Risk Assessment (Environmental)

1. INTRODUCTION

Originally the unwanted side effects on humans were the main concern regarding the use of pesticides. Only during the last 40 years, largely since Rachel Carson published her famous book “Silent Spring”[i], have environmental side effects also received significant attention. Many governments have therefore redrafted their pesticide legislation to include data requirements for environmental topics as well. In the mid 1970s, the pesticide industry started to carry out a significant amount of research on their newest active ingredients to cover these data requirements. Especially in the developed countries, methodologies were set up to estimate post-application pesticide concentrations, especially in the important environmental compartments of groundwater, surface water and soil. Estimated or Predicted Environmental Concentrations (EECs or PECs) were considered useful in the risk assessment process to identify possible effects on non-target organisms in the environment. In the last 10 to 15 years the methodology on environmental risk assessment has made enormous progress, not only through the development of fast and high capacity computers, but also through the development of equations describing physical and environmental processes based on sound science.

A fundamental starting point of all kinds of risk assessments is the following scheme[ii]:

2. DATA EVALUATION

For ecological risk assessment data on application, fate and ecotoxicity are used by the government registration authorities to prepare a policy decision on whether or not a substance should be allowed to be marketed in the country under consideration. In many cases industry first prepares a dossier on the substance in which the complete data set of information has been put together. The governmental authorities then use these data to estimate both the exposure and the hazard of the substance in the environment after a thorough evaluation of the data. All data are summarised and evaluated according to predefined, scientifically established criteria that are based on international test guidelines[iii]. The endpoints of a study are clearly determined and further used in the risk assessment where appropriate.

3. EXPOSURE ESTIMATION

Estimation of pesticide exposure for hazard assessment purposes is typically achieved either through (1) simple calculation based on previous information, (2) computer simulation modelling or (3) monitoring (field sampling and analysis) of pesticide residues. A field sample represents a pesticide residue level at single site at a single time. The very high cost of field monitoring often does not allow for sufficient samples to be collected and analysed to understand fully the range of exposure that results from pesticide application at a single site or across a region.

Computer modelling uses the relationship between previously monitored residues and field conditions at the monitoring site to estimate residues at other sites where no monitoring has been performed. Computer programs may also be used to estimate pesticide residue levels on or near a pesticide treated field at various times after application through knowledge of how weather interacts with specific pesticide properties to reduce the residue levels and concentrations. For computer modelling, data on application, physico-chemical properties and environmental fate and behaviour are used in the model to estimate the predicted or estimated environmental concentration (PEC or EEC).

Neither field monitoring nor computer simulation modelling alone is adequate to understand the impact of pesticide usage throughout an agricultural area. Monitoring by itself is usually too expensive because of the very large number of samples required. Computer modelling by itself may also be inadequate without some field monitoring as a check on the accuracy of the modelling. A combination of monitoring and modelled data may therefore provide the most accurate and cost-effective means of exposure assessment. Computer models that are based on data collected in other countries or regions may be acceptable without local monitoring data, if the model can be shown to represent the conditions of use where the model is being applied.

Ecotoxicity data, mainly single species information, are used to estimate the no-effect concentration for the ecosystem under consideration. The estimated concentrations and the no-effect concentrations are then compared to each other in the risk characterisation phase. If the ratio of both concentrations is considered not to meet the predefined standards, the substance cannot be registered unless additional information is made available to the authorities by the applicant. In the framework of risk management, authorities and applicant(s) may try to define circumstances in which a safe use of the substance is still possible. In this way the above mentioned scheme may be used again in a feedback loop. Generally, the risk assessment is carried out using a tiered approach. In the first Tier, reasonable, worst case assumptions are made to estimate the PEC or EEC and a comparison is made with the trigger value (the predefined standard) to evaluate the risk. If the comparison does not indicate risk, the substance may be approved for registration on this basis. If risk is perceived to be too high based upon this assessment, the evaluator proceeds to the next higher assessment Tier. At this level, additional information may be required or more specific, and therefore “less worst” case, assumptions are considered in order to derive a more realistic PEC or EEC. Again the comparison with the standard is carried out and a new risk quotient is determined. If this new quotient indicates no risk, the substance may be approved based upon this second tier evaluation. If risk is established, the next tier evaluation is entered and the process is repeated. This process is used and is described in several documents[iv]^{,[v] ,[vi] ,[vii] ,[viii] ,[ix]}.

4. HAZARD IDENTIFICATION

Environmental Risk Assessments are generally carried out for several non-target organisms. Usually these are:

• aquatic organisms (fish, daphnids, algae)
• birds and mammals
• terrestrial invertebrates
• honey bees
• earthworms
• soil micro-organisms
• non-target higher plants

As an example, for aquatic organisms, information on the ecotoxicity of the active substance and/or the formulated product to these organisms is required in the dossier submitted by the applicant. Depending on the results of the earlier tiers, additional information may be needed from higher tier, ecotoxicity studies. For example, in the first tier, acute and chronic ecotoxicity data for aquatic organisms are required. For the second tier, micro- or mesocosm studies should be carried out; whilst for the third tier, field data or monitoring data are required to finally conclude whether or not adverse effects on aquatic organisms are occurring. The result of an acute ecotoxicity study is an LC₅₀ or EC₅₀ value (lethal or effect concentration for 50% of the population studied). Chronic studies reveal a NOEC-value (no effect concentration) for the different species. Finally, micro- or mesocosm studies may be required to determine the NOEC value for the ecosystem studied. An assessment factor can be applied to these different LC, EC, NOEC(species) or NOEC(ecosystem) values. This factor depends on the amount and /or quality of the ecotoxicity data available. The purpose of the assessment factor is to have a more conservative approach to safety, because there is always a difference between the estimations in the risk assessment and the actual reality in the field. The assessment factor takes into account this uncertainty. The assessment factor may vary from 1000, e.g. if only one LC₅₀ or EC₅₀ value is available; to 100 if three acute values are available; or to 10, if three chronic data are available and sometimes even less, to 2 or 5, if appropriate field or monitoring data are available to characterise the risk.

The approach described above for aquatic organisms is also carried out for the other organisms. Sometimes the methods currently developed are not yet as clearly defined as for aquatic organisms, but the same assessment strategy applies.

To perform the environmental risk assessments requires a level of experience and expert judgement that may be built up slowly as a result of sufficient decisions having to be made. Several authorities have established guidance documents to support the risk assessor to carry out this task. Some guidance documents may be assessed by internet at the following sites:[x]^{,[xi] ,[xii] ,[xiii] ,[xiv] ,[xv] ,[xvi]}.

5. RISK CHARACTERISATION

At the risk characterisation stage, the results of the exposure estimation and the results of the hazard identification are compared with each other. Generally, it is assumed that if the result of the ratio between the predicted environmental concentration (PEC or EEC) and the NOEC (or equivalent) exceeds a value of 1.0, an additional safety factor taking into account the uncertainty may need to be incorporated. Authorities and the applicant may consider the consequences of such a situation. A tiered approach may apply for the exposure estimation but also for the hazard identification. Therefore, more realistic ecotoxicity data may need to be developed and a more sophisticated risk assessment may be performed. Finally, authorities may come to the conclusion that the use of the substance is considered safe with specific risk mitigation measures or the applicant may come to the conclusion that satisfying the data requirements is too costly. In the first case registration is granted in the other case the registration is denied[xvii].

REFERENCES