeil, R., Kneuer, C. and Marx-Stoelting, P. (2014). Combination effects of (tri)azole fungicides on hormone production and xenobiotic metabolism in a human placental cell line. International Journal of Environmental Research and Public Health 11, 9660-9679. Romdhane, S., Devers-Lamrani, M., Barthelmebs, L., Calvayrac, C., Bertrand, C., Cooper, J.-F., Dayan, F.E. and Martin-Laurent, F. (2016). Ecotoxicological impact of the bioherbicide Leptospermone on the microbial community of two arable soils. Frontiers in Microbiology 7, 775. Rueppel, M.L., Brightwell, B.B., Schaefer, J. and Marvel, J.T. (1977). Metabolism and degradation of glyphosate in soil and water. Journal of Agricultural and Food Chemistry 25 (3), 517-528. Sandmann, E.R.I.C. and Loos, M.A. (1984). Enumeration of 2,4-D-degrading microorganisms in soils and crop plant rhizospheres using indicator media; high populations associated with sugarcane (Saccharum officinarum). Chemosphere 13 (9), 1073- 1084. Schloter, M., Dilly, O. and Munch, J.C. (2003). Indicators for evaluating soil quality. Agriculture, Ecosystems and Environment 98, 255- 262. Shalini, K., Kumar, N., Drabu, S. and Kumar Sharma, P. (2011). Advances in synthetic approach to and antifungal activity of triazoles. Beilstein Journal of Organic Chemistry 7, 668-677. Singh, B.K., Walker, A. and Wright, D. (2002). Degradation of chlorpyrifos, fenamiphos, and chlorothalonil alone and in combination and their effects on soil microbial activity. Environmental Toxicology and Chemistry 21 (12), 2600-2605. Slattery, J.F., Coventry, D.R. and Slattery, W.J. (2001). Rhizobial ecology as affected by the soil environment. Australian Journal of Experimental Agriculture 41 (3), 289-298. Srinivasulu, M. and Rangaswamy, V. (2013). Influence of insecticides alone and in combination with fungicides on enzyme activities in soils. International Journal of Environmental Science and Technology 10, 341-350. Sorensen, S.R., Ronen, Z. and Aamand, J. (2001). Isolation from agricultural soil and characterization of a Sphingomonas sp. able to mineralize the phenylurea herbicide isoproturon. Applied and Environmental Microbiology 67 (12), 5403-5409. Storck, V., Karpouzas, D.G. and Martin-Laurent, F. (2016a). Towards a better pesticide policy for the European Union. Science of the Total Environment, http://dx.doi.org/10.1016/j.scitotenv.2016.09.167. Storck, V., Lucini, L., Mamy, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., Karas, A.P., Servien, R., Karpouzas, D., Trevisan, M., Benoit, P. and Martin-Laurent, F. (2016b). Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology. Environmental Pollution 208, 537-545. Topp, E. (2003). Bacteria in agricultural soils: Diversity, role and future perspectives. Canadian Journal of Soil Science 83 (3), 303- 309. Van Beelen, P. and Doelman, P. (1997). Significance and application of microbial toxicity tests in assessing ecotoxicological risks of contaminants in soil and sediment. Chemosphere 34, 455-499. Vasilyeva, N.V., Tsfasman, I.M., Suzina, N., Stepnaya, O.A. and Kulaev, I. (2008). Secretion of bacteriolytic endopeptidase L5 of Lysobacter sp. XL1 into the medium by means of outer membrane vesicles. The FEBS Journal 275 (15), 3827-3835. Vischetti, C., Monaci. E., Cardinali, A., Casucci, C. and Perucci, P. (2008). The effect of initial concentration, co-application and repeated applications on pesticide degradation in a biobed mixture. Chemosphere 72, 1739-1743. Von Tigerstrom, R.G. (1980). Extracellular nucleases of Lysobacter enzymogenes: production of the enzymes and purification and characterization of an endonuclease. Canadian Journal of Microbiology 26 (9), 1029-1037. Wang, C., Wang, F., Zhang, Q. and Liang, W. (2016). Individual and combined effects of tebuconazole and carbendazim on soil microbial activity. European Journal of Soil Biology 72, 6-13. Wang, F., Yao, J., Chen, H., Chen, K., Trebse, P. and Zaray, G. (2010). Comparative toxicity of chlorpyrifos and its oxon derivates to soil microbial activity by combined methods. Chemosphere 78, 319-326. Weaver, M. A., Krutz, L. J., Zablotowicz, R. M., and Reddy, K. N. (2007). Effects of glyphosate on soil microbial communities and its Mississippi soil. Pest Management Science 63, 388-393. Wessen, E., Hallin, S., and Philippot, L. (2010). Differential responses of bacterial and archaeal groups at high taxonomical ranks to soil management. Soil Biology and Biochemistry 42, 1759-1765. Zhang, Y. and Sun, Y. (2014). A method for de novo nucleic acid diagnostic target discovery. Bioinformatics 30 (22), 3174-3180.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 244

Conclusion

The aim of Chapter 4 was to test the employability of alternative techniques to investigate potential side effects of each of the three studied pesticides on soil microorganisms. The lab-to-field experiment carried out to test the ecotoxicological impact of CHL, IPU or TCZ (applied once at three different concentrations simulating a lower tier exposure scenario) showed that CHL and TCZ slightly but significantly modified soil bacterial diversity, while IPU did not have a significant impact, as evaluated by Illumina next-generation sequencing of 16S rDNA amplicons (Chapter 4.1). These results are in accordance with data contained in the risk assessment conclusion documents of responsible authorities indicating that the risk of each of these three pesticides is low for soil microrganisms. However, repeated exposure of soil microorganisms to CHL or TCZ (repeatedly applied at 10x agronomical doses in a microcosm study to simulate a higher tier exposure scenario) significantly modified the activity and diversity of soil microorganisms (Chapter 4.2). Indeed, repeated exposure to CHL led to the adaptation of the microbial community resulting in the enhanced mineralization of CHL, as observed by cumulative kinetics of 14CO2 evolved from 14C-labelled CHL in radiorespirometric measurements. No adaptation resulting in enhanced biodegradation of 14C-labelled TCZ was observed. As expected, radiorespirometry was found to be suitable to detect microbial adaptation to a pesticide resulting in enhanced biodegradation. The evolution of soil buffering functions (measured by the ability of a soil to degrade molecules that are known to be degradable by a large or small range of microorganisms) was found to be pesticide-specific and less sensitively detectable, making it rather difficult to be interpreted. Much more work has to be carried out to acquire enough reference enabling a more robust interpretation. In response to repeated pesticide exposure, Illumina next-generation sequencing of 16S rDNA amplicons allowed monitoring the evolution of the soil bacterial diversity. One could conclude that metagenomics may be more suitable for the detection of large shifts in the soil bacterial composition, while small shifts of rare species remain undetected. Consequences on ecosystem functions remain to be determined.

Chapter 4 Ecotoxicological impact of pesticides on soil microorganisms 246

Chapter 5

Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation

Veronika Storck Jérémie Béguet Nadine Rouard Sabir Hussain Marion Devers-Lamrani Fabrice Martin-Laurent

Manuscript in preparation for publication.

Foreword

Repeated exposure to pesticides can lead to the emergence of soil microbial populations able to use them as nutrient and/or energy sources for their growth. These populations harbor catabolic genes coding for enzymes that metabolize pesticides. The degrading microorganisms can be seen as environmentally friendly because they decrease the persistence of pesticides in soil. Moreover, pesticide-degrading microbes can be used in bioremediation techniques to depollute the environment. The detection and quantification of pesticide degradation-encoding genes by molecular approaches may therefore constitute an indicator of pesticide exposure and impact. Despite the promising advantages of using this approach, it remains challenging and needs to be studied by further research how reliable conclusions based on gene-bioindicator quantification can be drawn. With the research of Chapter 5, I aimed to study the possibility to use the quantification of pesticide degradation-encoding genes as bioindicators for the pesticide-degrading ability of a soil. This work was carried out in a microcosm and a field study by monitoring (i) the abundance of pdm genes responsible for the first step of the bacterial IPU degradation pathway, and (ii) the IPU-mineralizing ability of the soil microbial community.

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 250

Abstract

Despite their benefits, pesticides can persist in the environment where they can harm non-target organisms. Microbial degradation of pesticides can contribute to diminish their persistence and therefore to minimize their impact. The quantification of pesticide-degrading microorganisms by qPCR of catabolic genes may be used as bioindicator for the pesticide- degrading potential of a soil. To test this hypothesis, we quantified (i) the pdmA and pdmB genes involved in the first step of the bacterial degradation pathway of the phenyl-urea herbicide isoproturon (IPU), and (ii) and the plasmid pSH (harboring the pdmA and pdmB genes) of Sphingomonas sp. SH, an IPU-degrading strain, and compared the gene abundances to IPU mineralization in two French agricultural soils in a microcosm and a field study. In the microcosm experiment, the genes were quantified contemporary to IPU exposure, while the field soil was historically treated with IPU (> 6 months before soil sampling). The results showed that the abundance of pdmA and pdmB was not correlated to IPU mineralization observed in soil samples collected from the field experiment. In the microcosm experiment, we observed that pdmA and pdmB abundances concomitantly increased with the IPU-mineralizing activity, reaching their maximum at the highest IPU mineralization rate. The abundance of the plasmid pSH behaved similarly but at lower levels indicating that the IPU-degrading soil bacterial community possibly comprises other populations harboring the pdm genes on different genetic structures that remain unknown. As the IPU-degrading genetic potential increased in response to IPU exposure and rapidly decreased with time, we concluded the time point for soil sampling and DNA extraction to be crucial for the use of gene-bioindicators.

Figure 5.1. Graphical abstract of Chapter 5.

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 252

Introduction

The use of pesticides is an efficient method to protect crops from various pests in agriculture worldwide. Despite this advantage, pesticides can be harmful to the environment and human health (Hutzinger, 2013; Lewis and Maslin, 2015). Only a tiny part of spayed pesticides ends up in the target organism while big parts reach the soil, from where they can spread to water bodies and impact non-target organisms (Barbash, 2003; Looser et al., 2000; Smalling et al., 2013; Zhang et al., 2011). To minimize these issues, soil organisms such as bacteria, that degrade pesticides by using their components as nutrient and/or energy source (Topp, 2003), can be employed by mankind to counteract environmental pollution (Monard et al., 2013). For example, assessing the presence or absence of pesticide-degrading bacteria (via the presence of pesticide-degrading genetic potential) in an agricultural soil may be useful to predict pesticide degradation in soil or to monitor pollution. The development of molecular approaches able to detect and quantify genes that encode for pesticide-degrading enzymes has provided the possibility to assess the genetic potential of a soil to degrade a pesticide (Martin-Laurent et al., 2004). These approaches are based on direct soil DNA extraction (ISO 11063, 2012) and quantitative polymerase chain reaction (qPCR) (ISO 17601, 2016) or functional gene microarrays (Fenner et al., 2013) allowing the quantification or detection of pesticide-degrading genetic potential. Based on these techniques, the possibility to use catabolic genes as bioindicators for the pesticide-degrading genetic potential of soil microbial communities has been assessed for several herbicides (atrazine, 2,4-D and diuron), showing that the size of pesticide-degrading microbial populations was sensitive to pesticide exposure (Baelum et al., 2012; Pesce et al., 2013; Piutti et al., 2002). In addition, Monard et al. (2013) found a correlation between atrazine degradation in soil and the expression of an atzD gene and concluded a high potential to use atzD to assess atrazine biodegradation performance in soils. From this point of view, monitoring of the pesticide-degrading genetic potential in the environment can indicate the exposure of an environment to a pesticide (indicator of impact) but also its biodegradation capacity (indicator of resilience). A prerequisite to use these gene-based techniques, however, is the knowledge of degrading genes and their clear attribution to the pesticide biodegradation process (Bouskill et al., 2007; DeBryn et al., 2007; Kazy et al., 2010;

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 254

Ogunseitan et al., 2000; Peng et al., 2010). The usability of pesticide-degrading genes as bioindicators to estimate pesticide exposure on soil microorganisms and to predict the pesticide degradation potential requires profound research to enable reliable conclusions, as each gene may have its own reaction behavior to the presence of a specific pesticide. In this study, we focused on the quantification of isoproturon (IPU)-degrading pdmA and pdmB genes in two soils in Central Eastern France (‘Biobed soil’ of a biomixture and ‘Epoisses soil’ of an agricultural field) that were regularly exposed to this herbicide. The pdm genes code for two sub-units of the enzyme catalyzing the transformation of the substituted phenyl-urea herbicide IPU to mono-demethyl-IPU (MD-IPU) by N-demethylation. These genes have recently been discovered in an IPU- degrading Sphingobium sp. strain YBL2 isolated from an agricultural soil in China (GenBank accession number CP010958.1; Gu et al., 2013). They have also been found in an IPU-degrading Sphingomonas sp. strain SH isolated from the Epoisses soil in France (Hussain et al., 2011) and in an IPU-mineralizing Sphingomonas sp. strain SRS2 isolated from an agricultural soil in the United Kingdom (GenBank accession number LARW01000036.1; Nielsen et al., 2015). Interestingly, the nucleotidic sequences of the pdm genes in these three strains are 99 % similar. In Sphingomonas sp. strain SH, they are located in a transposon-like sequence in the plasmid pSH (GenBank accession number KU237244.1, unpublished data). In addition, the transposon-like sequence of 5295 bp of Sphingomonas sp. strain SH is 99 % similar to that of the Sphingobium sp. strain YBL2. The main aim of this study was not only to test the usability of the quantification of pdm genes as bioindicators for the IPU-mineralizing ability of soil microbial communities, but also to assess the ecological importance of the pdm-harboring transposon-like sequence of the plasmid pSH within the IPU-degrading bacterial communities of the Biobed soil and the Epoisses soil. Within this context, we measured the mineralization of 14C-labelled IPU to 14CO2 and the abundance of (i) the pdmA and pdmB gene sequences, (ii) a pdmA-down sequence specific for the pdm-harboring transposon-like sequence (targeting the pdmA gene and a sequence downstream of pdmA), and (iii) a pSH sequence located on the plasmid pSH, outside of the transposon- like sequence in the Biobed and the Epoisses soil. Our experimental set-up was based on the hypothesis that the time point of soil sampling for DNA extraction may be especially

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 255

crucial for a solid correlation between the abundance of a gene involved in pesticide degradation and the actual capability of an agricultural soil to degrade a pesticide. Therefore, two oppositional scenarios of IPU exposure were considered to distinguish the response of IPU-degrading communities to contemporary and historical exposure to this herbicide. In a first step (contemporary exposure scenario), we exposed the Biobed soil to IPU and monitored the contemporary responses (IPU mineralization and gene abundances) all along the incubation in microcosms. In a second step (historical exposure scenario), we collected soil of the Epoisses field over a 2.5 years period, when the last IPU treatment was 6 months, 1.5 years and 2.5 years ago and the soil had shown the ability to readily degrade IPU at any time (Hussain et al., 2013).

Material and methods

Experimental design

We based the experimental design on the lapse of time between IPU treatment and quantification of (i) degrading genes and (ii) IPU-mineralization in two oppositional scenarios of IPU exposure (contemporary or historical exposure) (Figure 5.2). The quantification of degrading genes and IPU mineralization contemporary to IPU exposure was realized in microcosms of the biomixture soil (Biobed soil) collected from a biobed receiving pesticide effluents (including IPU) from an experimental farm. The microcosms were treated with IPU and analyzed for their IPU-degrading genetic potential at five time points within 14 days after the treatment. The quantification of degrading genes and IPU mineralization after historical IPU exposure was realized in the agricultural field soil (Epoisses soil) rotationally cropped with wheat / rapeseed / barley that had last been treated with IPU 6 months, 1.5 years and 2.5 years ago at the soil sampling time point.

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 256

Figure 5.2. The timeline-based experimental design of the microcosm study (Biobac soil, contemporary exposure to IPU) and the field study (Epoisses soil, historical exposure to IPU).

Soil characteristics and sampling

The biomixture soil of the biobed (Biobed soil) was prepared with soil collected at the experimental farm of INRA Epoisses (Bretenière, 47°14'25.6"N 5°06'12.4"E, Google Maps) and was frequently treated with various pesticides. The biomixture is a silty clay loam soil and contains 36.7 % clay, 43.6 % silt and 19.7 % sand. It has an organic matter content of 10.2 % and a pH of 7.0. We sampled the Biobac soil in June 2015, sieved it (5 mm) and stored it at 4°C until use. Before starting the experiment, it was humidified to 80 % of its water-holding capacity (WHC = 70.19 %) and incubated at 20 °C in the dark. At day 0, IPU (99 % purity, Riedel-de-Haen) was applied at 2 mg/kg dwt to soil the microcosms. Samples were collected at 0 days before IPU treatment and at 2, 4, 6, 8, 10, 12 and 14 days after IPU treatment and immediately frozen at -20°C until DNA extraction. The Epoisses soil was sampled from the field of the experimental farm of INRA in April 2008, 2009 and 2010 (36 topsoil samples of 20 cm depth each year). Each sample was divided into three parts (i) to freeze the samples at -20 °C until DNA extraction, (ii) to analyze IPU mineralization of each sample and (iii) to sieve (5 mm) and characterize them as described by Hussain et al. (2013). The soil is a silty clay soil (47.7 % clay, 46.0 % silt, 8.1 % sand) with an organic matter content of 3.6 % and a pH of 7.7 (Hussain et al., 2011). It had been rotationally cultivated with winter wheat / rapeseed / barley

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 257

and had been periodically treated with pesticides for more than 10 years. During the 2.5 years of sampling (2008 - 2010), the soil was treated with IPU (October 2007), glyphosate (August 2008), napropamide, dimethachlor and clomazone (August 2009) and mefenpyr-dimethyl, iodosulfuron-methyl sodium and mesosulfuronmethyl (March 2010). IPU degradation parameters in the soil such as the maximum IPU mineralization

(A) [%], the maximum rate of mineralization (µm) [%/d] or the lag phase (λ) [d] were determined by Hussain et al. (2013).

Analysis of IPU mineralization in the soils

In the Biobed soil, IPU mineralization was measured in three independent samples of 20 g dwt). Each sample was treated with 2 mg/kg dwt soil IPU (99 % purity, Riedel-de-Haen) and 1.7 kBq of 14C-ring-labelled IPU (99 % radiochemical purity, specific activity = 666000 kBq/mmol, International Isotopes Munich). Samples were incubated at a constant humidity of 80 % of the WHC in the dark. At regular intervals (0,

2, 4, 6, 8, 10, 12 and 14 days after treatment), 14CO2 evolved from 14C-ring-labelled IPU was trapped in 5 mL 0.2 M sodium hydroxide and was analyzed by liquid scintillation counting (LS 6500 Multi-Purpose Scintillation Counter, Beckman) using 10 mL scintillation liquid (ACII scintillation fluid, Amersham). In the Epoisses soil, IPU mineralization was measured in the 36 soil samples collected from the field study each year. Each sample (40 g dwt) was treated with 1.5 mg/kg dwt soil IPU and 2 kBq of 14C-ring-labelled IPU. The samples were incubated for 46 days and analyzed as described for the Biobed soil.

DNA extraction and quantification

DNA was directly extracted from 250 mg soil samples according to the ISO 11063 (2012) method initially described by Martin-Laurent et al. (2001). The procedure involved the 3 main steps (i) microbial cell lysis by physical (bead beating) and chemical (sodium dodecyl sulfate at high temperature) actions, (ii) deproteination and (iii) DNA precipitation and washing. We determined DNA concentrations with a Thermo Fisher StepOnePlus thermocycler (Applied Biosystems) by using fluorescent dye of a Quant-iT™ PicoGreen® dsDNA assay kit (Invitrogen).

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 258

Primer design

Two primer sets targeting the pdmA (pdmA_F: 5’-TGG-CAG-TTC-AGC-TAT-GAT- GC-3’, pdmA_R: 5’-ATA-GTC-CCC-TTT-GCC-TTC-GT-3’) and pdmB (pdmB_F: 5’-TTC-GAC- GAT-GTC-AGA-ACG-AG-3’, pdmB_R: 5’- GTG-GTG-TGC-TTT-CAG-GGA-AT-3’) genes coding for the first step of the IPU degradation pathway (Gu et al., 2013) were designed. Another set of primers (targeting pdmA-down) was designed to target the transposon- like sequence (containing pdmA and pdmB) located on the plasmid pSH (position 36029 to 41475 bp) (Figure 5.3) of the IPU-degrading Sphingomonas sp. strain SH isolated from the Epoisses soil (Hussain et al., 2011). One of the primers (pdmA-down_F: 5’-ACG-AAG- GCA-AAG-GGG-ACT-AT-3’) is anchored on the pdmA sequence (position 38479 bp) and the other (pdmA-down_R: 5’-AAA-GCG-TCT-TTG-GTC-GAG-AA-3’) is located on the sequence downstream of the pdmA gene (position 38260 bp). Another primer set (targeting pSH) was designed to target a sequence on the plasmid pSH (without considering pdm genes) located at 8596 to 8792 bp on pSH (pSH_F: 5’-GGC-TTG-GGT- ACG-CTA-TGA-AA-3’, pSH_R: 5’-CGT-TCA-GGA-CTG-CCG-TAA-AG-3’).

Figure 5.3. Locations of the primer sets used in this study on the plasmid pSH. The sequences targeted by the different primer pairs (pdmA, pdmB, pdmA-down and pSH) are indicated in red. The transposon-like sequence (surrounded by tnpX) harboring the pdmA and pdmB genes is presented in grey.

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 259

Quantification of 16S rDNA, pdmA, pdmB, pdmA-down and pSH sequences

Prior to the quantification of target genes, the absence of PCR inhibitors in the DNA extracts was verified by spiking a known amount of a pGEM-T Easy plasmid vector (Promega) to the samples or to pure water as a control. We quantified the plasmid by qPCR (ISO 17601, 2016) in a StepOnePlus qPCR machine (Applied Biosystems) using universal SP6 and T7 primers as described by Henry et al. (2006). As PCR amplification was not inhibited, the DNA samples could be used to perform qPCR of the target gene sequences. The abundance of the total bacterial community was measured by qPCR targeting the 16S rDNA using the universal bacterial primers 341_F (5’-CCT-ACG-GGA-GGC-AGC-AG-3’) and 534_R (5’-ATT-ACC-GCG-GCT- GCT-GGC-A-3’) as described by Lopez-Gutierrez et al. (2004). The IPU-degrading genetic potential was monitored by qPCR on pdmA or pdmB. The abundances of the transposon- like sequence and of the plasmid pSH were respectively quantified by qPCR on pdmA- down or pSH. The qPCRs were done in 15 µL reaction volumes containing SYBR Green PCR Master Mix (Absolute QPCR SYBR Green Rox, ABgene), 100 ng of T4 gp 32 (Qbiogene), 2 µM of appropriate primers and 0.4 ng of template DNA. The qPCR conditions were 95 °C for 15 min for enzyme activation, followed by 35 cycles of denaturation at 95 °C for 15 sec, primer annealing at 60 °C for 30 sec and extension at 72 °C for 30 sec. Data acquisition was performed at 80 °C. After amplification, melting curves were obtained by increasing the temperature from 80 to 95 °C (+0.5 °C/sec). Calibration curves were established from previously cloned pdmA, pdmB, pdmA-down and pSH PCR products of the IPU-degrading Sphingomonas sp. strain SH (Hussain et al., 2011). Non-template controls were included in all qPCR assays.

Data analysis

IPU degradation parameters were determined as described by Hussain et al. (2013). We separately compared for each gene the quantities between all time points in the Biobed soil using ANOVAs followed by the Tukey HSD test (p < 0.05). In addition, one ANOVA was performed on the quantities of all genes at 7 days after IPU treatment. For the field samples collected at Epoisses, we separately compared for each gene, the quantities between all 3 years using ANOVAs followed by the Tukey HSD test (p < 0.05).

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 260

Results and discussion

Microcosm study - Contemporary exposure to IPU

The microcosm study revealed that the Biobed soil contemporarily exposed to IPU was found to be readily able to mineralize IPU (Figure 5.4). Within 14 days, around 50 % of the initially applied 14C-labelled IPU was transformed by microbial communities to 14CO2. IPU-mineralization followed first order kinetics typical for biodegradation by soil microbial communities already adapted to pesticide application (Soulas et al., 2001). The maximum IPU mineralization (A) was 53.6 ± 0.4 %, the maximum rate of mineralization (µm) was 12.8 ± 0.6 %/day, and the lag phase (λ) was 1.6 ± 0.0 days, according to the modified Gompertz growth model fitted to IPU-mineralization curves. The short lag phase suggests that IPU-degrading populations were initially present in the Biobed soil. This is in agreement with the qPCR analyses on the IPU- degrading pdmA and pdmB genes that report these genes to be present at day 0 before IPU treatment (at around 22 sequences per 104 16S rDNA sequences). This initial presence of the genes indicating the initial presence of IPU-degraders can most likely be attributed to the frequent IPU treatments that the Biobed soil was exposed to, as it has previously been observed for various herbicides such as IPU (El Sebai et al., 2004), atrazine (Piutti et al., 2002) or 2,4-dichlorophenoxyacetic acid (2,4-D) (Baelum et al., 2008). The acceleration of the IPU mineralization during the exponential phase (between 2 and 7 days after IPU treatment) suggests that IPU-degraders may have become more abundant after 2 days of IPU exposure, indicating the presence of highly responsive IPU-degrading microbial populations in the Biobed soil. This assumption is in agreement with results of pdmA and pdmB qPCR analyses showing that the IPU- degrading genetic potential and the IPU-mineralizing activity concomitantly reached their maximum after 7 days of IPU exposure (Figure 5.4), as the relative abundance of the pdm sequences significantly increased one order of magnitude reaching its maximum after 7 days (pdmA = 262 ± 91 and pdmB = 201 ± 29 copy numbers/104 copy numbers of 16S rDNA). This growth trait of pesticide-degraders has already been observed for atrazine, for which the abundance (Martin-Laurent et al., 2003; Piutti et al., 2002) and expression (Monard et al., 2013) of the atrazine-degrading genetic potential were found to increase in response to atrazine exposure.

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 261

Interestingly, already 14 days after IPU treatment, the relative abundance of pdmA and pdmB had decreased to its initial level (around 39 copy numbers/104 copy numbers of 16S rDNA) (Figure 5.4), suggesting that the abundance of IPU-degraders rapidly decreased when the amount of bioavailable IPU decreased. This is in agreement with previous studies showing that that the 2,4-D-, 2-methyl-4-chlorophenoxyacetic acid- (MCPA) (Baelum et al. 2008) and atrazine- (Martin-Laurent et al., 2003) degrading genetic potential rapidly decreased once the mineralization of the pesticides had ceased. The most abundant genes at day 7 were pdmA and pdmB (both > 200 copy numbers/104 copy numbers of 16S rDNA). Their detection at similar abundances indicates that they are similarly organized in the microbial community of the Biobed soil. This was expected as they were found to code for the α- and β-subunit of the PDM AB oxygenase responsible for the N-demethylation of IPU (Gu et al., 2013). Moreover, pdmA and pdmB are associated with each other in the genomes of the three known IPU- degrading strains Sphingobium sp. strain YBL2 (Gu et al., 2013), Sphingomonas sp. strain SRS2 (Sorensen et al., 2001) and Sphingomonas sp. strain SH (Hussain et al., 2011), respectively isolated from soils in China, UK and France. Although the relative abundances of the pdmA-down (around 100 copy numbers/104 copy numbers of 16S rDNA) and pSH (around 60 copy numbers/104 copy numbers of 16S rDNA) gene sequences were significantly lower (p < 0.05) than those of pdmA and pdmB at day 7, they followed a similar temporal pattern with abundances transiently increasing in response to IPU exposure. Indeed, the relative gene abundance of pdmA-down was lower before IPU treatment (around 10 copy numbers/104 copy numbers of 16S rDNA), increased to a maximum at day 7 (around 100 copy numbers/104 copy numbers of 16S rDNA) and decreased to around 19 copy numbers per 104 copy numbers of 16S rDNA 14 days after IPU treatment. This indicates that only a part of the IPU-degrading bacterial community (approximately 50 %) possesses the same genetic organization as the one of the three known IPU-mineralizing isolated bacterial strains. It further suggests that the pdmA and pdmB genes quantified in the Biobed soil may have the same genetic origin as those isolated strains from France, UK and China). This raises the question how such newly evolved degradation-encoding genes are dispersed all over the world (De Souza et al., 1998). The detection of the pSH sequence a similar abundance to pdmA-down further indicates that the pdmA and pdmB-harboring transposon-like sequence may be located

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 262

on the plasmid pSH in a part of the IPU-degrading community. However, as the relative abundances of pdmA-down and pSH were significantly lower than those of pdmA and pdmB at day 7, one could hypothesize that a fraction of the IPU-degrading community may harbor another pdmA and pdmB genetic organization remaining unknown.

Figure 5.4. Mineralization of 14C-ring-labelled IPU [% of initially applied amount] and relative gene abundance of pdmA, pdmB, pdmA-down and pSH to 16S rDNA abundance [copy number/104 copy numbers of 16S rDNA] measured in the Biobed soil microcosms contemporarily exposed to IPU. ANOVAs were followed by the Tukey HSD test (p < 0.05) separately grouping all time points per gene. Significant differences are indicated by letters.

Field study - Historical exposure to IPU

Soil samples of the agricultural field at Epoisses that had historically been exposed to IPU were collected 6 months (2008), 1.5 years (2009) and 2.5 years (2010) after the last IPU treatment (Hussain et al., 2013). Analyses of IPU mineralization curves followed first order kinetics showing that the microbial community of the Epoisses soil with IPU treatment history was readily able to mineralize IPU. The parameters

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 263

estimated by fitting the mineralization curves to the Gompertz growth model (maximum

IPU mineralization (A) = 38 ± 11 %, maximum rate of mineralization (µm) = 3 ± 2 %/day, and lag phase (λ) = 3 ± 2 days (Hussain et al., 2013)) were lower and more variable than those observed in the Biobed soil. In addition, pdmB, pdmA-down and pSH gene sequences were quantified at lower amounts than in the Biobed soil (Figure 5.5) (pdmA is not shown because it was previously shown to give identical results to pdmB). Indeed, the three tested genes were successfully detected, although close to the limit of quantification of each qPCR assay. Analogously to the results of the Biobed soil, the pdmB gene was the most abundant one (59 ± 74 copy numbers/106 16S rDNA copy numbers). However, the pdmB abundance in the Epoisses field soil historically exposed to IPU was not only two orders of magnitude lower than that quantified in the Biobed soil contemporarily exposed to IPU, but also much more variable. Part of this variability may be attributed to the spatial variability of the sampling spots as suggested by Hussain et al. (2013), who observed that spatial variability of IPU mineralization observed in the field can be explained by soil humidity and pH. The low IPU-degrading genetic potential recorded in the field of Epoisses can be explained by the tendency of IPU-degraders to rapidly decline in absence of IPU (as shown in the Biobed soil). This hypothesis is further reinforced by the fact that pdmB abundances (this study) and IPU mineralization activities (Hussain et al., 2013) measured in samples collected in year 2010 were significantly higher than that measured in years 2008 and 2009. As suggested by Hussain et al. (2013), this effect may be attributed to the application of three herbicides that were applied to the field one month before soil sampling in 2010 soil sampling. As observed in the Biobed soil, the relative abundances of pdmA-down (3 ± 8 copy numbers/106 16S rDNA copy numbers) and of pSH (7 ± 7 copy numbers/106 16S rDNA copy numbers) were significantly lower than that of pdmB (Figure 5.5). This difference was even more pronounced in the field of Epoisses than in the Biobed soil. It further confirmed that only a part of the IPU-mineralizing microbial community harboring pdmA and pdmB genes has the same transposable element as that of known IPU- mineralizing isolated bacterial strains. As pdmA-down and pSH show similar abundances, one could hypothesize that the pdmA- and pdmB-harboring transposable element may be located on the plasmid pSH in the soil of Epoisses, which is in agreement

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 264

with the isolation of Sphingomonas sp. strain SH from this field (Hussain et al., 2011). Nevertheless, as suggested for the Biobed soil, these results contibute to the hypothesis that a fraction of the IPU-degrading community of the Epoisses soil may harbor another unknown pdmA and pdmB genetic organization.

Figure 5.5. Relative gene abundance of pdm, pdmA-down and pSH [copy number/106 copy numbers of 16S rDNA] measured in the Epoisses soil historically exposed to IPU.

For each of the three quantified genes, no correlation between their relative abundance and IPU degradation parameters (A, µm and λ) could be observed regardless of the year (as shown in Figure 5.6 for the tentative correlation between the relative pdmB abundance and the maximum IPU mineralization (A)). This result is in agreement with Baelum at al. (2008) who reported no significant correlation between the gene abundances of tfdA (coding for an enzyme that degrades chlorophenoxy herbicides) and the mineralization rates of the two chlorophenoxy herbicides 2,4-D and MCPA. Similarly, Martin-Laurent et al. (2004) reported some discordances between the atrazine- degrading genetic potential (estimated by the quantification of atzA, atzB and atzC gene abundances) and atrazine degradation in soil.

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 265

Figure 5.6. Tentative of correlation between the relative abundance of pdmB [copy number/106 copy numbers of 16S rDNA] and the maximum IPU mineralization (A) [%] of all tested years (2008, 2009 and 2010).

Conclusions and perspectives

The quantification of the IPU-degrading genetic potential was in agreement with the IPU mineralization in the Biobed soil contemporarily exposed to this herbicide. In this context, the IPU-degrading genetic potential was found to be highly sensitive to IPU exposure being transiently increased, reaching its maximum when the IPU mineralization rate was highest and then decreased to its initial value. This rapid decrease of the IPU-degrading genetic potential suggests that IPU-degraders may decline to a minimum in the absence of IPU. This may explain why the quantification of the IPU-degrading genetic potential was not found to be correlated to IPU mineralization in the Epoisses soil historically exposed to IPU. Indeed, in Epoisses soil samples, pdmB was quantified in rather low amounts while substantial IPU mineralization was recorded regardless of the year of sampling. This discrepancy may most likely be explained by the decrease of the IPU-degrading genetic potential in response to the long time period without IPU exposure (6 months (2008), 1.5 years (2009) and 2.5 years (2010)),

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 266

emphasizing the importance of the soil sampling time point for gene-bioindicator quantification. Referring to the short-lived response of the IPU-degrading genetic potential to contemporary IPU exposure (Biobed soil), one could suggest that it would have been better to estimate the abundance of the IPU-degrading genetic potential in the Epoisses soil historically exposed to IPU directly after having it newly exposed to IPU for a few days under controlled conditions. In addition, this study provides new insights in the ecology of the IPU-degrading microbial community by showing that the pdmA- and pdmB-harboring transposable element (described in Sphingobium sp. strain YBL2) of the plasmid pSH (described in Sphingomonas sp. strain SH) is not the sole genetic element involved and that other (yet unknown) genetic elements may also harbor pdmA and pdmB genes. This observation underlines the fact that quantification of pesticide-degrading genes as a proxy of the pesticide-degrading genetic potential also requires a deep knowledge on the genetic organization of a pesticide-degrading gene in order to define the good gene target to be monitored. One could conclude that the quantification of the IPU-degrading genetic potential can be a good proxy of the IPU-mineralizing activity in case of contemporary exposure to this herbicide. However, it is a poor proxy of the IPU-mineralizing activity in case of historical exposure to IPU. To overcome this problem, one could suggest to quantify the IPU-degrading genetic potential in agricultural fields historically exposed to IPU in soil samples freshly exposed to IPU some days before DNA extraction.

Acknowledgements

The Ph.D. studies of Veronika Storck performed at the Doctoral School of Environment and Health (ED E2S) at the University of Burgundy (Grant number 2013- 31) were funded by the French Ministry of Education and Research (MESR).

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 267

References

Baelum, J., Nicolaisen, M.H., Holben, W.E., Strobel, B.W., Sorensen, J. and Jacobsen, C.S. (2008). Direct analysis of tfdA gene expression by indigenous bacteria in phenoxy acid amended agricultural soil. The ISME Journal 2, 677-687. Baelum, J., Prestat, E., David, M.M., Strobel, B.W. and Jacobsen, C.S. (2012). Modeling of phenoxy acid herbicide mineralization and growth of microbial degraders in 15 soils monitored by quantitative real-time PCR of the functional tfdA gene. Applied and Environmental Microbiology 78 (15), 5305-5312. Barbash, J.E. (2003). The geochemistry of pesticides. Treatise on Geochemistry 9, DOI: 10.1016/B0-08-043751-6/09056-3. Bouskill, N.J., Barnhart, E.P., Galloway, T.S., Handy, R.D. and Ford, T.E. (2007). Quantification of changing Pseudomonas aeruginosa soda, htpX and mt gene abundance in response to trace metal toxicity: A potential in situ biomarker of environmental health. FEMS Microbiology Ecology 60, 276-286. DeBryn, J.M., Chewning, C.S. and Sayler, G.S. (2007). Comparative quantitative prevalence of Mycobacteria and functionally abundant nidA, nahAc, and nagAc dioxygenase genes in coal tar contaminated sediments. Environmental Science and Technology 41, 5426-5432. Devers, M., Azhari, N.E., Kolic, N.-U. and Martin-Laurent, F. (2007). Detection and organization of atrazine-degrading genetic potential of seventeen bacterial isolates belonging to divergent taxa indicate a recent common origin of their catabolic functions. FEMS Microbiology Letters 273, 78-86. De Souza, M.L., Seffernick, J., Martinez, B., Sadowsky, M.J. and Wacket, L.P. (1998). The atrazine catabolism genes atzABC are widespread and highly conserved. Journal of Bacteriology 180 (7), 1951-1954. El Sebai, T., Lagacherie, B., Soulas, G. and Martin-Laurent, F. (2004). Isolation and characterisation of an isoproturon-mineralising Methylopila sp. TES from French agricultural soil. FEMS Microbiology Letters 239, 103-110. Fenner, K., Canonica, S., Wacket, L.P. and Elsner, M. (2013). Evaluating pesticide degradation in the environment: Blind spots and emerging opportunities. Science 341, 752-758. Gu, T., Zhou, C., Sørensen, S.R., Zhang, J., He, J., Yu, P., Yan, X. and Li, S. (2013). The novel bacterial N-demethylase pdmAB is responsible for the initial step of N,N-dimethyl-substituted phenylurea herbicide degradation. Applied and Environmental Microbiology 79 (24), 7846-7856. He, Z., Gentry, T.J., Schadt, C.W., Wu, L., Liebich, J., Chong, S.C., Huang, Z., Wu, W., Gu, B., Jardine, P., Criddle, C. and Zhou, J. (2007). GeoChip: a comprehensive microarray for investigating biogeochemical, ecological and environmental processes. The ISME Journal 1, 67-77. Henry, S., Bru, D., Stres, B., Hallet, S. and Philippot, L. (2006). Quantitative detection of the nosZ gene, encoding nitrous oxide reductase, and comparison of the abundances of 16S rRNA, narG, nirK, and nosZ genes in soils. Applied and Environmental Micribiology 72, 5181-5189. Hussain, S., Devers-Lamrani, M., El Azhari, N. and Martin-Laurent, F. (2011). Isolation and characterization of an isoproturon mineralizing Sphingomonas sp. strain SH from a French agricultural soil. Biodegradation 22, 637-650. Hussain, S., Devers-Lamrani, M., Spor, A., Rouard, N., Porcherot, M., Beguet, J. and Martin-Laurent, F. (2013). Mapping field spatial distribution patterns of isoproturon-mineralizing activity over a three-year winter wheat/rape seed/barley rotation. Chemosphere 90, 2499-2511. Hutzinger, O. (2013). The handbook of environmental chemistry – anthropogenic compounds, 3rd, ed., Springer, Berlin Heidelberg. ISO 11063 (2012). Soil quality - Method to directly extract DNA from soil samples. ISO 17601 (2016). Soil quality - Estimation of abundance of selected microbial gene sequences by quantitative PCR from DNA directly extracted from soil. Kazy, W.D., Monier, A.L. and Alvarez, P.J.J. (2010). Assessing the correlation between anaerobic toluene degradation activity and bssA concentrations in hydrocarbon-contaminated aquifer material. Biodegradation 21, 793-800. Kong, W.D. and Nakatsu, C.H. (2010). Optimization of RNA extraction for PCR quantification of aromatic compound degradation genes. Applied and Environmental Microbiology 76, 1282-1284. Lewis, S.L. and Maslin, M.A. (2015). Defining the Anthropocene. Nature, 519, 171-180. Looser, R., Froescheis, O., Cailliet, G.M., Jarman, W.M. and Ballschmiter, K. (2000). The deep-sea as a final global sink of semivolatile persistent organic pollutants? Part II: Organochlorine pesticides in surface and deep-sea dwelling fish of the North and South Atlantic and the Monterey Bay Canyon (California). Chemosphere, 40 (6), 661-70. Lopez-Gutierrez, J.C., Henry, S., Hallet, S., Martin-Laurent, F., Catroux, G. and Philippot, L. (2004). Quantification of a novel group of nitrate-reducing bacteria in the environment by real-time PCR. Journal of Microbial Ecology 57, 399-407. Martin-Laurent, F., Cornet, L., Ranjard, L., Lopez-Gutierrez, J.-C., Philippot, L., Schwartz, C., Chaussod, R., Catroux, G. and Soulas, G. (2004). Estimation of atrazine-degrading genetic potential and activity in three French agricultural soils. FEMS Microbiology Ecology 48, 425-435. Martin-Laurent, F., Philippot, L., Hallet, S., Chaussod, R., Germon, J.C., Soulas, G. and Catroux, G. (2001). DNA extraction from soils: old bias for new microbial diversity analysis methods. Applied and Environmental Microbiology 67, 2354-2359. Martin-Laurent, F., Piutti, S., Hallet, S., Wagschal, I., Philippot, L., Catroux, G. and Soulas, G. (2003). Monitoring of atrazine treatment on soil bacterial, fungal and atrazine-degrading communities by quantitative competitive PCR. Pest Management Science 59, 259-268. Monard, C., Martin-Laurent, F., Devers-Lamrani, M., Lima, O., Vandenkoornhuyse, P. and Binet, F. (2010). Atz gene expressions during atrazine degradation in the soil drilosphere. Molecular ecology 19, 749-759. Monard, C., Martin-Laurent, F., Lima, O., Devers-Lamrani, M. and Binet, F. (2013). Estimating the biodegradation of pesticide in soils by monitoring pesticide-degrading gene expression. Biodegradation 24, 203-213. Monard, C., Martin-Laurent, F., Vecchiato, C., Francez, A.-J., Vandenkoornhuyse, P. and Binet, F. (2008). Combined effect of bioaugmentation and bioturbation on atrazine degradation in soil. Soil Biology and Biochemistry 40, 2253-2259. Nielsen, T.K., Sorensen, S.R. and Hansen, L.H. (2015). Draft genome sequence of isoproturon-mineralizing Sphingomonas sp. SRS2, isolated from an agricultural field in the United Kingdom. Genome Announcements 3 (3), e00569-15. OECD-Guideline ENV/JM/MONO (2007)17 (2007). Guidance document on pesticide residue analytical methods. Series on pesticides no 39. Series on testing and assessment no 72, p. 34. Ogunseitan, O.A., Yang, S. and Ericson, J. (2000). Microbial deltaaminolevulinate dehydratase as a biosensor of lead bioavailability in contaminated environments. Soil Biology and Biochemistry 32, 1899-1906.

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 268

Peng, J.J., Cai, C., Qiao, M., Li, H. and Zhu, Y.G. (2010). Dynamic changes in functional gene copy numbers and microbial communities during degradation of pyrene in soils. Environmental Pollution 158, 2872-2879. Pesce, S., Beguet, J., Rouard, N., Devers-Lamrani, M. and Martin-Laurent, F. (2013). Response of a diuron-degrading community to diuron exposure assessed by real-time quantitative PCR monitoring of phenylurea hydrolase A and B encoding genes. Applied Microbiology and Biotechnology 97, 1661-1668. Pietramellara, G., Ascher, J., Borgogni, F., Ceccherini, M., Guerri, G. and Nannipieri, P. (2009). Extracellular DNA in soil and sediment: fate and ecological relevance. Biology and Fertility of soils 45, 219-235. Piutti, S., Hallet, S., Rousseaux, S., Philippot, L., Soulas, G. and Martin-Laurent, F. (2002). Accelerated mineralisation of atrazine in maize rhizosphere soil. Biology and Fertility of Soils 36, 434-441. Prosser, J.I. (2002). Molecular and functional diversity in soil micro-organisms. Plant and Soil 244, 9-17. Racke, K.D., Steele, K.P., Yoder, R.N., Dick, W.A. and Avidov, E. (1996). Factors affecting the hydrolytic degradation of chlorpyrifos in soil. Journal of Agricultural and Food Chemistry 44 (6), 1582-1592. Radicojac, P., Clark, W.T., Oron, T.R., Schnoes, A.M., Wittkop, T., Sokolov, A., Graim, K., Funk, C., Verspoor, K., Ben-Hur, A., Pandey, G., Yunes, J.M., Talwalkar, A.S., Repo, S., Souza, M.L., Piovesan, D., Casadio, R., Wang, Z., Cheng, J., Fang, H., Gough, J., Koskinen, P., Törönen, P., Nokso-Koivisto, J., Holm, L., Cozzetto, D., Buchan, D.W., Bryson, K., Jones, D.T., Limaye, B., Inamdar, H., Datta, A., Manjari, S.K., Joshi, R., Chitale, M., Kihara, D., Lisewski, A.M., Erdin, S., Venner, E., Lichtarge, O., Rentzsch, R., Yang, H., Romero, A.E., Bhat, P., Paccanaro, A., Hamp, T., Kaßner, R., Seemayer, S., Vicedo, E., Schaefer, C., Achten, D., Auer, F., Boehm, A., Braun, T., Hecht, M., Heron, M., Hönigschmid, P., Hopf, T.A., Kaufmann, S., Kiening, M., Krompass, D., Landerer, C., Mahlich, Y., Roos, M., Björne, J., Salakoski, T., Wong, A., Shatkay, H., Gatzmann, F., Sommer, I., Wass, M.N., Sternberg, M.J., Škunca, N., Supek, F., Bošnjak, M., Panov, P., Džeroski, S., Šmuc, T., Kourmpetis, Y.A., van Dijk, A.D., ter Braak, C.J., Zhou, Y., Gong, Q., Dong, X., Tian, W., Falda, M., Fontana, P., Lavezzo, E., Di Camillo, B., Toppo, S., Lan, L., Djuric, N., Guo, Y., Vucetic, S., Bairoch, A., Linial, M., Babbitt, P.C., Brenner, S.E., Orengo, C., Rost, B., Mooney, S.D. and Friedberg, I. (2013). A large-scale evaluation of computational protein function prediction. Nature Methods 10 (3), 221-227. Rocca, J.D., Hall, E.K., Lennon, J.T., Evans, S.E., Waldrop, M.P., Cotner, J.B., Nemergut, D.R., Graham, E.B. and Wallenstein, M.D. (2015). Relationships between protein-encoding gene abundance and corresponding process are commonly assumed yet rarely observed. The ISME Journal 9, 1693-1699. Smalling, K.L., Fellers, G.M., Kleeman, P.M. and Kuivila, K.M. (2013). Accumulation of pesticides in pacific chorus frogs (Pseudacris regilla) from California's Sierra Nevada Mountains, USA. Environmental Toxicology and Chemistry, 32 (9), 2026-2034. Sorensen, S.R., Ronen, Z. and Aamand, J. (2001). Isolation from agricultural soil and characterization of a Sphingomonas sp. able to mineralize the phenylurea herbicide isoproturon. Applied and Environmental Microbiology 67 (12), 5403-5409. Soulas, G. and Lagacherie, B. (2001). Modelling of microbial degradation of pesticides in soils. Biology and Fertility of Soils 33 (6), 551-557. Storck, V., Lucini, L., Mamy, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., Karas, P.A., Servien, R., Karpouzas, D.G., Trevisan, M., Benoit, P. and Martin-Laurent, F. (2016). Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology. Environmental Pollution 208, 537-545. Topp, E. (2003). Bacteria in agricultural soils: diversity, role and future perspectives. Canadian Journal of Soil Science 83, 303-309. Wagner, R. (1994). The regulation of ribosomal RNA synthesis and bacterial cell growth. Archives of Microbiology 161, 100-109. Zhang, W., Jiang, F. and Ou, J. (2011). Global pesticide consumption and pollution: with China as a focus. Proceedings of the International Academy of Ecology and Environmental Sciences 1(2), 125-144.

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 269

Conclusion

With the research of Chapter 5, I studied the interest of quantifying the IPU- degrading genetic potential as a proxy of IPU mineralization in soil. Under the contemporary scenario of exposure to IPU, the IPU-degrading genetic potential measured in the Biobed soil was found to transiently increase during the IPU mineralization activity and to decrease when IPU mineralization was completed. By tracking a transposon-like sequence harboring the pdmA and pdmB genes (involved in IPU degradation) and the plasmid pSH, we proposed that the IPU-degrading genetic potential does not only consist of the known IPU-degrading genetic organization. Under the historical scenario of exposure to IPU, the detection of IPU-degrading genetic potential was possible but remained rather low for all the samples collected over the three years during crop rotation. Contrariwise to the good correlation observed during contemporary exposure, the IPU-degrading genetic potential in the field with historical exposure to IPU did not show a correlation to the measured IPU mineralization. This study suggests that IPU-degrading genetic potential needs to be newly stimulated and quantified contemporary to IPU exposure to successfully use the genetic potential to predict IPU mineralization in a soil historically exposed to IPU. From this point of view, it may constitute a sensitive bioindicator of IPU exposure and of microbial resilience to IPU.

Chapter 5 Possibilities and limitations of using genes as bioindicators for potential pesticide degradation in soil - the case of pdm-encoded isoproturon degradation 272

Chapter 6

General discussion

Veronika Storck

Ways to improve the current pesticide policy in the EU

An existing regulation (EU-Regulation 1107/2009/EC) is responsible for the approval of active substances (used in PPPs) in the EU. Despite the high amounts of time, human resources and funding that are spent by the pesticide companies to comply with the pesticide policy, the overall pesticide authorization process is heavily criticized for being relatively inefficient and for compromising the evolution of a non-precautionary principle (EEA, 2013). As consequences of their extensive use worldwide, pesticide residues persist in agricultural soils from where they may be transported to non-directly exposed environments and where they may impact non-target organisms (Barbash, 2003; Dallaire et al., 2012; Looser et al., 2000; Multigner et al., 2010; Smalling et al., 2013). For instance, several new-generation insecticides were recently shown to have undesirable effects on honey bees among other insects (European Academies Science Advisory Council EASAC, 2015). Furthermore, also human health was shown to be impaired by exposure to pesticides. For instance, this was the case in the French West Indies where the pesticide chlordecone was associated to an increased frequency of prostate cancer and infantile delays in local exposed populations (Dallaire et al., 2012; Multigner et al., 2012). Improvement of the current pesticide policy can be a good leverage to tackle some of these environmental issues. Within this context, I aimed to understand the different steps of the complex pesticide authorization process to identify possible ways of improvement - notably by conducting a more reliable pesticide risk assessment during the pre- and post-authorization processes (Chapter 2). One major achievement of the opinion paper is the proposal of a simplified and intelligible view of the myriad steps involved in the pesticide authorization process. Indeed, it is a real challenge (for a non-expert) to obtain a clear view of the authorization process, as publicly available information is hardly deducible and needs to be complemented by many different sources of information. This was also perceptible from the reviewers’ comments on the opinion paper publication, which provided divergent views on this process (concerning details such as which authority is finally responsible for what). Another achievement is the identification of some weaknesses in the pesticide authorization process and the proposal of possible ways of improvement. Moreover, although the paper is an opinion paper, it was built on facts based on the current regulation.

Chapter 6 General discussion 276

Among possible pre-authorization improvements, one could cite (i) the reduction of the time lag between the market introduction of a new pesticide and the awareness of risks by enabling the contemporaneous initiation of private and public investigations, (ii) the commitment of only one authority regulating both active substances and PPPs, (iii) the assignment of environmental risk assessment studies to anonymous accredited subcontractors to avoid any kind of conflicts of interest, (iv) a limitation of the rapid evolution of the pesticide market by restricting new pesticide authorizations to those with an improved efficiency and environmental footprint than existing ones, (v) the redefinition of the criteria to qualify a pesticide transformation product as ‘relevant’ and (vi) the use of new methods to estimate the effect of pesticides on the diversity, abundance and activity of microbial communities supporting soil ecosystem services. The last two mentioned ideas were chosen and tackled by the research of this thesis. On the other hand, possible post-authorization improvements include (i) the immediate release of studies influencing environmental risk assessment to ease public follow-up research, (ii) a standardization of the data collection for monitoring studies to allow an easily accessible and comprehensive image of environmental pesticide contamination, (iii) the establishment of an action plan to speed up the post- authorization response to newly identified risks, (iv) the limitation of the number of exceptional authorizations for special uses of banned pesticides, (v) a new EU-law to ban the sale and export of EU-banned pesticides to other continents, and (vi) the implementation of the EU-Soil Framework Directive to preserve soil biodiversity and soil ecosystem functions (Van-Camp et al., 2004) that has been proposed to the EC in 2006 but was postponed several times. The latter idea is the keystone required to support the implementation of a more reliable estimation of pesticide effects on soil microorganisms important in ecosystem functioning, which was the main aim of my PhD research. One may criticize the proposed ways of improvement of being difficult to be realized, as this would imply drastic changes causing a deep increase in human resources and funding to be invested prior to the market introduction of pesticides. These changes (if applied) would implicate an entire re-thinking of the economic model of the pesticide business and consequently of agriculture, based on a ‘new deal’ established between the society, the pesticide companies and the farmers to protect ecosystem services provided by our vulnerable planet earth. One of the propositions of

Chapter 6 General discussion 277

the opinion paper was to improve the trustworthiness of the process by employing anonymously chosen subcontractors to perform the investigations on which the pesticide risk assessment is based (instead of relying on data from the pesticide industry as it is now). An argument against this possibility relies on the question who will pay the subcontractors. In order to avoid any conflicts of interest, subcontractors have to be paid by a third party but keeping it financed by the pesticide company. Furthermore, one may criticize the environmental orientation of the paper and argue that alternatives to the excessive use of pesticides are difficult to realize regarding the increasing demand of food on earth due to the important increase in the human population. However, in a globalized world where food and goods are exchanged all over, the food shortage in some regions of the world is more a problem of global food distribution and wasteful consumer habits than of global food production. A better organization would imply an optimization of the use of resources and thus a limitation of the efforts required to increase the production to comply with the growing world population. Moreover, an alternative to the excessive pesticide use may also be a restriction of the most harmful compounds and the expansion of a more sustainable pesticide use.

Critical review of used and developed techniques

My main aim was to propose and test new approaches to assess pesticide microbial ecotoxicology. To reach this objective, my research aimed to enhance the characterization of (i) the exposure scenario of soil microorganisms to pesticides and their TPs in soil and (ii) the consequences of pesticide exposure for the abundance, activity and diversity of soil microorganisms. The purpose of Chapter 3 was to define the environmental exposure scenario of the three pesticides (CHL (insecticide), IPU (herbicide), or TCZ (fungicide)) applied to North-Italian agricultural soil in a microcosm and a field study. Given the fact that this PhD thesis rather focused on the usability and development of methods for pesticide environmental risk assessment, we made the choice to apply each pesticide alone and not mixed together (pesticide cocktail). Therefore, a potential ‘cocktail effect’ (Relyea, 2009) on the environmental fate of the pesticides and on soil microorganisms was not assessed. One could notice that methods applied in my work to assess the fate and the

Chapter 6 General discussion 278

impact of three separated pesticides can be likewise used to study effects of pesticide cocktails. The dissipation of the pesticides was monitored by the classical and commonly used HPLC-PDA method able to quantify the pesticides and their main known TPs extracted from a soil matrix; sorption of the pesticides and their main TPs was assessed using the standard batch equilibrium method (OECD-Guideline 106, 2000) (Chapter 3.1). One of the originalities of the proposed approach is to consider lower and higher pesticide exposure scenarios (several applied pesticide concentrations) to study the environmental fate of the three pesticides. Another one is based on the use of PPPs to carry out the dissipation and sorption studies, and not only use active substances as recommended by EFSA to carry out environmental risk assessment. Indeed, in reality, active substances are always applied in formulated PPPs to the environment. Keeping in mind that several studies showed that the environmental risk of a PPP can be divergent from that of the active substance (EFSA, 2015a; IARC, 2015), splitting their risk assessment is not logical and one may be astonished that the environmental fate is only assessed for the active substance from the EU-regulatory point of view. To tackle this issue, the sorption study was performed with the active substance and the formulated PPP, which allowed the comparison of both. Moreover, pesticide degradation was assessed in a lab-to-field experiment by monitoring dynamics of known TPs. Studying the environmental fate of pesticides in field studies is important, as several abiotic (leaching, run-off, volatilization, photodegradation etc.) and biotic (plant-, animal-, or microorganism-interactions) processes can be considered that do not account to dissipation in soil microcosms. One of the weaknesses of the work I have carried out is that all my pesticide dissipation experiments have been conducted on one soil type and not on a range of soils differing in their physicochemical properties sampled from all over the world. However, even if one could expect that the use of various soils may have provided a stronger view of the fate and the microbial ecotoxicology of these three pesticides, having in mind that ‘everything is everywhere but the environment selects’ (O’Malley, 2008), one may hypothesize that observations obtained in a given soil provide general trends transposable to other soils by considering a range of modulation factors such as physicochemical parameters. One could recommend that environmental risk assessment of pesticides has to be carried out at least on three types of soils representative of the

Chapter 6 General discussion 279

three EU climatic zones (Scandinavian, temperate and Mediterranean soil) identified in the pesticide policy. This recommendation was followed within the framework of the program of the European ‘Love-to-Hate’ project but is not part of my PhD. Another weakness of this work may be attributed to the method chosen to assess pesticide soil sorption. Although the sorption studies were carried out according to existing standards required by the regulation to estimate pesticide fate in soil (i.e. batch study according to OECD-Guideline 106, 2000), one could regret to not have assessed pesticide sorption to soil under a more realistic scenario. Indeed, pesticide transfer through non-disturbed soil columns (Dagès et al., 2015), that takes into account preferential flow of water through soil macropores, may have given another estimate of pesticide sorption to soil. The exposure scenario of soil microorganisms to pesticides was assessed by quantifying pesticide residues and known TPs for which reference standards were already available (enabling their quantification). Although most TPs have a lower ecotoxicity than the parent molecule (Boxall et al., 2004), some of them may be more toxic and are often more soluble, so that they possess higher tendencies to be more bioavailable or more prone to migrate to water bodies (Fenner et al., 2013). Knowledge of the existence of TPs relevant for the environment is one of the biggest challenges in exposure scenario research. Past studies have shown that several pesticide TPs of relevant ecotoxicity typically emerged only 20 to 30 years after the pesticide’s market introduction (Fenner et al., 2013). To tend to fill this gap of knowledge on TPs formed in soils, I established a combined approach of suspect screening (UHPLC-ESI-QTOF-MS), molecular structure correlation (MSC program) and in silico molecular typology (TyPol) (Chapter 3.2). This approach was tested for the detection of TCZ TPs, as the degradation pathway and the TPs of this pesticide were the least experienced of the three pesticides. Moreover, triazole TPs possibly resulting from the transformation of TCZ are of particular importance, as they possibly interact with the hormone regulation network of non-target organisms (Rieke et al., 2014; Shalini et al., 2011). The major force of this approach lies on the establishment of a library made of empirical and theoretical TPs. The richness of the library is the keystone for TP search in organic extracts of environmental samples. Another of the forces of this approach is that the availability of reference standards of each TP stored in the library is not compulsory (albeit preferable, as they are needed for the identity confirmation and quantification of detected TPs). The classification of detected TPs by TyPol (Servien et al., 2014)

Chapter 6 General discussion 280

according to their estimated environmental parameters deduced from their physicochemical properties is a good possibility to identify important TPs to be synthesized to enable (i) their identity confirmation and quantification in environmental samples and (ii) the assessment of their risk to non-target organisms. One of the major weaknesses of this suspect screening approach is that TPs cannot be quantified and their identity cannot be confirmed without the use of reference standards. In addition, the suspect screening approach can only detect extractable compounds that are contained in the compound library. Moreover, the used TyPol approach has not yet experimentally proven to deliver reliable and robust deductions about the potential environmental behavior of classified compounds. Despite these weaknesses, this combined approach represents a large step towards a more comprehensive description of pesticide TPs formed in soil. In specific respect to the case of TCZ, one weakness of the used approach concerns that although TCZ and several of its TPs are known to have two enantiomers or even more than one chiral center for a few of them (Wang et al., 2012), the consequences of the chirality on stereoselective ecotoxicity and dissipation have not been addressed in this thesis. To overcome this gap of knowledge, several trials using chiral separation columns are currently ongoing in the research team of Dr. Luigi Lucini (Catholic University of the Sacred Heart, Piacenza, Italy). Aims of Chapter 4 were to test the usability of a range of methods for a better evaluation of potential pesticide impacts on soil microorganisms than it is actually done in pesticide risk assessments that solely rely on nitrogen and carbon mineralization tests (Martin-Laurent et al., 2013). One can recommend assessing the potential effects caused by pesticide exposure to soil microorganisms and their ecosystem functions via their microbial diversity, abundance and activity. This is in agreement with recent scientific opinions published by EFSA who proposed soil ecosystem services as subjects of specific protection goals in pesticide environmental risk assessments (EFSA, 2010) and functional microbial communities as entity to be protected (EFSA, 2016a). Within this context, I have tested different techniques on their ability to detect adverse pesticide effects on soil microorganisms. One technique was the assessment of potential pesticide effects on the diversity of soil microorganisms using the PCR-based DNA Illumina next-generation sequencing approach relying on soil 16S rDNA amplicons of soil DNA extracts (ISO 11063, 2012)

Chapter 6 General discussion 281

(Chapters 4.1 and 4.2). This DNA sequencing method is not yet standardized at the ISO level because of the lack of consensus on the most appropriate protocol to be used, which may be explained by its novelty as well as by its constant and rapid evolution inhibiting its standardization (Philippot et al., 2012). By the research of this thesis, the sequencing approach was found to be a suitable method for the detection of big shifts in the bacterial composition of a soil, whereas small shifts of rare species that may however have big impacts on ecosystem functions (such as accelerated pesticide degradation) were likely to remain undetected by this under-development approach. For instance, the abundance of pesticide degraders in soil usually remains quite low (around 103 to 104 cells per g of soil) (Piutti et al., 2003) making it difficult to detect them ‘diluted’ within dominant microbial populations (around 108 cells per g of soil). However, next-generation Illumina sequencing is one of the most sensitive sequencing techniques of today and thus remains the best alternative. A force of this research was that both the formulated PPP and the active substance were tested in microcosm and field studies at different doses (the smallest dose being the recommended agronomical dose) that were only once or repeatedly applied to the soil. Study results thus enabled a comprehensive conclusion on each pesticide. A weak point of this research may be that the sequencing technique only considered bacterial DNA. Pesticide impacts on fungi were thus excluded in the sequencing approach, although the technique could have been similarly performed on fungal DNA. The open question that remains after the detection of some pesticide effects on soil microorganisms is to which extent such impacts can be related to potential effects on soil ecosystem functions. This question is widely debated (Prosser, 2002), as it remains difficult to understand the functioning of all complex processes and interactions of microorganisms in soil. This is why it is a real challenge to drive reliable conclusions about the effect of any pollutant on non-target organisms in the environment. Another technique evaluated in this thesis was the assessment of potential pesticide impacts on microbial activity in soil, evaluated via pesticide effects on soil buffering functions (Chapter 4.2). As a proxy for soil buffering functions, I tested the capability of soil microbial communities to degrade a range of three compounds (sodium acetate, 2,4-D or glyphosate) that are thought to be degradable by a large, a moderate, or a small diversity of soil microorganisms. Sodium acetate was chosen because it is recommended to be used as readily biodegradable control substrate in the modified

Chapter 6 General discussion 282

Stürm test (OECD-Guideline 301, 1992). The herbicide 2,4-D was chosen because of its analogy to auxin (a natural phytohormone for plant growth), making it biodegradable by almost all soil microbial communities that are able to degrade auxin in the plant rhizosphere. It was the first herbicide for which accelerated biodegradation was reported (Audus, 1949). The herbicide glyphosate was chosen because it is commonly degraded in agricultural soil although it is currently under the spotlight of the society regarding the detection of its main degradate aminomethyl phosphonic acid (AMPA) in water resources (Kolpin et al., 2006). These three molecules possibly show different levels of degradation ease based on a potential functional redundancy following the trend: sodium acetate > 2,4-D > glyphosate. The degradation of these compounds in response to the exposure to two pesticides (CHL or TCZ) was found to be compound- specific, as only one of the two tested pesticides had significant impacts in these soil buffering functions. This first trial to attempt an estimation of the pesticide impact on soil buffering functions of soil microbial communities was a source of new questions. One question concerns the interpretation of the obtained results: Can we conclude that a given pesticide may have no impact on soil buffering functions, because it was not found to impact the degradation of the tested molecules? This would imply a non-compliance of the precautionary principle, as one cannot exclude that the capacity of soil microbial communities to degrade other pesticides may be impacted but ignored because not analyzed. To avoid this possible bias, it may be useful to re-think the choice of the molecules used in this test to cover a broader range of substrates (e.g. herbicides, fungicides and insecticides) using in silico typology to take into account their molecular and environmental characteristics. In addition to the two discussed techniques rather assessing adverse pesticide effects on soil microorganisms (Illumina next-generation sequencing assessing microbial diversity, or analysis of soil buffering functions assessing microbial activity), a third approach was used to study the positive effect of pesticide treatments on soil microbes, which was based on the adaptation of pesticide-degrading microorganisms to pesticide exposure (Chapter 4.2). The adaptation of microorganisms to pesticides is a commonly observed phenomenon (Sagarkar et al., 2015) based on the ability of some microbes to use pesticides as additional nutrient and/or energy source giving pesticide degraders an important advantage over other microorganisms in the densely populated soil ecosystem (Copley, 2009). Here, an increasing ability of soil microorganisms to

Chapter 6 General discussion 283

degrade a pesticide during repeated treatments signifies a (positive) pesticide effect. If adaptation and thus pesticide degradation is favored by repeated pesticide exposure, the exposure of pesticides to the environment is decreased and the accumulation of transformation products may be decreased. Adaptation is thus usually a desirable process although it may contribute to microbial diversity loss in soil. The approach used in this thesis to study such an adaptation of pesticide degraders to pesticide exposure consisted of the quantification of 14CO2 generated from 14C-labelled pesticides. It seemed to be a suitable approach, as it was able to sensitively detect an enhancement of pesticide mineralization due to microbial adaptation to repeated pesticide treatments. The purpose of Chapter 5 was to test if the quantification of a pesticide- degrading genetic potential can be a relevant bioindicator for the prediction of the potential pesticide degradation in soil as it has been proposed by Pesce et al. (2013). To achieve this objective, I quantified the abundance of the IPU-degrading genetic potential by qPCR (ISO 17601, 2016) targeting pdmA and pdmB sequences (coding for the first N- demethylation step in the bacterial IPU degradation pathway (Gu et al., 2013)) using as template soil DNA extracted either from an agricultural field regularly exposed to IPU or from an experimental soil (Biobed soil) receiving pesticide effluents (including IPU) from experimental farms. Our results showed that for the field experiment, the IPU- degrading genetic potential was not a good proxy of the measured IPU degrading potential. On the contrary, analyses carried out on the Biobed soil at different sampling times showed that (in response to IPU exposure) IPU-degrading genetic potential increased concomitantly with measured IPU mineralization. This is in agreement with previous studies done on different herbicides (atrazine, 2,4-D and diuron) showing that the size of the pesticide-degrading population increased in response to pesticide exposure (Baelum et al., 2008; Pesce et al., 2013; Piutti et al., 2002). Furthermore the findings of this thesis emphasized that the choice of the time point for soil sampling is crucial, as a correlation between the genetic potential and its resulting function is only achieved during the exponential phase of pesticide degradation. This may explain why quantification of IPU-degrading genetic potential was not a good proxy of the IPU microbial degradation in the field, where soil sampling time points were chosen after the IPU degradation was completed. The time-related variability of the abundance of a pesticide-degrading genetic potential is on one hand interesting because it is an indication for its sensitivity to pesticide exposure, but on the other hand it is a difficult to

Chapter 6 General discussion 284

deal with its time versatility. This makes it challenging to use catabolic pesticide degradation-encoding genes as bioindicators to predict pesticide degradation because it is an intermitting process that only occurs in sufficient amounts directly after pesticide treatment. However, despite these discrepancies, this approach may be of high interest to monitor the efficiency of bioremediation techniques or as good completion to other methods that investigate pesticide mineralization (Piutti et al., 2002) or pesticide concentrations (OECD-Guideline ENV/JM/MONO (2007)17, 2007) in the environment.

Pesticide environmental behavior - the case of the three studied pesticides

The environmental dissipation and the ecotoxicological impact on soil microorganisms of the three pesticides CHL, IPU and TCZ has been studied in a North- Italian agricultural soil in order (i) to use, develop and evaluate methods for a better pesticide environmental risk assessment, and (ii) to enhance the post-authorization risk assessment of the three chosen pesticides with new findings. To simulate a lower tier pesticide exposure scenario, CHL, IPU or TCZ were applied once (as formulated plant protection products) in a microcosm and a field study at three different doses (1x, 2x or 10x the agronomical dose (microcosms) or 1x, 2x or 5x the agronomical dose (field)). Their dissipation was assessed via quantification of the pesticides and their main TPs. Moreover, the detection of new TCZ TPs was achieved via a suspect screening approach. The ecotoxicological impact of the three pesticides on soil bacteria was assessed by Illumina next-generation sequencing of 16S rDNA amplicons generated from soil DNA extracts. To simulate a higher tier pesticide exposure scenario, CHL or TCZ were repeatedly applied (as active substances) in a microcosm study at 10x agronomical doses. The impact of repeated CHL or TCZ treatments on their mineralization to CO2 was quantified. Furthermore, their effects on soil buffering functions (evaluated by monitoring the mineralization of other compounds) and on soil bacteria (evaluated by Illumina next-generation sequencing of 16S rDNA amplicons) was assessed as a proxy for their ecotoxicological impact on soil microorganisms.

Chapter 6 General discussion 285

In addition, the abundance of IPU-degrading bacteria was assessed in two Central Eastern French soils (‘Biobed soil’ and ‘Epoisses soil’) (i) via quantification of IPU mineralization to CO2 and (ii) via quantification (by qPCR) of IPU degradation-encoding genes.

CHL fate and impact in agricultural soil

Regarding the lower tier CHL exposure scenario, the dissipation of CHL (applied as formulated PPP) and its main TP TCP was quantified in the microcosm and the field study at three different doses. The sorption affinity of CHL (active substance or formulated PPP) and TCP was assessed in a sorption batch study (OECD-Guideline 106,

2000). In addition, the mineralization of 14C-labelled CHL (active substance) to 14CO2 was quantified in the higher tier CHL exposure scenario, where 10x CHL doses were repeatedly applied in the microcosm study. CHL dissipation was biphasic in the microcosm and the field study (faster dissipation in the beginning) showing a moderate persistence with DT50-values varying between 28 to 120 days (depending on applied doses) that were within the range of

DT50-values reported in the literature (Jin and Webster, 1997; Laabs et al., 2002; Racke, 1993). CHL dissipation was found to be faster in the field study than in the microcosm study. Interestingly, CHL dissipation increased with increasing applied doses in the microcosm study, while the opposite trend was observed in the field study and generally in the literature (John and Shaike, 2015). This phenomenon may be attributed to the high sorption affinity of CHL (as found in this thesis). In static soil laboratory incubation systems such as microcosms, sorption may be one of the main processes explaining dissipation of CHL at low doses, resulting in its relatively low bioavailability but contributing to its higher persistence as bound residues. However, when applied at high doses, CHL may saturate soil sorption sites. Remaining non-sorbed CHL may be bioavailable and susceptible for transformation. In a field study, saturation of soil sorption sites is most likely not reached, wherefore we can observe the opposite trend. CHL mineralization results further support this hypothesis, as repeated exposure to CHL increased the capability of the soil microflora to mineralize it to CO2. We assume that this increase is the result of the adaptation of a fraction of the soil microbial community to CHL that became more abundant and more active in response to repeated exposure.

Chapter 6 General discussion 286

This hypothesis is supported by several studies reporting microbial adaptation and increasing degradation in soil repeatedly exposed to various pesticides (De Andrea et al., 2003; Topp et al., 2004; Udikovic-Kolic et al., 2012; Vischetti et al., 2008). Trials to purify CHL degraders from the soil by enrichment cultures led to the isolation of a consortium able to mineralize it, thereby indicating the existence of degraders in repeatedly exposed soil microcosms (data not shown). In both the microcosm and the field study, degradation of CHL appeared to happen via the transformation to its main transformation product TCP, which is consistent with other soil dissipation studies (Racke et al., 1996; Singh et al., 2003). The amount of generated TCP (present from the beginning) first increased to a maximum and then continuously decreased in the microcosm study (when treated at the 10x agronomical dose), while TCP amounts (present from the beginning) steadily decreased in the field study. A (temporary) accumulation of TCP in soil was assumed to typically occur due to the antimicrobial characteristics of TCP (Racke et al., 1990), supporting our findings of the microcosm study (only at the 10x dose). TCP has a low sorption affinity and a high water solubility (Baskaran et al., 2003), which may explain that TCP did not accumulate in the field study, where it could dissipate away from the sampled soil layer by leaching to groundwater or run-off to surface water. This hypothesis is further supported by the contemporary climatic data reporting a rain event at day 3 after CHL treatment, right before soil sample collections. In accordance with its hydrophobic character, CHL showed a high soil sorption affinity in the batch study. The sorption affinity of CHL applied as active substance was lower than that of formulated CHL (contained in a plant protection product), which may be attributed to the presence of organic additives in formulations (including dispersing or wetting agents and other surfactants) that potentially increased the sorption of CHL to soil, as it was as well found for IPU and TCZ and in the literature (Cadkova et al., 2012). As expected, the TCP soil sorption affinity was found to be low, which is in accordance with other studies (Baskaran et al., 2003) reporting TCP soil sorption affinity to be 100 times lower than CHL soil sorption affinity. Consequently, TCP may reach water resources and should receive more profound attention in both environmental risk assessment and post-authorization monitoring studies. The ecotoxicological impact of one-time CHL applications on soil bacteria was estimated in both the microcosm and the field soil exposed to CHL (applied in the lower

Chapter 6 General discussion 287

tier scenario as PPP at three different doses), evaluated by Illumina next-generation sequencing using soil DNA extracts as template. The impact of repeated CHL treatments (applied in the higher tier scenario at 10x doses as active substance) on soil microorganisms was evaluated in microcosms via its mineralization in soil, via the mineralization of various other organic compounds as a proxy of soil buffering functions, and via the soil bacterial composition. When applied only once, CHL had no clear impact on the soil bacterial diversity indices in both the microcosm and the field study. ANOSIM on PCoA revealed a slight but significant transient effect of CHL on the soil bacterial composition in the microcosm study. In addition, when repeatedly applied at 10x doses, CHL produced distinct effects on the soil microbial composition, as evaluated by 16S rDNA sequencing. This suggests that today available ‘next-generation sequencing’ methods may be suitable to detect large shifts of highly abundant species, while small shifts of rare species may be overshadowed by the dominant ones. Ongoing developments of sequencing facilities may rapidly increase the deepness of next-generation sequencing methods, which will give new insights on low abundant OTUs and thus facilitate the detection of pesticide effects on those. Although one could hypothesize that these methodologic issues can be overcome in the future, the ecological importance of the low abundant OTUs remains poorly understood for now (Bracken and Low, 2012; Lynch and Neufeld, 2015). According to the ecological insurance theory, keeping species diversity is the best way to keep the functionality of an ecosystem, wherein the rare species may be especially crucial (Jain et al., 2014). Repeated CHL treatments doubled its mineralization by soil microorganisms. By using CHL as an additional nutrient and/or energy source, degrading microorganisms had a selective advantage over other microorganisms leading to their propagation (Copley, 2009). This kind of adaptation of soil microorganisms to repeated pesticide treatments is known to lead to accelerated biodegradation of pesticides and was previously observed in a number of soils repeatedly treated with various pesticides (Crouzet et al., 2010; De Andrea et al., 2003; Hussain et al., 2011; Vischetti et al., 2008; Weaver et al., 2007). However, soil buffering functions were not deeply affected in response to repeated CHL exposure (only 2,4-D and sodium acetate degradation were slightly increased). Although the CHL degradation pathway diverges from that of 2,4-D or sodium acetate, one cannot exclude that CHL- degrading microorganisms may also degrade these compounds. Search for ‘multi-

Chapter 6 General discussion 288

degraders’ of various pollutants has often been aspired as a ‘Holy Grail’ by bioremediation researchers but was rarely achieved, wherefore this hypothesis has to be taken with care. In addition, respeated CHL exposure was found to have clear and significant effects on the soil bacterial composition. The abundance of six OTUs was significantly decreased in CHL-treated samples, while the abundance of nine OTUs was significantly increased. Among the sensitive species, two OTUs were affiliated with the genus Nitrospira or the order Rhizobiales. Species of the genus Nitrospira are important in ecosystem functioning as they are involved in the biogeochemical nitrogen cycle by oxidizing nitrite to nitrate (Altmann et al., 2003), a process especially important in global N2O greenhouse gas production (Ma et al., 2009; Philippot and Germon, 2005) or in aquatic ecosystems to counteract the accumulation of toxic nitrite concentrations (Hovanec et al., 1998). Moreover, one Nitrospira species has recently been found to perform complete nitrification from ammonia via nitrite to nitrate (Daims et al., 2015). Species of the order Rhizobiales are important key players for soil fertility in agriculture, as some of them (such as Rhizobium sp. or Bradyrhizobium sp.) fix nitrogen and are symbiotic with plant roots (Gage, 2004). The presence of anthropogenic chemicals was already found to reduce Rhizobium populations in soil (Slattery et al., 2001). Because of their importance for leguminous and other crops, Rhizobium species can be interesting bioindicators for the detection of potential pesticide threats to soil fertility. Among the species stimulated by CHL exposure, two OTUs were affiliated with the order Legionellales or the genus Lysobacter. Given the fact that several species of the order Legionellales are human pathogens such as Legionella pneumophila causing Legionnaires’ disease (Garrity et al., 2001), their increase in response to CHL is undesirable. Species of the genus Lysobacter are known to produce extracellular enzymes and antibiotics (Ahmed et al., 2003; Kimura et al., 1996; Von Tigerstrom, 1980), a feature that makes Lysobacter a good competitor against rival microorganisms (Vasilyeva et al., 2008) in the microbial fight for nutrients in the overpopulated soil (Fontaine et al., 2003). Thus, an expansion of Lysobacter in response to CHL exposure (potentially caused by weakening effects on a range of its rival microorganisms) may have indirect effects on the soil microbial diversity. This hypothesis emphasizes the importance that indirect pesticide effects may have on sensitive species, potentially influencing the equilibrium of the soil ecosystem.

Chapter 6 General discussion 289

One can conclude that lower tier exposure experiments (showing slight but significant transient CHL effects on the soil bacterial composition) were in accordance with the risk assessment of the EC (EC, 2005) concluding CHL to have low risks on soil microorganisms (evaluated by carbon and nitrogen mineralization in soil) The higher tier repeated exposure experiments showed that CHL caused changes in microbial activity and diversity, which is in accordance with a range of other studies having the similar conclusions (Chu et al., 2008; Dutta et al., 2010; Eisenhauer et al., 2009; Fang et al., 2009; Gilani et al., 2010; Gupta et al., 2013; Jastrzebska, 2011; Pozo et al., 1995; Srinivasulu and Rangaswamy, 2013; Wang et al., 2010).

IPU behavior in agricultural soil

The dissipation of IPU (applied as formulated PPP at the three doses of the lower tier IPU exposure scenario) and its main TPs MD-IPU, DD-IPU and 4-IA was assessed in the microcosm and the field study. The sorption affinities of IPU (active substance or formulated PPP), MD-IPU, DD-IPU and 4-IA were assessed in the sorption batch study (OECD-Guideline 106, 2000). IPU was rather lowly persistent in the soil. Its dissipation was found to be slightly faster in the field than in the microcosm study. It followed the SFO kinetic model in the microcosm study and a biphasic pattern (faster in the beginning) in the field study. The persistence of IPU in soil was found to increase with increasing applied doses.

Depending on the applied dose, DT50-values ranged between 16 and 25 days (microcosm study) or between 7 and 12 days (field study) (increasing with applied dose). They were within the range of values previously reported (Walker et al., 2001; Rodriguez-Cruz et al., 2016). Our findings are in accordance with a range of studies showing that the dissipation of pesticides in soil is often faster in field than in microcosm studies (Laabs et al., 2000; Dolaptsoglou et al., 2009). This may be explained by the concomitant occurrence of dissipation processes such as volatilization, photolysis, leaching or run-off contributing to pesticide dissipation in field studies (EEC, 2000). MD-IPU was found to be the most abundant TP, followed by DD-IPU, which was detected at lower amounts. A transitory increase in TP amounts was observed in both the microcosm and the field study. Especially in the field study, one could observe the sequential formation of IPU TPs, where the maximum amount of DD-IPU was found some days after the maximum

Chapter 6 General discussion 290

amount of MD-IPU. This is in agreement with previous studies reporting the successive N-demethylations of IPU to MD-IPU and then to DD-IPU (Sorensen et al., 2001). The TP 4-IA was not detected in our studies. Although 4-IA is often detected in liquid cultures of IPU-degrading microorganisms (Johannesen et al., 2003; Sorensen and Aamand, 2001), it is rarely observed in soil (Mudd et al., 1983), potentially because of its strong sorption affinity to soil, potentially making it unavailable for extraction. Supporting our findings on the dissipation of IPU and its TPs, IPU, MD-IPU and DD-IPU were found to have weak soil sorption affinities, potentially leading to their high bioavailability and rather low persistence. Moreover, IPU’s low sorption affinity may explain its common detection in surface and groundwater resources (Skark and Zullei- Seibert, 1995; Skark et al., 2004). IPU applied as a formulated PPP showed a higher soil sorption affinity than IPU applied as active substance alone, which is congruent with the findings for CHL and TCZ and was as well found in the literature (Cadkova et al., 2012), and which can be explained by the enhancing effect that other organic additives may have on the sorption affinity. The ecotoxicological impact of IPU (applied as formulated PPP at thee doses in the lower tier IPU exposure scenario) on non-target soil microorganisms was assessed in the microcosm and the field study (North-Italian soil). The impact of IPU on the soil bacterial diversity was assessed starting from soil DNA extracts used as template to generate 16S rDNA amplicons to be sequenced by Illumina next-generation sequencing. In addition, the abundance of IPU degraders was quantified in soil microcosms (‘Biobac soil’) and on an agricultural field (‘Epoisses soil’) from Central Eastern France exposed to this herbicide by using qPCR targeting (i) the pdm genes encoding for IPU- degradation (Gu et al., 2013) and (ii) sequences specific for the pdm-harboring plasmid pSH of an IPU-degrading bacterial strain isolated from this soil (Hussain et al., 2011). Regardless of the applied dose, a single application of (formulated) IPU to (North- Italian) microcosm or field soil did not affect the diversity of soil bacteria in both the microcosm and the field study. This finding is in agreement with the evaluation of the EFSA (2015b) who concluded that IPU only represents a low risk for soil microorganisms, based on the pesticide environmental risk assessment reporting small differences in carbon and nitrogen mineralization. In addition, this in agreement with a range of studies showing that IPU had only small to moderate (sometimes temporary) effects on soil microorganisms (evaluated by microbial activity and biomass) (Harden et

Chapter 6 General discussion 291

al., 1993; Kuriyal and Pandey, 1994; Perrin-Ganier et al., 2001; Schuster and Schröder, 1990; Tag-El-Din, 1982). However, it has been clearly described in the literature that repeated application of IPU may lead to the selection and emergence of IPU-degrading populations (Walker et al., 1999; El Sebai et al., 20014). Within this context, I monitored the abundance of IPU-degraders by qPCR targeting pdm genes and a plasmid pSH harboring the pdm genes in Sphingomonas sp. SH, testing the hypothesis if IPU exposure may induce an increase of pdm- and pSH-harboring bacterial communities. Although the pdm genes and pSH were detected in DNA extracts of an experimental field (Epoisses soil of Central Eastern France) regularly exposed to IPU, no significant correlation between their abundance and IPU mineralization was observed. One may hypothesize that this was due to a relatively low abundance of IPU degraders in soil, as it has already been reported for other biodegradation genes (Piutti et al., 2003). This hypothesis was further supported by the experiment carried out in soil microcosms (Biobed soil of Central Eastern France repeatedly exposed to IPU and other pesticides), where an IPU treatment induced an increasing abundance of IPU degraders in soil, reaching its maximum during the exponential phase of IPU mineralization. This observation suggests that, although IPU exposure did not modify the diversity of the bacterial community (as estimated by Illumina next-generation sequencing in the North-Italian microcosm and field soil), IPU can lead to the emergence of degraders which may be present in too low amounts to be identified by next-generation sequencing. Having this in mind, one cannot exclude that IPU exposure may induce changes in the soil bacterial composition that remained undetected by the sequencing approach because of its relatively limited deepness.

TCZ behavior in agricultural soil

In the same manner as for the two other pesticides, the dissipation of TCZ (applied as formulated PPP at three doses in the lower tier TCZ exposure scenario) was monitored in the microcosm and the field study. As reference standards of the potentially most interesting TCZ TPs are not commercially available, they were not quantified. Nonetheless, I searched for known and new TCZ TPs by suspect screening of soil sample extracts from the field plots treated with the 5x agronomical dose. In addition, the mineralization of repeatedly applied TCZ (active substance, 10x doses) to

Chapter 6 General discussion 292

CO2 was quantified in the microcosm study of the higher tier TCZ exposure scenario. The sorption affinity of TCZ (active substance or formulated plant protection product) was assessed in a sorption batch study (OECD-Guideline 106, 2000). TCZ was found to be rather persistent in the soil of both the microcosm and the field experiment, showing a biphasic dissipation pattern. However, the microcosm and field study differed from each other, as they did not fit to the same dissipation model. In accordance with previous studies (Munoz-Leoz et al., 2011; Wang et al., 2016), the persistence of TCZ increased with increasing applied doses in both studies. As commonly found in previous studies (Herrero-Hernandez et al., 2011; Wang et al.,

2015), TCZ was more persistent in the microcosm than in the field study. DT50-values of TCZ ranged between 60 and 97 days in the microcosm study, which is in the range of

DT50-values found in the literature (Strickland et al., 2004; EFSA, 2014). In contrast, they were much lower (1 to 3 days) in the field study, where around 70 % of applied TCZ dissipated during the first days. This drastic dissipation may be explained by the occurrence of a rain precipitation during this period of time, which may have contributed to (i) TCZ leaching towards groundwater or to soil layers below the studied 10 cm topsoil layer or to (ii) TCZ transport to surface water by run-off. This hypothesis is strengthened by the frequent detection of TCZ in surface and groundwater (Herrero- Hernandez et al., 2013; Sancez-Gonzales et al., 2013). In addition, the drastic dissipation of TCZ during the first days is very unlikely to be fully attributed to TCZ degradation, as its persistence was much higher in the microcosm study, where microbial biodegradation is supposed to constitute the main dissipation process. This argument was further supported by the very weak TCZ mineralization observed in soil microcosms of the same soil (i.e. 10 % of applied TCZ mineralized to CO2 within 150 days of incubation). Herrero-Hernandez et al. (2011) proposed that the rapid formation of bound TCZ residues can be the cause of its rapid dissipation in a field study. However, regarding its rather low to moderate soil sorption affinity (EFSA, 2014), this hypothesis seems not very robust. The rapid initial concentration decline (of around 70 %) of TCZ was followed by a slow dissipation phase until the end of the experiment, where TCZ loss was found to be limited, suggesting a much higher persistence of TCZ than assumed by exclusively considering DT50-values. For this reason, DT90-values may be a more accurate measure of TCZ persistence in soil, which were modeled to be between 200 and 600 days in the field study. This secondary persistence of TCZ further

Chapter 6 General discussion 293

confirms that the initial fast TCZ dissipation may not fully be attributed to TCZ degradation. It may be attributed to a strongly sorbed TCZ fraction that potentially was less accessible to the dissipation process (Munoz-Leoz et al., 2011; Herrero-Hernandez et al., 2011). This hypothesis is supported by the mass balance of TCZ in the mineralization study, where > 20 % of applied 14C-labelled-TCZ were found to remain as (probably non-bioavailable) bound TCZ residues in the soil. TCZ sorption affinity to soil was found to be moderate, which is in agreement to other studies (Cadkova et al., 2013). By comparing TCZ applied as active substance or as formulated PPP, one can observe that the formulated pesticide showed a higher soil sorption affinity than the active substance, which was already observed for CHL and IPU and in the literature (Cadkova et al., 2012) and which may be explained by the organic additives present in the formulated pesticides, potentially favoring its sorption affinity to soil. The suspect screening study searching for 47 empirical and 29 theoretical TCZ TPs in the soil of the field study led to the detection of 22 known and 12 new TCZ TPs. These 34 TPs were detected with different levels of identification confidence. Among them, some have exactly the same mass and retention times. 12 TPs passed all steps of identification confidence, from non-target MS-only identification to target tandem MS and structure correlation. However, all detected TPs remain inferred until confirmed by the retention times and fragmentation patterns of pure reference standards that, in most cases, are not commercially available and expensive to synthesize. The lack of reference standards was also the reason why detailed quantifying data on the temporal patterns of detected TPs could not be established. It is noteworthy that 31 of the 34 detected TPs had an intact triazole-ring, while only three TPs had a cleaved, differentiated or degraded triazole-ring. These findings were somehow in accordance with the mineralization study, where less than 1 % of TCZ (14C-labelled on the triazole-ring) was mineralized to CO2 after 150 days. TPs with intact triazole-ring may be relevant, as they are not only persistent and but also possibly impair the hormone regulation network of non-target organisms by inhibiting the cytochrome P450-dependent conversion of lanosterol to ergosterol (Rieke et al., 2014; Shalini et al., 2011). The 34 TPs were categorized into three groups of different estimated environmental behaviors. 23 TPs are estimated to have a similar behavior in the environment as their parent compound TCZ, while 11 TPs may be more mobile and volatile than TCZ, suggesting a higher risk

Chapter 6 General discussion 294

for groundwater and air contamination (FOCUS, 2008; McCall et al., 1980). However, their estimated lower persistence and bioaccumulation potential than TCZ may counterbalance this risk. In order to verify the accuracy of this in silico classification approach, some TPs of each cluster should be synthesized to enable their quantification in environmental samples and to investigate their actual impact on non-target organisms. This data should then be included in future environmental risk assessments of TCZ. The ecotoxicological impact of TCZ (one-time applied as formulated PPP at three doses in a microcosm and a field study (lower tier exposure), or repeatedly applied as active substance at the 10x agronomical dose in microcosms (higher tier exposure)) on non-target soil microorganisms was assessed via (i) the soil bacterial diversity by Illumina next-generation sequencing of 16S rDNA amplicons generated from soil DNA extracts and (ii) the potential of the soil to mineralize TCZ and other molecules (as a proxy of soil buffering functions). Lower tier TCZ exposure was found to have slight but significant transient effects on the soil bacterial composition in the field study. However, repeated exposure to TCZ (applied at 10x doses) significantly modified the diversity of soil bacteria. It has to be noticed that the significance of shifts in bacterial communities remains a subject of discussion, in particular when lowly abundant OTUs are implied. Repeated exposure to TCZ significantly modified the bacterial diversity at all time points (evaluated by ANOSIM on PCoA ordinations of OTU weighted unifrac matrices) and even seemed to increase with treatment repetitions. Repeated exposure to TCZ led to an increased abundance of 11 OTUs and to a decreased abundance of 12 OTUs. One of the decreased OTUs was a species of the nitrite-oxidizing genus Nitrospira (important in global nitrogen cycling (Altmann et al., 2003; Ma et al., 2008; Philippot and Germon, 2005)) that was also found to be decreased by CHL exposure. These two observations suggest that Nitrospira may be sensitive to various pesticides. If this is the case, the common application of a range of different pesticides in agriculture all along a cropping cycle (Faasen, 1995) may be deleterious to Nitrospira abundance. This hypothesis has to be investigated on a range of pesticides. Five of the increased OTUs were affiliated to species of the genus Kaistobacter, the genus Lysobacter and three species of the family Saprospiraceae. The Kaistobacter genus belongs to the family of Sphingomonadaceae that is known to comprise several species able to degrade some aromatic compounds

Chapter 6 General discussion 295

(Balkwill et al., 2006), such as Sphingomonas sp. SRS2 (Sorensen et al., 2001) or Sphingomonas sp. SH3 (Hussain et al., 2011) able to degrade IPU. Thus, it is not impossible that Kaistobacter may be involved in TCZ degradation, although this hypothesis needs to be proven by further research. As observed for repeated CHL exposure, TCZ caused an increased abundance of Lysobacter (known for its features to effectively compete against rival microorganisms by using extracellular enzymes and antibiotics (Ahmed et al., 2003; Kimura et al., 1996; Von Tigerstrom, 1980)). It thus reinforces the hypothesis that Lysobacter profited from a potential weakening effect of these pesticides on rival microorganisms. Moreover, this hypothesis is supported by the increasing effects that TCZ was found to have on three species of the family Saprospiraceae, as one species of this family (Saprospira grandis) is known for its ability to capture other microorganisms by the production of slime and to lyse them by the production of exotoxins and enzymes with the purpose of using released nutrients (Lewin, 1997). Similarly to the strengthening effects on Lysobacter, species of this family may thus have been indirectly strengthened by the weakening effects that TCZ may have had on potential victims of these predators. These observations open the discussion on the interest of analyzing the bacterial diversity as a network of interactions, and also of performing risk assessment at multi-trophic levels in order to take into account pesticide effects on the microbial loop. One could conclude that at the lower tier scenario of exposure, TCZ ecotoxicological impacts on the soil bacterial diversity may be low, which is in accordance with results of TCZ’s risk assessments by the responsible authorities (EFSA, 2008; EFSA, 2014) indicating a low risk of TCZ on soil microorganisms (evaluated via carbon and nitrogen mineralization). At the higher tier scenario of exposure, TCZ was found to significantly modify the soil bacterial diversity in accordance with a range of previous studies evaluating the impact of this fungicide on soil microorganisms by various means (microbial biomass, diversity, activity or community analyses) (Anuradha et al., 2016; Bending et al., 2007; Cycon et al., 2006; Ferreira et al., 2009; Munoz-Leoz et al., 2011; Wang et al., 2016).

Chapter 6 General discussion 296

Combination of approved and new methods for a more reliable pesticide environmental risk assessment

My PhD thesis was part of the European project ‘Love-to-Hate’ aiming a better understanding of the ecotoxicological impact of pesticides on soil microorganisms. This subject is of primary interest as emphasized by the recent scientific opinions published by EFSA identifying soil microorganisms as specific protection goals to be protected (EFSA, 2010) by considering the concept of ecological recovery (EFSA, 2016a) and proposing a list of possible microbial indicators to be implemented in the environmental risk assessment of PPPs (EFSA, 2016b). My thesis concerned several aspects of pesticide microbial ecotoxicology as a potential input to pesticide environmental risk assessments. Within this framework, (i) I described weaknesses of the pesticide authorization process and proposed possible ways of improvements and (ii) I tested a number of the available techniques to estimate the pesticide exposure scenario and the ecotoxicological impact of pesticides on soil microorganisms (evaluated via the bacterial diversity and the microbial degradation activity as proxy of soil buffering functions). The general idea of the project was to carry out a comprehensive study to assess the impact of pesticides on soil microorganisms (i) using a lower and a higher tier pesticide exposure scenario, and (ii) applying a broad combination of chemical, biochemical and molecular techniques. The aim was to characterize the scenario of pesticide exposure on soil microorganisms and its potential consequences on the abundance, activity and diversity of microbial communities. This broad approach consists of two principal steps (Figure 6.1). The first step (determination of the pesticide exposure scenario) consists in the identification of pesticide TPs in the treated soil (by suspect screening), followed by their classification according to their estimated environmental risk (by TyPol). Based on this, one could recommend the synthesis of potentially relevant TPs to further confirm their identity, to quantify their abundance, and to assess their ecotoxicological impact in the soil. This way, a comprehensive and more reliable scenario of exposure of soil microorganisms to pesticide residues could be established. The second step (determination of the pesticide ecotoxicity scenario) consists in the evaluation of the ecotoxicological impact of pesticide residues on soil microorganisms by a combination of techniques able to assess the activity (e.g. pesticide

Chapter 6 General discussion 297

mineralization, soil buffering functions or enzyme activity), the abundance (e.g. qPCR targeting bioindicator genes or indicator species) and the diversity (e.g. next-generation DNA sequencing) of soil microorganisms. Although the studies of this project have been carried out on one single soil, they generated a huge amount of data, raising the question on how to organize, update, analyze and share these data. Similar questions emerged through the opinion paper (Chapter 2) on how to make publicly available data of pesticide environmental risk assessments and additional studies. In today’s digital era, one possibility may be to engage the construction of publicly available databases collecting data from (i) pesticide environmental risk assessment studies produced by the pesticide companies (pre- authorization studies), (ii) environmental monitoring data produced by a range of national agencies, and (iii) research results of post-authorization studies. Suggestions for the establishment of such a comprehensive and responsive database consisting of comprehensive research findings are visualized in Figure 6.1. The existence of such a database would simplify public research and enable the deep insight into collected monitoring data. Furthermore, a legal obligation to consider all public research and monitoring data in (post-)authorization pesticide environmental risk assessment should be established at EU-level.

Chapter 6 General discussion 298

Figure 6.1. Suggestions for the improvement of pesticide environmental risk assessment: Acquisition of data by combining chemical, biochemical and molecular techniques and subsequent collection and publication of obtained knowledge.

Chapter 6 General discussion 299

Conclusion and perspectives

The results of this thesis may enrich future post-authorization risk assessments of CHL (due in 2020), IPU (due in 2017), and TCZ (due in 2023) with additional robust information. Future research perspectives may be the identification of new CHL and IPU TPs and the synthesis of detected TPs of all three pesticides for their identity confirmation, their quantification in the environment and to assess their impact on non- target organisms. Furthermore, as TCZ and some of its TPs have two enantiomers or even possess more than one chiral center, it may be of interest to test if TCZ biodegradation and ecotoxicity may be stereoselective. Moreover, the potential ecotoxicological impact of IPU on soil microorganisms may be investigated in a higher tier IPU exposure scenario by techniques evaluated in this PhD (as done for CHL and TCZ). The unexplored spot of this study was the potential pesticide impact on fungal communities which should be assessed as well, in particular arbuscular mycorrhiza that have been suggested to be monitored within the frame of specific protection goals by EFSA (EFSA, 2016b). Another perspective may be the isolation and characterization of CHL degraders from the repeatedly treated soil, as they could be useful for the bioremediation of CHL-polluted environments. Last but not least, it may be investigated if impacted genera (Nitrospira, Rhizobium, Lysobacter and Saprospira) have the potential to act as indicator genera for screening or monitoring purposes testing the ecotoxicity of organic pesticides of various groups in agricultural soils of various geographical zones. The findings of this thesis open a new door for the enhancement of environmental risk assessment of PPPs to soil microorganisms. From this point of view, they may have the potential to improve the current pesticide environmental risk assessment. Suggestions for the improvement of the current pesticide policy are provided and several alternative approaches were tested, developed and evaluated. Moreover, to increase the possibility that this thesis may contribute to the establishment of a more reliable pesticide authorization process (or – to fix more realistic aims – that this thesis may contribute to a broader public knowledge about hidden pesticide effects in the environment), one perspective is the vulgarization of parts of this thesis in science outreach activities.

Chapter 6 General discussion 300

References

Ahmed, K.., Chohnan, S., Ohashi, H., Hirata, T., Masaki, T. and Sakiyama, F. (2003). Purification, bacteriolytic activity, and specificity of β-lytic protease from Lysobacter sp. IB-9374. Journal of Bioscience and Bioengineering 95, 27-34. Altmann, D., Stief, P., Amann, R., De Beer, D. and Schramm, A. (2003). In situ distribution and activity of nitrifying bacteria in freshwater sediment. Environmental Microbiology 5 (9), 798-803. Anuradha, B., Rekhapadmini, A. and Rangaswamy, V. (2016). Influence of tebuconazole and copper hydroxide on phosphatase and urease activities in red sandy loam and black clay soils. 3 Biotech (6), 78. Audus, L.J. (1949). The biological detoxication of 2,4-dichlorophenoxyacetic acid in soil. Plant and Soil 2 (1), 31-36. Baelum, J., Nicolaisen, M.H., Holben, W.E., Strobel, B.W., Sorensen, J. and Jacobsen, C.S. (2008). Direct analysis of tfdA gene expression by indigenous bacteria in phenoxy acid amended agricultural soil. The ISME Journal 2, 677-687. Balkwill, D.L., Fredrickson, J.K. and Romine, M.F. (2006). The prokaryotes. A handbook of the biology of bacteria. Volume 7 – Proteobacteria: Delta and epsilon subclasses. Deeply rooting bacteria. Sphingomonas and related genera. ISBN 978-0-387- 30747-3. Barbash, J.E. (2003). The geochemistry of pesticides. Treatise on Geochemistry 9, DOI: 10.1016/B0-08-043751-6/09056-3. Baskaran, S., Kookana, R.S and Naidu R. (2003). Contrasting behaviour of chlorpyrifos and its primary metabolite, TCP (3,5,6- trichloro-2-pyridinol), with depth in soil profiles. Austrian Journal of Soil Science 41, 749-760. Bending, G.D., Rodriguez-Cruz, M.S. and Lincoln, S.D. (2007). Fungicide impacts on microbial communities in soils with contrasting management histories. Chemosphere 69, 82-88. Boxall, A.B.A., Sinclair, C.J., Fenner, K., Kolpin, D. and Maund, S.J. (2004). When synthetic chemicals degrade in the environment. Environmental Science and Technology 38 (19), 368A-375A. Bracken, M.E.S. and Low, N.H.N. (2012). Realistic losses of rare species disproportionately impact higher trophic levels. Ecology Letters 15, 461-467. Čadková, E., Komárek, M., Kaliszová, R., Koudelková, V., Dvořák, J. and Vaněk, A. (2012). Sorption of tebuconazole onto selected soil minerals and humic acids. Journal of Environmental Science and Health, Part B47, 336-342. Čadková, E., Komárek, M., Kaliszová, R., Vaněk, A. and Balíková, M. (2013). Tebuconazole sorption in contrasting soil types. Soil and Sediment Contamination 22, 404-414. Chu, X., Fang, H., Pan, X., Wang, X., Shan, M., Feng, B. and Yu, Y. (2008). Degradation of chlorpyrifos alone and in combination with chlorothalonil and their effects on soil microbial populations. Journal of Environmental Sciences 20, 464-469. Copley, S.D. (2009). Evolution of efficient pathways for degradation of anthropogenic chemicals. Nature Chemical Biology 5, 559-566. Crouzet, O., Batisson, I., Besse-Hoggan, P., Bonnemoy, F., Bardot, C., Poly, F., Bohatier, J. and Mallet, C. (2010). Response of soil microbial communities to the herbicide mesotrione: a dose-effect microcosm approach. Soil Biology and Biochemistry 42, 193-202. Cycon, M., Piotrowska-Seget, Z. and Kaczynska, A. (2006). Microbiological characteristics of a sandy loam soil exposed to tebuconazole and λ-cyhalothrin under laboratory conditions. Ecotoxicology 15, 639-646. Dages, C., Samouëlian, A., Storck, V., Huttel, O. and Voltz, M. (2015). Seepage patterns of diuron in a ditch bed during a sequence of flood events. Science of the Total Environment 537, 120-128. Daims, H., Lebedeva, E.V., Pjevac, P., Han, P., Herbold, C., Albertsen, M., Jehmlicht, N., Palatinszky, M., Vierheilig, J., Bulaev, A., Kirkegaard, R.H., von Bergen, M., Rattei, T., Bendinger, B., Nielsen, P.H. and Wagner, M. (2015). Complete nitrification by Nitrospira bacteria. Nature 528, 504-509. Dallaire, R., Muckle, G., Rouget, F., Kadhel, P., Bataille, H., Guldner, L., Seurin, S., Chajes, V., Monfort, C., Boucher, O., Thome, J.P., Jacobson, S.W., Multigner, L. and Cordier, S. (2012). Cognitive, visual, and motor development of 7-month-old Guadeloupean infants exposed to chlordecone. Environmental Research 118, 79-85. De Andrea, M.M., Peres, T.B., Luchini, L.C., Bazarin, S., Papini, S., Matallo, M.B. and Tedeschi Savoy, V.L. (2003). Influence of repeated applications of glyphosate on its persistence and soil bioactivity. Pesquisa Agropecuaria Brasileira 38 (11), 1329-1335. Dolaptsoglou, C., Karpouzas, D., Menkissoglu-Spiroudi, U., Eleftherohorinos, I. and Voudrias, E.A. (2009). Influence of different organic amendments on the leaching and dissipation of terbuthylazine in a column and a field study. Journal of Environmental Quality 38, 782-791. Dutta, M., Sardar, D., Pal, R. and Kole, R.K. (2010). Effect of chlorpyrifos on microbial biomass and activities in tropical clay loam soil. Environmental Monitoring and Assessment 160, 385-391. EC (2005). Review report for the active substance chlorpyrifos. Finalised in the Standing Committee on the Food Chain and Animal Health at its meeting on 3 June 2005 in view of the inclusion of chlorpyrifos in Annex I of the Directive 91/414/EEC. SANCO/3059/99 - rev. 1.5. EEA (2013). Late lessons from early warnings: Science, precaution, innovation - summary. European Environment Agency 2013 - 44 pp. - 21 x 29.7 cm, ISBN 978-92-9213-349-8, doi:10.2800/70069. EEC (2000). Guidance Document No 9188/VI/97-ver. 8, Persistence in soil, Commission of the European Communities, Directorate General for Agriculture, GG VI B II-1, Brussels, Belgium, Issued on July 2000. EFSA (2008). Conclusion on the peer review of tebuconazole. EFSA Scientific Report 176, 1-109. EFSA (2010). Scientific opinion on the development of specific protection goal options for environmental risk assessment of pesticides, in particular in relation to the revision of the guidance documents on aquatic and terrestrial ecotoxicology (SANCO/3268/2001 and SANCO/10329/2002), EFSA Panel on Plant Protection Products and their Residues (PPR). EFSA Journal 2010, 8 (10): 1821. EFSA (2014). Conclusion on the peer review of the pesticide risk assessment of the active substance tebuconazole. EFSA Journal 12 (1), 3485. http://dx.doi.org/10.2903/j.efsa.2014.3485, 98. EFSA (2015a). Conclusion on the peer review of the pesticide risk assessment of the active substance glyphosate. EFSA Journal 2015; 13 (11): 4302 (107 pp.). EFSA (2015b). Conclusion on the peer review of the pesticide risk assessment of the active substance isoproturon. EFSA Journal 13 (8), 4206. doi:10.2903/j.efsa.2015.4206. EFSA (2016a). Recovery in environmental risk assessments at EFSA, EFSA Scientific Committee. EFSA Journal 2016, 14 (2): 4313. EFSA (2016b). Scientific opinion addressing the state oft he science on risk assessment of plant protection products for in-soil organisms. Draft scientific opinion for public consultation. Eisenhauer, N., Klier, M., Partsch, S., Sabais, A.C.W., Scherber, C., Weisser, W.W. and Scheu, S. (2009). No interactive effects of pesticides and plant diversity on soil microbial biomass and respiration. Applied Soil Ecology 42, 31-36.

Chapter 6 General discussion 301

EU-Regulation 1107/2009/EC (2009). Regulation (EC) no 1107/2009 of the European parliament and of the council of 21 October 2009 concerning the placing of plant protection products on the market and repealing council directives 79/117/EEC and 91/414/EEC. Official Journal of the European Union 309, 1.Faasen, R. (1995). Agricultural pesticide use – a threat to the European environment? Auropean Water Pollution Control 5 (2), 34-40. Fang, H., Yu, Y., Chu, X., Wang, X., Yang, X. and Yu, J. (2009). Degradation of chlorpyrifos in laboratory soil and its impact on soil microbial functional diversity. Journal of Environmental Sciences 21, 380-386. Fenner, K., Canonica, S., Wackett, L.P. and Elsner, M. (2013). Evaluating pesticide degradation in the environment: Blind spots and emerging opportunities. Science 341, 752-758. Ferreira, E.P.B., Dusi, A.N., Costa, J.R., Xavier, G.R. and Rumjanek, N.G. (2009). Assessing insecticide and fungicide effects on the culturable soil bacterial community by analyses of variance of their DGGE fingerprinting data. European Journal of Soil Biology 45, 466-472. FOCUS (2008). Pesticides in Air: Considerations for Exposure Assessment. Report of the FOCUS Working Group on Pesticides in Air, EC Document Reference SANCO/10553/2006 Rev 2 June 2008. 327 pp. Fontaine, S., Mariotti, A. and Abbadie, L. (2003). The priming effect of organic matter: a question of microbial competition? Soil Biology and Biochemistry 35 (6), 837-843. Gage, D.J. (2004). Infection and invasion of roots by symbiotic, nitrogen-fixing Rhizobia during nodulation of temperate legumes. Microbiology and Molecular Biology Reviews 68 (2), 280-300. Garrity, G.M., Boone, D.R. and Castenholz, R.W. (2001). Bergey’s Manual of Systematic Bacteriology. 2nd edition, vol. 1, Springer New York. Gilani, S.T.S., Ageen, M., Shah, H. and Raza, S. (2010). Chlorpyrifos degradation in soil and its effect on soil microorganisms. The Journal of Animal and Plant Sciences 20 (2), 99-102. Gu, T., Zhou, C., Sorensen, S. R., Zhang, J., He, J., Yu, P., Yan, X. and Li, S. (2013). The novel bacterial N-demethylase pdmAB is responsible for the initial step of N,N-dimethyl-substituted phenylurea herbicide degradation. Applied and Environmental Microbiology 79 (24), 7846-7856. Gupta, S., Gupta, R. and Sharma, S. (2013). Impact of chemical- and bio-pesticides on bacterial diversity in rhizosphere of Vigna radiata. Ecotoxicology 22, 1479-1489. Harden, T., Joergensen, R.G., Meyer, B. and Wolters, V. (1993). Soil microbial biomass estimated by fumigation-extraction and substrate-induced respiration in two pesticide-treated soils. Soil Biology and Biochemistry 25 (6), 679-683. Herrero-Hernández, E. Andrades, M.S., Alvarez-Martina, A., Pose-Juana, E., Rodriguez-Cruz, M.S. and Sánchez-Martín, M.J. (2013). Occurrence of pesticides and some of their degradation products in waters in a Spanish wine region. Journal of Hydrology 486, 234-245. Herrero-Hernández, E., Andrades, MS., Marín-Benito, J.M., Sánchez-Martín, M.J. and Rodríguez-Cruz, M.S. (2011). Field-scale dissipation of tebuconazole in a vineyard soil amended with spent mushroom substrate and its potential environmental impact. Ecotoxicology and Environmental Safety 74, 1480-1488. Hovanec, T.A., Taylor, L.T., Blakis, A. and Delong, E.F. (1998). Nitrospira-like bacteria associated with nitrite oxidation in freshwater aquaria. Applied and Environmental Microbiology 64 (1), 258-264. Hussain, S., Devers-Lamrani, M., El Azhari, N. and Martin-Laurent, F. (2011). Isolation and characterization of an isoproturon mineralizing Sphingomonas sp. strain SH from a French agricultural soil. Biodegradation 22, 637-650. Hussain, S., Devers-Lamrani, M., El Azhari, N. and Martin-Laurent, F. (2011). Isolation and characterization of an isoproturon mineralizing Sphingomonas sp. strain SH from a French agricultural soil. Biodegradation 22, 637-650. Hussain, S., Sorensen, S. R., Devers-Lamrani, M., El-Sebai, T. and Martin-Laurent, F. (2009). Characterization of an isoproturon mineralizing bacterial culture enriched from a French agricultural soil. Chemosphere 77, 1052-1059. IARC (2015). IARC Monographs on the evaluation of carcinogenic risks to humans. Some organophosphate insecticides and herbicides: Diazinon, Glyphosate, Malathion, Parathion, and Tetrachlorvinphos. Glyphosate. IARC monograph 112. ISO 11063 (2012). Soil quality - Method to directly extract DNA from soil samples. ISO 17601 (2016). Soil quality - Estimation of abundance of selected microbial gene sequences by quantitative PCR from DNA directly extracted from soil. Jain, M., Flynn, D.F.B., Prager, C.M., Hart, G.M., DeVan, C.M., Ahrestani, F.S., Palmer, M.I., Bunker, D.E., Knops, J.M.H., Jouseau, C.F. and Naeem, S. (2014). The importance of rare species: a trait-based assessment of rare species contributions to functional diversity and possible ecosystem function in tall-grass prairies. Ecology and Evolution 4 (1), 104-112. Jastrzebska, E. (2011). The effect of chlorpyrifos and teflubenzuron on the enzymatic activity of soil. Polish Journal of Environmental Studies 20 (4), 903-910. Jin H. and Webster G.R.B. (1997). Dissipation of chlorpyrifos, oxon and 3,5,6-tricholor-2-pyridinol in litter and elm forest soil. Journal of Environmental Science and Health Part B 32, 879-900. Johannesen, H., Sørensen, S.R. and Aamand, J. (2003). Mineralization of soil-aged isoproturon and isoproturon metabolites by Sphingomonas sp. strain SRS2. Journal of Environmental Quality 32, 1250-1257. John, E. M. and Shaike, J.M., (2015). Chlorpyrifos: pollution and remediation. Environmental Chemistry Letters 13, 269-291. Keesstra, S.D., Geissen, V., Mosse, K., Piiranen, S., Scudiero, E., Leistra, M. and van Schaik, L. (2012). Soil as a filter for groundwater quality. Current Opinion on Environmental Sustainability 4 (5), 507-516. Kimura, H., Izawa, M. and Sumino, Y. (1996). Molecular analysis of the gene cluster involved in cephalosporin biosynthesis from Lysobacter lactamgenus YK90. Applied Microbiology and Biotechnology 44, 589-596. Kolpin, D., Thurman, E.M., Lee, E.A., Meyer, M.T., Furlong, E.T. and Glassmeyer, S.T. (2006). Urban contributions of glyphosate and its degradate AMPA to streams in the United States. Science of the Total Environment 354 (2-3), 191-197. Kuriyal, S.K. and Pandey, B.N. (1994). Studies of the effects of isoproturon on soil microflora. Journal of the Indian Botanical Society 73, 47-49. Laabs, V., Amelung, W., Pinto, A., Altstaedt, A. and Zech, W. (2000). Leaching and degradation of corn and soybean pesticides in anoxisol of the Brazilian Cerrados. Chemosphere 41, 1441-1449. Lewin, R.A. (1997). Saprospira grandis: A flexibacterium that can catch bacterial prey by ‘ixotrophy’. Microbial Ecology 34 (3), 232- 236. Looser, R., Froescheis, O., Cailliet, G.M., Jarman, W.M. and Ballschmiter, K. (2000). The deep-sea as a final global sink of semivolatile persistent organic pollutants? Part II: Organochlorine pesticides in surface and deep-sea dwelling fish of the North and South Atlantic and the Monterey Bay Canyon (California). Chemosphere, 40 (6), 661-70. Lynch, M.D.J. and Neufeld, J.D. (2015). Ecology and exploration of the rare biosphere. Nature Reviews Microbiology 13, 217-229.

Chapter 6 General discussion 302

Ma, W.K., Bedard-Haughn, A., Siciliano, S.D. and Farrell, R.E. (2008). Relationship between nitrifier and denitrifier community composition and abundance in predicting nitrous oxide emissions from ephemeral wetland soils. Soil Biology and Biochemistry 40 (5), 1114-1123. Martin-Laurent, F., Kandeler, E., Petric, I., Djuric, S. and Karpouzas, D. (2013). ECOFUN-MICROBIODIV: an FP7 European project for developing and evaluating innovative tools for assessing the impact of pesticides on soil functional microbial diversity: towards new pesticide registration regulation? Environmental Science and Pollution research 2, 1203-1205. McCall, P.J., Swann, R.L., Laskowski, D.A., Unger, S.M., Vrona, S.A. and Dishburger, H.J. (1980). Estimation of chemical mobility in soil from liquid chromatographic retention times. Bulletin of Environmental Contamination and Toxicology 24, 190-195. Mudd, P.J., Hance, R.J. and Wright, S.J.L. (1983). The persistence and metabolism of isoproturon in soil. Weed Research 23, 239-246. Multigner, L., Ndong, J.R., Giusti, A., Romana, M., Delacroix-Maillard, H., Cordier, S., Jegou, B., Thome, J.P. and Blanchet, P. (2010). Chlordecone exposure and risk of prostate cancer. Journal of Clinical Oncology 28 (21), 3457-3462. Munoz-Leoz, B., Ruiz-Romera, E., Antigëdad, I. and Garbisu, C. (2011). Tebuconazole application decreases soil microbial biomass and activity. Soil Biology and Biochemistry 43, 2146-2183. OECD-Guideline 106 (2000). OECD Guidelines for the testing of chemicals. Adsorption – desorption using a batch equilibrium method. Doi 10.1787/9789264069602-en. OECD-Guideline 301 (1992). OECD-Guideline for testing of chemicals. Adopted by the Council on 17th July 1992. Ready Biodegradability. OECD-Guideline ENV/JM/MONO (2007)17 (2007). Guidance document on pesticide residue analytical methods. Series on pesticides no 39. Series on testing and assessment no 72, p. 34. Papadopoulou, E.S., Lagos, S., Spentza, F., Vidiadakis, E., Karas, P.A., Klitsinaris, T. and Karpouzas, D.G. (2016). The dissipation of fipronil, chlorpyrifos, fosthiazate and ethoprophos in soils from potato monoculture areas: first evidence for the enhanced biodegradation of fosthiazate. Pest Management Science 72, 1040-1050. Perrin-Ganier, C., Schiavon, F., Morel, J.-L. and Schiavon, M. (2001). Effect of sludge-amendment or nutrient addition on the biodegradation of the herbicide isoproturon in soil. Chemosphere 44, 887-892. Pesce, S., Margoum, C., Rouard, N., Foulquier, A. and Martin-Laurent, F. (2013). Freshwater sediment pesticide biodegradation potential as an ecological indicator of microbial recovery following a decrease in chronic pesticide exposure: A case study with the herbicide diuron. Ecological Indicators 29, 18-25. Philippot, L. and Germon, J.C. (2005). Contribution of bacteria to initial input and cycling of nitrogen in soils. Microorganisms in soils: Roles in genesis and functions 3, 159-176. Philippot, L., Ritz, K., Pandard, P., Hallin, S. and Martin-Laurent, F. (2012). Standardisation of methods in soil microbiology: progress and challenges. FEMS Microbiology Ecology 82, 1-10. Piutti, S., Hallet, S., Rousseaux, S., Philippot, L., Soulas, G. and Martin-Laurent, F. (2002). Accelerated mineralisation of atrazine in maize rhizosphere soil. Biology and Fertility of Soils 36, 434-441. Piutti, S., Semon, E., Landry, D., Hartmann, A., Dousset, S., Lichtfouse, E., Topp, E., Soulas, G. and Martin-Laurent, F. (2003). Isolation and characterisation of Nocardioides sp SP12, an atrazine degrading bacterial strain possessing the gene trzN from bulk- and maize rhizosphere soil. FEMS Microbiology Letters 221, 111-117. Pozo, C., Martinez-Toledo, V., Salmeron, V., Rodelas, B. and Gonzalez-Lopez, J. (1995). Effect of chlorpyrifos on soil microbial activity. Environmental Toxicology and Chemistry 14 (2), 187-192. Prosser, J.I. (2002). Molecular and functional diversity in soil micro-organisms. Plant and Soil 244, 9-17. Racke, K.D. (1993). Environmental fate of chlorpyrifos. Reviews of Environmental Contamination and Toxicology 131, 1-150. Racke, K.D., Laskowski, D.A. and Schultz, M.R. (1990). Resistance of chlorpyrifos to enhanced biodegradation in soil. Journal of Agricultural and Food Chemistry 38, 1430-1436. Racke, K.D., Steele, K.P., Yoder, R.N., Dick, W.A. and Avidov, E. (1996). Factors affecting the hydrolytic degradation of chlorpyrifos in soil. Journal of Agricultural and Food Chemistry 44 (6), 1582-1592. Relyea, R.A. (2009). A cocktail of contaminants: how mictures of pesticides at low concentrations affect aquatic communities. Oecologia 159 (2), 363-376. Rieke, S., Koehn, S., Hirsch-Ernst, K., Pfeil, R., Kneuer, C. and Marx-Stoelting, P. (2014). Combination effects of (tri)azole fungicides on hormone production and xenobiotic metabolism in a human placental cell line. International Journal of Environmental Research and Public Health 11, 9660-9679. Rodríguez-Cruz, M.S., Jones, J.E. and Bending, G.D. (2006). Field-scale study of the variability in pesticide biodegradation with soil depth and its relationship with soil characteristics. Soil Biology and Biochemistry 38, 2910-2918. Sagarkar, S., Bhardwaj, P., Storck, V., Devers-Lamrani, M., Martin-Laurent, F. and Kapley, A. (2015). s-triazine degrading bacterial isolate Arthrobacter sp. AK-YN10, a candidate for bioaugmentation of atrazine contaminated soil. Applied Microbiology and Biotechnology 100 (2), 903-913. Sancez-Gonzalez, S., Pose-Juan, E., Herrero-Hernandez, E., Alvarez-Martin, A., Sancez-Martin, M.J. and Rodriguez-Cruz, S. (2013). Pesticide residues in groundwaters and soils of agricultural areas in the Águeda River Basin from Spain and Portugal. International Journal or Environmental and Analytical Chemistry 93, 1585-1601. Schuster, E. and Schröder, D. (1990). Side-effects of sequentially-applied pesticides on non-target soil microorganisms: Field experiments. Soil Biology and Biochemistry 22 (3), 367-373. Servien, R., Mamy, L., Li, Z., Rossard, V., Latrille, E., Bessac, F., Patureau, D. and Benoit, P. (2014). TyPol – A new methodology for organic compounds clustering based on their molecular characteristics and environmental behavior. Chemosphere 111, 613-622. Shalini, K., Kumar, N., Drabu, S. and Kumar Sharma, P. (2011). Advances in synthetic approach to and antifungal activity of triazoles. Beilstein Journal of Organic Chemistry 7, 668-677. Singh, B.K., Walker, A. and Wright, D. (2002). Degradation of chlorpyrifos, fenamiphos, and chlorothalonil alone and in combination and their effects on soil microbial activity. Environmental Toxicology and Chemistry 21 (12), 2600-2605. Singh, B.K., Walker, A., Alun, J., Morgan, W. and Wright, D.J. (2003). Effects of soil pH on the biodegradation of chlorpyrifos and isolation of a chlorpyrifos-degrading bacterium. Applied and Environmental Microbiology 69, 5198-5206. Skark, C. and Zullei-Seibert, N. (1995). The occurrence of pesticides in groundwater - Results of case studies. International Journal for Environmental and Analytical Chemistry 58, 387-396. Skark, C., Zullei-Seibert, N., Willme, U., Gatzemann, U. and Schlett, C. (2004). Contribution of non-agricultural pesticides to pesticide load in surface water. Pest Management Science 60, 525-530. Slattery, J.F., Coventry, D.R. and Slattery, W.J. (2001). Rhizobial ecology as affected by the soil environment. Australian Journal of Experimental Agriculture 41 (3), 289-298.

Chapter 6 General discussion 303

Smalling, K.L., Fellers, G.M., Kleeman, P.M. and Kuivila, K.M. (2013). Accumulation of pesticides in pacific chorus frogs (Pseudacris regilla) from California's Sierra Nevada Mountains, USA. Environmental Toxicology and Chemistry, 32 (9), 2026-2034. Sorensen, S. R, Ronen, Z. and Aamand, J. (2001). Isolation from agricultural soil and characterization of a Sphingomonas sp. able to mineralize the phenylurea herbicide isoproturon. Applied and Environmental Microbiology, 5403-5409. Srinivasulu, M. and Rangaswamy, V. (2013). Influence of insecticides alone and in combination with fungicides on enzyme activities in soils. International Journal of Environmental Science and Technology 10, 341-350. Strickland, T.C., Potter, T.L. and Joo, H. (2004). Tebuconazole dissipation and metabolism in Tifton loamy sand during laboratory incubation. Pest Management Science 60, 703-709. Tag-El-Din, A. (1982). Herbicides and respiration flush of soil biomass. Medical Faculty Landbouww University Gent 47, 191-199. Topp, E., Martin-Laurent, F., Hartmann, A. and Soulas, G. (2004). Bioremediation of atrazine-contaminated soil. Pesticide Decontamination and Detoxification 11, 141-154. Udikovic-Kolic, N., Scott, C. and Martin-Laurent, F. (2012). Evaluation of atrazine-degrading capabilities in the environment. Applied Microbiology and Biotechnology 95 (5), 1175-1189. Van-Camp, L., Bujarrabal, B., Gentile, A.-R., Jones, R.J.A., Montanarella, L., Olazabal, C. and Selvaradjou, S.-K. (2004). Reports of the technical working groups established under the thematic strategy for soil protection. EUR 21319 EN/4, 872 pp. Office for Official Publications of the European Communities, Luxembourg. Vasilyeva, N.V., Tsfasman, I.M., Suzina, N., Stepnaya, O.A. and Kulaev, I. (2008). Secretion of bacteriolytic endopeptidase L5 of Lysobacter sp. XL1 into the medium by means of outer membrane vesicles. The FEBS Journal 275 (15), 3827-3835. Vischetti, C., Monaci. E., Cardinali, A., Casucci, C. and Perucci, P. (2008). The effect of initial concentration, co-application and repeated applications on pesticide degradation in a biobed mixture. Chemosphere 72, 1739-1743. Von Tigerstrom, R.G. (1980). Extracellular nucleases of Lysobacter enzymogenes: production of the enzymes and purification and characterization of an endonuclease. Canadian Journal of Microbiology 26 (9), 1029-1037. Walker, A., Jurado-Exposito, M., Bending, G.D. and Smith, V.J.R. (2001). Spatial variability in the degradation rate of isoproturon in soil. Environmental Pollution 111, 407-415. Wang, C., Wang, F., Zhang, Q. and Liang, W. (2016). Individual and combined effects of tebuconazole and carbendazim on soil microbial activity. European Journal of Soil Biology 72, 6-13. Wang, F., Yao, J., Chen, H., Chen, K., Trebse, P. and Zaray, G. (2010). Comparative toxicity of chlorpyrifos and its oxon derivates to soil microbial activity by combined methods. Chemosphere 78, 319-326. Wang, X., Wang, X., Zhang, H., Wu, C., Wang, X., Xu, H., Wang, X. and Li, Z. (2012). Enantioselective degradation of tebuconazole in cabbage, cucumber and soils. Chirality 24, 104-111. Wang, Y., Wang, C., Gao, J., Liu, C., Cui, L. and Li, A. (2015). Dissipation, residues, and safety evaluation of trifloxystrobin and tebuconazole on ginseng and soil. Environmental Monitoring and Assessment 187, 344-356. Weaver, M. A., Krutz, L. J., Zablotowicz, R. M., and Reddy, K. N. (2007). Effects of glyphosate on soil microbial communities and its Mississippi soil. Pest Management Science 63, 388-393.

Chapter 6 General discussion 304

Summary

English summary

Pesticides are mainly applied to cultivable soils in agriculture to protect crops from various pests. Only a tiny part of sprayed pesticides ends up in their target organisms, while the rest persists on plants or in soil from where it can disperse to surrounding environments and harm non-target organisms. To minimize environmental damage caused by pesticides, their environmental fate and potential impact on non-target organisms is evaluated in an environmental risk assessment procedure by EU-authorities before their authorization and introduction to the market. The pesticide approval process is criticized for compromising a non- precautionary principle, as numerous formerly used pesticides are now banned due to the late emergence of environmental issues after years of their usage. Especially the formation of pesticide transformation products and the ecotoxicological impact of pesticide residues on soil microorganisms supporting numerous ecosystem functions are not sufficiently evaluated during the environmental risk assessment process. Within this context, my PhD aimed to support pesticide environmental risk assessment by the application and development of innovative methods of analytical chemistry and molecular biology to study the environmental dissipation (degradation / sorption / transformation) of pesticides in soil and their ecotoxicological impact on soil microorganisms. The three pesticides chlorpyrifos (CHL), isoproturon (IPU) or tebuconazole (TCZ) were chosen to establish lower tier and higher tier pesticide exposure scenarios in lab-to-field experimental designs. A research highlight of my PhD is the development of a combined approach of suspect screening and molecular typology to detect and classify new TCZ transformation products in soil and to estimate their environmental parameters. Moreover, findings of my PhD partly confirmed risk assessment studies, as the lower tier pesticide exposure scenario suggested a low but significant impact of CHL and TCZ (and no impact of IPU) on the soil bacterial composition, estimated by next-generation sequencing of 16S rDNA amplicons. However, in the higher tier pesticide exposure scenario, CHL and TCZ were found to induce significant changes in the microbial activity (evaluated via pesticide mineralization as a proxy for soil buffering functions) and the bacterial diversity (evaluated via next-generation sequencing of 16S rDNA amplicons). The results of my PhD have the potential to enrich future post-authorization risk assessments of CHL (due in 2020), IPU (due in 2017) and TCZ (due in 2023) with additional findings. In addition, the research of my PhD opens a new door for the enhancement of pesticide environmental risk assessment, especially concerning the methods used to assess the ecotoxicological impact of pesticides on soil microorganisms.

Summary 307

French summary (Résumé français)

Les pesticides sont principalement appliqués sur les sols cultivés dans l'agriculture pour protéger les cultures. Seule une infime partie des pesticides pulvérisés atteint les organismes cibles, tandis que la plus grande partie reste sur couvert végétal ou dans le sol d'où il peut se disperser dans les milieux environnants et nuire aux organismes non-cibles. Pour réduire au minimum les dommages environnementaux causés par les pesticides, leur devenir dans l'environnement et l'impact potentiel sur les organismes non-cibles sont évalués dans une procédure d'évaluation des risques environnementaux par des autorités de l’UE avant leur autorisation et l'introduction sur le marché. Le processus d'approbation des pesticides est critiqué car il remet en cause le principe de précaution, étant donné que de nombreux pesticides autorisés auparavant sont maintenant interdits en raison de l'émergence tardive des problèmes environnementaux après des années d’utilisation. En particulier, la formation de produits de transformation des pesticides et l'impact écotoxicologique de résidus des pesticides sur les microorganismes du sol supportant de nombreuses fonctions de l'écosystème ne sont pas suffisamment pris en compte au cours du processus d'évaluation des risques environnementaux des pesticides. Dans ce contexte, ma thèse visait à évaluer les risques environnementaux des pesticides par l'application et le développement de méthodes innovantes de la chimie analytique et de la biologie moléculaire pour étudier la dissipation (dégradation / sorption / transformation) des pesticides dans le sol et l’impact écotoxicologique qu’ils causent sur les microorganismes du sol. Les trois pesticides chlorpyrifos (CHL), isoproturon (IPU) ou tebuconazole (TCZ) ont été choisis pour les tester à différents niveaux d’exposition dans des dispositifs expérimentaux allant du laboratoire-au-champ. Un résultat marquant de ma thèse est le développement d'une approche combinant le criblage ciblé et le typage moléculaire pour détecter et classifier des nouveaux produits de transformation du TCZ dans le sol et d'estimer leurs paramètres environnementaux. En outre, mes expérimentations ont partiellement confirmé les études d'évaluation du risque, montrant pour le scénario d’exposition le plus bas un impact faible mais significatif de l’exposition à CHL et TCZ sur la composition bactérienne du sol, estimée par le séquençage de nouvelle génération d’amplicons du gène de l’ARNr16S (et aucun impact du IPU). Toutefois, pour le scénario d'exposition le plus élevé, l’exposition à CHL et TCZ a induit des changements importants de l'activité microbienne (évaluée par la minéralisation des pesticides comme proxy de la fonction de filtration du sol) et la diversité bactérienne (évaluée par le séquençage de nouvelle génération d’amplicons du gène de l’ARNr 16S). Les résultats acquis au cours de ma thèse pourront enrichir l'évaluation des risques à posteriori qui sera conduite dans le cadre de la révision de l’autorisation du CHL (attendue en 2020), de l’IPU (attendue en 2017) et du TCZ (attendue 2023). En outre, mes recherches ouvrent de nouvelles perspectives pour l’évaluation de l'impact écotoxicologique des pesticides sur les microorganismes qui soutiennent des fonctions écosystémiques du sol.

Summary 308

Acknowledgements

Everything began in a Master’s lecture of Environmental Microbiology by Professor Hans-Curt Flemming at the University of Duisburg-Essen (Germany) in 2012, where he taught us about the various processes of bioremediation and the approach to use microorganisms to counteract environmental pollution. As I was interested in environmental pollution topics since I was little, the ideas of this lecture fascinated me and I decided to do an internship in a research group dealing with the transformation, biodegradation and environmental impact of pollutants. After reading some of his publications, I sent an internship application to my future supervisor Dr. Fabrice Martin- Laurent without imagining the amazing adventure I was about to face. After a great internship in his research group in Dijon (France) in 2012, I decided to come back and started my PhD studies in his laboratory in 2013.

So the first person I want to thank are you Fabrice, for giving me the great opportunity of doing my PhD under your supervision, for being willing to give advice at any time, for your knowledge, ideas, guidance and good mood, but also for giving me space to develop, to gain autonomy and to realize my ideas. You are not only a great supervisor but also a motivating and reassuring friend.

The second wave of thanks goes to my supervisors of my six months stay within the research teams at Aeiforia srl. and at the Catholic University of the Sacred Heart in Piacenza (Italy) – Dr. Federico Ferrari, Dr. Luigi Lucini and Professor Marco Trevisan. Federico, many thanks for your heartwarming contribution to my very pleasant stay in Piacenza, patiently translating Italian during my first weeks, advising me which wines to buy in the supermarket (when I realized that you knew ALL of them), supplying warm boots in my mini shoe size (to keep my feet warm while sampling frozen field soil), and helping me to realize my idea of discovering new pesticide transformation products by introducing me to Luigi. Luigi, 1000 thanks for all your time, knowledge and motivation! Although you were not implied in the project from the beginning, you passionately jumped into the idea of discovering new pesticide transformation products and spent hours to teach me how to use the UHPLC-ESI-QTOF-MS/MS (already the name…). Fortunately, green ice tea and chocolate were a good currency to pay for all your time and advice.

Acknowledgements 311

And then Marco, I want to especially thank you for your kindness, your belief in my research, and for the amazing adventures full of happiness that we had at the IUPAC congress in San Francisco and especially during our road trip through California!

The next group of people I would like to thank are all of you who I met at work or in my free time during my three years of PhD. To the members of the research group in Dijon: Marion (my dear co-supervisor), with your friendly, open-minded, efficient and awake attitude, you have been a source of inspiration, laughter and good advice for me! Nadine, thank you for your friendly state of mind, and for your help with my experiments in the radioactivity sector. Aymé, thanks for the time that you spent teaching me bioinformatics to enable the analysis of my DNA sequencing data. Luiz, Ilonka and Sara, you have become true friends of mine and I will badly miss you in my daily life when our ways will separate. David, Laurent, Jérémie, Marie-Christine, Kadiya and Chloé, thanks for all the good mood, ideas, technical help, advices and for your friendship. To my beloved flatmates, housemates or building mates in Dijon, Piacenza and Lyon and to all the other friendly souls that I am grateful to have come across: Aurélie, Loreleï, Yuko, Ambre, Benj, Camille, Yann, Gabriella, Emy, Tiffaine, Gino, Antonio, Gennaro, Lucio, Najoi, Evangelia and Sofia, you represented to me a real equilibration and source of fun and sympathy!

I would also like to thank the committee members of my PhD thesis, Dr. Stéphane Pesce and Dr. Isabelle Batisson, for their friendly advices throughout the three years of my thesis, and the reviewers of this thesis, Dr. Edward Topp and Dr. Kathrin Fenner, for their time and helpful contribution to the finalization of my dissertation.

Last but certainly not least, I want to thank the people in my life that contribute to my personal well-being, that support and inspire me, and that are true friends: Insa, Steffi, Dini, Adri, Wölli, Mark, Susi, Yannick, Simon, Anna, mum, dad and sister!

Acknowledgements 312

List of abbreviations

2,4-D 2,4-dichlorophenoxyacetic

4-IA 4-isopropyl-aniline

ANOSIM analysis of similarity

ANOVA analysis of variance

ANSES National Agency of Sanitary Security of Alimentation, Environment and Work

ARDRA amplified ribosomal DNA restriction analysis atz atrazine-degrading gene

BCF bioconcentration factor

BfR Federal Institute for Risk Assessment

BUND Federation for Environmental Studies and Nature Conservation

BVL Federal Office of Consumer Protection and Food Safety

CFS candidate for substitution

CHL chlorpyrifos

CHL-oxon chlorpyrifos-oxon

DAR draft assessment report

DD-IPU di-desmethyl-isoproturon

DFOP double first order in parallel model

DGGE denaturing gradient gel electrophoresis

DNA deoxyribonucleic acid

DT50 time needed to deplete 50 % of compound

DT90 time needed to deplete 90 % of compound

EC European Commission

ED E2S Doctoral School of Environment and Health

EEA European Environment Agency

List of abbreviations 315

ERA environmental risk assessment

ESI electron spray ionization

EU European Union

FAO Food and Agriculture Organization of the United Nations

FOMC first order multi-compartment model

FOR functional operating range

FT-MS Fourier transform mass spectrometry

HPLC-PDA High Pressure Liquid Chromatography with photodiode array detector

HPPD 4-hydroxyphenylpyruvate dioxygenase

IARC International Agency for Research on Cancer

INIA National Institute of Investigation, Agrarian Technology and Alimentation

IPU isoproturon

ISO International Standard Organization

JKI Julius Kühn Institute

KOC sorption coefficient

MCPA 2-methyl-4-chlorophenoxyacetic acid

MD-IPU mono-desmethyl-isoproturon mdp chlorpyrifos-degrading gene

MESR French Ministry of Education and Research

MRL maximum residue level

MS mass spectrometry

MSC molecular structure correlation

NMR nuclear magnetic resonance

List of abbreviations 316

NOR normal operating range

NROSO National Register of Sprayer Operators

OECD Organisation for Economic Co-operation and Development opd chlorpyrifos-degrading gene

OpenTea Open Training and Education Association

PAN Pesticide Action Network

PCoA principal coordinate analysis

PCR polymerase chain reaction pdm isoproturon-degrading gene

PEC predicted environmental concentration

PPP plant protection product pSH plasmid of Sphingomonas sp. strain SH puh isoproturon-degrading gene

Pvap vapor pressure qPCR quantitative polymerase chain reaction

QTOF-MS quadrupole time-of-flight mass spectrometry

RISA ribosomal intergenic spacer analysis

RMS Rapporteur Member State

RNA ribonucleic acid

SCOPAFF Standing Committee on Plants, Animals, Food and Feed

SFO single first order kinetic model

SIP stable isotope probing

TCP 3,5,6-trichloro-2-pyridinol

TCZ tebuconazole

List of abbreviations 317

tfd phenoxy acid-degrading gene

TGGE temperate gradient gel electrophoresis

TOPPS Training the Operators to prevent Pollution from Point Sources

TP transformation product tri atrazine-degrading gene trz atrazine-degrading gene

TyPol typology of pollutants

U.S. EPA United States Environmental Protection Agency

UBA Federal Environment Agency

UHPLC-ESI-QTOF-MS ultra high performance liquid chromatography electron spray ionization quadrupole time-of-flight mass spectrometry

UN United Nations

WHC water holding capacity

WHO World Health Organization

List of abbreviations 318

List of authors

Author

Veronika Storck French Institute of Agronomic Research (INRA), Dijon, France [email protected]

Co-authors:

Baguelin, Céline Enoveo srl., Lyon, France [email protected]

Béguet, Jérémie French Institute of Agronomic Research (INRA), Dijon, France [email protected]

Benoit, Pierre French Institute of Agronomic Research (INRA), Thiverval-Grignon, France [email protected]

Devers-Lamrani, Marion French Institute of Agronomic Research (INRA), Dijon, France [email protected]

Ferrari, Federico Aeiforia srl., Fidenza, Italy [email protected]

Hussain, Sabir University College Dublin, Dublin, Ireland [email protected]

Karas, Panagiotis A. University of Thessaly, Larissa, Greece [email protected]

Karpouzas, Dimitrios G. University of Thessaly, Larissa, Greece [email protected]

Lucini, Luigi Catholic University of the Sacred Heart, Piacenza, Italy [email protected]

Malandain, Cedric Enoveo srl., Lyon, France [email protected]

List of authors 321

Mamy, Laure French Institute of Agronomic Research (INRA), Thiverval-Grignon, France [email protected]

Martin-Laurent, Fabrice French Institute of Agronomic Research (INRA), Dijon, France [email protected]

Nikolaki, Sofia University of Patras, Agrinio, Greece [email protected]

Papadopoulou, Evangelia S. University of Thessaly, Larissa, Greece [email protected]

Perruchon, Chiara University of Thessaly, Larissa, Greece [email protected]

Pertile, Giorgia Catholic University of the Sacred Heart, Piacenza, Italy [email protected]

Rouard, Nadine French Institute of Agronomic Research (INRA), Dijon, France [email protected]

Servien, Remi French Institute of Agronomic Research (INRA), Toulouse, France [email protected]

Sibourg, Olivier Enoveo srl., Lyon, France [email protected]

Spor, Aymé French Institute of Agronomic Research (INRA), Dijon, France [email protected]

Trevisan, Marco Catholic University of the Sacred Heart, Piacenza, Italy [email protected]

Tsiamis, George University of Patras, Agrinio, Greece [email protected]

List of authors 322

List of publications

Paper publications

Storck, V., Karpouzas, D.G. and Martin-Laurent, F. (2016). Towards a better pesticide policy for the European Union. Science of the Total Environment, http://dx.doi.org/10.1016/j.scitotenv.2016.09.167.

Papadopoulou, E.S., Karas, P.A., Nikolaki, S., Storck, V., Ferrari, F., Trevisan, M., Tsiamis, G., Martin-Laurent, F. and Karpouzas, D.G. (2016). Dissipation and adsorption of isoproturon, tebuconazole, chlorpyrifos and their main transformation products under laboratory and field conditions. Science of the Total Environment 569-570, 86-96.

Storck, V., Lucini, L., Mamy, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., Karas, P.A., Servien, R., Karpouzas, D.G., Trevisan, M., Benoit, P., Martin-Laurent, F. (2016). Identification and characterization of tebuconazole transformation products in soil by combining suspect screening and molecular typology. Environmental Pollution 208, 537-545.

Manuscripts in preparation for publication

Storck, V., Nikolaki, S., Perruchon, C., Pertile, G., Baguelin, C., Karas, P.A., Spor, A., Sibourg, O., Malandain, C., Trevisan, M., Ferrari, F., Karpouzas, D.G., Tsiamis, G. and Martin-Laurent, F. (2016). Lab-to-field evaluation of the ecotoxicological impact of three pesticides on the soil bacterial diversity by Illumina nexy-generation sequencing. Manuscript in preparation for publication.

Storck, V., Rouard, N., Béguet, J., Spor, A., Devers, M. and Martin-Laurent, F. (2016). Impact of repeated treatments of the pesticides chlorpyrifos or tebuconazole on soil microorganisms using alternative methods. Manuscript in preparation for publication.

Storck, V., Béguet, J., Rouard, N., Hussain, S., Devers, M. and Martin-Laurent, F. (2016). Possibilities and limitations of using genes as bioindicators for potential pesticide

List of publications 325

degradation in soil - the case of pdm-encoded isoproturon degradation. Manuscript in preparation for publication.

Presentations

Storck, V., Lucini, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., Karas, P.A., Karpouzas, D.G., Trevisan, M. and Martin-Laurent, F. (2016). Evaluation of the environmental fate and ecotoxicological impact of the pesticide chlorpyrifos in soil for improvement of its risk assessment. 7th SETAC World Congress/ 37th SETAC North America Annual Meeting, Orlando, USA. (Oral.)

Storck, V., Lucini, L., Mamy, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., Karas, P.A., Servien, R., Karpouzas, D.G., Trevisan, M., Benoit, P. and Martin-Laurent, F. (2015). Evidence for the interest of suspect screening to identify known and unknown pesticide transformation products in soil. 15th Symposium in Pesticide Chemistry, Piacenza, Italy. (Poster.)

Papadopoulou, E.S., Karas, P., Nikolaki, S., Storck, V., Ferrari, F., Trevisan, M., Martin- Laurent, F. and Karpouzas, D.G. (2015). A tiered-based approach to study the dissipation and adsorption of isoproturon, tebuconazole and chlorpyrifos in soil. 15th Symposium in Pesticide Chemistry, Piacenza, Italy. (Poster.)

Pertile, G., Baguelin, C., Ferrarini, A., Fornasier, F., Karas, P.A., Papadopoulou, E.S., Nikolaki, S., Storck, V., Ferrari, F., Trevisan, M., Tsiamis, G., Sibourg, O., Malandain, C., Martin-Laurent, F. and Karpouzas, D.G. (2015). Assessment of the impact of isoproturon, chlorpyrifos and tebuconazole on soil microbial functions using a lab-to-field tiered approach. 15th Symposium in Pesticide Chemistry, Piacenza, Italy. (Poster.)

Storck, V., Lucini, L., Mamy, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., Karas, P.A., Servien, R., Karpouzas, D.G., Trevisan, M., Benoit, P. and Martin-Laurent, F. (2015). New approach to identify and categorize pesticide transformation products in soil combining

List of publications 326

suspect screening with in silico molecular typology. 5th International Conference on Environmental Pollution and Remediation, Barcelona, Spain. (Oral.)

Storck, V., Baguelin, C., Rouard, N., Beguet, J., Bru, D., Devers, M. and Martin-Laurent, F. (2014). Evaluation of the ecotoxicological impact of pesticides on the diversity and the abundance of soil microorganisms. 5th INRA Seminar of ecotoxicology, Biarritz, France. (Oral.)

Storck, V., Pertile, G., Papadopoulou, E.S., Rouard, N., Devers, M., Beguet, J., Ferrari, F., Trevisan, M., Karpouzas, D.G. and Martin-Laurent, F. (2014). Fate and transformation of the herbicide isoproturon in soil microcosms and its impact on soil microbial communities. 13th IUPAC International Congress of Pesticide Chemistry, San Francisco, USA. (Poster & Oral.)

Papadopoulou, E.S., Karas, P.A., Nikolaki, S., Storck, V., Ferrarini, A., Fornasier, F., Ferrari, F., Trevisan, M., Martin-Laurent, F., Tsiamis, G., Karpouzas, D.G. (2014). A lab-to-field experimental approach to study the dissipation, metabolism and soil microbial ecotoxicity of isoproturon, tebuconazole and chlorpyrifos. 13th IUPAC International Congress of Pesticide Chemistry, San Francisco, USA. (Poster.)

Pertile, G., Storck, V., Papadopoulou, E.S., Ferrari, F., Karpouzas, D.G. and Martin- Laurent, F. (2014). Microcosm assessment of the dissipation and soil microbial ecotoxicity of chlorpyrifos and tebuconazole using standardized advanced molecular tools. 24th SETAC Europe Annual Meeting, Basel, Switzerland. (Poster.)

List of publications 327

List of tables

Chapter 1 Table 1.1. Available ISO methods in soil microbiology (modified from 28 Martin-Laurent, 2013 and Philippot et al., 2012).

Table 1.2. List of bacterial and fungal strains and genes coding for 40 degradation enzymes that are involved in the biodegradation of CHL, IPU or TCZ (modified from Chishti et al., 2013; Hussain et al., 2015).

Chapter 3.1 Table 3.1.1. The dissipation parameters of IPU, TCZ, CHL and their main 99 transformation products in soil in the laboratory experiment. Dissipation kinetic parameters were calculated with the single first order (SFO) kinetic model or with the biphasic first order multi compartment (FMOC) model.

Table 3.1.2. The dissipation parameters of IPU, TCZ, CHL and their main 102 transformation products in soil in the field experiment. Dissipation kinetic parameters were calculated with the single first order (SFO) kinetic model or with the biphasic hockey stick (HS) model.

Table 3.1.3. The adsorption coefficients of IPU, TCZ, CHL, determined 104 either with the use of the active substance or the commercial formulation, and of their main transformation products MD-IPU, DD-IPU, 4-IA, and TCP.

Supplementary Table 3.1.1. The physicochemical characteristics of the 111 soil used in the study.

Supplementary Table 3.1.2. Dose levels (mg/kg soil dwt) of the model 112 pesticides used in the laboratory and in the field study. The expected pesticide concentrations in soil were calculated based on the assumption of a soil bulk density of 1.3 kg/L, diffusion of pesticide residues to a soil depth of 5 cm for IPU and TCZ, and to 10 cm for CHL (soil incorporated).

Supplementary Table 3.1.3. Soil dissipation kinetics of IPU, TCZ, and 112 CHL in the laboratory experiment.

Supplementary Table 3.1.4. Soil dissipation kinetics of IPU, TCZ, and 114 CHL in the field experiment.

Chapter 3.2 Table 3.2.1. TCZ TPs detected in this study by suspect screening analysis. 130 Reference to the degree of confidence on identification (a: detected by MS; b: selected by target tandem MS; c: proposed molecular structure by MSC). The name, structure, mass and reference if empirical of each TP are provided. The time points when a TP was detected and the cluster into which it was categorized are given. Blue boxes group TPs with the same mass.

Supplementary Table 3.2.1. TCZ TP library containing TCZ and 76 TPs 138 (47 empirical TPs and 29 theoretical TPs). TPs with the same mass are indicated by blue boxes.

Supplementary Table 3.2.2. TyPol clustering results for TCZ and its TPs: 148 Median values of molecular descriptors and median values of environmental parameters estimated from compounds with similar molecular descriptors and known environmental parameters.

Supplementary Table 3.2.3. TyPol clustering of the 116 pesticides and 149 19 TPs into six clusters (Pesticide-clusters 1 to 6) according to their molecular descriptors and KOC: median values of molecular descriptors

List of tables 331

and median values of KOC. The list of compounds in each cluster is indicated in Table 3.2.4. The comparison of these median molecular descriptor values to those of TCZ and its TPs (Supplementary Table 3.2.2) showed that cluster 1 can be assimilated to Pesticide-cluster 4, cluster 2 to Pesticide-cluster 5, and cluster 3 to Pesticide-cluster 1.

Supplementary Table 3.2.4. TyPol clustering of the 116 pesticides and 152 19 TPs into six clusters (pesticide-clusters 1 to 6): example of clustering according to their molecular descriptors and KOC values (see Supplementary Table 3.2.3 and Supplementary Figure 3.2.3). TPs are indicated in italics.

Chapter 4.1 Table 4.1.1. Microcosm study - Bacterial α-diversity indices at various 174 time points (0, 7, 42, 56 and 100 days after treatment) in soil exposed to each of the three studied pesticides (CHL, IPU or TCZ) applied at different doses (0x, 1x, 2x or 10x). Diversity indices of the bacterial community are 'observed species', 'PD whole tree', 'chao1' and 'simpson reciprocal'. For each diversity index, ANOVAs followed by the Tukey HSD test, (p < 0.05) were done by grouping (for each tested pesticide) the five sampling time points and the four pesticide doses. Significant differences are indicated by letters.

Table 4.1.2. Field study - Bacterial α-diversity indices at various time 175 points (0, 14, 35, 70 and 105 days after treatment) in soil exposed to each of the three studied pesticides (CHL, IPU or TCZ) applied at different doses (0x, 1x, 2x or 5x). Diversity indices of the bacterial community are 'observed species', 'PD whole tree', 'chao1' and 'simpson reciprocal'. For each diversity index, ANOVAs followed by the Tukey HSD test, (p < 0.05) were done by grouping (for each tested pesticide) the five sampling time points and the four pesticide doses. Significant differences are indicated by letters.

Supplementary Table 4.1.1. Microcosm and field study - ANOSIM of the 197 PCoA ordinations of OTU weighted unifrac distance matrices for untreated (control) and treated (CHL, IPU or TCZ at 1x, 2x or 10x (microcosms) or 1x, 2x or 5x (field) doses) soil samples at different time points (0, 7, 42, 56 and 100 days (microcosms) or 0, 14, 35, 70 and 105 days (field)). Significant differences between groups are indicated by p < 0.01.

14 14 Chapter 4.2 Table 4.2.1. Amount of CO2 evolved from the C-labelled pesticides 219 (14C-ring-labelled CHL or 14C-phenyl-ring-labelled TCZ) after 150 days in % of initially applied radioactivity (A150d), and time needed to mineralize 14 25 % (CHL) or 5% (TCZ) of the initially applied radioactivity to CO2 in days (t25% or t5%) in soil untreated (control) or repeatedly treated with pesticides (CHL or TCZ). Mineralization was monitored after 1, 2 and 3 months of incubation (i.e. one month after the 1st, 2nd and 3rd pesticide treatment). ANOVA on four replicates were performed followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

14 14 14 Table 4.2.2. Amount of CO2 evolved from C-labelled compounds ( C- 222 ring-labelled 2,4-D, 14C-labelled glyphosate or 14C-labelled sodium acetate) after 150 days in % of initially applied radioactivity (A150d), and time needed to mineralize 25 % (2,4-D and glyphosate) or 50 % (sodium 14 acetate) of the initially applied radioactivity to CO2 in days (t25% or t50%) in soil untreated (control) or repeatedly treated with pesticides (CHL or TCZ). Mineralization was monitored after 1, 2 and 3 months of incubation (i.e. one month after the 1st, 2nd and 3rd pesticide treatment). ANOVA on four replicates were performed followed by the Tukey HSD test

List of tables 332

(p < 0.05). Significant differences are indicated by letters.

Table 4.2.3. Diversity indices of the soil bacterial community repeatedly 225 exposed to CHL or TCZ (at 10x agronomical doses) or untreated (control). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVA of each diversity index were followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

Supplementary Table 4.2.1. Details of the sampled soil amounts of the 235 established mesocosms per treatment (CHL, TCZ or control) and time (1 month, 2 months, 3 months). TCZ-P abbreviates for 14C-phenyl-ring- labelled TCZ. TCZ-T abbreviates for 14C-triazole-ring-labelled TCZ.

Supplementary Table 4.2.2. ANOSIM of the PCoA ordinations of OTU 238 weighted unifrac distance matrices for soil repeatedly treated (once per month) with CHL or TCZ or untreated (control). Significant differences between groups are indicated by p < 0.1.

List of tables 333

List of figures

Thesis outline Thesis outline Figure. Graphic presentation of the thesis outline. 13

Chapter 1 Figure 1.1. Worldwide pesticide use (amount of pesticide in kg per ha 19 per year) between 2005 and 2009 (Enserink et al., 2013).

Figure 1.2. Main abiotic and biotic processes involved in pesticide 20 dissipation in the environment (modified from Barriuso et al., 1996).

Figure 1.3. Presentation of the importance of soil microbial diversity for 22 the maintenance of ecosystem functions under changing environmental conditions.

Figure 1.4. Presentation of the ecological recovery of an ecological metric 23 (e.g. population size) after exposure to a pollutant (e.g. pesticide residues) above the effect threshold (modified from EFSA, 2016a).

Figure 1.5. Microbial ecotoxicology, a multidisciplinary science 26 combining microbial toxicology, microbial ecology and chemistry/physics to understand and describe the complex interactions between pollutants, microorganisms and the ecosystem (modified from Ghiglione et al., 2016).

Figure 1.6. Conceptual presentation of the ecotoxicological impact 27 caused by pesticides on the abundance, activity and diversity of the soil microbial community and possible consequences on supported ecosystem functions.

Figure 1.7. Illustration of the main research aim of the European Marie 32 Curie ‘Love-to-Hate’ project: Establishment and development of innovative methods to better assess the ecotoxicological impact of pesticides on soil microorganisms.

Figure 1.8. Molecular structures of CHL and its known environmental 36 TPs. References for each molecule are given in the reference section.

Figure 1.9. Molecular structures of IPU and its known environmental 37 TPs. References for each molecule are given in the reference section.

Figure 1.10. Molecular structures of TCZ and its known environmental 38 TPs. References for each molecule are given in the reference section.

Chapter 2 Figure 2.1. Graphical abstract of Chapter 2. 56

Figure 2.2. Schematic presentation of the stepwise process of pesticide 64 authorization in the EU.

Figure 2.3. Presentation of a typical pesticide life cycle. 69

Chapter 3.1 Figure 3.1.1. Graphical abstract of Chapter 3.1.1. 88

Figure 3.1.2. The laboratory dissipation of (a) IPU, (b) TCZ and (c) CHL in 98 soil samples treated with x1, x2, or x10 the recommended dose of each of these compounds. Each value is the mean of three replicates ± standard deviation.

List of figures 337

Figure 3.1.3. The formation and dissipation patterns of the 99 transformation products of IPU ((a) MD-IPU and (b) DD-IPU) and of CHL((c) TCP) in the soil samples treated with x1, x2, or x10 the recommended dose of the pesticide in the laboratory. Each value is the mean of three replicates ± the standard deviation.

Figure 3.1.4. The field dissipation of (a) IPU, (b) TCZ and (c) CHL in soil 101 samples treated with x1, x2, or x5 the recommended dose of each of these compounds. Each value is the mean of three replicates ± standard deviation.

Figure 3.1.5. The formation and dissipation patterns of the 102 transformation products of IPU ((a) MD-IPU and (b) DD-IPU) and of CHL((c) TCP) in the soil samples treated with x1, x2, or x5 the recommended dose of the pesticide in the field. Each value is the mean of three replicates ± the standard deviation.

Figure 3.1.6. Adsorption isotherms of (a) IPU, (b) TCZ and (c) CHL, 103 prepared with the use of the active substance or the commercial formulation, and (d) their main transformation products MD-IPU, DD- IPU, 4-IA and TCP.

Supplementary Figure 3.1.1. Location of the North Italian agricultural 111 field.

Supplementary Figure 3.1.2. Meteorological data recorded during the 113 field experiment: (a) Mean daily precipitation [mm], (b) mean daily air and soil temperature [°C], (c) mean daily solar radiation [W/m2], (d) mean daily air relative humidity [%] and (e) mean daily wind speed [m/s].

Chapter 3.2 Figure 3.2.1. Graphical abstract of Chapter 3.2. 120

Figure 3.2.2. Categorization of TCZ and its 34 TPs into three clusters 136 according to their molecular descriptors on the two main components of the PLS regression (PLS 1 and PLS 2), and their estimated environmental parameters. Clusters are represented by different color codes. Each TP has its respective number. The inserted graph is an enlargement of a part of cluster 1.

Supplementary Figure 3.2.1. Chromatogram examples of TPs with the 146 molecular formula C16H20ClN3O2. (a) MS chromatogram, total ion current (TIC), (b) MS chromatogram, extracted ion current (EIC), (c) tandem MS chromatogram.

Supplementary Figure 3.2.2. MSC of the constitutional isomers TP 13 147 and 15.

Supplementary Figure 3.2.3. TyPol clustering of the 116 pesticides and 151 19 TPs into six clusters on the two main components of the PLS (PLS1 and PLS2): Example of clustering according to their molecular descriptors and KOC values.

Chapter 4.1 Figure 4.1.1. Graphical abstract of Chapter 4.1. 164

Figure 4.1.2. Microcosm study - Relative abundance (%) of bacterial 176

List of figures 338

phyla identified in the untreated (control) and treated (CHL, IPU or TCZ at 10x dose) soil samples at different sampling time points (0, 7, 42, 56 and 100 days after treatment). Each color represents a phylum. 16S rDNA sequences that could not be assigned to a phylum are grouped as ‘unassigned’.

Figure 4.1.3. Field study - Relative abundance (%) of bacterial phyla 177 identified in the untreated (control) and treated (CHL, IPU or TCZ at 5x dose) soil samples at different sampling time points (0, 14, 35, 70 and 105 days after treatment). Each color represents a phylum. Bacterial phyla, whose abundance sum of all samples was < 0.05 % are grouped as ‘others’. 16S rDNA sequences that could not be assigned to a phylum are grouped as ‘unassigned’.

Supplementary Figure 4.1.1. Microcosm study - Bacterial diversity in 183 soil treated with CHL at various doses (0x, 1x, 2x, 10x) at various time points after treatment (0, 7, 42, 56 and 100 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Supplementary Figure 4.1.2. Microcosm study - Bacterial diversity in 184 soil treated with IPU at various doses (0x, 1x, 2x, 10x) at various time points after treatment (0, 7, 42, 56 and 100 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Supplementary Figure 4.1.3. Microcosm study - Bacterial diversity in 185 soil treated with TCZ at various doses (0x, 1x, 2x, 10x) at various time points after treatment (0, 7, 42, 56 and 100 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Supplementary Figure 4.1.4. Field study - Bacterial diversity in soil 186 treated with CHL at various doses (0x, 1x, 2x, 5x) at various time points after treatment (0, 14, 35, 70 and 105 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Supplementary Figure 4.1.5. Field study - Bacterial diversity in soil 187 treated with IPU at various doses (0x, 1x, 2x, 5x) at various time points after treatment (0, 14, 35, 70 and 105 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five time points and four doses per each diversity index. Significant differences are indicated by letters.

Supplementary Figure 4.1.6. Field study - Bacterial diversity in soil 188 treated with TCZ at various doses (0x, 1x, 2x, 5x) at various time points after treatment (0, 14, 35, 70 and 105 days). Diversity indices are ‘observed species’, ‘PD whole tree’, ‘chao1’ and ‘simpson reciprocal’. ANOVAs were followed by the Tukey HSD test (p < 0.05) grouping the five

List of figures 339

time points and four doses per each diversity index. Significant differences are indicated by letters.

Supplementary Figure 4.1.7. Microcosm study - PCoA ordinations of 189 OTU weighted unifrac distance matrices for untreated (control) and CHL- treated (1x, 2x or 10x doses) soil samples at different time points (0, 7, 42, 56 and 100 days). Replicates of the same treatment are represented in same colors.

Supplementary Figure 4.1.8. Microcosm study - PCoA ordinations of 190 OTU weighted unifrac distance matrices for untreated (control) and IPU- treated (1x, 2x or 10x doses) soil samples at different time points (0, 7, 42, 56 and 100 days). Replicates of the same treatment are represented in same colors.

Supplementary Figure 4.1.9. Microcosm study - PCoA ordinations of 191 OTU weighted unifrac distance matrices for untreated (control) and TCZ- treated (1x, 2x or 10x doses) soil samples at different time points (0, 7, 42, 56 and 100 days). Replicates of the same treatment are represented in same colors.

Supplementary Figure 4.1.10. Microcosm study - PCoA ordinations of 192 OTU weighted unifrac distance matrices for untreated (control) soil samples at different time points (0, 7, 42, 56, 100 days). Replicates of the same treatment are represented in same colors.

Supplementary Figure 4.1.11. Field study - PCoA ordinations of OTU 193 weighted unifrac distance matrices for untreated (control) and CHL- treated (1x, 2x or 5x doses) soil samples at different time points (0, 14, 35, 70 and 105 days). Replicates of the same treatment are represented in same colors.

Supplementary Figure 4.1.12. Field study - PCoA ordinations of OTU 194 weighted unifrac distance matrices for untreated (control) and IPU- treated (1x, 2x or 5x doses) soil samples at different time points (0, 14, 35, 70 and 105 days). Replicates of the same treatment are represented in same colors.

Supplementary Figure 4.1.13. Field study - PCoA ordinations of OTU 195 weighted unifrac distance matrices for untreated (control) and TCZ- treated (1x, 2x or 5x doses) soil samples at different time points (0, 14, 35, 70 and 105 days). Replicates of the same treatment are represented in same colors.

Supplementary Figure 4.1.14. Field study - PCoA ordinations of OTU 196 weighted unifrac distance matrices for untreated (control) soil samples at different time points (0, 14, 35, 70 and 105 days). Replicates of the same treatment are represented in same colors.

Supplementary Figure 4.1.15. Microcosm study - Abundance 197 [number/sample] of OTUs responsible for significant differences of ANOSIM, detected by ‘pamR’ in untreated (control) and treated (CHL at 1x, 2x or 10x doses) soil samples at 56 days. ANOVAs were followed by the Bonferroni test per each taxon. Significant differences are indicated by letters.

Supplementary Figure 4.1.16. Field study - Abundance 198 [sequence/sample] of an OTUs responsible for significant differences of ANOSIM, detected by ‘pamR’ in untreated (control) and treated (TCZ at

List of figures 340

1x, 2x or 10x doses) soil samples at 35 days. ANOVAs were followed by the Bonferroni test per each taxon. Significant differences are indicated by letters.

Chapter 4.2 Figure 4.2.1. Graphical abstract of Chapter 4.2. 206

Figure 4.2.2. Flow chart of the experimental procedure. 211

14 14 Figure 4.2.3. Kinetics of CO2 evolution from C-labelled pesticides and 218 mass balance of 14C-labelled residues (mineralized, extractable and non- extractable fraction) in soil untreated (control) or repeatedly treated with pesticides (CHL or TCZ). Mineralization of 14C-labelled CHL or 14C- labelled TCZ was monitored after 1, 2 and 3 months of incubation (i.e. one month after the 1st, 2nd and 3rd pesticide treatment). Initial radioactivity was applied to soil as 14C-ring-labelled CHL, as 14C-phenyl-ring-labelled TCZ (TCZ-P) or as 14C-triazole-ring-labelled TCZ (TCZ-T). Mass balance 14 fractions represent the radioactivity mineralized during 150 days ( CO2), the radioactivity extracted by methanol (extractable residues), and the radioactivity released during soil combustion (non-extractable residues). Error bars represent standard errors between four replicates.

14 14 Figure 4.2.4. Kinetics of CO2 evolution from C-labelled compounds (in 221 % of initially applied radioactivity) in soil microcosms untreated (control) or repeatedly treated with pesticides (CHL or TCZ). Mineralization of 14C-ring-labelled 2,4-D, 14C-labelled glyphosate, or 14C- labelled sodium acetate was monitored after 1, 2 and 3 months of incubation (i.e. one month after the 1st, 2nd and 3rd pesticide treatment). Error bars represent standard errors between four replicates.

Figure 4.2.5. PCoA ordinations of OTU weighted unifrac distance 226 matrices for soil microcosms repeatedly exposed to CHL or TCZ or untreated (control). Replicates of the same treatment per month (1st, 2nd or 3rd pesticide treatment) are represented in the same colors.

Supplementary Figure 4.2.1. Location of 14C atoms in 14C-ring-labelled 235 CHL, 14C-phenyl-ring-labelled TCZ, 14C-triazole-ring-labelled TCZ, 14C- ring-labelled 2,4-D, 14C-labelled glyphosate and 14C-labelled sodium acetate.

Supplementary Figure 4.2.2. Setup of mineralization jars (closed 236 system). A 14C-labelled compound is applied to a soil microcosm and 14 generated CO2 is trapped in sodium hydroxide that can be exchanged to quantify the generated radioactivity.

Supplementary Figure 4.2.3. Mass balance in % of initially applied 236 radioactivity. The fractions represent the radioactivity mineralized 14 during 150 d ( CO2), the radioactivity extracted by methanol, and the radioactivity released during soil combustion (non-extractable fraction). Initial radioactivity was applied as 14C-ring-labelled 2,4-D, as 14C-labelled glyphosate or as 14C-labelled sodium acetate to soil untreated (control) or repeatedly treated (once per month) with pesticides (CHL or TCZ). Error bars represent standard errors between four replicates.

Supplementary Figure 4.2.4. Bacterial diversity (observed species, PD 237 whole tree, chao1, simpson reciprocal) in soil repeatedly treated (once per month) with CHL or TCZ or untreated (control). ANOVA for each diversity index were followed by the Tukey HSD test (p < 0.05).

List of figures 341

Significant differences are indicated by letters.

Supplementary Figure 4.2.5. Relative abundance (%) of prokaryotic 239 phyla in soil repeatedly treated (once per month) with CHL or TCZ or untreated (control). Each color represents a phylum. Phyla whose abundance sum of all samples was < 0.05 % are grouped as ‘others’. Sequences that could not be assigned to a phylum are grouped as ‘unassigned’.

Supplementary Figure 4.2.6. Venn diagrams indicating the number of 239 shared and unique genera over time in soil repeatedly treated with CHL or TCZ or untreated.

Supplementary Figure 4.2.7. Number of sequences per sample of OTUs 240 altering between soil samples repeatedly treated with CHL (once per month) or untreated (control). Each bar represents four replicates. ANOVA for each taxon were followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

Supplementary Figure 4.2.8. Number of sequences per sample of OTUs 241 altering between soil samples repeatedly treated with TCZ (once per month) or untreated (control). Each bar represents four replicates. ANOVA for each taxon were followed by the Tukey HSD test (p < 0.05). Significant differences are indicated by letters.

Chapter 5 Figure 5.1. Graphical abstract of Chapter 5. 252

Figure 5.2. The timeline-based experimental design of the microcosm 257 study (Biobac soil, contemporary exposure to IPU) and the field study (Epoisses soil, historical exposure to IPU).

Figure 5.3. Locations of the primer sets used in this study on the plasmid 259 pSH. The sequences targeted by the different primer pairs (pdmA, pdmB, pdmA-down and pSH) are indicated in red. The transposon-like sequence (surrounded by tnpX) harboring the pdmA and pdmB genes is presented in grey.

Figure 5.4. Mineralization of 14C-ring-labelled IPU [% of initially applied 263 amount] and relative gene abundance of pdmA, pdmB, pdmA-down and pSH to 16S rDNA abundance [copy number/104 copy numbers of 16S rDNA] measured in the Biobed soil microcosms contemporarily exposed to IPU. ANOVAs were followed by the Tukey HSD test (p < 0.05) separately grouping all time points per gene. Significant differences are indicated by letters.

Figure 5.5. Relative gene abundance of pdm, pdmA-down and pSH [copy 265 number/106 copy numbers of 16S rDNA] measured in the Epoisses soil historically exposed to IPU.

Figure 5.6. Tentative of correlation between the relative abundance of 266 pdmB [copy number/106 copy numbers of 16S rDNA] and the maximum IPU mineralization (A) [%] of all tested years (2008, 2009 and 2010).

Chapter 6 Figure 6.1. Suggestions for the improvement of pesticide environmental 299 risk assessment: Acquisition of data by combining chemical, biochemical and molecular techniques and subsequent collection and publication of obtained knowledge.

List of figures 342