UNIVERSITÉ DU QUÉBEC À MONTRÉAL

GLOBAL DEFINITIONS OF ENVIRONMENTAL AND HEALTH TOXICITY OF PESTICIDES: KEY CONCEPTS, HOT TOPICS AND CURRENT

CONTROVERSIES OVER ATRAZINE AND GLYPHOSATE

RAPPORT DE SYNTHÈSE ENVIRONNEMENTALE

PRÉSENTÉ

COMME EXIGENCE PARTIELLE

DU DOCTORAT EN SCIENCES DE L’ENVIRONNEMENT

PAR LOUISE HÉNAULT-ETHIER

DÉCEMBRE 2013

1

Table of content1.  Abstract 2 2.  Introduction 3 

2.1.  Pesticide consumption 3 2.2.  Basics of toxicology 4 2.3.  Pesticide registration 6 

3.  Goal 7 4.  Pesticide definitions and legislation worldwide 7 

4.1.  Regional definition for pesticides 7 4.2.  North American legislations on pesticides and chemicals 8 4.3.  European legislation on plant protection products, biocides and chemicals 9 4.4.  Divergences and similarities of American and European pesticides legislation 10 4.5.  International organizations 10 

5.  Defining toxicity 11 

5.1.  Common human toxicity requirements 11 5.2.  Common ecotoxicity requirements 12 5.3.  (Eco)toxicity classifications 13 5.4.  (Eco)toxicity indicators 14 

6.  Key aspects of toxicity testing 14 

6.1.  Observable and quantifiable toxicity end-points 14 6.2.  Human data and model organisms 15 6.3.  Dose-response relationships 15 6.4.  Test animals, concurrent and historical controls 16 6.5.  Statistics 16 

7.  Hot topics in pesticides (eco)toxicity 17 

7.1.  GLP vs peer-review 17 7.2.  Discounting and conflicts of interest 18 7.3.  Active vs Inert ingredients 18 7.4.  Animal welfare and ethics 19 7.1.  Proving safety and safety factors 20 7.2.  Mixtures 20 7.1.  Subtle effects 21 

8.  Most widely used pesticides controversies 22 

8.1.  Atrazine 22 8.2.  Glyphosate 28 

9.  Conclusion: Controversies and resolutions 34 10.  References 37 

2

Global Definitions of Environmental and Health Toxicity of Pesticides: Key Concepts, Hot Topics and Current Controversies

Over Atrazine and Glyphosate

Louise Hénault-Ethier†*

† GÉOTOP, Institut des Sciences de l’Environnement, Université du Québec à Montréal, Montréal, Canada

1. Abstract Worldwide pesticide consumption is increasing, and the greatest share of the market is held by herbicides. The US is the world’s largest consumer of pesticides and Glyphosate (GLY) surpassed Atrazine’s (ATZ) leadership in American sales in 2001. In Europe, GLY is among the most popular plant protection products, but ATZ is not registered since 2001. The US Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) and the EU Plant Protection Products Regulation (2009/1107/EC) are the main statutes regulating pesticides Active Ingredients (AI) and formulation registration. International harmonization efforts are made for testing guidelines (OECD) and human and environmental toxicity classifications and tolerances settings (WHO/UNEP/FAO). Human toxicity testing is similar in the US and EU, with more extensive testing of AI compared to formulations. Ecosystemic impacts of pesticides are characterized through environmental fate and behavior studies, as well as ecotoxicity testing on animals. Key aspects of (eco)toxicity testing reviewed include toxicity end-points, dose-response relationships, number of test animals, controls and statistics. Hot topics in pesticides (eco)toxicity testing and risk assessments, at the root of current pesticide (eco)toxicity controversies, are reviewed. They include Good Laboratory Practices vs peer-review, Conflicts of interest and discounting, Subtle effects, Animal welfare and ethics, Active vs inert ingredients, Mixtures and Safety. ATZ controversies reviewed include court litigation, industrial influence on registration authorities, carcinogenesis, reproductive toxicity and endocrine disruption and toxicity of widespread environmental mixtures. GLY’s low toxicity is challenged with alleged reproductive toxicity and teratogenicity and its rapid degradation and low leaching potential are confronted with persistence, widespread contamination and potential crop hazards.Transparency is essential for the credibility and confidence in pesticide (eco)toxicity risk assessments and regulation, and this is more important than ever with the upcoming registration reviews of GLY and ATZ.

Key Words Pesticides, Toxicity, Ecotoxicity, Controversy, Atrazine, Glyphosate, Carcinogenesis, Teratogenesis, Endocrine Disruption, FIFRA, Plant Protection Products * The author declares no competing financial interests

3

2. Introduction

2.1. Pesticide consumption In 2007, the world pesticide expenditures was $39.4 billion, 32% of that was spent in the US alone, representing 22% of global pesticide consumption by weight, ranking it world’s first for sales and use.1 It is estimated that 1 to 2.5 million tons of Active Ingredients (AIs), formulated as pesticides, are used worldwide each year.1, 2 In the United States (US), 1055 AIs are registered.1 Agriculture is by far the most important sector of pesticides use, it accounts for an average of 70% of pesticides used among the Economic Co-Operation and Development (OECD) members,3 and 80% in the US.1 Annual pesticide use intensity in arable lands and permanent crops averages 4.22 tonnes of AIs per 1000 ha.4 There is no global compilation precisely detailing pesticides uses,3 and use varies widely across the globe. For instance, occidental countries preferably use herbicides while the global South uses more insecticides.5 Globally, approximately 40% of pesticides are used as herbicides, followed by insecticides (17%) and fungicides (10%)1. This general ranking applies to the US, but in the European Union (EU) fungicides precede herbicides and insecticides though trends are reversing.1, 6 Pesticides statistics are not uniformly compiled in time on a worldwide basis, and monitoring indexes of agricultural environmental pressure were changed from pesticide use7 to sales3 recently. But pesticide sales are only an imperfect proxy of pesticide use,8-10 perhaps simpler and cheaper to compile.8 Even within a country, pesticide sales compilations are neither detailed, nor very reliable, nor uniform, nor current.1, 8-12 Although expenses for pesticides have been on the rise in the US, the consumption of agricultural pesticides (Kg AI) has been decreasing, particularly for organophosphate insecticides.1 Common chemical classes of pesticides ranked in decreasing order of global consumption include : dithiocarbamates, organophosphate, phenoxy alkanoic acids, amides, bipyridyls, triazines, di- and triazoles, carbamates, urea derivatives and pyrethroids.2, 13 For greater details on chemical classes, structural motifs, and modes of action refer to Fenner13, Sternerson14 and Marrs.15 GLY and ATZ are amongst the most common conventional pesticides AIs used in the agriculture sector in the word. They have ranked 1st and 2nd respectively since 2001 in pesticides sales, averaging 82-84 million kg (180-185 million lbs) and 33-35 million kg (73-78 million lbs) in 2007 (most recent data available) in the US, which is the world greater consumer of pesticides.1 In Europe, besides the dominant inorganic sulfur fungicides, organophosphorous herbicides (GLY and Glufosinate) are ranked second in pesticide volumes with 10% of all sales.6 However, triazine herbicides (i.e. ATZ) rapidly declined from 5th position to 12th position from 1992 to 2003,16 at which point reregistration was denied in EU.17

4

2.2. Basics of toxicology

2.2.1. Toxicity and ecotoxicity of pesticides Global impacts of pesticide use such as their putative roles in cancer epidemics18 or alleged role in reducing global biodiversity19-21 will not be weighed here against the benefits they may have provided in feeding mankind.22 Instead, for the remainder of the introduction we will mechanistically depict how human health and environmental protection are defined in toxicity assessments. Toxicity refers to the intrinsic capacity of a chemical or mixture to injure an organism.16 Throughout this review we will sometimes distinguish public health (toxicity) from ecosystems protection (ecotoxicity), or use the indiscriminate term (eco)toxicity.

2.2.2. General approaches to toxicity testing: in vivo, in vitro and in silico Descriptive toxicology needs to clearly establish the causal effect of a pesticide, or in other words, associate toxicant-toxicity relationship.23 Tests are classically performed in vivo, using live organisms, but in vitro tests, in ‘’glass’’ or test tubes, may use cultured cells or tissues.23 For example, short or medium-term bioassays are useful in establishing mechanisms and mode of actions (MoA) of carcinogens, by targeting neoplasia or pre-neoplasic lesions, but not all available tests have equal significance to help predict human carcinogenicity24 and developmental toxicity.25 Nowadays, toxicity testing may even be done in silico, using extensive databases and computer modeling through Quantitative Structure Activity Relationships (QSAR),26 though QSAR is not yet perfectly reliable to replace laboratory testing, for example biodegradability,26 carcinogenicity,27, 28 eye/skin irritation, reproductive toxicology, developmental toxicity and neurotoxicity28. The US is a strong proponent of QSAR which can decrease costs of toxicity testing, decrease the use of test animals and increase throughput,29 the EU prefers a precautionary approach,28 but also recognizes the alternative approaches to reduce the number of animals required for (eco)toxicological testing.30-32

2.2.3. Route of exposure and duration Toxicity depends on duration and location of exposure, and varies according to target organisms.26 Exposure is usually defined as short term (28d), sub-chronic (90d) and long-term (>12mo).33 In acute toxicity, symptoms appear rapidly (≤24h oral/dermal or ≤4h inhalation) and cellular damages may be reversible. Chronic toxicity usually involves continuous or long-term exposure and may yield permanent cellular impairment which may result in loss of function and death in severe cases. It is sensible to model human lifelong exposure by spanning the majority of the lifespan of shorter lived organisms; i.e., 18 or 24 months for mices or rats respectively;34, 35 extending to 24 and 30 months respectively for weaker carcinogens,27, 36 some even advocating until natural death to encompass cancer latency.37 Reproductive studies (preconception, gestation, lactation and F1 generation exposure) typically last 30 to 36 months.27

5

Toxicokinetics (absorption, distribution, storage, biotransformation and elimination) will result in local or systemic toxicity.23 The time required to reduce the amount of a pesticide from the body by half is called the biological half-life (T1/2). Target organ toxicity refers to adverse effects exhibited in a specific organ;23, 34 and organs or tissues affected by pesticides include blood cells (hematotoxicity), liver (hepatotoxicity), kidney (nephrotoxicity), central and peripheral nervous systems (neurotoxicity), skin (dermatotoxicity) and lungs (pulmonotoxicity). Neonicotinoid insecticides are neurotoxic, and so are the additionally dermatotoxic Pyrethrins.14, 15 Metamsodium fumigant is pulmonotoxic.38 DDT is hematotoxic, hepatotoxic and neurotoxic.23 Not all organs respond to a toxicant with the same intensity, and a toxicant may or may not affect more than one organ, though several toxicants can affect the same organ; for instance 68 AI registered in EU are neurotoxins.23, 39 Toxicodynamics is concerned with specific cellular effects of a toxicant within an organism. Pesticide toxicity is exerted through different biochemical modes of action (MoA) which Sternerson graciously refers to as the seven routes to death14: (1) enzyme inhibitors; (2) endocrine disturbers (antagonists or agonists); (3) free-radical-producing poisons; (4) membrane pH gradient degraders; (5) membrane structure disturbers; (6) electrolytic, osmotic or pH balance disturbers; and (7) tissue, DNA or proteins destroyers. While target organ toxicity is generally well characterized, data on MoAs are often limited or even inexistent.39 Carcinogens cause uncontrolled proliferation of cells resulting in either benign or malignant tumors (neoplasms).16, 34 Cancers may be induced via genotoxic (direct or indirect action on DNA) or epigenetic mechanisms, the latter referring to various actions on cells, mitosis, immunity, endocrine systems, etc.16 Mutagens cause mutations in somatic or germ cell lines, involving point mutations in a gene or a large-scale change at the chromosome level. Teratogens produce malformed fetuses by altering cellular differentiation and growth processes in utero.23 Substances that mimic, block or disrupt in any other way the normal functioning of hormones (i.e. estrogen, angrogen, thyroid) are considered endocrine disruptors (ED).40 Over 120 pesticides are known ED, and these belong to numerous chemical classes.41

2.2.4. Treshold doses The threshold dose refers to the dose at which the first test organism will exhibit toxicity symptoms. The highest dose at which no effects are observed is the No Observable (Adverse) Effect Level (NOEL/NOAEL). The lowest dose causing changes in the treated groups is termed the Lowest Observable (Adverse) Effect Limit (LOEL/LOAEL) and Treshold Limit Value (TLV).16 Between both extremes, we may observe an effective (ED), toxic (TD) or lethal dose (LD) at which the desirable response is observed, toxicity symptoms are displayed or death occurs, respectively.23 Subscripts facilitate comparison amongst studies and pesticides, i.e., LD50 denotes the dose at which 50% of the test organisms died.23 When comparing equivalent indicators, the most toxic or potent compound is the one with the lowest ED50, TD50 and LD50.23

6

2.2.5. Risk assessment Risk assessment refers to the probability that a hazard, such as (eco)toxicity, will be encountered upon exposure : Risk = Hazard x Exposure.16 Hence, reducing either hazard or exposure minimizes risk. Hazard identification, requires conclusions on a pesticide causal link to health effects, and a positive dose-response relationship is the best evidence that a chemical is hazardous.16 Evaluation of real-life exposure is more difficult to ascertain then exposure of laboratory animals. Food exposure to pesticide residues is calculated based on Maximum Residue Limits (MRL), under Good Agricultural Practices (GAP) and is expressed as a Theoretical Maximum Daily Intake (TDMI). Inherent to GAP is the assumption that pesticides are used in accordance with the label instructions, thereby decreasing risks, but it is not a precautionary assumption owing to the unpredictable nature of human behaviors. A National TDMI (NTDMI) can also be calculated taking into account regional diets (common food eaten, average portion size). Regional diets represent long term consumption habits compiled with food production, import and export data.42 Dietary exposure to pesticide residues also occur via drinking water.43 Aggregated exposure sums all the possible dietary, environmental and domestic sources of exposure to a specific pesticide. Occupational exposure (at work) is considered separately (Acceptable Operator Exposure Level; AOEL).15 Cumulative exposure combines all the individual AIs which have a similar mode of action.15

2.2.6. Toxicity risk management Risk management interprets the risk assessment in order to develop regulatory actions that address public health, social, economic and environmental concerns. Three components of risk are considered in risk management (1) its’ probability (likeliness), its’ magnitude (spatial or temporal) and its’ uncertainty (quality of information).44 Comparative risk analysis, is a procedure for ranking health or environmental problems by their seriousness or relative risk, with the intended purpose of assigning them program priorities (Cleland-Hamnett 1993), i.e. EU preferential registration of safer pesticides45 or USGS prioritization for pesticides analytical developments.46 (Eco)toxicity risks are disseminated to workers and consumers via pesticide labels (standardized warning phrases and color-coded hazard pictograms) and Material Safety Data Sheets (personal protection, first aid, spill, elimination, etc.).26

2.3. Pesticide registration Pre-registration toxicity testing is generally the burden of industrial pesticide fabricants benefiting from the sale of pesticides.15, 47 To obtain a commercialization authorization (registration or homologation) for a given AI or formulation product, the industry needs to realize tests and put together a dossier containing toxicity testing results, a justification for the omission of a result, as well as acceptable peer-reviewed literature, epidemiological studies and occupational exposures data.15 Regulatory authorities will eventually evaluate the dossier and grant or deny registration.47

7

For registration, the EU15 and US48 have similar general data requirements: (1) Identification of the AI, (2) physico-chemical properties, formulation, impurities, (3) analytical methods for enforcement, (4) toxicology, (5) food and feed residue chemistry, (6) environmental fate and behavior, (7) ecotoxicology. The EU specifically requires a proposed classification and labeling and a summary and evaluation of data.15 The EU and Canada generally require efficacy or product performance studies, while the US requires it only for public health use pesticides.15 The US additionally require spray drift studies.48 The data provided should suffice to establish NAOEL, AOEL, ADI and MRL.15 3. Goal The aim of the current paper is twofold: to introduce the reader to the world of pesticide human and environmental toxicity, visiting both key and hot topics, before unraveling critical controversies. To appreciate the subtle nuances in pesticide’s (eco)toxicity definitions as they vary around the world, pesticides definitions and legislations from the new world (US) and the old continent (EU) will be compared and contrasted. Though pesticides (eco)toxicity assessments and registrations process is mainly national, international organization embrace an overarching role and put forward harmonization efforts which will be synthesized herein. The simplified initiation to (eco)toxicity provided in the introduction will be deepened for selected key concepts, pivotal to comprehend current contentious issues in pesticides risk assessment. As detailed publications on testing methodologies and hazard assessment abound,16, 49 the upcoming discussion will not be exhaustive. The last section will expose controversies over the world’s most commonly used pesticides, GLY and ATZ. The concluding statement will review the philosophical roots of pesticide polemics and address practical recommendations for their resolution. Throughout the text, parallels between pesticides and other chemicals will be drawn, as both rely on similar, albeit legally distinct, testing methodologies and risk assessment frameworks which mutually influence each other. 4. Pesticide definitions and legislation worldwide

4.1. Regional definition for pesticides In the simplest expression, pesticides are defined as substances that are used to kill unwanted organisms.50 The most common categorization of pesticides refers to their intended end-points: insecticides, rodenticides, nematicides, fungicides, herbicides and acaricides. Legal terminology and functional definitions vary across the globe.50 For instance, under the American Federal Insecticide, Fungicide and Rodenticide Act (FIFRA), pesticides definition is precise, but encompasses numerous substances and intended uses:

(1) any substance or mixture of substances intended for preventing, destroying, repelling, or mitigating any pest [insect, rodent, nematode, fungus, weed, other forms of terrestrial or aquatic plant or animal life or viruses, bacteria, or other micro-organisms, except viruses, bacteria, or other micro-organisms on or in living man or other animals, which the

8

Administrator declares to be a pest] and (2) any substance or mixture of substances intended for use as a plant regulator, defoliant or desiccant.(p.192)51

To encompass all substances regulated under FIFRA, the EU has two distinct regulations on plant protection products and biocides:

‘Plant protection products’ are defined as chemical or biological products intended to: protect plants or plant products against harmful organisms; influence the life processes of plants, other than as a nutrient (e.g. growth regulators); preserve plant products; destroy undesired plants or parts of plants; and check or prevent undesired growth of plants. Biocidal products cover a wide range of products, including: disinfectants (bacteria and viruses), preservatives (mould, fungi, and insects), public hygiene insecticides (e.g. flies, mosquitoes, ants), rodenticides (rats, mice), and antifouling preparations.(p.2)15

Note the distinction from the medical and veterinarian pharmacopeia, explicitely stated in the American definition, which is of interest as some similar molecules may be used in both the curing and eradicating domains.52, 53 A critical highlight of both definitions is the intent to cause harm to targeted living organisms, which substantiates the possibility of adverse toxic effects on man and the environment.50 Sometimes, the mode of action affecting target pests is the same as that harming man and wildlife, but in many cases, it is through a different and co-incidental physical or biochemical property.50 The reality of both small-scale and large-scale undesirable side effects exposes the crucial importance of pesticide assessment by competent authority relying on both independent and ethical professionals.50

4.2. North American legislations on pesticides and chemicals Four statutes anchor the legal framework for chemicals and pesticides regulations in the US.28 The foundation was laid with the 1972 Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA).28 Chemicals were only regulated subsequently with the 1976 Toxic Substances Control Act (TSCA).28 Due to widespread concern about the expansion of chemicals and pesticides, the US Congress granted authority to the Environmental Protection Agency (EPA) over testing, risk assessment and management for existing and novel chemicals and pesticides.28 The Emergency Planning and Community Right-To-Know Act (EPCRA) led to the establishment of toxic release inventory (TRI) to follow-up on the releases and transfers of chemicals from the industry28. FIFRA was amended and strengthened in 1988. In 1990, the Pollution Prevention Act (PPA) was adopted. A Chemical Safety Improvement Act was recently proposed in 2013 as a bi-partisan effort to distinguish high versus low priority chemicals, and help overcome the ‘no unreasonnable risk provision’ of the current TSCA. Contrary to pesticides which are now all subject to periodic toxicity assessment reviews where the manufacturer is obligated to fulfill toxicity data requirements, 33 000 industrial chemicals in use as of 1976 have been ‘grandfathered’, with the burden to prove toxicity relying on EPA.54 In 35 years, only 5 chemicals were deregulated due to their widely recognized toxicity.55 Among these are dioxins, which nevertheless appear as regulated contamination in several pesticides formulations widely applied to agricultural lands across the globe.56 In the US, different legislations bear different statutory

9

mandates on risk, ranging from ‘zero-risk’ to ‘no unreasonable risk’. For instance, the Delaney Clause of the Food, Drug, and Cosmetic Act is a ‘zero-risk’ provision as it prohibits the use of any food additives (including pesticide residues above the MRL) which have been found to induce cancer in humans. On the other hand, FIFRA requires the EPA to register pesticides which will not cause ‘’unreasonable adverse effects on the environment’’, referring to ‘’any unreasonable risk to man or the environment taking into account the economic, social and environmental costs and benefits of the use of any pesticide’’.57 A substance is no longer regulated under TSCA when it is incorporated in a pesticide formulation as an ‘inert ingredient’.58 The EPA does not fully pre-empt states from regulating pesticides on their territory. The 1992 North American Free Trade Agreement (NAFTA), signed by the US, Canada and Mexico led to the creation of the Technical Working Group on Pesticides in 1996 which aimed at (1) sharing regulation work, (2) harmonizing (eco)toxocity evaluation and policy considerations and (3) facilitating pesticide trade.15

4.3. European legislation on plant protection products, biocides and chemicals Historically, each EU Member State had their own legislation and registration policies concerning plant protection products. Until recent reviews, two directives governed pesticides in EU, under the previously defined denomination of plant protection product (PPP) (91/414/EEC)59 and biocides (98/8/EC).60 For the purpose of this review, we will concentrate on PPPs, now under Regulation (2009/1107/EC)30 which oversees test guidelines, classification and labeling, and MRLs. A rapporteur Member State is appointed for each substance sponsored by the industry, and then writes a Draft Assessment Report (DAR) based on the industry’s dossier and produces a risk assessment. Active substances are initially evaluated at the EU level, then substances without significant risk for humans, animals or the environment (water, soil, air, non-target organisms, plants) are included in Annex I. Individual Member States must then authorize (or register) each acceptable product, but once registered by one nation, other Member states cannot deny registration unless agricultural or environmental considerations differ significantly. New products are only registered for 10 years, and existing products should not exceed 15 years before review.30 European Food Safety Authority (EFSA) was charged with peer-review of plant protection products as of 2003 and completed the first round of review in 2009.61 Of the 1000 AI sold in EU before 1993, 67% were removed from Annex I because of absence of industry sponsor, incomplete dossier submission or voluntary withdrawal by the industry; and another 7% was denied registration due to lack of evidence for safe use on the health of humans and the environment.61 In 2006, a new EU strategy to reduce the impact of pesticides on health and the environment was adopted; it encourages low-input or pesticide-free cultivation. The new directive 2009/1107/EC sets stricter registration criteria for Plant Protection Products, denying registration for carcinogens, mutagens, endocrine disruptors, reproduction toxicants or highly persistent pesticides, unless limited human exposure can be demonstrated, and proposes a mechanism whereby safer alternatives should substitute toxic substances.30

10

In 2007, the EU enacted the Registration, Evaluation, Authorization and Restriction of Chemical substances legislation (REACH)32 which burdens the industrials considerable toxicity testing for registration of novel commercial chemicals and periodic reviews for existing chemicals. Pesticides ‘’Inert’’ ingredients are controlled under REACH, and appropriate REACH data is used during PPP registration.58, 62

4.4. Divergences and similarities of American and European pesticides legislation Pesticide legislations vary across the globe, and two major trends oppose dominant cultures in addressing environmental and human risks of agbiotechs: ‘’Innocent until proven guilty’’28, 63 or ‘’wait and see attitude’’27 versus the precautionary principle.63 For example the US permits chemical pollution until scientific evidence confirms risk, while the EU sets maximal tolerances a priori as a precautionary measure.64, 65 In the US and EU pesticides AI and inert ingredients are regulated under FIFRA and PPP directive respectively, but only the EU also regulates inerts under its chemical directive REACH even they enter pesticides formulations.58 Despite policies divergence, there is a planetary trend towards harmonization of (eco)toxicity testing and risk assessment for pesticides. Standardization facilitates fulfillment of commercialization registrations by industrials, permits exhaustive analyses without limiting throughput allowing timely market-entry of novel substances with potential lower (eco)toxicity, aid pesticide registration for less fortunate countries and rationalizes limited time and financial resources through sharing of dossiers within organizations such as EU or NAFTA.26 In the EU, the Uniform or Common Principles aid in technical harmonization.15 Under REACH, industrials may jointly request a registration and are obligated to share toxicity testing data, with a financial compensation in the first 10 years, and through a mechanism for public disclosure after that delay.58 However, the chemicals and pesticide information collected under the REACH agreement is claimed Confidential Business Information (CBI) and cannot be shared outside the EU.55 As EU is the largest importer and exporter of chemicals, the upcoming revision of TSCA is expected to be influenced by REACH, and acceptance by industrials might be simpler in the US reform as they already produce toxicity data for EU registration.55, 58

4.5. International organizations International organizations concerned with harmonization or pesticide review procedures and reducing the risks of pesticide use regrouped in 1992 under the Pesticide Programme, directed by the Working Group on Pesticides. Members include the OECD, the United Nations Environment Programme (UNEP), the Food and Agriculture Organisation (FAO), the World Health Organization (WHO) and the European Commission (EC) as well as pesticide industry and public interests organization observers.16 Three important international agreements affect registration and international movements of pesticide.28 The Rotterdam Convention requires explicit consent for importation of restricted or banned pesticides in developing countries; the Stockholm Convention controls Persistent Organic Pollutants (POPs) and the Cartagena Biosafety Protocol

11

manages GMOs which also affects Plant Incorporated Protectants (PIPs). The FAO is responsible to set the Acceptable Daily Intake (ADI), akin to Reference Dose (RfD) in the US. The FAO/WHO establish Maximum Residue Limits (MRL) on food.66 The most important contribution of the WHO to defining pesticide toxicity is the Recommended Classification of Pesticides by Hazard now under the Globally Harmonized System (GHS)26 which will be detailed further. In an attempt to harmonize procedures for toxicity evaluation around the globe (as pesticides are often marketed in various countries), the OECD compiled methodologies to assess the (eco)toxicity of various pesticides which are recognized and authorized by most national regulatory authorities.67 The OECD Test Guidelines cover physico-chemical properties, human and wildlife health, environmental accumulation and degradation. The OECD also publishes documents on Good Laboratory Practices (GLP) and Compliance Monitoring and Risk Management.16 5. Defining toxicity

5.1. Common human toxicity requirements Toxicity testing requirements in the US48, EU68 and UN GHS26 are similar for the AIs: (1) acute toxicity (oral, dermal, inhalation); acute irritation (dermal, eyes); sensitization (dermal); sub-chronic toxicity (oral); chronic toxicity; carcinogenicity; reproductive toxicity (teratogenicity and fertility); genotoxicity (in vitro and in vivo assays); metabolism and toxicokinetics; and neurotoxicity (acute, short-term, delayed). US requires additional tests on sub-chronic toxicity (additional dermal and inhalation routes), developmental neurotoxicity and immunotixicity. In the most recent revision of the PPP regulation (2013/183/EC),68 EU now requires multigeneration and metabolite studies, but abandoned two testing routes (intraperitoneal and percutaneous) which were in the original regulation (91/414/EEC).15 Testing guidelines for carcinogenicity studies have been designed by numerous regulatory agencies (EPA34 OECD/EU35). The 1981 OECD guidelines for carcinogenicity tests have been revised recently.35 The EU now recommends to conduct carcinogenicity screening jointly with chronic toxicity tests, and equivocal data may then require specific additional testing.68 Adequately tested, all human carcinogens have led to positive results in minimally one of the tested model animals27 and although animal carcinogenicity does not guarantee human carcinogenicity, the likeliness of its risk, unless otherwise proven, should be taken into account in regulatory decisions based on both OECD and EPA guidelines.36 The EPA initiated the development of an Endocrine Disruptor Screening Program (EDSP) in 1996 and launched a first screening in 2009 which included GLY and ATZ (see controversy details in ATZ section). EU integrated ED screening in chronic toxicity tests, with MoA studies conditionally required;68 and specific scientific ED criteria should be adopted in December 2013.30

12

Toxicity requirements for the pesticide formulation (AI and co-formulants together) are limited in both the US48 and EU62: acute toxicity, acute irritation, sensitization and dermal absorption. Toxic effects on livestock and pets are generally required in the EU, while only pets are conditionally required in the US, but testing is either on AI or formulation (if data from AI is insufficient for extrapolation). A summary table for EU and US with requirement contingencies, exposure route and preferred testing species is provided in Supplementary materials Table S1.

5.2. Common ecotoxicity requirements Ecostudies on pesticides are generally subdivided in environmental fate and behavior studies and ecotoxicity tests on various organisms. As for toxicity studies, in vivo, in vitro and in silico methods may be used. For example, SAR data may substitute for ecotoxicity and bioaccumulation testing, but not for easy degradation tests.26 Environmental fate and behavior studies refer to three compartments (soil, water and air). Aspects studied include rate, route and process of degradation (aerobic and anaerobic biodegradation, photodegradation, hydrolysis), identification of metabolites, adsorption and desorption (soil and sediments), leaching and ground water monitoring. Only the AI is typically studied, except for specified requirements in the US legislation for the field soil dissipation studies and ground water monitoring; and conditionally required field aquatic sediments dissipation studies. For biodegradability, a BOD5/COD ≥ 0.5 (Biological Oxygen Demand over 5 days/Chemical Oxygen Demand) generally indicates easy degradability, but standardized OECD degradability tests are preferred.26 Easy degradability usually reaches >70% degradation (based on dissolved organic carbon) within 28 days.26 Refer to Fenner13 for a discussion on the limitations of current environmental pesticide degradation knowledge. Details of required fate and environmental behavior studies, with EU and US specificities, are presented in Supporting Information Table S2. Typically required ecotoxicity testing include acute (48-96h), dietary, chronic, reproduction, growth and bioaccumulation on various target animal models. For both the US48 and EU,68 required ecotoxicity studies focus on birds, fish, aquatic invertebrates (Daphnia sp.), terrestrial invertebrates (honeybees) and algae; and conditionally required ecotoxicity tests include wild fauna and terrestrial plants. Only the US requires vascular plants growth tests for AI or formulation. Only the EU requires additional toxicity test for other beneficial arthropods (i.e. predators), earthworms and other soil macro-organisms. Both the US and EU conditionally require assessing the effects of formulations on biological methods for water filtration, but the EU also conditionally requires additional microbiological tests for example on soil nitrogen transformation and other formulations effects on microflora. Details on ecotoxicity tests on various organisms, with EU and US specificities, are presented in Supporting Information Table S3.

13

In the UN GHS, environmental hazards pertinent to pesticides focus on aquatic ecosystems.26 Standardized OECD techniques (accessible online 67) are preferred, but other GLP variants of aquatic ecotoxicity may be acceptable, for example guidelines by the EPA,69 EU,33 ASTM (American Society for Testing and Materials)70 and ISO(International Organization for Standardization).71

5.3. (Eco)toxicity classifications In 1975, the WHO first approved Recommended Classification of Pesticides by Hazard (RCPH), a classification that has evolved and gained wide acceptance.26 The goal of this classification is to accurately reflect acute health risk (single or multiple exposures over a short period of time) in the case of an accidental contact during normal handling of the product, storage and transportation. The WHO organizes pesticides AIs on the basis of acute oral (or dermal if the value is lower) LD50 on rats (unless a more sensitive or man-resembling model data is available). The most recent update of the WHO classification has now been unified with the UN GHS of Classification and Labeling of Chemicals. Briefly, GHS categories 5 to 1 correspond to WHO epithets ranging from Unlikely to present acute hazard up to Extremely hazardous and they are coherently assigned to a pre-determined range of oral LD50 between ≥5000 and ˂5mg/kg, to which hazard statements correspond in a standardized crescendo culminating at Fatal if swallowed. A similar organization, with different ranges, qualifies dermal hazard statements also culminating at Fatal in contact with skin. Formulations should be classified according to, in order of preference, (1) the manufacturer’s toxicological data, (2) the most toxic constituent, or (3) based on the ratio of classification of the different AIs entering in the formulation. Table S4 outlines the harmonized classifications and labeling requirements of the RCPH and GHS. EU also has standardized hazard symbols and phrases for example R25: Toxic if swallowed or R61: May cause harm to the unborn child.15 The US EPA also has an acute toxicity classification (classes I to IV) for oral, percutaneous and inhalation based on rats LD50 or LC50,15 but classes limits vary from those of the UN system. In addition, the US EPA classifies (I to IV) pesticides for their cutaneous and ocular irritancy.15 Carcinogens are categorized by both the International Agency for Research on Carcinogenesis (IARC) and the American Conference of Governmental Hygienists.15 Although the nomenclature varies, both recognize five classes which range from confirmed human carcinogen with sufficient clinical and epidemiological evidence to probably not carcinogenic in humans due to lack of scientific evidence. The GHS acute and chronic ecotoxicity categories are based on the most sensitive organism tested.26 Acute and chronic toxicity (distinguishing rapidly biodegradable from persistent substances) is subdivided into a maximum of 3 categories depending on LC50 or EC50 to standardize communication when referring to very toxic (1), toxic (2) or harmful (3) for aquatic organisms (acute toxicity) and resulting in long term detrimental effects (chronic toxicity). Mixtures should also be classified, taking into account pertinent components or based on normalized additive principles if no specific ecotoxicity data for mixtures exists. Table S5 outlines the GHS classifications for acute and chronic ecotoxicity.

14

5.4. (Eco)toxicity indicators Finally, several indicators have been developed to help risk competent authorities, as well as pesticide users, rank the relative health and environmental risks of various pesticides. A reduction of the overall quantity of pesticides used in the world is desirable,7, 9 but may not necessarily reflect a reduction in risk to human and wildlife, as not all pesticides were created equal.9 Indicators are strongly dependent on available (eco)toxicological data, which is often restricted to AI (as opposed to formulations), and perhaps that newer, and less tested pesticides, may benefit from lack of data and obtain better indices in this type of ranking system. In 2010, the EU launched the HAIR indicators (HArmonised Environmental Indicators for Pesticide Risk) targeting aquatic and terrestrial organisms, ground water protection, public (including pregnant women) and applicator health.72 Starting in 2013, the OECD will revisit all existing indicators with the hope of developing globally harmonized indicators.73 Amongst a wide diversity of national indexes, the human health (in French IRS: Indice the risque pour la santé) and environmental toxicity (IRE: Indice de risque pour l’environnement) indices used in Quebec (Canada) are interesting as they attempt to consider the global effect of formulations for most pesticides.74 They aim at tracking desired reduction of the pesticide impact and guide farmer’s and agronomists choices.74 For a comparable concentration of 480g AI/l in formulation, ATZ has one of the highest health and environmental score (IRS = 466, IRE = 240) while GLY amine salts ranks amongst the safest pesticides (IRS = 17, IRE = 5)(www.sagepesticides.qc.ca75). From 1997 to 2009, while the global consumption of pesticides (kg AI) has rised (17,5%), the cumulative health (-19.9%) and environmental impact (-24.7%) of the pesticides used in Quebec has declined.9 However, remember that indices are subjective aggregates with varying assumptions, greatly dependent on the quality of the available (eco)toxicological data, and may not change as rapidly as science progresses in finding new (eco)toxicity effects. 6. Key aspects of toxicity testing To maximizes the reader’s appreciation of the subsequent hot topics and controversies sections, seven key aspects of classical toxicity testing will be reviewed, by expanding on the basic toxicity assessment framework laid in the introduction: (1) Observable and quantifiable end-points; (2) Human data and model organisms; (3) Dose-response relationships; (4) Test animals, concurrent and historical controls; and (5) Statistics.

6.1. Observable and quantifiable toxicity end-points First, observable and quantifiable responses, or end effect, must be selected. For example, typical quantifiable responses of reproductive toxicity may include changes in the oestrus cycle, spermatogenesis, oogenesis, abortions, morphological malformations of progeny such as unossified bones.68 Not all responses

15

observed in a study represent toxicity, some simply represent physiological changes; and the discrimation should be made with discernment and via competent advice in face of uncertainty.16 An adverse response results in functional impairment, impairment to cope with additional stress, or increased susceptibility to environmental changes.16 Remember that an effect which is not seeked for or monitored, perhaps in a standard protocol, cannot be demonstrated.

6.2. Human data and model organisms Pesticide toxicological tests on humans are unacceptable for ethical reasons,26, 68 and because we have reasonable certainty that animal models suffice; for example cancer detected in laboratory animals should also occur in humans (despite occasional lobbying by industrial interests to convince regulatory agencies that rodents cannot be used as human surrogates when positive findings occur).27 Epidemiologic studies are considered,26, 68 but occupational or accidental exposure68 rarely supports firm causality conclusion (anecdotic nature).76 Mices and rats are preferred mammal test animals due to their small size, rapid multiplication, relatively short lifespan and the extent of our knowledge about their physiology and biochemistry.34-36 Acute oral or dermal tests are preferably conducted on rats, while rabbits may also be used for dermal toxicity.16, 26 Ecotoxicity may use fish, birds, invertebrates, algae, bacteria or several other model organism;68, 69, 77 though we ignore if the small set of domestic or laboratory organisms used really model wildlife realistically.19 Gender, age, diseases and strains are also considered in the choice of test organisms.23, 34, 35 Different strains may be insensitive,78, 79 overly sensitive16, 80 or exhibit MoA not relevant for humans.81 For instance Sprague-Dawley (SD) rats are commonly used80 and recommended78 for carcinogenicity testing, except to study mammary tumors for which they exhibit a high background incidence.27, 80 Discounting based on strain historical data82 or limited understanding of carcinogenesis MoAs is not precautionary,27 and was deemed instrumental in ATZ and GLY controversies. Using two species in carcinogenicity tests might reduce uncertainty.34, 36

6.3. Dose-response relationships Maximal test doses should minimize the chance of false-negative while not compromise the biological interpretability of responses;16 for example not significantly affecting growth and survival or skin integrity, while still challenging test organisms Maximal Tolerated Doses (MTD) in carcinogenic studies,27, 34 or refined based environmental concentrations or pharmacokinetics parameters saturation for endocrine disruption studies.16, 36, 83 A minimum of three data points in addition to the control is required to establishing a linear dose-response relationship,16 in carcinogenicity studies36 and to establish a NOEL.35

16

Paracelsus wrote that the dose makes the poison about 500 years ago, and until recently most regulatory agencies and industrial toxicity specialists applied this dogma litteraly,16 not to say blindly. The dose-response relationship requires direct causality, with the proportion of organisms exhibiting toxicity mathematically related to pesticide exposure. This regulatory pre-requisite is not only limited by our ability to correctly discern and measure responses, but also by thresholds appearing in carcinogenicity studies (no neoplasm induction below a certain dose; reviewed in ref36), and endocrine effects unpredictable from higher doses and exhibiting nonmonotonic responses (U or ∩ shape; reviewed in ref83). A current controversy over ATZ stems from this.

6.4. Test animals, concurrent and historical controls A greater number of test animals increases statistical power, but also challenges budgets, logistics and ethics.30, 36 A minimum of 50 animals per sex per group is required for carcinogenicity tests,34, 35 and up to 100 or 1000 for weaker carcinogens or widespread contaminants.27 Survival milestones may also be required (50% in all groups,36 25% in control,35 or 50% at 18 months and 25% at 24 months34). Fewer subjects may impede statistical analysis36 and even spark controversy (refer to GLY section).82 Concurrent controls are unconditionally required16, 27 and though historical control data may frame means and ranges for natural variability of rare tumors,27 their statistical comparison with treatment groups is inappropriate as biological parameters, including spontaneous lesions incidence, can vary significantly over time.16 Likewise, pooling species, strains, age cohorts, sex groups, laboratories, diets, and exposure routes is inappropriate,68 unless perhaps statistical equivalence was demonstrated.80 Inappropriate use of historical data hinders the power of carcinogenic effect detection in laboratory animals27 and should not be abusively used to dismiss positive findings.36 The recent revision of the EU PPP guidelines finally specifically addresses this topic.68

6.5. Statistics Specific statistical tests are not typically required by the OECD and EPA, though mean, standard deviations and a priori selected significance levels (p values) should be reported under GLP.36 EU mentions that appropriate statistical tests should be conducted and methodological details reported.68 Inappropriate statistical adjustments for age and survival may hinder the power of carcinogenic screens.27, 36 Contrary to other agencies, the FDA does formulate specific reporting and statistical guidelines, emphasizing that statistical significance (p≤0.05) does not imply biological significance,84 and even further reducing thresholds (p≤0.025 for rare tumors and p≤0.005 for common tumors) to minimize the rate of false positives.36 Statistically, in trying to increase the specificity of a test (control for Type I errors), a greater number of false negatives may be found over that of false positives.44 While appropriate statistics are desirable, it is certainly

17

more dangerous for public health to accept a false negative and more precautionary to accept a false positive.27 Finally, unstandardized statistics opens the door to controversies when industrials and contrarian scientists accept or reject significance using different thresholds.79 7. Hot topics in pesticides (eco)toxicity Several aspects of pesticide (eco)toxicology testing and risk assessments are, to say the least, technically contentious. These selected hot topics comprise scientific disagreements or policies divergence which are at the root of ongoing controversies on specific pesticides: (1) GLP vs peer-review; (2) Discounting and conflicts of interest; (3) Subtle effects; (4) Animal welfare and ethics; (5) Active vs Inert ingredients; (6) Mixtures; (7) Proving safety and Safety Factors.

7.1. GLP vs peer-review Standardized laboratory management and reporting rules, referred to as Good Laboratory Practices (GLP) were instituted to prevent fraud in industry supported toxicity assessments.85 A fraudulent report on GLY toxicity by Industrial Biotest Laboratories in 1976 pushed the FDA (1979) and EPA (1983) to adopt a GLP framework which was in place when a second counterfeit test on GLY residues was unmasked at Craven Labs.86, 87 In both cases, Monsanto was victim of dishonest tests on Roundup, and had to redo all studies which support current registrations.86-88 GLP are now required by OECD and US guidelines36 and encouraged in the EU under REACH32 and PPP regulation.30 GLP are at the heart of intense debates in the credibility of industrial testing. The industry claims GLP guarantee the reliability and verifiability of data,88, 89 while independent researchers argue GLP do not prove scientific validity and outdated standardized guidelines lack flexibility to explore yet unidentified mechanisms or end-points, and are insensitive to realistic exposure dosages.85, 90 Peer-reviewed studies have often been dismissed from industries registration dossiers when diverging from GLP or OECD guidelines.36, 85, 90 Finally, EU laid a regulatory framework for acceptable non-GLP (eco)toxicology data.30 Peer-review is a process whereby anonymous specialists analyze the experimental design, results and conclusions of a study (or a review), and suggest modifications, improvements or limitations to authors prior to publication, keeping an open mind for innovation.91 The public nature of peer-reviewed literature is an advantage over confidentiality of industrial research. In response to independent scientists and citizens concerns over industrials conflict of interest in pesticide toxicity testing, registration authorities now insist on the inclusion of independent peer-reviewed literature in dossiers for commercialization authorization;30 though peer-reviewed studies are then funneled through conservative quality control guidelines,36, 92, 93 sometimes explicitly favoring standard procedures and GLP, dismissing novelty.85 While peer-review is a useful quality control tool, is by no means a perfect.91 At least publications set the stage for sound and transparent debates over the scientific validity of disputable findings (i.e. letters to the editor, publication

18

rebuttals).94 A highly controversial toxicity publication on RoundUp was retracted by the editors, not after the first round of peer-review, but after publication and subsequent pressure from the industry and certain members of the scientific community.94

7.2. Discounting and conflicts of interest Discounting carcinogenic and other toxicity effects of pesticides based of certain biological issues is becoming a more prominent and dangerous trend.27, 95 Conspicuous rationales often claimed to minimize the relevance of a model organism response to human populations include: (1) disregarding rodent-specific organs exhibiting carcinogenicity; (2) neglecting carcinogenic responses based on hypothetical MoA discrepancy between test animals and humans;27 (3) blaming tumor responses as indirect effects of lesions or correlating fetuses malformations to maternal toxicity instead of correlating tumors and malformations directly to chemicals; or (4) simply disregarding minor tumors, rodent sensitive organs tumors; or (5) claiming absence of pertinence to human exposure when cancers are induced in laboratory animals at high exposure (reviewed in Huff 27). Beyond discounting for biological reasons, conflicts of interests cast a shadow on credibility. Placing the burden of demonstrating no unreasonable harm in the hands of those who aim to profit from the manufacturing and sale of pesticides is delicate. Competitive advantages keep industries alive, the economic imperatives of trade secrets should not preclude transparency, essential for thorough evaluations of pesticides.50 Is the tendency to discount meaningless or inappropriate results for predicting potential human risk escalating in studies sponsored by industries with vested interests?27, 85, 96 In the famous and highly controversial case of the chemical bisphenol A (BPA), a troubling 11 chemical industry sponsored publications (100% concluding no harm) kept the EPA and FDA assuming safety of bisphenol A when 109 independent publications reported significant effect at low doses.97 This specific case led to changes of perspective in regulatory toxicity, including increased transparency and use of peer-reviewed literature. Fortunately, bisphenol A is now banned in children’s baby bottles in the US and EU but BPA is still used abundantly in other products.55 Impartiality of pesticide toxicity evaluation has been questioned in the face of industry studies and reviews consistently failing to detect toxicity, while numerous independent studies demonstrate toxicity end-points and MoAs for pesticides such as ATZ79 and GLY.85 But perhaps some discrepancies may be explained by the differential toxicity of AI and formulations.

7.3. Active vs Inert ingredients The Europeans refer to active substance and adjuvants mixed in a preparation, while in a formulation the Americans combine active and inert ingredients. But the US terminology is misleading, at least for consumers, because something inert should be devoid of physical, chemical and biological activity.98 The EPA recognizes that inerts can possess biological or chemical reactivity and exert human toxicity. Moreover,

19

several inerts (>500) were once used or are presently used as AIs, an apparently arbitrary decision which fluctuates from one formulation to another; and a formulation may contain more inerts than AIs (averages of >50-86%).98 Pesticide labels do not list ‘’inerts’’ as they are CBI.99 Fortunately EU registration authorities do have access to the exact composition and toxicity of co-formulats for their risk assessment, and they list restricted substances.30 In the US though over 3000 substances, 50% of which are at least moderately risky, are used as inerts.98 But what exactly are these inerts or adjuvants? Adjuvants are added in pesticides formulations or tank mixes to aid or modify the action of an AI, or change the physical properties of a mixture.100 While AI may cost >100$/kg, adjuvants may represent only 1-3% of a herbicide cost, but enable a profitable 20-50% reduction of volume usage.100 Adjuvants are defined and may be tested with ASTM standards,70 and more recently under REACH.32 Adjuvants may have lipophilic, anionic, cationic and amphoteric properties.100 Oils, Phytobland (paraffin), solvents, polymers, salts, diluents, and plain water may all be used as adjuvants.100 Adjuvants include surfactants (surface acting agents) which help to lower the surface tension of a liquid to help wetting or spreading on the target surface.100 They usually have a hydrophobic tail and hydrophilic head, and may be variants of the same class of chemicals with varying carbon-chain ranges. Adjuvants may be qualified as wetting agents (reduce surface tension), spreading agents (increase area of a droplet), sticking agents (increase adhesion and persistence), humectants (increase drying time) and penetration agent (enhance surface absorption), etc.100 The EPA recognizes that formulated products may be more toxic than the AI alone, so formulations are generally tested independently.99 Remember that only a limited range of short-term tests are conducted on formulations, testing waivers may be granted if toxicity of a formulation can be extrapolated from AI tests, and required ecotoxicity tests may only target the species most sensitive to AI.48, 62 Beyond enhanced dermal absorption, surfactants also increase protective gloves permeability and resist washing on clothes.98

7.4. Animal welfare and ethics Reaching statistical significance for various end-points and understanding pesticides’ MoAs requires massive animal testing. EU targets minimization of vertebrate animal testing by encouraging novel and reliable alternatives and obligating data sharing among industrials to avoid test duplications.30, 32 Tier I testing of the US EDSP for 47 chemicals required 27 731 test amphibians, fish and rats, but Other Scientifically Relevant Information (OSRI) waivers saved 3325 animals. PETA and Humane Society researchers alleged inconsistencies in EPA’s review of Weight of Evidence: more OSRI were granted for positive ED evidence while negative findings required further testing.40 But isn’t this precautious knowing that safety can never be proven?

20

7.1. Proving safety and safety factors Science cannot prove safety, the best it can do is to repeatedly find no evidence of adverse effects.16, 44 Legislation guidelines however use the misleading term safety assurance. For instance, a chemical safety assessment includes assessment of human health and environmental hazards, physico-chemical properties, leading to persistence, bioaccumulation and toxicity.33 The ADI of a pesticide represents the dose of pesticide residue in food which can be [safely] consumed by humans everyday for a lifetime without appreciable risks.16, 42 It is derived from the lowest NOAEL in the most sensitive test species. A Safety Factor (SF), known as Uncertainty Factor in the EU, ranging from 1000 to 5000 is used to calculate ADI: 10X for extrapolating from animals to humans, 10X to cover individual variability in human sensitivity, and an optional 10X or more to cover special risk groups (i.e. infants) or 10-50X when toxicological data is poor or incomplete (the latter referred to as Modulating Factor MF in the US). The Reference Dose (RfD) is a US variant not exactly synonymous to ADI, as it is not defined for carcinogenic pesticides. Not all doses below the RfD are acceptable, neither are all doses above the RfD unacceptable, but the overall frequency and/or magnitude of exposure increases the probability of adverse effects on human populations. RfD = NOAEL or LOAEL/(UFxMF) and is given in mg/kg bw/d.16 The maximum Tolerated Daily Intake (TDI) in water is distinguished from the Acceptable Daily Intake (ADI) in food, as the presence of pesticides residues in drinking water serves no intended function.43 Exposure to multiple sources of various pesticides is discussed below.

7.2. Mixtures Exposure to pesticide mixture is a very sensitive aspect in defining (eco)toxicology. Pesticide mixtures are much more common than individual pesticides: 90% of the time, streams from developed US watersheds had ≥2 pesticides, and ~20% of the time ≥10 pesticides. Groundwater is less frequently contaminated by mixtures though shallow wells, 47% and 37% in agricultural and urban areas respectively, had ≥2 pesticides.101 Sediments sampled in most streams also contain mixtures of organochlorines (OC) and degradates from historic use.102 Fish from urban or agricultural/mixed land use watersheds had ≥2 pesticides in their tissues 90% and 75% of the time respectively, and ≥10 pesticides 33% and 10% of the time respectively.102 Mixtures of up to 9 pesticides have been found in amphibian habitat.103 Consequently, environmental mixtures also contaminates humans. Different organophosphate insecticide metabolites were detected in 33-83% of US pregnant women tested, and banned but persistent OC pesticides, DDE (metabolite of DDT) and hexachlorobenzene, were detected in the serum of 100% of US pregnant women tested. Of 13 OC pesticides measured, mixtures of 5, 6, 7, 8 and 9 compounds were detected in >15%, ~30%, ~20%, ˂25% and ˂5% of pregnant women respectively.104 To draw political attention on environmental mixtures, the WWF analyzed blood samples of 14 EU Ministers. On average 37 chemicals were found in each individual, with a maximum of 43; and out of the 103 looked for, 25 chemicals were found in 100% of subjects, including the OC pesticides DDE and hexachlorobenzene.105

21

The diversity of mixtures is tremendous, for instance in US agricultural watersheds where 6000 unique mixtures containing 5 different pesticides were observed. Hence, it is impossible to test all possible permutations of AIs and concentrations to assess toxicity of mixtures. Some studies confirm that concentration addition (CA) can predict the effect of pesticide mixtures with similar MoAs, and is an acceptable starting point in Cumulative Risk Assessment (CRA), but a few studies highlight enhanced toxicity of mixtures (synergy) (reviewed in refs 39, 102, 106, 107). CRA grouping criteria may include chemical structure, pesticidal MoA, mammalian toxicity MoA, and common toxic effects.39 The WHO estimates that aggregated exposure cannot be predicted reliably (in part because of potential synergism) and assumes SF provide a sufficient margin of protection in the calculation of ADI.42 However, aggregated and cumulative exposure assessments are mandatory in the US since 1996.15 In 2008, EFSA adopted a new methodology for CRA of pesticide residues on food (to establish MRLs), which may eventually be applicable to non-dietary exposures (i.e. workers and residents).39 It is based on grouping compounds with similar toxicity on target organs or systems (Cumulative Assessment Groups, CAGs); a grouping which can be achieved even when the exact underlying biochemical MoAs are not clearly understood, simply by extracting oral toxicology data from the DARs which support all EU pesticides registrations. An initial exercise on cumulative risk assessment was achieved in 2009 targeting the well studied triazoles. In 2013, the first CAGs targeted the thyroid and central nervous systems (CNS).39 Limitations of this methodology include the fact that lack of data, for example on cognitive or developmental neurotoxicity, precludes the inclusion of these effects in CAGs targeting the CNS. Eventually, all pesticides to be registered in EU will require cumulative risk assessment, and CRA should also include non-approved pesticides to which consumers are nevertheless exposed, although this cumulative risk will be restricted to pesticides and not yet expanded to other environmental or food contaminants. As for human toxicity, a multi-tiered approach, initially based on estimations of CA, with clear decision criteria (thresholds) and relying on minimal data (acute ecotoxicity with 3 species) could be used to optimize resource allocation to assess an ecosystem risk quotient (ratio between the toxicity of a pesticide mixtures and expected environmental exposure), whereby mixtures with a greater potential risk would be prioritized for further testing.106 CRA based on CA can provide interesting insight on the effect of mixtures, but further animal testing of mixtures will nevertheless be required.

7.1. Subtle effects Unfortunately, not all pesticides result in obvious target organ damage, some effect occur in the absence of clinical symptoms or obvious illnesses. For instance, chlorpyrifos insecticide is linked with developmental delays and pervasive developmental disorders (a form of autism).108 Cognitive effects (learning and memory) are not routinely assessed during neurotoxicity studies of pesticides for registration, and may only be found in a higher tier of assessment.39 Pesticides have also recently been identified as a compounding factor in obesity and type 2 diabetes epidemics,109 and critical research is required to fully understand these effects and adjust pesticide toxicity assessments and registration accordingly. In a recent past, recognition of ‘’silent’’

22

lead poisoning, producing an IQ decline of 0.5-1.0pt per 3μg per dl blood in children, was not even the root reason to ban lead additives in gasoline: lead damage to catalytic converters of cars justified EPA’s decision.55 How subtle yet widespread effects of pesticides will be addressed legally remains an open question. 8. Most widely used pesticides controversies Now that (eco)toxicity testing and legal frameworks have been defined, bearing in mind key concepts and hot topics, let’s review the current controversies over Atrazine and Glyphosate, and try to discern the moral of the story. To visualize physico-chemical and toxicological properties of both herbicides, refer to the summary Table S6.

8.1. Atrazine

8.1.1. Herbicidal effect and usage Atrazine (ATZ)(2-chloro-4-ethylamino-6-isopropylamino-s-triazine), from the Triazines family, is a herbicide to control broadleaf weeds and some grassy weeds in monocot cultures.110 It is an inhibitor of photosynthesis for most plants, but monocots (corn, sorghum) exhibit resistance due to increased metabolism via the glutathione reaction and excretion.111, 112 So while ATZ is mainly applied in fields pre-emergence, it can be directly applied to a corn field to kill the weeds but not the corn.110, 112 ATZ was the world’s best selling pesticide until GLY ravished its title around 2001,113 but it is still widely sold pesticide in 80 countries,110 with steadily increasing sales in Asia.114 ATZ’s marketing success is based on its cheap cost and high efficiency.115 However, ATZ has not been (legally) used in the EU since 2003,17 and even before that in Italy and Germany who deny yield losses, and even claim surpassing US yields.116, 117 Should triazines be banned in the US, 1.4% to 9.5% maize yield losses (depending on the region and herbicidal alternative), increased consumer prices and decreased margin of profits for agricultors are predicted by a Syngeta funded study.118 USDA’s predictions are closer to 1.2% yield losses,119 so is farmer’s health unjustifiably at stake for marginal yield gains?64, 117

8.1.2. Registration history: court litigation, public and industrial influences First appearing in the US market in 1958, ATZ was initially marketed by Syngenta (anciently CIBA then NOVARTIS), an agro-chemical company based in Switzerland.64 It was last reregistered in 2006120 and a new Reregistration Review started in 2012 (www.epa.gov120). Contrary to the US, the EU denied ATZ registration (access to Annex I) in 2001 because respect of the uniform pesticide limit (0.1 μg/l) in drinking and groundwater could not be ascertained.17, 64 Though the term ban is commonly used,64 the industry121

23

points that the EU simply claimed they had insufficient monitoring data17 and that registration could theoretically be granted should all required data be provided (www.atrazinefacts.com121). The last registration review for ATZ, and other pesticides, in the US was polemic. In 1999, environmental, consumer and public health advocacy groups filed a lawsuit against the EPA for delaying risk assessment of higher-risk chemicals (including ATZ) (Natural Resources Defense Council v. Whitman, No. C993701CAL122). In 2002, a court issued a consent Decree in the litigation extending the deadline for Review of ATZ and including an extended mandate of the EPA to bring to the FIFRA Scientific Advisory Panel further evidence concerning (1) potential association between exposure and cancer (prostate) incidence in humans, (2) amphibian endocrinology and development effects and (3) monitoring and mitigation of ecosystemic pollution of ATZ at the watershed level. To encourage transparency, the court also granted a 90-day public comment period upon submission of EPA’s Review to the registrant. The terms of the consent Decree included mandatory annual reports by the EPA to follow up on risks or tolerances characterized in Interim (IRED) and Registration Eligibility Decisions (RED), new regulations, as well as a CRAs for certain pesticides including ATZ (and compounds with a similar MoA). A Memorandum of Agreement was signed between EPA and Syngenta in 2003 to follow up on the implementation of community drinking water protection measures, and included a provision for deregistration of ATZ should quarterly reports on drinking water monitoring fail to be transmitted to the EPA. In 2003, the EPA held a public panel, due to pressure groups requesting a ban of ATZ in the US.110 But EPA nevertheless reconducted ATZ registration in 2006. With the re-registration process reopening, heated actions are rising again. In 2011, supporters of the Natural Resources Defense Council sent over 50 000 letters to the EPA and a conservation organization submitted over 10 000 signatures in a petition requesting an immediate ban on ATZ in the US.123, 124 EPA denied the petition concluding that the non-urgent matter was best dealt with under FIFRA's standard reregistration process.124 But, authors have argued that Syngenta influenced that exact risk assessment process, through submission of flawed scientific data evidencing no harm and repetitive private negotiations with the EPA.64 One of the greatest whistle-blower on ATZ’s endocrine toxicity to amphibians even allegated that 100% of Syngenta’s studies on ATZ claimed no effect, even when he could observe toxicity in two of Syngenta’s publications,79 a dismissal of significant toxicity which is corroborated by yet another team of reviewers.125 Finally, Syngenta just recently ended a decade long lawsuit through a major voluntary monetary settlement. A class-action lawsuit representing approximately 2000 communities in the Corn-Belt was filed against Syngenta in 2000 to recover costs of purifying water to remove ATZ (Greenville vs Syngenta, no. 310cv00188JPG-PMF126). Syngenta denied any liability, but agreed to a settlement of $ 105 million in 2012, thereby protecting itself from any further related lawsuits.

8.1.3. Atrazine‘s disputed toxicity The general consensus among regulatory bodies is that ATZ has only low to moderate acute toxicity,110, 127-

24

129 However, several chronic or low-dose toxicity endpoints have been observed by independent researchers, and a non-exhaustive list of examples is presented here. ATZ has been linked with prostate and mammary cancer81, 130 (reviewed in ref 24), immune system failures,114 neural damage,131 and abortion132, 133 in rodents. In multigenerational rat studies, ATZ exerted impaired mammary development in F1 females134, 135 and enlarged prostate development in F1 males136 born from pregnant mothers exposed to ATZ (though not always137). Furthermore, the F2 offspring of filial descent suffered impaired growth and development due to decreased lactation of the F1 mothers.134 In amphibians, ATZ was identified as an inhibitor of sexual differentiation,113 inducer of hermaphroditism,113, 138 and an immunosuppressant.139 ATZ has been described as a gonadotoxin across several vertebrate classes (fish, amphibians, reptiles, and mammals).140 Except perhaps for mammary cancers restricted to one strain of rats, all other observations have been the subject of controversy or some form of scientific rebuttle or regulatory dismissal.24, 110, 137, 141-143 Frogs are perhaps the organism over which most of ATZ’s reproductive toxicity debates have focused, and it can be caricatured as David against Goliath, with a whistle-blower, Hayes, against the consensual agencies and industrial establishment. Early divergence concerned male frogs (Xaenopus laevis) voice box growth inhibition at 1μg/l ATZ,113 questionably refuted by Syngenta funded field observations in corn and non corn growing regions144, 145 as reported ATZ concentrations of the two regions overlapped and spanned above and below Hayes’ threshold. ATZ was subsequently linked with hermaphroditism, testes demasculinization and partial or complete feminization in frogs, reptiles and fish at low environmental concentrations.138, 140 But deniers claimed replication of these results was impossible, perhaps as Hayes points out, because of control water contamination, high mortality rates and other major flaws.79 In response to EPA’s clarification request,110 Kloas, funded by Syngenta, directed a final experimental proof denying sexual interference of ATZ,141 or so he claimed. Under blinded conditions and GLP, two separate laboratories replicated experiments and evaluated sex ratio, mixed sex (male and female tissue in a gonad) or intersex (left and right gonads of different gender) and testicular oocytes (ovarian follicles in testicles). Their use of a positive estrogen control to finally demonstrate the frogs’ sensitivity to estrogenic induction was certainly an improvement over previous studies failing to observe gonadotoxic effects.83 However, no intersex nor testicular oocytes were observed; although in one laboratory, treated females were statistically smaller than negative controls, a slight difference dismissed as not dose-dependent and therefore random.141 Knowing that endocrine mediated effects are not necessarily monotonic,83 the authors claims appear unjustified; especially when closer observation of the data means and reported standard error in both laboratories reveal smaller ATZ treated females compared to the control, approaching male sizes. So has this study finally nailed the coffin of ATZ’s gonadotoxicity? Even the EPA doubts it.83

8.1.4. To be or not to be…carcinogenic From a concerning ‘’possible human carcinogen’’ in 1983,110 ATZ slipped to an animal carcinogen ‘’not classifiable for human carcinogenicity’’ in 199924 and now rests as ‘’not likely to be carcinogenic to humans’’,

25

based on reasonable certainty of no harm verbalized in 2006.110 Indeed ATZ induces prostate and mammary cancers in SD rats,24, 81, 130 through a putative endocrine mechanism irrelevant for humans24, 81, 110. But human epidemiology is unclear. Elevated prostate cancer incidence (1.8X, significant) in Syngenta ATZ plants workers and elevated prostate cancer mortality (2X)146 is dismissed for small sample sizes, absence of dose-dependence and intensive screening of employees compared to control populations.110 Unjustified dismissal?79 While occupational pesticide exposure correlates with prostate cancer, causality cannot be established for ATZ.147 Direct causality in breast cancer higher incidence (p˂0.0001) for women drinking ATZ contaminated well water also fails to be demonstrated,148 and subsequent follow ups drown the apparent correlation.149 Slightly elevated ovarian and thyroid cancers among 57 310 licensed pesticide applicators also correlated with ATZ, but below significance tresholds.142 In summary, trends suggesting increased cancers (lung, bladder, Non-Hodgkins Lymphoma (NHL) and multiple myeloma;150 ovarian and thyroid142) may be considered preliminary and methodologically insufficient to conclude on human carcinogenicity of ATZ, but warrant further studies.110 Two extensive recent reviews insist: ATZ does not bear a carcinogenic risk for humans, neither from an animal toxicology standpoint nor from epidemiological data.143, 151 Can we once and for all consider ATZ non-carcinogenic for humans? Perhaps not quite yet. A recent epidemiological study points to a novel possible interaction between nitrate and ATZ correlating with 2.5 times greater chances of NHL.152 Neither nitrate alone nor ATZ alone increased the odds of NHL, but their combination in drinking water could be linked with formation of N-nitrosoatrazine in vivo, and nitrosamines are known carcinogens. So would the controversial carcinogenicity of ATZ hide in mixture effects?

8.1.5. Endocrine disruption as a Mode of Action Endocrine disruption is the postulated MoA in putative carcinogenesis and reproductive toxicity of ATZ.127, 139 ED were not an ecotoxicity regulatory endpoint during EPA’s last review of ATZ, and available evidence was insufficient to draw conclusions on ED.110 EPA’s Endocrine Substance Screening Program launched in 2009 granted almost all testing waivers for OSRI requested by the industry, precluding the necessity to search for novel evidence.40 A recent review of low-doses responses and non-monotonic dose-response relationships, which includes examples on ATZ, Bisphenol A and Dioxins, highlights that important changes in toxicity and safety regulatory assessments are necessary to adequately protect human health.83 A report from Australia’s pesticide regulation authority reviewed the current evidence on 3 ATZ endocrine MoA summarized below.127 (1) In SD rats, ATZ affects luitenising hormone (LH) which acts on prolactin in mammary tissues to induce mammary tumors, and suppresses ovulation; a mechanism apparently related to premature ageing of SD female rats unlikely related to human epidemiological data.24, 81, 110 (2) In the brain, ATZ may influence the hypothalamus and pituitary, depressing the production of gonadotropin releasing hormones (GnH).153 The hypothalamus-pituitary-adrenal axis, can influence the hypothalamus-pituitary-gonadal axis, and plausibly lead to reproductive toxicity, i.e. full litter resorption or abortions in test animals;132, 133 not yet evaluated for human relevance.127 (3) ATZ decreases androgens,154 and increases estrogens140 possibly by activating the

26

aromatase enzyme, responsible for the conversion of testosterone in estrogen (in reptiles155, mammals156, 157, fish156 and amphibians138) though aromatase induction is disputed (see ref 141 and references therein). But ATZ, other triazines, and primary metabolites of ATZ were shown to activate the aromatase enzyme expression in human cell lines;157 and increases in ovarian and non-ovarian estrogen production could be important in estrogen dependent human breast cancer.158 Absence of relevant benchmark studies and limitations to in vitro studies, preclude assessments on relevance for human toxicity.127

8.1.6. Monitoring, mitigation, and mixture of unknown toxicity ATZ is moderately110 water soluble. It resists abiotic hydrolysis and aqueous photolysis, and biodegrades moderately or slowly under aerobic and anaerobic aqueous conditions respectively.110 Its’ low adsorption to soil particles further compounds its high leaching potential.110 Despite low volatility, ≤14% may volatilize after field application under conventional tillage159 and may redeposit with rain some 500 km further.160 ATZ is the most commonly detected pesticide in agricultural streams (±90%), and ATZ and DEA are the two most common pesticides in groundwater (±40%), directly correlating with corn growing regions.102 The Maximum Concentration Limit (MCL) was recently increased from 2 and 3 μg/l by the WHO and US respectively, to 100 μg/l;110, 128 but maintained at 0.1 μg/l in the EU.65 There was a ≥5% chance of exceeding the historic MCL of 3 μg/l annual average in streams around the US Corn Belt; and nationwide, an average of 1% of all ATZ applied to land will reach the exit of the watershed through flowing streams.102 Remember that a court ordered the development of an ATZ environmental monitoring program from 2004 to 2006 as a pre-requisite for reregistration. As a part of this program, it was established that if ATZ exceeded a Level of Concern (LOC) of 10 ug/l over a 60 days period, implementation of mitigation measures by the registrant would be necessary. The LOC was established based on 5% changes in aquatic plant communities established through 25 micro and mesocosm studies. High priority watersheds (40 corn and sorghum growing regions spanning 10 states) were monitored every 4 days during the growing season, while a subset of 10 were monitored daily throughout the year (to establish seasonal fluctuations). Critiques have argued that monitoring 40 watersheds among the >10 000 watersheds at risk for ATZ contamination is insufficient.64 By 2011, EPA had added 25 sites (www.epa.gov120). Despite these monitoring efforts and Syngenta’s watershed stewardship and outreach activities, some sites regularly exceeded the LOC (i.e. Big Blue River, 5 of 6 years; www.epa.gov120). However, despite the abovementioned monitoring efforts, ATZ’s registration did not extensively consider the complicated impact of pesticide mixtures and interactions with environmental conditions.110 This is not surprising. In fact, 80% of studies on pesticides aquatic ecotoxicity are concerned with single pesticides161 and most studies focus on gross toxicological effects (lethality, external malformations, etc.) at relatively high doses (mg/l), even though aquatic ecotoxicity may be exhibited at lower doses (μg/l) and through more subtle

27

effects.139 ATZ was identified in 30% of all pesticides mixtures found in agricultural surface and groundwater, and ATZ frequently exceeded ecotoxicity benchmarks (vascular and non-vascular plants acute, aquatic community effects and invertebrate chronic).102 Much of the debate in the ATZ’s toxicity controversy has revolved around freshwater vertebrates.125 Pesticide mixtures, including ATZ, are shown to exert a complex effect on frogs139. In mesocosm experiments, the herbicide ATZ (20 μg/l) was not more toxic to Hyla versicolor frogs tadpoles when combined to Carbaryl (2.5 mg/l) insecticide, but the presence of nitrate (10mg/L) with ATZ increased survival through metamorphosis.162 Changes in the food-web or chemical interactions may thus be important in masking the effect of ATZ.162 Invertebrates may also be affected by mixtures containing ATZ. In a behavioural study, ATZ alone did not have any effect on chironomids (aquatic invertebrate) immobility even at 200 μg/l, but a fixed quantity of Chlorpyrifos insecticide (0.17μg/l) synergistically increased ATZ’s effect, augmenting the percent of time spent immobile in a dose-dependent fashion (40-200 μg/l); conversely ATZ acted as a synergist to Diazinon insecticide increasing its’ toxicity by up to 400% for midges and amphipods (reviewed in ref102). In a laboratory experiment on X. laevis, ATZ’s growth and development retardation effects were increased in mixtures with S-metolachlor herbicide, which had no effect on its own (acts as an effector). But Bicep II Magnum which contains both S-metolachlor and ATZ plus added surfactants appeared less toxic. Growth and development were further reduced in complex mixtures of 9 pesticides containing ATZ.139 This is significant because delays to metamorphosis can mean drying of temporary ponds before adulthood, and smaller size may equate ability to catch only smaller preys, increased possibility of being eaten by a predator and reduced reproductive fitness (bigger females produce more eggs and bigger males are preferred mates)(reviewed in ref139). Furthermore, although controls carried flavobacteria asymtomatically, frogs exposed to mixtures exhibited an unexpected increased disease rate, which correlated with damages to the thymus and elevated corticosterone levels in the 9 pesticide mix, indicators of ED at low concentrations of pesticide mixtures.139 In a meta-analysis of freshwater fish and amphibians, ATZ exposure resulted in a 77% decrease in immunity end points assessed, indifferently for the AI both alone (21/27) or in mixtures (12/16). Concerning susceptibility to trematode, nematode, viruses or bacteria though, ATZ in mixtures (9/9) slightly increased infection endpoints compared to ATZ alone (4/7). However, interpretation is not always obvious as higher concentrations of ATZ could be damaging to viruses and trematodes while ecologically common concentrations could increase amphibian susceptibility125. Clearly, more studies are needed on the effect of pesticides mixtures on complex communities and this may affect pesticide regulations.161, 162 This controversial portrait of ATZ can only be concluded with one undisputable statement over reproductive toxicity, carcinogenicity and ecotoxicity of ATZ: ‘’While high uncertainty may obscure both the probability of a risk and the magnitude of harm, uncertainty does not eliminate risk. Unrecognized risks are still risks; uncertain risks are still risks; and denied risks are still risks.’’ 44 Can more conflicts be resolved for GLY, ATZ

28

biggest competitor?

8.2. Glyphosate A portrait of the world leading herbicide (1) usage and (2) registration status will anchor current polemics about Glyphosate. Environmental controversies reviewed here include (3) its persistence in soils, (4) effect on plants, (5) altered water quality and aquatic biota. The health controversies of GLY reviewed here relate to (6) general human toxicity; (7) reproductive toxicity and endocrine disruption; and (8) carcinogenicity.

8.2.1. Herbicidal effect and usage Glyphosate (GLY), better known under its commercial denomination of Round-Up is a non selective herbicide frequently used in genetically modified cultures (Roundup Ready) such as soy and maize. It’s main herbicidal MoA is through inhibition of the ESPS enzyme (5-enolpyruvyl-shikimate-3-phosphate synthase), thereby blocking the synthesis of aromatic amino acids.163 The pathway for biosynthesis of aromatic amino acids is indeed not expressed in any other kingdom than that of plants, making this mechanism somewhat exclusive,164 but some fungi and microbes also use this pathway,165 and such claims cannot be used to imply safety for other organisms as they may be impacted via a different MoA. The various forms of GLY used include free acid form as well as various salt forms (sodium, potassium, isopropylamine, mono- and di-ammonium, ethanolamine) all of which are active in the dissociate acid form, such that acid equivalents (ae) represent a common denominator for application rates comparisons.99 More than 400 formulations are registered in the US with authorized application on more than 100 food crops.166 From 2004 to 2011, the US crops that globally received the largest quantities of GLY were soybean (˂40 000 mt AI), corn (˂25 000 mt AI) and cotton (˂ 8 000 mt AI) (average per year).166 The most extensively treated crops are soybeans (95%), oranges (90%), cotton, pistachios, almonds and grapefruits (80%).166 The environmental (IRE) and health risk (IRS) of the active substance and formulations is relatively low, making it a preferred choice among agricultors (www.sagepesticides.qc.ca75).

8.2.2. Registration history GLY was first patented by Monsanto and registered in the US in 1974.167 Following patent expiry, other companies (Syngenta, Cheminova, BASF, Dow, etc.) now market GLY based pesticides.167 The last registration of GLY was granted in 1993, and the US EPA started GLY’s registration review in 2009 (industry dossier, identification of data gaps and data call-in process), will be expecting public comments on the Proposed Review Decision in 2014 and should finalize its decision in 2015.167 The Pest Management Regulatory Agency of Health Canada is also currently reviewing GLY registration, and is coordinating with the US EPA to harmonize assessments and scheduling.167 In 1998, Germany, as EU’s appointed rapporteur state, submitted a draft assessment report on GLY. EU approved GLY inclusion in Annex 1 in 2002,168 and

29

was due to revise approval in 2012, but this was delayed until 2015.169 With both the US and EU approaching their re-registration deadlines, there is currently a surge of activity concerning GLY’s benefits and hazards in the literature.

8.2.3. Glyphosate persistence in soil GLY is generally sprayed on foliage, but a fraction may be directly deposited on soil or carried by the wind to adjacent water bodies.170 It is quite soluble in water but not very volatile.99 GLY has a strong tendency to sorb to soil particles, and because GLY is less negatively charged at lower pH,171 it binds more to soil particles in acid soils than in alkaline soils, in fact pH is the best predictor for soil sorption.172, 173 Adsorption of GLY to variable-charge surfaces of soil particles is mediated via the phosphonate moiety,171 and is related to the content of aluminum and iron oxides, aluminium silicate, goethite and to a lesser extent soil organic matter.165, 171, 173 The initial soil concentration of inorganic phosphate has also been been regarded as important for sorption,174 and in certain soils, phosphate fertilization can induce resolubilization of GLY, making it available for plant reabsorbtion, leaching or microbial degradation.175, 176 However, the very existence of competitive interaction between PO43- and GLY is questioned, and minimized in its importance and duration in a recent review.165 Degradation rate depends on microbial activity, soil type, pH and temperature176 and it is primarily a microbiological process.173 GLY degrades in the soil via the sarcosine (non-toxic) or AMPA (toxic) pathways.173 The persistence of GLY in soil is a controversial subject in itself. Laboratory studies, with optimal and perfectly controlled conditions generally measure shorter half life than field studies which may yield extremely variable dissipation time (DT50) results ranging from 1.2 to 197.3 (reviewed in ref177) or 9 days176 to 151 days,178 reviewed in Duke et al.165 Even after 748 days, 59% of GLY (and AMPA) remained in a clay soil despite heavy precipitation.178 On railway embankments, DT50 values between 2 and 5 months were common, and 8 months was the maximum. However, DT90 was 1 to 2 years.179 Colder climates have longer persistence times.165

8.2.4. Glyphosate effect on plants GLY is little degraded in plants, but may be metabolized to AMPA, 180-182 and it can be released in the soil via root exudates183 or upon senescence and death of the plant.184 In its 1993 RED, the EPA had concluded that aquatic plants were expected to be at risk under certain conditions, but information was insufficient for non-target terrestrial plants, though risk was not dismissed for endangered species.99 Since then, it was demonstrated that GLY affects indigenous plant populations185 and sometimes threatens vulnerable plants.186 Even in agricultural settings, GLY may have some deleterial effects. GLY influences key functions of the rhizosphere, by reducing root nodulation necessary for nitrogen fixation in legumes and reducing the

30

bioavailability of essential micronutrients.187, 188 Even Round-up Ready soy could be inhibited by AMPA,182 leading to a transient yellow flash occurring shortly after GLY application not to be confused with mineral deficiencies according to a recent review by Duke et al.165 In fact, alleged mineral deficiencies due to the chelating power of GLY have been criticized as unlikely due to the much larger concentration of mineral nutrients in soil and soil solution as opposed to the concentration of applied GLY.165 However, this reductionist bulk approach does not address heterogeneous interactions occurring in the rhizosphere. By reducing interactions between antagonistic bacteria and phytopathogenic ones,187, 188 yields and culture vulnerability to diseases could also be affected,189 although this mechanism is also critiqued.190 The appearance of resistant weed now encourages applications of higher doses of GLY191-193 and a return to previously reduced or abandoned pesticides.194, 195 In conclusion, weed resistance is accepted as an important issue, but mineral deficiencies of plants and vulnerability to diseases seems to polarize the scientific community into believers and deniers and conflicting conclusions based on Weight of Evidence have been drawn by both groups.165, 187, 189

8.2.5. Glyphosate alters water quality and endangers aquatic biota Despite being described as a rapidly biodegradable and relatively low mobility pesticide,196 a fraction of GLY invariably enters water bodies, often at concentrations superior to MCL.197 Leaching potential depends on soil type and precipitations.173 Typically, less than 1% of the GLY applied to a land will enter surface water.198 However, surface contamination is widespread, 90.6% of samples in streams of agricultural regions of Quebec (Canada) have measurable concentrations of GLY.199 Ground water contamination is generally estimated to be less likely, and less important with GLY compared to other herbicides,180 however, concentrations above EU standards (0.1 μg/l) have been observed in tile drains of certain soils where measured leaching represented 0.12% of the annual load of GLY (reviewed in ref 200). In the US, GLY has also been detected in groundwater, though less frequently than in surface waters.201 In water, Glyhosate has a half life of 7-14 days,177 up to 91 days in other sources.202 Under aerobic and anaerobic conditions, GLY DT50 equals 14 and 248 days respectively in water and sediments systems.99 Globally, the ecological risk of GLY was found to be minimal for mammals, birds and aquatic biota, and the toxicity of AMPA is assumed to be even lower.99 It has been said that the quality of vulnerable watersheds may improve if ATZ was replaced by GLY;177 but the biodiversity and productivity of aquatic communities is also affected by GLY203, notably amphibians204 and phytoplankton,205 even at doses below the chronic aquatic life criteria (65μg/l)(Smedbol et al. Unpublished). The Houston Toad and California Red Legged Frog may be at risk from specific GLY usages, but GLY is only practically non-toxic to moderately acutely toxic to fishes according to the EPA, except for formulations with POEA which are highly toxic to the rainbow trout.99 Further acute avian toxicity and acute marine organisms toxicity is required by the EPA for the upcoming registration review.99

31

GLY, a highly water soluble pesticide, does not properly wet hydrophobic surfaces of plants covered in a waxy layer. Inerts used in the formulations of GLY all have inherent toxicity, some of which are considered severe or extreme: ammonium sulfate, benzisothiazolone, 3-iodo-2-propynyl butylcarbamate (IPBC), isobutane, methyl pyrrolidinone, pelargonic acid, polyethoxylated tallowamine (POEA), potassium hydroxide, sodium sulfite, sorbic acid and Isopropylamine (see review by Cox87 and references therein). Roundup is more toxic than GLY in fish, amphibians and microorganisms,98 as well as in human cell cultures,206, 207 and is at the root of controversies. However, except for POEA, the EPA has little data on other surfactants, and some other non-POEA surfactants appear more toxic than GLY alone.99

8.2.6. Is Glyphosate toxic or safe for humans? According to the EPA, GLY has low acute toxicity via oral, dermal and inhalation exposure routes, it is not carcinogenic, not toxic for reproduction or development and not neurotoxic.76 Based on SAR, AMPA and N-acetyl-AMPA (plant metabolite) are considered equally toxic to GLY, but are currently not regulated in food and feeds, regardless of their residue levels (this may be revisited during registration review).76 However, several studies published since the precedent EPA registration review (1993) have revealed various types of previously unrecognized toxicities. These studies are regarded as marginal contrarian science, but some of them led to impedance (resistance by industrials) and scientific dissent (rebuttals by contrarian scientists). GLY and its technical formulations including adjuvants and surfactants are suspected to perturb the endocrine system,208, 209 to be carcinogenic210-212 (genotoxic211, 212 and mutagenic213), to increase the risk of spontaneous abortions208, 214 or congenital diformities215 and to affect the nervous system216 of both man and animals. For the regulatory decision on GLY, no human toxicity studies were used or relied upon.76 GLY has been involved in 289 accidents between 2002 and 2008 in the US, including 10 children, and the most commonly reported adverse health effects were neurological (34.2%), dermal (30.1%), ocular (13.8%) and upper-respitatory (10.3%).76 But the EPA does not considers GLY as neurotoxic (further studies called-in) and does not require inhalation studies.76 GLY has been detected in home of agricultors,217 in the urine agricultural neighbourhood families,218 and residue has been found in food.98 As no common toxicity MoA for GLY and other pesticides has been identified, cumulative risk assessments are not currently conducted by the EPA, and a novel residential exposure study will eventually allow aggregate exposure risk.76 To address concerns over neurotoxicity and immunotoxicity, the EPA has estimated that limited assessment via routine developmental or reproductive studies could be refined with a specific protocol allowing detection of more subtle effects, and has called-in appropriate research data.167

8.2.7. Reproductive toxicity and endocrine disruption Both the EPA76 and EU219 agree that GLY is not significantly toxic for human reproduction or development. The US is so convinced that the SF to address the potential enhanced sensitivity of infants and children was

32

decreased from 10X to 1X in 1998, based on no evidence of increased susceptibility in juvenile rats and rabbits.76 However, peer-reviewed publications highlight increased risk of spontaneous abortions208, 214 or congenital diformities.85, 215 Weren’t these effects evidenced in registration toxicity studies? Antoniou and colleagues85 argue that teratogenicity evidence in industry supported studies has been minimized and dismissed by Germany, EU rapporteur’s state for GLY. Several studies sponsored by GLY manufacturers have showed trends (linear dose-response relationships) or statistically significant evidence of teratogenicity in rabbits and rats, such as thoracic, craniofacial, heart, lung and other visceral malformations, including reduced ossification, as well as death and resorption of embryos and fetuses.85 However, both the industry and Germany discounted these results by claiming: (1) absence of dose-response relationships; (2) natural variability upon comparison with historical control data; (3) ‘’equivocal’’ observations; (4) differential effect of gavage compared to ad libitum feeding; (4) developmental variations rather than malformation; and (5) developmental retardation that would correct themselves briefly (consult refs 219, 220 and references therein). Antoniou et al.85 counter argued that: (1) several of Germany’s retained NOAEL and LOAEL were unfounded; (2) absence of ‘’other soft organ malformation’’ was invalid to dismiss heart malformations because toxicity can be organ-specific; (3) lacking dose-response relationship could really be found in the data even though a linear dose-response relationship is not necessarily observed in endocrine-mediated effects; (4) gavage versus oral feeding differential effects are not defensible; (5) defining unossified sternebrae as developmental variation is scientifically unjustifiable as it is a recognized skeletal deformity;49 (6) eventual correction of developmental delay is not supported by evidence; and (7) the use of historical control data is of lower relevance than concurrent controls, and minimally the unpublished historical control data should be made available for review. The ADI should be established based on the most sensitive species in the most appropriate study.219 However, Antoniou and colleagues85 observed that Germany dismissed the lowest NOAEL found in a sub-chronic toxicity study, alleging that long-term toxicity studies are superior to determine safe chronic levels of exposure, and Germany’s general claim that rats are more sensitive indicators than rabbits, despite evidence of the contrary, are unfounded and led to the establishment of an erroneous ADI. Instead of the EU accepted ADI of 0.3 mg/kg bw/d, Antoniou pleaded85 that it should have been set at 0.1 mg/kg bw/d if the lowest LOAEL from the sub-chronic rabbit study sponsored by the industry had been retained (Suresh221 unpublished industry study reported in ref 219, which is discounted by Kimmel and colleagues222). Conservative inference of NOAEL (2.5 mg/kg bw/d) based on peer-reviewed studies (LOAEL ±5mg/kg in Romano209 and Benedetti,223 with the customary 100-fold safety factor, would have resulted in an ADI twelve-fold lower (0.025 mg/kg bw/d). Monsanto, Dow and Syngenta jointly published a letter claiming that even at high doses, GLY does not cause adverse reproductive effects in adult animals not birth defects in offspring of exposed adults.89 More recently, the occurrence of cardiovascular malformations in offspring of exposed rabbits and rats was reviewed222 as a rebuttal addressed to Antoniou.85 Kimmel and colleagues forwardly disclosed funding and ties with the

33

industry.222 The unpublished data they analyzed was quite old (1980 to 1996 for rabbits and 1980 to 2002 for rats), arguing that more recent peer-reviewed publications were not located. Despite major differences in the reviewed tests (3 and 2 strains of rabbits and rats respectively, doses tested, duration of studies, number of animals per group and most likely type of diet due to the large time interval and varying location of studies), the authors saw it fit to pool all the rabbits and rats data in distinct databases to review the overall occurrence of heart malformations. Expectedly, the strong background noise of pooled controls (akin to usage of historical control data instead of concurrent controls), made it impossible to distinguish statistically significant effects. Historical control data should only be used if experimental conditions are comparable (same laboratory, same strain, time proximity).36 Furthermore, when removing doses which led to maternal toxicity (≥300 mg/kg/d), only 0.1% of rabbit fetuses (2/1388) revealed cardiovascular malformations upon maternal exposure. But Antoniou95 had warned us that discounting foetal effects based on the confounding effects of maternal toxicity is not precautionary. At maternally toxic doses, whether toxicity is a direct effect of GLY exposure, an indirect effect of maternal toxicity, or even both, may be mechanistically important, but it does not negate the absolute observation of foetotoxicity. It merely segregates the definition of feotoxicity’s relative exclusivity to doses below maternal NOAEL.224 This exclusivity, as a disqualifying criterion of foetotoxicity for pesticide registration, is critical for industrials as foetotoxic pesticides cannot be sold, for example, in Brazil,225 and require consumer discouraging risk labeling in the EU (‘’may impair fertility’’ or ‘’may cause harm to unborn child’’).77 Hence, focus should be directed on discerning the direct vs. indirect mechanisms without discounting the importance of observed effects,224 and this once again warrants further studies. The reproductive toxicity of GLY is possibly related to endocrine disruption.209 GLY was first recognized as a potential endocrine disruptor in a 2000 study on mouse cell cultures where it was shown to interfere with aromatase and induce a decrease of the expression of the Steroidogenic Acute Regulatory Protein (STAR).226 In 2005, the Séralini laboratory published another study showing disruption of the aromatase enzyme at low non-toxic concentrations of GLY, this time in human placental cell lines206 followed in 2007 by a study on human embryonic and placental cell lines.207 In 2010, Romano et al. suggested that Roundup is a potent in vivo ED, decreasing testosterone production and altering testis morphology in rats.209 Though the EPA does not currently regard GLY as an ED, the EPA Endocrine Disrupters Screening Program (EDSP) has targeted GLY and will include screening on both AI and ‘’inert’’ ingredients for estrogenic, androgenic and thyroid hormone systems effects.76

8.2.8. Carcinogenicity GLY is classified as a chemical with no evidence of human carcinogenicity in the US based on lack of convincing evidence in adequate studies in mice and rats and humans.227 In the Agricultural Health Study (AHS), a prospective epidemiologic study launched in 1993 which had studied 89 658 people as of 2005, incidence of cancers (or cancer subtypes) was not linked to GLY exposure, except for a suggested association with a few cases of multiple myeloma which should motivate further research.228 A novel

34

publication confirms that the current weight of evidence does not suggest genotoxicity of GLY AI and formulations.229 Why not make waves in this apparent calm sea of consensus, tip our smoothly sailing vessel and dive in the recent controversy of GLY carcinogenicity? An important paper has been at the heart of one of the most visible controversies in the history of pesticides toxicity. A team of researchers investigated the toxicity of both genetically modified corn and a GLY AI and formulation on rats. The study entitled “Long term toxicity of a Roundup herbicide and a Roundup-tolerant genetically modified maize” was published in Food and Chemical Toxicology in September 2012.230 Séralini and colleagues reported earlier death, mammary tumors, disabling of the pituitary gland and endocrine disruption, hepatotoxicity and nephrotoxicity.230 Soon after its publication, it has been the subject of intense scientific debate over the validity of its findings. Finally, after being cited 137 times in the scientific literature and countless times in the popular media (see examples of both pros231-234 and cons235-238) and being the subject of a publication rebuttle by the original authors,82 Elsevier decided to retract the publication following a 15 months re-evaluation of the raw data and peer-review process. While the publisher states that no fraud or intentional misinterpretation of the data was found, and commands the corresponding author for his willingness and openness to participate in the editor’s investigation, Elsevier concludes that due to the low number of rats used in the study (10 per group) and the strain used (Sprague-Dawley) no conclusion on the incidence of tumors or overall mortality can be linked to the NK603 corn or GLY.94 SD rats, though sensitive carcinogenicity subjects,78, 82 may be overly sensitive to study mammary tumors16, 80 (revisit model organisms section above for details). Some authors argued that the number of test animals was justifiable according to explicated goals of the study and the effective OECD guidelines when the experiment began,232, 239 bearing in mind that industrials demonstrating the safety of GM corn used a similar number of animals for histopathologic and biochemical analyses.240 Both support232 and criticism241 of the statistical approach used have been verbalized. Indeed, the limited number of test animals is a serious concerns,82 put perhaps this study should simply be considered as an interesting first step to encourage further statistically powerful research. Some even argue that by dismissing Sérallini’s study, EFSA defended its earlier conflictual decision on the safety of GM maize.242 The publisher retraction94 may very well be seen as suppression of adverse findings,231 or censorship, because it arrived at a timely moment in GLY registration review process in the US and EU. 9. Conclusion: Controversies and resolutions Controversies abound in the world of toxicology. Thalidomide, a nausea medication given to pregnant women which led to diformities in newborns, was certainly one of the earliest pharmaceutical controversy243 which profoundly impacted toxicological assessments. In the world of food additives, it is worth mentioning the controversy over aspartame’s carcinogenic potential at half of the current ADI based on a novel testing protocol244 which has been deemed methodologically flawed, invalid and uncomparable to standardized

35

protocols by the US FDA and EFSA.36 In the chemical domain, Bisphenol A’s endocrine disruption potential was certainly one of the pivotal point in rethinking toxicity assessments.97 Finally, the world of pesticides, designed to kill by industrials who proselytize about their safety and dogmatize their unsubstitutability in modern agriculture,27 is unsurprisingly the scene of several lasting controversies. The earliest and most famous controversy on DDT, exposed to the public in the Silent Spring novel245 was the precursor of several other controversies culminating with ATZ and GLY described in this review. We could have explicated countless other controversies: Genetically modified Bt corn or cotton harming useful lady beetles and green lacewing larvae;246 Neonicotinoids which made a flash appearance on the market and now coat 100% of the seeds sown in Quebec (Canada)9 and which are suspected to affect bees critical for human nutrition,247 etc. But simply listing controversies does not enlighten us, we also need to understand their roots. While polarizing the debate is not a way to solve controversies44 over agbiotech, severe asymmetry in politics, economy and institutions engaged in controversies precludes perfect symmetry in the analysis of the heterogeneity of dissent.248 Dissent underlines the minority report of contrarian science to dominant scientific trajectories, and it is stronger than a benign disagreement.248 ‘’When scientists and other interested parties challenge contrarian science [impedance], the first sparks of controversy appears.’’ 248 Without bias in favor of proponents or opponents of current pesticides (eco)toxicity polemics, it is important to underline immoderate activities of involved stakeholders trying to rally public opinion or silence inconvenient evidence. There is a history of assail on studies and whistle-blower researchers that report risk,79, 82, 246 which is sometimes dubiously orchestrated by public relation firms hired by pesticide manufacturers,248 and sometimes even escalating to psychological threats and physical violence.248, 249 The popular media may fail to balance criticism on a research;235-237 although balanced arguments are published in other media.250 Media reporting may be misleading or untruthful, for example when blaming Séralini230 for poor experimental design, but failing to report that the design was strongly inspired from an industry sponsored study.235-237, 240,

251 Exceptional measures such as editorial notes to withdraw support on original publications252 (concerning Quist & Chapela253 findings of transgenic DNA introgressed in traditional maize reserves in Mexico) or plain retraction of articles (Elsevier94 on Séralini et al.254) remind us of the importance of research accessibility and credibility. Regulatory authorities and governments have also been pointed for consistently dismissing or minimizing evidence of risk85 or even changing toxicity requirements to circumvent inconvenient evidence.246 Finally, in trying to rally the public opinion (in the face of important power differentials248) against potential danger of pesticides, alleged fraudulent intentions of industrials and blindness of regulators, some authors254,

255 or media citations of author’s work250 instill fear and dissipate trust of the public in a system which, despite evident limitations, has for primary purpose the protection of the people and the environment. Toxicity testing depending on pesticide sponsor companies bears inherent transparency and credibility risks. Protection of CBI is essential, but should never compromise health and safety of workers or consumers, nor

36

the protection of the environment.64 Toxicity testing sponsored by industry (financial responsibility) in independent or government laboratories (scientific credibility), with public disclosure of results (minus essential limitations of CBI), is one of the proposed solutions.85, 242, 256 Transparency is recognized as critical in the process of toxicity assessment and toxicity management, as it is one of the aspect which can lead to credibility and confidence.242, 256 Publication of supplemental information on the web is believed to be one way of increasing transparency, and so is publication of negative results from both industry and academic researchers.256 However, increased access to raw data is a challenge when it comes to protection of proprietary information, protection of ongoing research for future publication and excessive or faulty reinterpretation of original data. Time has come to stop shooting at the messengers or dogmatically denying novel findings, and start to build an ethical framework around toxicity risk assessment to ensure that the people and the planet are preserved from the putative toxic threats of pesticides.

Acknowledgements Thank you to M. M. Lucotte, M. M. Labrecque, Y. Gélinas and M. Larocque for insightful comments in reviewing this manuscript. Scholarship and Strategic grant by the National Science and Engineering Research Council of Canada.

Supporting Information Tables S1 to S6 in Supporting Information detail pesticide toxicity tests required in the US and EU; Fate and environmental behavior testing requirements for pesticides; Ecotoxicity testing requirements on various organisms for the active substance or formulation of pesticides; Recommended Classification of Pesticides by Hazard (WHO) and Globally Harmonized System of Classification (GHS) and Labeling of Chemicals (UN) for acute toxicity; Globally Harmonized System of Classification (GHS) acute and chronic ecotoxicity for aquatic biota; and Important physico-chemical properties of Atrazine and Glyphosate and related (eco)toxicity considerations.

37

10. References 1. EPA, Pesticide Industry Sales and Usage 2006 and 2007 Market Estimates. In Office of Pesticide Programs: 2011; p 41. 2. FAO, FAOSTAT. UN. faostat.fao.org (Accessed 2013-11-15). 3. OECD, Compendium of Agri-environmental indicators. In OECD Publishing: 2013; p 185. 4. FAO, FAOSTAT. UN Active ingredient use in Arable Land & Permanent Crops. FAOSTAT. UN. http://faostat.fao.org (Accessed 2013-11-15) 5. British Medical Association, The BMA guide to Pesticides, Chemicals and Health. 2nd ed.; Edward Arnold: London, 1992; p 215. 6. Eurostat and European Comission The Use of Plant Protection Products in the EU Data 1992-2003. Nadin, P. In Eurostat Statistical Books.

http://epp.eurostat.ec.europa.eu/cache/ITY_OFFPUB/KS-76-06-669/EN/KS-76-06-669-EN.PDF (Accessed 2013-11-15; p215), 7. OECD, Environmental Performance of Agriculture in OECD Countries Since 1990. OECD Publishing: 2008; Vol. 4, p 210. 8. Brimble, S.; Bacchus, P.; Caux, P.-Y., Pesticide Utilization in Canada: A compilation of Current Sales and Use Data Env.Can. 2005; p 144. 9. Gorse, I.; Balg, C., Bilan des ventes de pesticides au Québec pour l'année 2009 In MDDEP DSE, Gouv. Québec: 2012, p.82. 10. Health Canada, Rapport concernant les ventes de produits antiparasitaires en 2007 et 2008. In ARLA. Gouv. Canada: 2011; p 33. 11. OECD Survey on Countries' Approaches to the Collection and Use of Agricultural Pesticides Sales and Usage Data. Survey Results;

ENV/JM/MONO(2009)37; Paris, 2009-10-23, 2009; p 34. 12. Québec, Loi sur l'accès aux documents des organismes publics et sur la protection des renseignements personnels. In Éditeur officiel du

Québec ed.; Publications Québec: 2011; Vol. L.R.Q. A-2.1, p 56. 13. Fenner, K.; Canonica, S.; Wackett, L. P.; Elsner, M., Evaluating Pesticide Degradation in the Environment: Blind Spots and Emerging

Opportunities. Science 2013, 341, (6147), 752-758. 14. Sternerson, J. H. V., Chemical Pesticides Mode of Action and Toxicology. CRC Press: Boca Raton, 2004; p 276. 15. Marrs, T. C.; Ballantyne, B., Pesticide Toxicology and International Regulation. John Wiley & Sons, West Sussex, England, 2004; p 554. 16. OECD, Guidance notes for analysis and evaluation of chronic toxicity and carcinogenicity studies. OECD Publishing: Paris, 2002; p 112. 17. EC Review report for the active substance atrazine; SANCO/10496/2003-final; Health & Cons Prot Dir.Gen.: 2003-09-10, 2003; p 4. 18. Bassil, K. L.; Vakil, C.; Sanborn, M.; Cancer health effects of pesticides: Systematic review. Can. Fam. Physician 2007, 53, (10), 1704-1711. 19. Kohler, H.-R.; Triebskorn, R., Wildlife Ecotoxicology of Pesticides: Can We Track Effects to the Population Level and Beyond? Science

2013, 341, (6147), 759-765. 20. Pimentel, D.; Stachow, U.; Takacs, D. A.; Brubaker, H. W.; Dumas, A. R.; Meaney, J. J.; O'Neil, J. A. S.; Onsi, D. E.; Corzilius, D. B.,

Conserving Biological Diversity in Agricultural/Forestry Systems. BioScience 1992, 42, (5), 354-362. 21. Simeonov, L. I.; Macaev, F. Z.; Simeonova, B. G.; Mrema, E.; Rubino, F.; Colosio, C., Obsolete Pesticides - A Threat to Environment,

Biodiversity and Human Health. In Environmental Security Assessment and Management of Obsolete Pesticides in Southeast Europe, Simeonov, L. I.; Macaev, F. Z.; Simeonova, B. G., Eds. Springer Netherlands: Dordrecht, The Netherlands, 2013; pp 1-21.

22. Popp, J.; Petö, K.; Nagy, J., Pesticide productivity and food security. A review. Agron. for Sust. Dev. 2013, 33, (1), 243-255. 23. Hughes, W. W., Essentials of Environmental Toxicology - The Effects of Environmentally Hazardous Substances on Human Health. Taylor &

Francis: Philadelphia, 1996; p 176. 24. IARC, Some Chemicals that Cause Tumours of the Kidney or Urinary Bladder in Rodents and Some Other Substances - Atrazine. IARC

Monogr Eval Carcinogen Risk Hum 1999, 73, 59-113. 25. Piersma, A. H.; Bosgra, S.; van Duursen, M. B. M.; et al., Evaluation of an alternative in vitro test battery for detecting reproductive toxicants.

Reprod Toxicol 2013, 38, 53-64.

38

26. UN (United Nations), Système Général harmonisé de classification et d'étiquetage des produits chimiques (SGH). 3rd edition ed.; Nations Unies: New York & Genève, 2009; p 611.

27. Huff, J., Long-Term Chemical Carcinogenesis Bioassays Predict Human Cancer Hazards: Issues, Controversies, and Uncertainties. Annals

NY Acad Sciences 1999, 895, (1), 56-79. 28. Goldman, L. R., Preventing Pollution? U.S. Toxic Chemicals and Pesticides Policy and Sustainable Development. ELR News & Analysis

2002, 32, (9), 11018-11041. 29. Russom, C. L.; Bradbury, S. P.; Broderius, S. J.; Hammermeister, D. J.; Drummond, R. A.; Veith, G. D., Predicting modes of toxic action

from chemical structure. Environ Toxicol and Chem 2013, 32, (7), 1441-1442. 30. EC, Regulation (EC) No 1107/2009 of the European Parliament and of the Council concerning the placing of plant protection products on the

market and repealing Council Directives 79/117/EEC and 91/414/EEC. In Off J of the EU: 2009; p 50. 31. EC, Alternative approaches can reduce the use of test animals under REACH; 2004-11, 2004; p 18. 32. EC, REGULATION (EC) No 1907/2006 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL concerning the Registration,

Evaluation, Authorisation and Restriction of Chemicals (REACH). In Off J EU: 2006; Vol. 2006/1907/EC, p 849. 33. CCE (Commission des Communautés Européennes), RÈGLEMENT (CE) no 440/2008 DE LA COMMISSION établissant des méthodes

d'essai conformément au règlement (CE) no 1907/2006 du Parlement européen et du Conseil concernant l'enregistrement, l'évaluation et l'autorisation des substances chimiques, ainsi que les restrictions applicables à ces substances (REACH). In J off EU: 2008; p 739.

34. EPA, Health Effects Test Guidelines OPP & Toxic Substances 870.4200 Carcinogenicity. In Regulations.gov: 1998; p 17. 35. OECD, Test No. 451: Carcinogenicity Studies. OECD Publishing: Paris, 2009; p 15. 36. Hayes, A. W.; Dayan, A. D.; Hall, W. C.; Kodell, R. L.; Williams, G. M.; Waddell, W. D.; Slesinski, R. S.; Kruger, C. L., A review of

mammalian carcinogenicity study design and potential effects of alternate test procedures on the safety evaluation of food ingredients. Regulatory Toxicology and Pharmacology 2011, 60, (1), S1-S34.

37. Maltoni, C.; Soffritti, M.; Belpoggi, F., The scientific and methodological bases of experimental studies for detecting and quantifying carcinogenic risks. Annals of the New York Academy of Sciences 1999, 895, (1), 10-26.

38. Harrison, J. L., Pesticide Drift and the Pursuit of Environmental Justice. The MIT Press: Cambridge, 2011; p 277. 39. EFSA (European Food Safety Authority), Scientific Opinion on the identification of pesticides to be included in cumulative assessment

groups on the basis of their toxicological profile. EFSA Journal 2013, 11, (7:3293), 131pp. 40. Bishop, P. L.; Willett, C. E., The Use and Acceptance of Other Scientifically Relevant Information (OSRI) EPA Endocrine Disruptor

Screening Program. Birth Defects Research Part B: Developmental and Reproductive Toxicology 2013, DOI: 10.1002/bdrb.21077. 41. McKinlay, R.; Plant, J. A.; Bell, J. N. B.; Voulvoulis, N., Endocrine disrupting pesticides: Implications for risk assessment. Environment

International 2008, 34, (2), 168-183. 42. WHO, Guidelines for predicting dietary intake of pesticide residues. In Global Environment Monitoring System - Food Contamination

Monitoring and Assessment Programme (GEMS/Food) in collaboration with Codex Committee on Pesticide Residues. Programme on Food Safety and Food Aid - World Health Organization: 1997; p 41.

43. WHO, Guidelines for Drinking-water Quality. In Global Environment Monitoring System - Food Contamination Monitoring and Assessment Programme (GEMS/Food) in collaboration with Codex Committee on Pesticide Residues. Programme on Food Safety and Food Aid - World Health Organization: 2006; Vol. Volume 1 - Recommendations, p 595.

44. Cairns, J. J., Absence of certainty is not synonymous with absence of risk. Env Health Persp 1999, 107, (2), A56. 45. Marshall, E., Supreme Court to Review the Scope of Monsanto's Seed Patents. Science 2013, 339, (6120), 639.

39

46. Norman, J.; Kuivila, K.; Nowell, L. Prioritizing Pesticide Compounds for Analytical Methods Development; USGS: Reston, V., 2012; p 206. 47. Damalas, C. A.; Eleftherohorinos, I. G., Pesticide exposure, safety issues, and risk assessment indicators. Int J of Env Res and Pub Health

2011, 8, (5), 1402-1419. 48. US (Government of the United States of America), Code of Federal Regulations - 40 Protection of the Environment - Part 158 Data

requirements for pesticides. In US Government Printing Office: 2013; Vol. CFR 40. 49. Krieger, R., Handbook of Pesticide Toxicology, Vol I: Principles. Academic Press: San Diego, CA, 2001; Vol. 1, p 1185. 50. Ballantyne, B.; Marrs, T. C., Chapter 1 - Pesticides : An Overview of Fundamentals. In Pesticide Toxicology and International Regulation,

Ballantyne, B.; Marrs, T. C., Eds. John Wiley & Sons, Ltd.: West Sussex, England, 2004; p 23. 51. US (Government of the United States of America), Federal Insecticide, Fungicide and Rodenticide Act (FIFRA). In Office of the Federal

Register National Archives and Records Administration: 1987; Vol. 40 CFR 162.3 CFR Protection of Environment p652. 52. Duke, S. O., Herbicide and Pharmaceutical Relationships. Weed Science 2009, 58, (3), 334-339. 53. Delaney, J.; Clarke, E.; Hughes, D.; Rice, M., Modern agrochemical research: a missed opportunity for drug discovery? Drug Discovery

Today 2006, 11, 839-845. 54. Johnson, B. L., Will Testing of Chemicals for Safety Assessment and Management Finally Become a Reality in the U.S.? Some Good

News? Maybe. Human and Ecological Risk Assessment: An International Journal 2013, 19, (6), 1419-1421. 55. Landrigan, P. J.; Goldman, L. R., Childrens Vulnerability To Toxic Chemicals: A Challenge And Opportunity To Strengthen Health And

Environmental Policy. Health Affairs 2011, 30, (5), 842-850. 56. Holt, E.; Weber, R.; Stevenson, G.; Gaus, C., Polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) impurities in pesticides: a

neglected source of contemporary relevance. Env Sc & Tech 2010, 44, (14), 5409-5415. 57. Patton, D. E., The ABCs of Risk Assessment. EPA J. 1993, 19, 10. 58. Motaal, D. A., Reaching REACH: The Challenge for Chemicals Entering International Trade. J Int Economic Law 2009, 12, (3), 643-662. 59. European Commission, COUNCIL DIRECTIVE of 15 July 1991 concerning the placing of plant protection products on the market

(91/414/EEC). In Off J EU, 1991; p 32. 60. European Commission, DIRECTIVE 98/8/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 16 February 1998 concerning

the placing of biocidal products on the market. In Official Journal of the European Union, 1998; p 63. 61. EC (European Comission) Factsheet - EU Action on Pesticides; Brussels, 2009-03, 2009; p 2. 62. EC (European Comission), Comission regulation (EU) no 284/2013 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. In Official Journal of the European Union: 2013; Vol. 2013/284/EU, p 68.

63. Van den Belt, H., Debating the precautionary principle:''guilty until proven innocent''or ''innocent until proven guilty''? Plant Physiology 2003, 132, (3), 1122-1126.

64. Sass, J. B.; Colangelo, A., European Union bans atrazine, while the United States negotiates continued use. International journal of

occupational and environmental health 2006, 12, (3), 260-267. 65. European Community Council, Council Directive 98/83/EC of 3 November 1998 on the quality of water intended for human consumption. In

Official Journal of the European Communities: 1998; p 23. 66. WHO and FAO. 3RD FAO/WHO JOINT MEETING ON PESTICIDE MANAGEMENT and 5TH SESSION OF THE FAO PANEL OF EXPERTS

ON PESTICIDE MANAGEMENT; Rome, 2009-10-6, 2009; p 48. 67. OECD Test Guidelines. http://www.oecd.org/env/testguidelines (Accessed 2013-11-15),

40

68. EC (European Comission), Comission regulation (EU) no 283/2013 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. In Official Journal of the European Union: 2013; Vol. 2013/283/EU, p 84.

69. EPA (United States Environmental Protection Agency), OCSPP Harmonized Test Guidelines Series 850 - Ecological Effects Test Guidelines. In Office of Chemical Safety and Pollution Prevention (OCSPP) 2012.

70. ASTM Standards & Publications. http://www.astm.org/ (Accessed 2013-12-19), 71. ISO Standards. http://www.iso.org/iso/home/standards.htm (Accessed 2013-12-09), 72. Kruijne, R.; Deneer, J.; Lahr, J.; Vlaming, J. HAIR 2010 Documentation - Calculating risk indicators related to agrcultural use of pesticides

within the European Union; Alterra-report 2113.1; Wageningen Alterra: 2011; p 202. 73. OECD Pesticides: Risk Indicators. oecd.org (Accessed 2013-12-15) 74. Samuel, O.; Dion, S.; St-Laurent, L.; April, M. H., Indicateur de risques des pesticides du Québec In MAPAQ MDDEP INSPQ. In 2 ed.;

Québec: 2012; p 48. 75. Québec SAgE pesticides - Traitements phytosanitaires et risques associés.

http://www.sagepesticides.qc.ca/Recherche/RechercheTraitements.aspx (Access date 2013-12-15), 76. EPA, Glyphosate Registration Review Human-Health Assessment Scoping Document. In Office of Pesticide Programs, Ed. regulations.gov:

2009; p 82. 77. EC, Council Directive of 5th June 1992 amending for the 7th time Directive 67/548/EEC. In Off. J. Eur. Commun.: 1992; Vol. L.154/1. 78. King-Herbert, A. P.; Sills, R. C.; Bucher, J. R., Commentary: update on animal models for NTP studies. Tox pathology 2010, 38, 180-181. 79. Hayes, T., Chapter 10 - Atrazine Has Been Used Safely for 50 Years? In Wildlife Ecotoxicology - Forensic Approaches, Elliot, J. E.; Bishop,

C. A.; Morrissey, C. A., Eds. Springer: Oak Ridge, TN, USA, 2011; p 476. 80. Dinse, G. E.; Peddada, S. D.; Harris, S. F.; Elmore, S. A., Comparison of NTP Historical Control Tumor Incidence Rates in Female Harlan

Sprague Dawley and Fischer 344/N Rats. Toxicologic Pathology 2010, 38, (5), 765-775. 81. Stevens, J. T., A Risk Characterization for Atrazine: Oncogenicity Profile. J Tox & Env Health, Part A 1999, 56, (2), 69-109. 82. Séralini, G.-E.; Mesnage, R.; Defarge, N.; Gress, S.; Hennequin, D.; Clair, E.; Malatesta, M.; de Vendômois, J. S., Answers to critics: Why

there is a long term toxicity due to a Roundup-tolerant genetically modified maize and to a Roundup herbicide. Food and Chemical

Toxicology 2013, 53, 461-468. 83. Vandenberg, L. N.; Colborn, T.; Hayes, T. B.; Heindel, J. J.; Jacobs, D. R.; Lee, D.-H.; Shioda, T.; Soto, A. M.; vom Saal, F. S.; Welshons,

W. V.; Zoeller, R. T.; Myers, J. P., Hormones and Endocrine-Disrupting Chemicals: Low-Dose Effects and Nonmonotonic Dose Responses. Endocrine Reviews 2012, 33, (3), 378-455.

84. FDA (United States Food and Drug Administration) FDA Redbook Toxicological principles for the safety assessment of food ingredients. Section IV.B.4 Statistical considerations in toxicity studies. http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocuments RegulatoryInformation/IngredientsAdditivesGRASPackaging/ucm078320.htm (Accessed 2013-12-14),

85. Antoniou, M.; Habib, M.; Howard, C.; Jennings, R.; Leifert; Nodari, R.; Robinson, C.; Fagan, J., Teratogenic Effects of Glyphosate-Based Herbicides: Divergence of Regulatory Decisions from Scientific Evidence. J Environ Anal Toxicol 2012, S4:006.

86. Novak, R., The long arm of the lab laws. Today's Chemist at Work 2001, 10, 45-46. 87. Cox, C., Glyhphosate (Roundup) - Herbicide factsheet. Journal of pesticide reform 1998, 18, 3-17. 88. Monsanto Backgrounder Testing Fraud: IBT and Craven Laboratories; Monsanto: June 2005, 2005; p 1. 89. Saltmiras, D.; Bus, J. S.; Spanogle, T.; Hauswirth, J.; Tobia, A.; Hill, S., Letter to the editor regarding the article by Paganelli et Al. Chem Res

41

Toxicol 2011, 24, (5), 607-8. 90. Séralini, G.-E.; Mesnage, R.; Clair, E.; Gress, S.; de Vendômois, J. S.; Cellier, D., Genetically modified crops safety assessments: present

limits and possible improvements. Environmental Sciences Europe 2011, 23, (1), 1-10. 91. Horrobin, D. F., The philosophical basis of peer review and the suppression of innovation. J Am Med Ass 1990, 263, (10), 1438-1441. 92. EPA Guidelines for Ensuring and Maximizing the Quality, Objectivity, Utility, and Integrity, of Information Disseminated by the Environmental

Protection Agency; EPA/260R-02-008; 2002-10, 2002; p 61. 93. Klimisch, H. J.; Andreae, M.; Tillmann, U., A systematic approach for evaluating the quality of experimental toxicological and ecotoxicological

data. Regulatory toxicology and pharmacology 1997, 25, (1), 1-5. 94. Elsevier. Elsevier Announces Article Retraction from Journal Food and Chemical Toxicology - “Long term toxicity of a Roundup herbicide

and a Roundup-tolerant genetically modified maize,” by Gilles Eric Séralini et al. has been retracted by the journal Food and Chemical Toxicology. http://www.elsevier.com/about/press-releases/research-and-journals/el…nnounces-article-retraction-from-journal-food-and-chemical-toxicology (Accessed 2013-12-01),

95. Antoniou, M.; Habib, M. E. E.-D. M.; Howard, C. V.; Jennings, R. C.; Leifert, C.; Onofre Nodari, R.; Robinson, C.; Fagan, J., Roundup and

birth defects - Is the public being kept in the dark? Earth Open Source: 2011; p 52. 96. Huff, J., Industry Influence on Occupational and Environmental Public Health. Int J Occ & Env Health 2007, 13, (1), 107-117. 97. vom Saal, F. S.; Welshons, W. V., Large effects from small exposures. II. The importance of positive controls in low-dose research on

bisphenol A. Environmental Research 2006, 100, (1), 50-76. 98. Cox, C.; Surgan, M., Unidentified inert ingredients in pesticides: implications for human and environmental health. Environmental Health

Perspectives 2006, 114, (12), 1803. 99. EPA (United States Environmental Protection Agency), Registration Review - Preliminary Problem Formulation for the Ecological Risk and

Drinking Water Exposure Assessments for Glyphosate and Its Salts EPA-HQ-2009-0361-0007. In regulations.gov: 2009; p 46. 100. Hanzen, J. L., Adjuvants - Terminology, Classification, and Chemistry. Weed Technology 2000, 14, 773-784. 101. Gilliom, R. J., Pesticides in US streams and groundwater. Env Sc & Tech 2007, 41, (10), 3408-3414. 102. Gilliom, R. J.; Barbash, J. E.; Crawford, C. G.; Hamilton, P. A.; Martin, J. D.; Nakagaki, N.; Nowell, L. H.; Scott, J. C.; Stackelberg, P. E.;

Thelin, G. P.; Wolock, D. M. The Quality of Our Nation's Waters - Pesticides in the Nation's Streams and Ground Water; 2006; p 172. 103. Smalling, K.; JL, O.; Cahoun, D.; Battaglin, W. A.; Kuivila, K. Occurence of Pesticides in Water and Sediments Collected from Amphibian

Habitats Located Throughout the Unites States; USGS: 2012; p 36. 104. Woodruff, T. J.; Zota, A. R.; Schwartz, J. M., Environmental chemicals in pregnant women in the United States: NHANES 2003-2004.

Environmental health perspectives 2011, 119, (6), 878. 105. WWF (World Wildlife Fund) Bad Blood? A Survey in the Blood of European Ministers; Brussels, 2004; p 40. 106. Backhaus, T.; Faust, M., Predictive Environmental Risk Assessment of Chemical Mixtures: A Conceptual Framework. Env Sc & Tech

2012, 46, (5), 2564-2573. 107. Buchanan, I.; Liang, H. C.; Liu, Z.; Razaviarani, V., Pesticides and Herbicides. Water Env Res 2011, 83, (10), 1549-1597. 108. Mascarelli, A., Growing Up With Pesticides. Science 2013, 341, (6147), 740-741. 109. Thayer, K. A.; Heindel, J. J.; Bucher, J. R.; Gallo, M. A., Role of environmental chemicals in diabetes and obesity: a National Toxicology

Program workshop review. Env Health Persp 2012, 120, (6), 779-789. 110. EPA, Decision Documents for Atrazine In Office of Prevention, Pesticides and Toxic Substances. In 2006; p 323. 111. Shimabukuro, R. H.; Swanson, H. R., Atrazine metabolism, selectivity, and mode of action. J Agr Food Chem 1969, 17, (2), 199-205.

42

112. Solomon, K. R.; Giesy, J. P.; LaPoint, T. W.; Giddings, J. M.; Richards, R. P., Ecological risk assessment of atrazine in North American surface waters. Env Toxicol Chem 1996, 32, (1), 10-11.

113. Hayes, T. B.; Collins, A.; Lee, M.; Mendoza, M.; Noriega, N.; Stuart, A. A.; Vonk, A., Hermaphroditic, demasculinized frogs after exposure to the herbicide atrazine at low ecologically relevant doses. PNAS 2002, 99, (8), 5476-5480.

114. Chen, J. Y.; Song, Y.; Zhang, L. S., Immunotoxicity of atrazine in Balb/c mice. J Env Sc & Health, Part B 2013, 48, (8), 637-645. 115. Coursey, D. Illinois Without Atrazine: Who Pays? Economic Implications of an Atrazine Ban in the State of Illinois; Chicago, 2007-02-27,

2007; p 32. 116. Ackerman, F., Chapter 5 - The Economics of Atrazine. In Poisoned for Pennies: The Economics of Toxics and Precaution, Island Press:

Washington, DC, 2007; pp 95-112. 117. Ackerman, F., The Economics of Atrazine. Int J Occ & Env Health 2007, 13, (4), 437-445. 118. Mitchell, P. D., Market-Level Assessment of the Economic Benefits of Atrazine in the United States. Pest Man Sc 2013, DOI:

10.1002/ps.3703. 119. Ribaudo, M. O.; Bouzaher, A. Atrazine: Environmental Characteristics and Economics of Management, Report number 699; Washington,

DC, 1994; p 22. 120. EPA (United States Environmental Protection Agency) Atrazine Updates. http://www.epa.gov/pesticides/reregistration/atrazine/atrazine-

2010-25-new-sites-aemp.pdf (Accessed 2013-12-15), 121. Syngenta Crop Protection Atrazine Regulation - International - European Union. http://www.atrazinefacts.com/atrazine-

regulation/international/european-union/ (Accessed 2013-12-15), 122. Natural Resources Defense Council v. Whitman, Case Number C -99-3701 CAL. In N.D. California: 2002. 123. SAVE THE FROGS!, Submission to the US Environmental Protection Agency. In Petition to Ban Atrazine, Supportng & Related Material

document issued by the Environmental Protection Agency (EPA), Kriger, K., Ed. EPA on Regulations.gov: 2011; p 413. 124. EPA, Response to Save the Frogs May 6, 20 I I Petition Requesting a Federal Ban on the Use and Production of Atrazine. In Office of

Chemical Safety and Pollution Prevention. In regulations.gov: 2011; p 3. 125. Rohr, J. R.; McCoy, K. A., A qualitative meta-analysis reveals consistent effects of atrazine on freshwater fish and amphibians. Env Health

Persp 2010, 118, (1), 20. 126. City of Greenville vs Syngenta Crop Protection Inc. Settlement Agreement Case No.: 3:10-cv-00188-JPG-PMF. In US District Court for the

Southern District of Illinois: 2012; p 107. 127. Government of Australia Atrazine toxicity: Analysis of potential modes of action; Canberra, 2010-01, 2010; p 49. 128. WHO (World Health Organisation), Guidelines for Drinking-water Quality. In 4rd ed. ed.; GEMS/Food Codex Committee on Pesticide

Residues. Programme on Food Safety and Food Aid: 2011; Vol. Chapter 8 - Chemical aspects, p 541. 129. WHO (World Health Organisation) and FAO (Food and Agriculture Organization) Pesticide residues in food - 2007 - Evaluations - Part II -

Toxicological; Geneva, 2007; p 544. 130. Wetzel, L. T.; Luempert, L. G.; Breckenridge, C. B.; Tisdel, M. O.; Stevens, J. T.; Thakur, A. K.; Extrom, P. J.;, Chronic effects of atrazine

on estrus and mammary tumor formation in female Sprague-Dawley and Fischer 344 rats. J Toxicol & Env Health 1994, 43, (2), 169-182. 131. Coban, A.; Filipov, N. M., Dopaminergic toxicity associated with oral exposure to the herbicide atrazine in juvenile male C57BL/6 mice. J

Neurochem 2007, 100, (5), 1177-1187. 132. Narotsky, M. G.; Best, D. S.; Guidici, D. L.; Cooper, R. L., Strain comparisons of atrazine-induced pregnancy loss in the rat. Reprod

Toxicol 2000, 15, (1), 61-69.

43

133. Cummings, A. M.; Rhodes, B. E.; Cooper, R. L., Effect of Atrazine on Implantation and Early Pregnancy in 4 Strains of Rats. Toxicological

Sciences 2000, 58, (1), 135-143. 134. Rayner, J. L.; Wood, C.; Fenton, S. E., Exposure parameters necessary for delayed puberty and mammary gland development in Long-

Evans rats exposed in utero to atrazine. Toxicol & Appl Pharmacol 2004, 195, (1), 23-34. 135. Rayner, J. L.; Enoch, R. R.; Fenton, S. E., Adverse Effects of Prenatal Exposure to Atrazine During a Critical Period of Mammary Gland

Growth. Toxicological Sciences 2005, 87, (1), 255-266. 136. Rayner, J. L.; Enoch, R. R.; Wolf, D. C.; Fenton, S. E., Atrazine-induced reproductive tract alterations after transplacental and/or lactational

exposure in male Long-Evans rats. Toxicol & Appl Pharmacol 2007, 218, (3), 238-248. 137. Rosenberg, B. G.; Chen, H.; Folmer, J.; Liu, J.; Papadopoulos, V.; Zirkin, B. R., Gestational Exposure to Atrazine: Effects on the Postnatal

Development of Male Offspring. Journal of Andrology 2008, 29, (3), 304-311. 138. Hayes, T.; Haston, K.; Tsui, M.; Hoang, A.; Haeffele, C.; Vonk, A., Atrazine-induced hermaphroditism at 0.1 ppb in American leopard frogs

(Rana pipiens): laboratory and field evidence. Env Health Persp 2003, 111, (4), 568. 139. Hayes, T. B.; Case, P.; Chui, S.; Chung, D.; Haeffele, C.; Haston, K.; Lee, M.; Mai, V. P.; Marjuoa, Y.; Parker, J.; Tsui, M., Pesticide

Mixtures, Endocrine Disruption, and Amphibian Declines: Are We Underestimating the Impact? Env Health Persp 2006, 114, (S-1). 140. Hayes, T. B.; Anderson, L. L.; Beasley, V. R.; de Solla, S. R.; Iguchi, T.; Ingraham, H.; Kestemont, P.; Demasculinization and feminization

of male gonads by atrazine: consistent effects across vertebrate classes. J Steroid Biochem & Mol Biol 2011, 127, (1), 64-73. 141. Kloas, W.; Lutz, I.; Springer, T.; Krueger, H.; Wolf, J.; Holden, L.; Hosmer, A., Does Atrazine Influence Larval Development and Sexual

Differentiation in Xenopus laevis? Toxicological Sciences 2009, 107, (2), 376-384. 142. Beane Freeman, L. E.; Rusiecki, J. A.; Hoppin, J. A.; Lubin, J. H.; Koutros, S.; Andreotti, G.; Zahm, S. H.; Hines, C. J.; Coble, J. B.;

Barone-Adesi, F.; Sloan, J.; Sandler, D. P.; Blair, A.; Alavanja, M. C. R., Atrazine and Cancer Incidence Among Pesticide Applicators in the Agricultural Health Study (1994–2007). Env Health Persp 2011, 119, (9), 1253-1259.

143. Jowa, L.; Howd, R., Should Atrazine and Related Chlorotriazines Be Considered Carcinogenic for Human Health Risk Assessment? J Env

Sc & Health, Part C 2011, 29, (2), 91-144. 144. Smith, E. E.; Du Preez, L. H.; Gentles, A.; Solomon, K. R.; Tandler, B.; Carr, J. A.; Van Der Kraak, G. L.; Kendall, R. J.; Giesy, J. P.;

Gross, T. S., Assessment of laryngeal muscle and testicular cell types in Xenopus laevis (Anura Pipidae) inhabiting maize and non-maize growing areas of South Africa. Afr J Herpet 2005, 54, (1), 69-76.

145. Hecker, M.; Giesy, J. P.; Jones, P. D.; Jooste, A. M.; Carr, J. A.; Solomon, K. R.; Smith, E. E.; Van Der Kraak, G.; Kendall, R. J.; du Preez, L., Plasma sex steroid concentrations and gonadal aromatase activities in African clawed frogs (Xenopus laevis) from South Africa. Env

Toxicol & Chem 2004, 23, (8), 1996-2007. 146. MacLennan, P.; Delzell, E.; Sathiakumar, N.; Myers, S., Mortality Among Triazine Herbicide Manufacturing Workers. J Toxicol & Env

Health, Part A 2003, 66, (6), 501-517. 147. Van Maele-Fabry, G.; Libotte, V.; Willems, J.; Lison, D., Review and Meta-analysis of Risk Estimates for Prostate Cancer in Pesticide

Manufacturing Workers. Cancer Causes & Control 2006, 17, (4), 353-373. 148. Kettles, M. K.; Browning, S. R.; Prince, T. S.; Horstman, S. W., Triazine herbicide exposure and breast cancer incidence: an ecologic study

of Kentucky counties. Env Health Persp 1997, 105, (11), 1222. 149. Hopenhayn-Rich, C.; Stump, M. L.; Browning, S. R., Regional Assessment of Atrazine Exposure and Incidence of Breast and Ovarian

Cancers in Kentucky. Arch Env Cont & Toxicol 2002, 42, (1), 127-136. 150. Rusiecki, J. A.; Roos, A. D.; Lee, W. J.; Dosemeci, M.; Lubin, J. H.; Hoppin, J. A.; Blair, A.; Alavanja, M. C. R., Cancer Incidence Among

44

Pesticide Applicators Exposed to Atrazine in the Agricultural Health Study. J Nat Cancer Inst 2004, 96, (18), 1375-1382. 151. Boffetta, P.; Adami, H. O.; Berry, S. C.; Mandel, J. S., Atrazine and cancer: a review of the epidemiologic evidence. Eur J Cancer Prev

2013, 22, (2), 169-80. 152. Rhoades, M. G.; Meza, J. L.; Beseler, C. L.; Shea, P. J.; Kahle, A.; Vose, J. M.; Eskridge, K. M.; Spalding, R. F., Atrazine and nitrate in

public Drinking Water supplies and non-Hodgkin Lymphoma in nebraska, UsA. Env Health Insights 2013, 7, 15. 153. Cooper, R. L.; Stoker, T. E.; Tyrey, L.; Goldman, J. M.; McElroy, W. K., Atrazine disrupts the hypothalamic control of pituitary-ovarian

function. Toxicological Sciences 2000, 53, (2), 297-307. 154. Tavera-Mendoza, L.; Ruby, S.; Brousseau, P.; Fournier, M.; Cyr, D.; Marcogliese, D., Response of the amphibian tadpole (Xenopus laevis)

to atrazine during sexual differentiation of the testis. Env Toxicol & Chem 2002, 21, (3), 527-531. 155. Crain, D. A.; Guillette, L. J., Reptiles as models of contaminant-induced endocrine disruption. Animal Reprod Sc 1997, 53, (1), 77-86. 156. Sanderson, J. T.; Letcher, R. J.; Heneweer, M.; Giesy, J. P.; van den Berg, M., Effects of chloro-s-triazine herbicides and metabolites on

aromatase activity in various human cell lines and on vitellogenin production in male carp hepatocytes. Environ Health Perspect 2001, 109, (10), 1027-31.

157. Sanderson, J. T.; Seinen, W.; Giesy, J. P.; van den Berg, M., 2-Chloro-s-triazine herbicides induce aromatase (CYP19) activity in H295R human adrenocortical carcinoma cells: a novel mechanism for estrogenicity? Toxicol Sci 2000, 54, (1), 121-7.

158. Fan, W.; Yanase, T.; Morinaga, H.; Gondo, S.; Okabe, T.; Nomura, M.; Komatsu, T.; Morohashi, K.-I.; Hayes, T. B.; Takayanagi, R., Atrazine-induced aromatase expression is SF-1 dependent: implications for endocrine disruption in wildlife and reproductive cancers in humans. Env Health Persp 2007, 115, (5), 720.

159. ATSDR (Agency for Toxic Substances and Disease Registry), Toxicological Profile for Atrazine. In U.S. Public Health Service. In U.S. Department of Health and Human Services 2003; p 262.

160. Calamari, D.; Jones, K.; Kannan, K.; Ledoux, A.; Mats Olsson; Thurman, M.; Zannetti, P., Monitoring as an Indicator of Persistence and Long-Range Transport. In Evaluation of persistence and long-range transport of organic chemicals in the environment, Klecka, G.; Boethling, R.; Franklin, J.; Grady, L.; Graham, D.; Howard, P. H.; Kannan, K.; Larson, R.; Mackay, D.; Muir, D., Eds. SETAC Press: Pensacola, Fla, 2000; p 36.

161. Relyea, R.; Hoverman, J., Assessing the ecology in ecotoxicology: a review and synthesis in freshwater systems. Ecology Letters 2006, 9, (10), 1157-1171.

162. Boone, M. D.; Bridges-Britton, C. M., Examining multiple sublethal contaminants on the gray treefrog (Hyla versicolor): Effects of an insecticide, herbicide, and fertilizer. Env Toxicol & Chemistry 2006, 25, (12), 3261-3265.

163. Boocock, M. R.; Coggins, J. R., Kinetics of 5-enolpyruvylshikimate-3-phosphate synthase inhibition by glyphosate. FEBS Letters 1983, 154, (1), 127-133.

164. Williams, G. M.; Kroes, R.; Munro, I. C., Safety evaluation and risk assessment of the herbicide Roundup and its active ingredient, glyphosate, for humans. Reg Toxicol & Pharmacol 2000, 31, (2), 117-165.

165. Duke, S. O.; Lydon, J.; Koskinen, W. C.; Moorman, T. B.; Chaney, R. L.; Hammerschmidt, R., Glyphosate effects on plant mineral nutrition, crop rhizosphere microbiota, and plant disease in glyphosate-resistant crops. J Agr & Food Chem 2012, 60, (42), 10375-10397.

166. EPA (United States Environmental Protection Agency), Update Screening Level Usage Analysis (SLUA) Report in Support of Registration Review of Glyphosate. In Office of Pesticide Programs, E. P. A., Ed. regulations.gov: 2012; p 4.

167. EPA (United States Environmental Protection Agency), Glyphosate Final Work Plan (FWP) Registration Review Case No. 0178. In Office of Pesticide Programs, E. P. A., Ed. regulations.gov: 2009.

45

168. European Commission Review report for the active substance glyphosate. In Health & Cons Protection Dir-General: 2002; p 56. 169. European Commission, Directive 2010/77/EU of 10 November 2010 amending Council Directive 91/414/EEC as regards the expiry dates

for inclusion in Annex I of certain active substances. Off J EU 2010, 48-57. 170. Haney, R.; Senseman, S.; Krutz, L.; Hons, F., Soil carbon and nitrogen mineralization as affected by atrazine and glyphosate. Biology and

Fertility of Soils 2002, 35, (1), 35-40. 171. Sprankle, P.; Meggitt, W. F.; Penner, D., Adsorption, Mobility, and Microbial Degradation of Glyphosate in the Soil. Weed Science 1975,

23, (3), 229-234. 172. Gimsing, A. L.; Borggaard, O. K.; Bang, M., Influence of soil composition on adsorption of glyphosate and phosphate by contrasting

Danish surface soils. European Journal of Soil Science 2004, 55, (1), 183-191. 173. Borggaard, O. K.; Gimsing, A. L., Fate of glyphosate in soil and the possibility of leaching to ground and surface waters: a review. Pest

Man Sc 2008, 64, (4), 441-456. 174. Hill, H., Competitive sorption between glyphosate and inorganic phosphate on clay minerals and low organic matter soils. J Radioanal &

Nucl Chem 2001, 249, (2), 385-390. 175. Borggaard, O., Does Phosphate Affect Soil Sorption and Degradation of Glyphosate?A Review.Trends Soil Sc & Plant Nut 2011,2,16-27. 176. Simonsen, L.; Fomsgaard, I.; Svensmark, B.; Spliid, N. H., Fate and availability of glyphosate and AMPA in agricultural soil. J Env Sc &

Health, Part B 2008, 43, (5), 365-375. 177. Wauchope, R. D.; Estes, T. L.; Allen, R.; Baker, J. L.; Hornsby, A. G.; Jones, R. L.; Richards, R. P., Predicted impact of transgenic,

herbicide-tolerant corn on drinking water quality in vulnerable watersheds of the mid-western USA. Pest Man Sc 2002, 58, (2), 146-160. 178. Bergström, L.; Börjesson, E.; Stenström, J., Laboratory and Lysimeter Studies of glyphosate and Aminomethylphosphonic Acid in a Sand

and a Clay Soil. J Env Qual 2011, 40, 207-216. 179. Torstensson, L.; Börjesson, E.; Stenström, J., Efficacy and fate of glyphosate on Swedish railway embankments. Pest Man Sc 2005, 61,

(9), 881-886. 180. Cerdeira, A. L.; Duke, S. O., The Current Status and Environmental Impacts of Glyphosate-Resistant Crops. J Env Qual 2006, (35), 1633-

1652. 181. Arregui, M. C.; Lenardón, A.; Sanchez, D.; Maitre, M. I.; Scotta, R.; Enrique, S., Monitoring glyphosate residues in transgenic glyphosate-

resistant soybean. Pest Man Sc 2004, 60, (2), 163-166. 182. Reddy, K. N.; Rimando, A. M.; Duke, S. O., Aminomethylphosphonic Acid, a Metabolite of Glyphosate, Causes Injury in Glyphosate-

Treated, Glyphosate-Resistant Soybean. J Agr & Food Chem 2004, 52, (16), 5139-5143. 183. Coupland, D.; Caseley, J. C., Presence of 14C Activity in Root Exudates and Guttation Fluid from Agropyron repens Treated with 14C-

Labelled Glyphosate. New Phytologist 1979, 83, (1), 17-22. 184. Neumann, G.; Kohls, S.; Landsberg, E.; Stock-Oliveira Souza, K.; Yamada, T.; Romheld, V., Relevance of glyphosate transfer to non-

target plants via the rhizosphere. Journal of Plant Diseases and Protection 2006, 20, 963. 185. Heard, M. S.; Hawes, C.; Champion, G. T.; Clark, S. J.; Firbank, L. G.; Haughton, A. J.; Parish, A. M.; Perry, J. N.; Rothery, P.; Scott, R. J.;

Skellern, M. P.; Squire, G. R.; Hill, M. O., Weeds in fields with contrasting conventional and genetically modified herbicide-tolerant crops. I. Effects on abundance and diversity. Phil Trans Roy Soc London. Series B: Biol Sc 2003, 358, (1439), 1819-1832.

186. Matarczyk, J. A.; Willis, A. J.; Vranjic, J. A.; Ash, J. E., Herbicides, weeds and endangered species: management of bitou bush (Chrysanthemoides monilifera ssp. rotundata) with glyphosate and impacts on the endangered shrub, Pimelea spicata. Biol Cons 2002, 108, (2), 133-141.

46

187. Kremer, R. J.; Means, N. E., Glyphosate and glyphosate-resistant crop interactions with rhizosphere microorganisms. Eu J Agron 2009, 31, (3), 153-161.

188. Bohm, G. M. B.; Alves, B. J. R.; Urquiaga, S.; Boddey, R. M., Short communication: Glyphosate- and imazethapyr-induced effects on yield, nodule mass and biological nitrogen fixation in field-grown glyphosate-resistant soybean Soil Biol & Biochem 2009, 41, (2), 420-422.

189. Johal, G. S.; Huber, D. M., Glyphosate effects on diseases of plants. European Journal of Agronomy 2009, 31, (3), 144-152. 190. Powell, J. R.; Swanton, C. J., A critique of studies evaluating glyphosate effects on diseases associated with Fusarium spp. Weed Res

2008, 48, (4), 307-318. 191. Benbrook, C. M. Genetically Engineered Crops and Pesticide Use in the United States: The First Nine Years; BioTech InfoNet: October

2004, 2004; p 53. 192. Benbrook, C. M. Genetically Engineered Crops and Pesticide Use in the United States: The First Thirteen Years; BioTech InfoNet:

November 2009, 2009; p 61. 193. Beckie, H. J., Herbicide-Resistant Weeds: Management Tactics and Practices1. Weed Technology 2006, 20, (3), 793-814. 194. Green, J. M.; Hazel, C. B.; Forney, D. R.; Pugh, L. M., New multiple-herbicide crop resistance and formulation technology to augment the

utility of glyphosate. Pest Management Science 2008, 64, (4), 332-339. 195. Powles, S. B., Evolved glyphosate-resistant weeds around the world: lessons to be learnt. Pest Man Sc 2008, 64, (4), 360-365. 196. Québec, SAgE pesticides Fiche Toxicologie de la matière active: glyphosate. In MAPAQ MDDEP INSPQ: 2011. 197. Giroux, I.; Robert, C.; Dassylva, N., Présence de pesticides dans l'eau au Québec. In MDDEP, Ed. Québec: 2006; p 21. 198. Coupe, R.; Kalkhoff, S.; Capel, P.; Gregoire, C., The Fate and Transport of Glyphosate and AMPA into Surface Waters of Agricultural

Watersheds. Pest management science 2011, 1, 1003. 199. Giroux, I.; Pelletier, L., Présence de pesticides dans l'eau au Québec - Bilan des quatre cours d'eau de zones en culture de maïs et de

soya en 2008, 2009 et 2010. In MDDEP - DSE - Gouvernement du Québec: 2012; p 86. 200. Kjaer, J., Leaching of glyphosate and amino-methylphosphonic acid from Danish agricultural field sites. J Env Qual 2005, 34, (2), 608. 201. Scribner, E. A.; Battaglin, W. A.; Gilliom, R. J.; Meyer, M. T. Concentrations of glyphosate, its degradation product, aminomethyl-

phosphonic acid, and glufosinate in ground-and surface-water, rainfall, and soil samples collected in the US, 2001-06; USGS: 2007; p 111. 202. Miller, A.; Gervais, J. A.; Luukinen, B.; Buhl, K.; Stone, D. Glyphosate Technical Fact Sheet; Oregon State University Extension Services:

2010; p 14. 203. Relyea, R. A., The impact of insecticides and herbicides on the biodiversity and productivity of aquatic communities. Ecol Appl 2005, 15,

(2), 618-627. 204. Relyea, R. A., The lethal impact of Roudup on aquatic and terrestrial amphibians. Ecol Appl 2005, 15, (4), 1118-1124. 205. Pérez, G. L.; Torremorell, A.; Mugni, H.; Rodriguez, P.; Vera, M. S.; Nascimento, M. d.; Allende, L.; Bustingorry, J.; Escaray, R.; Ferraro,

M.; Izaguirre, I.; Pizarro, H.; Bonetto, C.; Morris, D. P.; Zagarese, H., Effects of the Herbicide Roundup on Freshwater Microbial Communities: A Mesocosm Study. Ecol Appl 2007, 17, (8), 2310-2322.

206. Richard, S.; Moslemi, S.; Sipahutar, H.; Benachour, N.; Seralini, G.-E., Differential effects of glyphosate and Roundup on human placental cells and aromatase. Env Health Persp 2005, 113, (6), 716.

207. Benachour, N.; Sipahutar, H.; Moslemi, S.; Gasnier, C.; Travert, C.; Séralini, G. E., Time-and dose-dependent effects of roundup on human embryonic and placental cells. Arch Env Cont & Toxicol 2007, 53, (1), 126-133.

208. Benachour, N.; Séralini, G.-E., Glyphosate Formulations Induce Apoptosis and Necrosis in Human Umbilical, Embryonic, and Placental Cells. Chem Res Toxicol 2008, 22, (1), 97-105.

47

209. Romano, R.; Romano, M.; Bernardi, M.; Furtado, P.; Oliveira, C., Prepubertal exposure to commercial formulation of the herbicide glyphosate alters testosterone levels and testicular morphology. Archives of Toxicology 2010, 84, (4), 309-317.

210. Bellé, R.; Le Bouffant, R.; Morales, J.; Cosson, B.; Cormier, P.; Mulner-Lorillon, O., L'embryon d'oursin, le point de surveillance de l'ADN endommagée de la division cellulaire et les mécanismes à l'origine de la cancérisation. J. Soc. Biol. 2007, 201, (3), 317-327.

211. Bolognesi, C.; Bonatti, S.; Degan, P.; Gallerani, E.; Peluso, M.; Rabboni, R.; Roggieri, P.; Abbondandolo, A., Genotoxic Activity of Glyphosate and Its Technical Formulation Roundup. J Agr & Food Chem 1997, 45, (5), 1957-1962.

212. Cavalcante, D. G. S. M.; Martinez, C. B. R.; Sofia, S. H., Genotoxic effects of Roundup® on the fish Prochilodus lineatus. Mutation

Res/Gen Toxicol & Env Mutagenesis 2008, 655, (1-2), 41-46. 213. Kale, P. G.; Petty, B. T.; Walker, S.; Ford, J. B.; Dehkordi, N.; Tarasia, S.; Tasie, B. O.; Kale, R.; Sohni, Y. R., Mutagenicity testing of nine

herbicides and pesticides currently used in agriculture. Env & Mol Mutagenesis 1995, 25, (2), 148-153. 214. Arbuckle, T.; Lin, Z.; Mery, L., An exploratory analysis of the effect of pesticide exposure on the risk of spontaneous abortion in an Ontario

farm population. Environ Health Perspect 2001, (109), 851-57. 215. Paganelli, A.; Gnazzo, V.; Acosta, H.; LoÌ pez, S. L.; Carrasco, A. s. E., Glyphosate-Based Herbicides Produce Teratogenic Effects on

Vertebrates by Impairing Retinoic Acid Signaling. Chem Res Toxicol 2010, 23, (10), 1586-1595. 216. Anadon, A.; Del Pino, J.; Martinez, M.; Cabalerro, V.; Ares, I.; Nieto, I.; Martinez-Marranaga, M., Neurotoxicological effects of the herbicide

glyphosate. Toxicol Letters 2008, (180S), S164. 217. Curwin, B. D.; Hein, M. J.; Sanderson, W. T.; Nishioka, M. G.; Reynolds, S. J.; Ward, E. M.; Alavanja, M. C., Pesticide contamination

inside farm and nonfarm homes. J Occ & Env Hygiene 2005, 2, (7), 357-367. 218. Curwin, B. D.; Hein, M. J.; Sanderson, W. T.; Striley, C.; Heederik, D.; Kromhout, H.; Reynolds, S. J., Urinary pesticide concentrations

among children, mothers and fathers living in farm and non-farm households in Iowa. Annals Occ Hygiene 2007, 51, (1), 53-65. 219. Rapporteur member state-Germany, Monograph on Glyphosate. Annex B-5: Toxicology and Metabolism: Volume 3-1, Glyphosate. In

1998; Vol. 5, p 12. 220. Rapporteur member state-Germany, Monograph on Glyphosate. List of Tests and Studies - Annex A. In 1998; Vol. 2, pp 95-101. 221. Suresh, T., Teratogenicity study in winstar rats - test compound: glyphosate technical (FSG 03090 H/05 March 90) Study No.

TOXI:ES.883.TER-R. In 1993. 222. Kimmel, G. L.; Kimmel, C. A.; Williams, A. L.; DeSesso, J. M., Evaluation of developmental toxicity studies of glyphosate with attention to

cardiovascular development. Critical Reviews in Toxicology 2013, 43, (2), 79-95. 223. Benedetti, A. s. L.; de Lourdes Vituri, C.; Gonçalves Trentin, A.; Custódio Domingues, M. A.; Alvarez-Silva, M., The effects of sub-chronic

exposure of Wistar rats to the herbicide Glyphosate-Biocarb®. Toxicology Letters 2004, 153, (2), 227-232. 224. Paumgartten, F. J. R., Influence of Maternal Toxicity on the Outcome of Developmental Toxicity Studies. J Toxicol & Env Health, Part A

2010, 73, (13-14), 944-951. 225. Brazilian government, Brazilian Pesticide Law - Federal Law In 1989; Vol. no. 7802. 226. Walsh, L. P.; McCormick, C.; Martin, C.; Stocco, D. M., Roundup inhibits steroidogenesis by disrupting steroidogenic acute regulatory

(StAR) protein expression. Env Health Persp 2000, 108, (8), 769. 227. EPA (United States Environmental Protection Agency), Glyphosate Summary Document - Registration Review: Initial Docket. In Office of

Pesticide Programs, E. P. A., Ed. regulations.gov: 2009; p 16. 228. De Roos, A. J.; Blair, A.; Rusiecki, J. A.; Hoppin, J. A.; Svec, M.; Dosemeci, M.; Sandler, D. P.; Alavanja, M. C., Cancer Incidence among

Glyphosate-Exposed Pesticide Applicators in the Agricultural Health Study. Environ Health Perspect 2005, 113, (1), 49-54.

48

229. Kier, L. D.; Kirkland, D. J., Review of genotoxicity studies of glyphosate and glyphosate-based formulations. Critical reviews in toxicology

2013, 43, (4), 283-315. 230. Séralini, G.-E.; Clair, E.; Mesnage, R.; Gress, S.; Defarge, N.; Malatesta, M.; Hennequin, D.; de Vendômois, J. S., Long term toxicity of a

Roundup herbicide and a Roundup-tolerant genetically modified maize. Food & Chem Toxicol 2012, 50, 4221-4231 (Retracted by Editor). 231. Bardocz, S.; Clark, A.; Ewen, S.; Hansen, M.; Heinemann, J.; Latham, J.; Pusztai, A.; Schubert, D.; Wilson, A. Seralini and Science: an

Open Letter. http://www.ask-force.org/web/Seralini/Seralini-Independent-Science-News-Open-Letter-20121002.pdf (Accessed 2013-12-03), 232. De Heuvels, P. Étude de Séralini sur les OGM: pourquoi sa méthodologie est statistiquement bonne.

http://leplus.nouvelobs.com/contribution/646458-etude-sur-les-ogm-une-methodologie-statistiquement-bonne.html (Accessed 2013-11-15), 233. Demortain, D., L'étude Séralini et ce qu'elle nous apprend sur la toxicologie réglementaire. Natures Sciences Sociétés 2013, 21, 84-87. 234. Vélot, C.; Morvan, C., Les conséquences de l'affaire Séralini. Statistique et Société 2013, 1, (1), 17-22. 235. Carmen, T., French scientists question safety of GM corn. Washington Post 2012-09-19, 2012. 236. Mackenzie, D., Study linking GM crops and cancer questioned. New Scientist 2012-09-19, 2012. 237. Pollack, A., Foes of modified corn find support in a study. New York Times 2012-09-12, 2012. 238. Arjo, G.; Portero, M.; Pinol, C.; Vinas, J.; Matias-Guiu, X.; Capell, T.; Bartholomaeus, A.; Parrott, W.; Christou, P., Plurality of opinion,

scientific discourse and pseudoscience: an in depth analysis of the Séralini et al. study claiming that RoundupTM Ready corn or the herbicide RoundupTM cause cancer in rats. Transgenic research 2013, 1-13.

239. OECD, Test No. 452: Chronic Toxicity Studies. OECD Publishing: 2009; p 16. 240. Hammond, B.; Dudek, R.; Lemen, J.; Nemeth, M., Results of a 13 week safety assurance study with rats fed grain from glyphosate tolerant

corn. Food and Chemical Toxicology 2004, 42, (6), 1003-1014. 241. Ollivier, L., A Comment on "Séralini, G.-E., et al., Long term toxicity of a Roundup herbicide and a Roundup-tolerant genetically modified

maize. Food Chem. Toxicol. (2012)," http://dx.doi.org/10.1016/j.fct.2012.08.005. Food Chem. Toxicol. 2012, 53, 458. 242. Robinson, C.; Holland, N.; Leloup, D.; Muilerman, H., Conflicts of interest at the European Food Safety Authority erode public confidence.

J Epidemiol & Comm Health 2013, 10.1136/jech-2012-202185. 243. Annas, G. J.; Elias, S., Thalidomide and the Titanic: reconstructing the technology tragedies of the twentieth century. American journal of

public health 1999, 89, (1), 98-101. 244. Soffritti, M.; Belpoggi, F.; Degli Esposti, D.; Lambertini, L.; Tibaldi, E.; Rigano, A., First experimental demonstration of the multipotential

carcinogenic effects of aspartame administered in the feed to Sprague-Dawley rats. Env Health Persp 2006, 114, (3), 379. 245. Carson, R., Silent spring. Houghton Mifflin Harcourt: New York, 1962; p 381. 246. Hilbeck, A.; Meier, M.; Trtikova, M., Underlying reasons of the controversy over adverse effects of Bt toxins on lady beetle and lacewing

larvae. Environmental Sciences Europe C7 - 9 2012, 24, (1), 1-5. 247. Stokstad, E., Field Research on Bees Raises Concern About Low-Dose Pesticides. Science 2012, 335, (6076), 1555. 248. Delborne, J. A., Transgenes and transgressions: scientific dissent as heterogeneous practice. Social Studies of Science 2008, 38, (4),

509-541. 249. Amesty International Argentina: Threats deny community access to research UA: 173/10 Index: AMR 13/005/2010.

http://www.amnesty.org/en/library/asset/AMR13/005/2010/en/303e9ee6-9138-405f-97fc-ed58965b76d0/amr130052010en.html (Accessed 2013-12-03),

250. Poulter, S. Cancer row over GM foods as study says it did THIS to rats... and can cause organ damage and death in humans. Cancer-row-GM-foods-French-study-claims-did-THIS-rats--cause-organ-damage-early-death-humans (Accessed 2012-09-19),

49

251. Science Media Center Study on cancer and GM maize - experts respond. http://www.sciencemediacentre.co.nz/2012/09/20/study-on-cancer-and-gm-maize-experts-respond/ (Accessed 2013-12-03),

252. Nature Editor, Editorial Note. Nature 2002, 416 (April 11), 601. 253. Quist, D.; Chapela, I. H., Transgenic DNA introgressed into traditional maize landraces in Oaxaca, Mexico. Nature 2001, 414, 541-543. 254. Séralini, G.-É., Tous cobayes! OGM, pesticides, produits chimiques. Flammarion Document: Mayennes, 2012; p 256. 255. Jaud, J.-P.; Torrenton, P., Tous Cobayes. In J+B Séquences Distribution: France, 2012; p 1h55m. 256. Schreider, J.; Barrow, C.; Birchfield, N.; Dearfield, K.; Devlin, D.; Henry, S.; Kramer, M.; Schappelle, S.; Solomon, K.; Weed, D. L.,

Enhancing the credibility of decisions based on scientific conclusions: transparency is imperative. Toxicol Sci 2010, 116, (1), 5-7.