General enquiries on this form should be made to: Defra, Science Directorate, Management Support and Finance Team, Telephone No. 020 7238 1612 E-mail: research.competitions@defra.gsi.gov.uk SID 5 z Research Project Final Report Note In line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The SID 5 (Research Project Final Report) is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website. A SID 5 must be completed for all projects. A SID 5A form must be completed where a project is paid on a monthly basis or against quarterly invoices. No SID 5A is required where payments are made at milestone points. When a SID 5A is required, no SID 5 form will be accepted without the accompanying SID 5A. • z Project identification 1. Defra Project code 2. Project title PS2302 Acceptability of Pesticide Effects on Non-Target Species 3. Contractor organisation(s) This form is in Word format and the boxes may be expanded or reduced, as appropriate. Crane Consultants Horticulture Research International Virginia Insitute of Marine Sciences University of Warwick University of Reading University of Edinburgh Imperial College MORI Social Research Institute INFORM Training and Communication ACCESS TO INFORMATION The information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000. Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors. SID 5 (2/05) 4. Total Defra project costs 5. Project: Page 1 of 30 £ 143750.00 start date ................ 01 November 2002 end date ................. 31 October 2004 6. It is Defra’s intention to publish this form. Please confirm your agreement to do so....................................................................................YES NO (a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow. Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer. In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000. (b) If you have answered NO, please explain why the Final report should not be released into public domain Executive Summary 7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work. EU Directive 91/414/EEC requires there to be no unacceptable effects on the environment from the use of pesticides. This report examines the ecological and socioeconomic bases for establishing Ecologically Acceptable Concentrations (EACs), with the following specific objectives: 1. To produce a comprehensive literature review on current knowledge and understanding of the ecological significance of long-term risks from pesticides to non-target species. 2. To run focus groups with particular stakeholders and to perform a statistically robust public opinion survey to determine the spectrum of views on EACs and the factors that influence these views. 3. To combine information from 1 and 2 above within a formal decision analysis framework to provide PSD with a coherent, logical, yet flexible decision tool for identifying EACs under Directive 91/414/EEC. Review of the ecological literature showed that although much is known about the potential for pesticides to cause long-term risks to non-target organisms, there is little information on the frequency or extent that these risks are realised in the natural environment. Current risk assessment methods would allow reasonable predictions of longterm effects of pesticide application if three changes were instituted. First, more population-based laboratory studies should be applied in predictive pesticide risk assessment. Second, risk assessment should include as much effort on collating and integrating ecological knowledge into the assessment in Tier 1 as is currently expended on gathering chemical and toxicological information on exposure and effects. Production of a formal conceptual ecological risk assessment model for each product or active substance for which authorisation is sought would provide an appropriate framework for integrating and applying such knowledge. Third, in acknowledgment of the uncertainties in the predictive risk assessment process, more post-authorisation monitoring should be done. The review clearly identified the existence of large uncertainties when extrapolating from laboratory tests to environmental effects, and the continuing need for elements of judgement to be applied when authorising active substances. The next stage of the project examined what stakeholder and public values should be taken into account when these judgements are made. Stakeholders in focus groups were almost entirely concerned with the potential effects of pesticides on animal and plant population viability, and micro-organism function. Stakeholders in focus groups also recognised that a trade-off exists between the potential economic advantages of responsible pesticide use and the potential disadvantages of individual poisoning events. A subsequent public opinion survey of 2000+ respondents showed that although pesticides are widely used in homes and gardens, their use on farm crops remains of concern to the public. Concerns are greatest on issues of human health and food quality, but potential SID 5 (2/05) Page 2 of 30 environmental effects are also an issue for a substantial number of people, particularly if attractive species could be affected. In contrast to the stakeholders in focus groups, substantial numbers of the public would remain very concerned about the effects of pesticides even if they affected only individual organisms and not populations. However, the desire to purchase food that has not been produced with the use of pesticides depends on its price relative to the total available budget, with only one-fifth of the public, or fewer, prepared to pay a substantial additional amount for this. Pesticide risk managers need to ensure that long-term adverse pesticide effects on populations of plants and animals, or on microbial function, do not occur. This can be achieved by use of improved predictive effects and exposure models combined with more effective post-authorisation monitoring. Individual vertebrate deaths should be avoided, if at all possible, even if these deaths are unlikely to affect population size or viability, and particularly if the organisms at risk are those that the public care about. However, a substantial proportion of the public is unlikely to be satisfied by any risk mitigation for pesticides, even if evidence shows that it results in negligible risks to the environment. The information from earlier project components was used to construct a formal decision tree to assist risk managers and their advisors in coming to consistent decisions on pesticide authorisations, informed by public values. Scientific evidence takes precedence over opinion in this framework. However, public values are used when decisions are value-laden rather than evidence-based. Project Report to Defra 8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: z the scientific objectives as set out in the contract; z the extent to which the objectives set out in the contract have been met; z details of methods used and the results obtained, including statistical analysis (if appropriate); z a discussion of the results and their reliability; z the main implications of the findings; z possible future work; and z any action resulting from the research (e.g. IP, Knowledge Transfer). SID 5 (2/05) Page 3 of 30 1. INTRODUCTION EU Directive 91/414/EEC (EC 1993) harmonises the registration of pesticides throughout the European Union. The environmental objectives of the Directive are contained in the preamble which states that: ‘…plant protection products can have non-beneficial effects upon plant production…[and]…their use may involve risks and hazards for humans, animals and the environment, especially if placed on the market without having been officially tested and authorised and if incorrectly used. …the provisions governing authorisation must ensure a high standard of protection, which, in particular, must prevent the authorisation of plant protection products whose risks to health, groundwater and the environment and human and animal health should take priority over the objective of improving plant production. ….it is necessary, at the time when plant protection products are authorised, to make sure that, when properly applied for the purpose intended, they are sufficiently effective and have no unacceptable effect on plants or plant products, no unacceptable influence on the environment in general, and, in particular, no harmful effect on human or animal health or on groundwater.’ 'Animals' are defined in the Directive as 'belonging to species normally fed and kept or consumed by man,' so this does not include wild flora or fauna. 'Environment' is defined as 'water, air, land and wild species of fauna and flora, and any interrelationship between them, as well as any relationship with living organisms.' A key phrase in the Directive, which is quoted above, is that pesticides should have 'no unacceptable influence on the environment in general…' Crane and Giddings (2004) argue that the acceptability or otherwise of an environmental influence (i.e., an adverse effect on the environment) clearly involves social values and should be based upon consultation with a wide group of stakeholders, including environmental scientists, government regulators and pesticide manufacturers. They also argue that representatives from the wider community and environmental Non-Governmental Organisations should be consulted (Crowfoot and Wondolleck 1990). However, to date, the concept of the ‘acceptability’ of pesticide effects in Europe, through definition of Ecologically Acceptable Concentrations (EACs), has been developed and refined largely by scientists from the regulatory and business communities, who may not represent the full spectrum of views on what constitutes an acceptable environmental effect. At one time scientists were seen as experts who could provide an unbiased opinion on scientific matters, and whose advice would be accepted by decision makers on the basis of that acknowledged expertise. The unbiased nature of this knowledge was generally accepted until some damaging decisions, particularly in the agricultural sector, caused questions to be asked about the decision-making process and the nature of scientific evidence. Research in the social sciences has also turned up complex relationships between scientific results and assessment, trust and public perception (Douglas 2000). With an increasingly sceptical society it is important that scientists and non-specialists communicate and, in particular, discuss complex ideas. This is necessary because the public’s perception of risks may well diverge significantly from that of specialists (Frewer 2004, Hansen et al. 2003). An individual’s perception of risk depends upon an often intuitive judgement of the probability of occurrence and the severity of the consequences of that risk. This perception is usually a judgement that is made without consideration of associated benefits, and risks only become acceptable to an individual when they are able to balance them with these benefits. However, even if individuals agree on the degree of risk, they may still disagree on its acceptability because of differences in their level of expertise and education, their gender or their personal values. For example, the wholesale rejection of Genetically Modified Crop technology by the British public was significantly at odds with the views of many scientists who chose to see the technology as safe and controllable (Frewer 2003, Frewer et al. 2004, Tait 2001a). Additionally, of course, the examination of motives within science can itself be questioned as the ongoing pressures of funding, essential to the continuation of particular research lines, and the requirement for novelty in research in general, essential to publication and career development in science, means that there is a strong science agenda which may be at considerable variance with wider societal wants and needs. This report describes a study in which the ecological literature was reviewed to identify evidence for the effects of pesticides in natural environments. The views of specific stakeholders were then elicited in focus groups on the acceptability of pesticide effects from normal agricultural use on crops, and a public opinion survey of more than 2000 respondents across the UK was run to determine what the public value and their attitudes towards the use of pesticides. Finally, information from these activities was combined to develop a coherent decision framework for pesticide authorisation that consistently incorporates public values, when appropriate. 2. AIMS AND OBJECTIVES The overall aim of this project was to establish the ecological and socioeconomic bases for defining Ecologically Acceptable Concentrations (EACs). The specific objectives were: 1. To produce a comprehensive literature review on current knowledge and understanding of the ecological significance of long-term risks from pesticides to non-target species. 2. To run focus groups with particular stakeholders and to perform a statistically robust public opinion survey to determine the spectrum of views on EACs and the factors that influence these views. 3. To combine information from 1 and 2 above within a formal decision analysis framework to provide PSD with a coherent, logical, yet flexible decision tool for identifying EACs under Directive 91/414/EEC. SID 5 (2/05) Page 4 of 30 3. REVIEW OF SIGNIFICANCE OF LONG-TERM ECOLOGICAL RISKS FROM PESTICIDES TO NON-TARGET ORGANISMS 3.1 Long term implications of pesticide exposure for ecosystems 3.1.1 Habitat and resource modification Can plausible inferences be made about long-term habitat or resource modifications caused by the approved use of pesticides, based on evidence in the literature? Without doubt, agricultural activities have produced long-lasting changes on the British landscape and associated biota. The specific suggestion that pesticides play a significant role in habitat and ecological resource change is supported by the following theory and evidence. 1. The community conditioning hypothesis (Matthews et al. 1996) suggests that community structure may remain in an altered state for a long time after pesticide residues have dropped to toxicologically insignificant levels. Sheffield and Lochmiller (2001) provide an example of modified species interactions that persisted for longer than pesticide residues. 2. The removal of a keystone or dominant species alters habitat qualities (Power et al. 1996). Although it is important to prevent this (Mills et al. 1993), current toxicity tests performed during the authorisation of pesticides cannot ensure that keystone species populations remain extant. Therefore habitat or resource alteration due to unintentional removal of keystone species could occur. 3. Although commonly invoked, the general application of the ecosystem redundancy hypothesis has not been adequately addressed and an equally plausible hypothesis, the rivet popper hypothesis, according to which the loss of any species is of concern, might be a more accurate and conservative depiction of ecosystem consequences of pesticide exposures (Pratt and Cairns 1996). 4. British studies of grey partridge (Perdix perdix) provide strong evidence that herbicide use has diminished habitat quality, and consequently, decreased partridge population densities (Blus and Henny 1997, Ewald and Aebischer 1999). 5. Population densities of thirteen farmland bird species have fallen 30% on average in the United Kingdom from 1968 to 1995 but twenty-nine generalist species have increased (Krebs et al. 1999). This suggests a decrease in habitat or resource quality for bird species that prefer farmlands. 6. Other instances of medium-term adverse effects of agricultural practice on aquatic species (e.g., Furse et al. 1995) and terrestrial species (e.g., Martin et al. 2000) have been reported, suggesting that longer-term effects are plausible. In contrast to the above, the following theory and evidence detracts from any definitive statement that pesticides play an important role in general changes to the farming landscape and associated biota. 1. 2. 3. 3.1.2 Most communities display a degree of resistance to change and an ability to recover after stress-induced change (Cairns and Niederlehner 1993, Pratt and Cairns 1996). Although current toxicity tests in support of pesticide authorisations generate imprecise predictions of population or community vitality or viability, there are indications that effects information from these tests are, on average, conservative relative to population effects (e.g., Caslin and Wolfe 1999, Bishop et al. 2000, Forbes and Calow 2003). Other plausible causal agents co-occur with pesticides, such as physical habitat modification, introduced species, and increased fertilizer use. The presence of these other potential causes confounds the identification of pesticides as a major factor contributing to long-term habitat or resource modifications. Local extinction and reduced biodiversity Evidence can also be compiled from the literature about the risk of local extinction and reduced biodiversity. Support for the view that pesticides contribute to local extinctions or reduced biodiversity includes the following: 1. Laboratory studies have identified mechanisms capable of disrupting species interactions and, consequently, species persistence and community structure. 2. Publications addressing widespread declines in bird (Beaumont 1997, Krebs et al. 1999) and amphibian (Wake 1991) species suggest that pesticides might play a role. Sparling et al. (2001) provide one example of amphibian decline that is directly linked to pesticide use. 3. Mineau (2002) suggests that bird mortality likely occurs frequently in Canadian agricultural fields. 4. Pesticides can modify endocrine function in ways that impact population viability (Gray et al. 1998). For example, changes in alligator sexual characteristics and reproductive viability in Florida lakes are notionally linked to endocrine-modifying pesticides (Crain et al. 1998, Guillette et al. 1996, 1999, Milnes et al. 2002). 5. Pesticides can adversely influence vital rates, and changes in these vary in complex ways among species (Forbes and Calow 2003). SID 5 (2/05) Page 5 of 30 6. Many indirect effects, not quantified in conventional tests, are documented for diverse taxa exposed to pesticides. These effects could lead to local extinctions and, gradually, to reduced diversity (Preston 2002). 7. Although commonly invoked, the general applicability of the ecosystem redundancy hypothesis has not been adequately ascertained. The equally plausible, and more conservative, rivet popper hypothesis might provide a more appropriate framework for protecting ecosystems (Pratt and Cairns 1996). 8. Long-term risk to mammalian species has been documented (Muller et al. 1981). 9. Some mesocosm studies (Van den Brink 1996, Woin 1998) document species abundance shifts as a consequence of pesticide exposure. 10. Bird populations are impacted by both legacy pesticides (e.g., Hickey and Anderson 1968) and modern pesticides (e.g., Schroeder and Sturges 1975, Savidge 1978, Blus and Henny 1997, Martin et al. 2000, Mineau 2002, Sotherton and Holland 2003). Some impacts have long-term consequences for bird populations. In contrast to the above, theory and evidence detracting from any conclusion that pesticides influence local extinction and reduced biodiversity are the three provided above for habitat and resource modification, and the following, 1. 2. 3. 3.1.3 Phenotypic plasticity and life history strategies are important features of most populations (Stearns 1992) and could lessen the effects of pesticides on populations (Stearns and Crandall 1984, Sibly and Calow 1989). Harrington and MacDonald (2002) saw no indication in the sparse data available that the decline in UK mammalian species is linked to pesticides. Some bird studies suggest that sublethal effects on nestling (e.g., Bishop et al. 2000) and adult birds (Blus and Henny 1997) occur, but compensation results in these effects having no apparent adverse consequences on overall fitness. Recovery and recolonisation Evidence supporting a negative view about community recovery and recolonisation after pesticide exposure includes the following. 1. 2. 3. 4. There can be a significant delay between the removal of a keystone species and its replacement in a community (Ernst and Brown 2001). Negative impacts on mammalian populations can persist longer than pesticide residues do in treated environments (Sheffield and Lochmiller 2001). Although life history traits can help identify mammal species that rely on dispersal to remain viable (Fagan et al. 2001), such traits are not formally considered in present risk assessment methods. The community conditioning hypothesis (Matthews et al. 1996) suggests that community structure can remain altered after pesticide residues have dropped to toxicologically insignificant levels. In contrast to the above, evidence detracting from any conclusion about the adverse effects of pesticides on community recovery and recolonisation include all of those provided above for habitat and resource modification, and local extinction and reduced biodiversity. 3.2 An ecological vantage on pesticide risk assessment 3.2.1 Organismal Most of the lower tier information applied in pesticide risk assessment is interpreted from an autecological vantage. Such a vantage is directly useful for endangered or threatened species because the taking of even one of these individuals is illegal, so information about direct effects on individuals is essential. Such information might also be relevant for charismatic species for which the taking of an individual would be very undesirable. For example, a 1984 poisoning of Atlantic brant (Branta bernicla) led to the US regulatory authorities discontinuing diazinon use on turf grasses despite the lack of any evidence of risk to populations of this unlisted species (Bascietto 1998). Application of diazinon to golf courses and sod farms was regarded as constituting an unreasonable risk to individuals of a charismatic species. The autecological vantage is also useful, but insufficient, for predicting effects on species populations. This approach has a long standing in ecology where it was used to examine the relationship between individual organisms or species and their physical, chemical, and biological environment. Liebig’s law of the minimum (Liebig 1840), Shelford’s law of tolerances (Shelford 1911, 1913), and the concept of a fundamental niche grow out of this premise that knowledge of the tolerances or requirements of individuals can be used to predict species distributions and abundances. However, this vantage is insufficient for explaining or predicting all aspects of population or community ecology, and the synecological vantage is an essential one for general prediction or description of ecological systems (Preston 2002). This view can perhaps be summed up by a quote from over twenty years ago: Although this discussion may appear hostile to single species toxicity testing efforts, it is not intended to be. Single species tests are exceedingly useful and are presently the major and only reliable means of estimating probable damage from anthropogenic stress. Furthermore, a substantial majority, perhaps everyone in this meeting is certainly aware of the need for community and system level toxicity testing. How then does one account for the difference between awareness and performance? (Cairns 1984) SID 5 (2/05) Page 6 of 30 3.2.2 Population The EC Guidance Document on Aquatic Ecotoxicology (EC 2002a) and EC Guidance Document on Terrestrial Ecotoxicology (EC 2002b) explicitly state that species populations are to be protected by the EC risk assessment process for pesticides. Although the issue is gradually being remedied, the tools most often used in pesticide risk assessment provide imperfect insights into population level effects. For example, Grant (1998) indicates that substantial changes in some vital rates such as those often measured in toxicity tests can have little impact on populations because density-dependent population dynamics shift to compensate for the toxicant’s effects. Newman (2001) identified four reasons why current toxicity tests might not provide sufficiently accurate predictions of consequences for field populations. First, toxicity testing often focuses on the most sensitive stage in an organism’s life cycle based on the assumption that protection of individuals in this most sensitive stage will ensure protection of the associated species population. But the most sensitive stage of a life history may not be the most important one relative to maintaining a viable population, as illustrated by Kammenga et al. (1996). Second, metrics such as the LC50, LOEC or NOEC cannot be incorporated directly into demographic models used to project population change through time. Third, postexposure mortality is ignored in most laboratory-derived metrics but field populations can experience significant mortality after exposure to a toxicant ends (e.g., Newman and McCloskey 2000, Zhao and Newman 2004). Fourth and finally, the underlying assumptions for some conventional concentration-effect models have not been resolved. As an example, the probit model is based on the concept of the individual effective dose or concentration. It has been hypothesized that each individual has a unique concentration above which it will die and below which it will survive exposure (Bliss and Cattell 1943, Finney 1947). The distribution of tolerances in any population is assumed to be lognormal. An alternate hypothesis is that each individual has the same chance of dying as any other and whether it dies depends on chance alone. Which of the hypotheses dominates in a particular population exposure scenario currently remains ambiguous. The consequences of this ambiguity are significant because the two hypotheses predict very different population consequences with repeated exposures (Newman and McCloskey 2000). A risk assessment for atrazine by Solomon et al. (1996) provides another illustration of subtle but crucial population-related issues left unexplored in current ecological risk assessment (ERA) approaches. Although not considered in the very thorough ERA by Solomon et al. (1996), Hayes et al. (2002a,b, 2003) and Withgott (2002) indicate that atrazine acts as an endocrine modifier and, more controversially, is suspected to affect the reproductive fitness of amphibians at environmentally realistic concentrations. Dodson et al. (1999) found that atrazine also influenced the production of males in Daphnia pulicaria cultures, and that D. pulicaria fecundity and survival were much less sensitive than male production. Both of these effects could significantly change population vitality but were not considered in the original ERA for atrazine. Solomon et al. (1996) focused on the effects of atrazine on primary production, maintenance of macrophyte community structure, and long-term viability of fish populations in their higher tier ERA. In their effects characterisation, they commented on the lack of information for amphibians but concluded that, “the limited data suggest that amphibians are tolerant of atrazine.” Appropriately designed laboratory (e.g., Van Der Hoeven and Gerritsen 1997, Snell and Serra 2000), mesocosm (e.g., Van Den Brink 1996, Sherratt et al. 1999), enclosure (e.g., Caslin and Wolfe 1999, Wang et al. 2001), and field (e.g., Schroeder and Sturges 1975, Savidge 1978) studies can provide valuable insights into the population effects of pesticides, and more studies of these types are included in both predictive and retrodictive pesticide risk assessments each year. As more such tests are applied to ERA, the strength of associated inferences about long-term effects of pesticides to populations will improve. As is occurring now in conservation biology, more and more emphasis is slowly being placed in ERA on the risk of local extinction under specified exposure conditions (Tanaka 2003). 3.2.3 Community and species assemblage As discussed already, interactions among species populations are also essential in any ERA to understanding the long-term consequences of pesticide use. The value of individual-based effect metrics from the laboratory is ambiguous for this purpose and, in some cases, demonstrably inadequate. Laboratory-based designs for quantifying community effects exist (e.g., Cairns et al. 1986) but lack the realism of field or mesocosm studies. Field studies have tremendous value for this purpose and methods exist for extracting insights about effects on species assemblages or communities (e.g., Savidge 1978). Studies involving mesocosms or enclosures (e.g., Liber et al. 1992, Sheffield and Lochmiller 2001) may also provide valuable information about these effects but the practical temporal scale is shorter for most mesocosm and enclosure studies than for field studies so that, although such studies are more realistic than laboratory studies, they still may not reflect the true field situation with adequate accuracy or precision (Crane 1997). 3.3 A conservation biology vantage on pesticide risk assessment 3.3.1 Pesticides and non-target species of conservation concern Changes in the way that farmland is managed have unquestionably reduced insect abundance over the last 30 years in the UK and other countries (Wilson et al. 1999, Benton et al. 2002). Pesticides could have contributed to this decline both through direct poisoning by insecticides (Dover et al. 1990; Cardwell et al. 1994) and the removal of food plants by herbicide application (Marshall et al. 2003). Many bird species declining on farmland are reliant on insects and other invertebrates during the breeding season as staple nestling food (Cramp 1988, 1992, Cramp and Perrins 1993, 1994a, 1994b, Benton et al. 2002, Gruar et al. 2003). Furthermore, there is a correlation between the declining abundance of insects on farmland and the abundance of several bird species (Benton et al. 2002). Such a correlation, however, is not definitive proof that pesticide SID 5 (2/05) Page 7 of 30 usage is the principal cause of the decline. There is ample evidence that agricultural intensification is responsible for reducing bird populations on farmland (Chamberlain et al. 1999; Henderson et al. 2000) and, while pesticide usage has evolved during the intensification process, it is not the only practice that has changed. For example, the UK corncrake (Crex crex) population crashed during the 20th century through the intensification of farmland practices (Green and Gibbons 2000), but here the driving factor was the shift from hay cutting to silage production (Green and Stowe 1993, Stowe et al. 1993). Silage is cut much earlier in the year than hay with the consequence that corncrakes, which nest in long grass, are unable to produce fledglings before the nests are destroyed during mechanical cutting of grass (Broyer 1994, Green 1996). The skylark, Alauda arvensis, and lapwing, Vanellus vanellus, provide further examples of species that have declined on farmland through reasons other than pesticide application (Chamberlain et al. 2000, Wilson et al. 2001). There are few examples where pesticides are known to have affected the population sizes of non-target species in the long term. Exceptions include the bioaccumulation of organochlorine (OC) pesticides through food chains into top avian predators, such as the peregrine falcon, Falco peregrinus (Ratcliffe 1980) and the sparrowhawk, Accipiter nisus (Newton 1986), and the effect of the loss of insects in field margins on the grey partridge (Chiverton 1999), as described earlier. Once the effect of OC pesticides was realized, appropriate legislation was introduced in developed countries and the bird of prey populations recovered to earlier levels (Millsap et al. 1998, Wingfield Gibbons et al. 1993, Horne and Fielding 2002). For the grey partridge, the introduction of conservation headlands conclusively demonstrated a link between pesticide application (both herbicide and insecticide) and a reduction in insect food required by the precocious grey partridge chicks (Green 1984, Borg and Toft 2000, Southwood and Cross 2002). The headland experiments further showed that pesticide application could be sympathetically designed to achieve both acceptable agricultural and conservation targets. Where implemented, headlands have resulted in a dramatic increase in the local abundance of grey partridges (Rands 1985, 1986; Chiverton 1999). In addition, a remarkable increase has also been recorded in the numbers of certain very rare species of farmland flowers, such as pheasant’s-eye, Adonis annua, shepherds needle, Scandix pectin-veneris, and cornflower, Centaurea cyanus. Headlands are also more heavily used by butterflies (Dover et al. 1990) and small mammals (Tew et al. 1992). Despite a general lack of direct evidence, pesticides and veterinary medicines are frequently cited as possible factors causing population declines in currently scarce species. For example, the hornet robberfly, Asilus crabroniformis, breeds on farmland and is associated with large grazing animals (Holloway et al. 2003a). In Britain, the species is Biodiversity Action Plan listed and scarce, and exists as a series of fragmented populations (Smith 2000, Clements and Skidmore 2002). Ivermectin, a veterinary medicine, is cited as a factor contributing to the decline of hornet robberflies (Smith 2000). While research has shown that ivermectin can influence the numbers of coprophagous flies emerging from dung (McCracken and Foster 1993, Wardhaugh et al. 2001), the effect is considerably less dramatic for dung beetles (Kadiri et al.1999). By extension, it is therefore assumed that ivermectin can negatively influence the hornet robberfly (Clements and Skidmore 1998, Smith 2000, 2001), even though neither a direct nor indirect effect has ever been shown. There is a general assumption within the conservation community that pesticides and other anthropogenic chemicals are detrimental to wildlife. Concern about pesticide application extends mainly to the effects of spray drift. For example, where pesticide application is occurring close to a Site of Special Scientific Interest, a barrier such as scrub, is often allowed to develop between the two sites to intercept any drift (D. Sheppard, English Nature, pers. comm.) as a precaution against any possible adverse effects. Many species have become rare on farmland and beyond as a result of habitat loss and degradation. Consequently, many populations are highly fragmented across the landscape substantially altering plant (Jacquemyn et al. 2003, Luoto et al. 2003) and animal (Sherman and Runge 2002, Schmiegelow and Monkkonen 2002, Ryttman 2003, Bellamy et al. 2003, Watson et al 2003) population dynamics. Some populations, particularly of insect species such as silver spotted skipper, Hesperia comma (Hill et al. 1999), Glanville fritillary, Melitaea cinxia (Saccheri et al. 1998) and bog fritillary, Proclossiana eunomia (Sawchik et al. 2002), but also some vertebrate species such as red squirrels, Sciurus vulgaris (van Apeldoorm et al. 1994) and nuthatches, Sitta europaea (van Langevelde 2000), function as metapopulations, i.e., a series of populations only loosely connected through occasional dispersion of individuals among populations (Hanski 1999). A situation that is also common in the wild is where local populations have gone extinct and the distance separating populations is too great to allow exchange of individuals through dispersion (Hill et al. 1996, Epperson 2000, Kudoh 2001, Doebeli and Killinback 2003). When this happens, genetic variation in an isolated population begins to erode at a rate negatively correlated with the size of the population (Frankham et al. 2002). Small populations go through population bottlenecks and, particularly if the population remains small for several generations (inevitable if a species is rare), a substantial amount of genetic variation can be lost (Frankham et al. 2002). K-strategy species may be at more risk from extended bottlenecks than r-strategists due to their slower rates of recovery (Begon et al. 1986). Notionally, lower genetic variation results in a decline in adaptability, an increased probability that an entire population will be lost as a result of a single event, and a general reduction in fitness (Ralls and Ballou 1983, Frankham et al. 2002). There are examples of species that survive in the wild with apparently very little genetic variation, for example the northern elephant seal, Mirounga angustirostrus (Hoelzel et al. 1993, Hoelzel 1999) and the fallow deer, Dama dama (Pemberton and Smith 1985) in Britain but, of course, those populations that have disappeared as a result of reduced genetic variation no longer exist to allow their levels of variation to be examined. Madsen et al (1999) working with adders, Vipera berus, unequivocally demonstrated that low genetic variation is associated with low fitness for this species. With rare, fragmented populations in agricultural landscapes, pesticides could play a role in the longterm reduction in fitness of some populations by repeatedly driving them through bottlenecks. Each time this happens, the genetic variation would be reduced more and more until the population enters an extinction vortex (Tanaka 2000, Lane and Alonso 2001, Rowe and Beebee 2003, McGinnity et al. 2003) from which there is no prospect of recovery. In addition to this process, if pesticide application results in the loss of one or more populations within a metapopulation, increasing the distance among populations to beyond the maximum for successful emigration, populations could become isolated and enter a process of genetic erosion. SID 5 (2/05) Page 8 of 30 3.3.2 Future considerations The risk of pesticide application to, for example, vertebrate wildlife is currently assessed using small numbers of bird (usually Japanese quail, Coturnix coturnix, or Northern bobwhite quail, Colinus virginianus, and mallard ,Anas platyrhynchos) or mammal (e.g., rat, Rattus norvegicus or mouse, Mus spp.) species under laboratory conditions. It is fair to say that pesticide risk assessment requires a leap of faith to extend the results from the small number of species tested under artificial conditions to a natural and complex field scenario. Furthermore, proposals for the safe use of pesticides under field conditions often focus on species of economic value. For example, to protect honeybees, Apis melifera, it is recommended that spraying be carried out early in the morning or late in the evening to coincide with the cooler temperatures before and after flight. However, bumblebees Bombus sp. and many species of solitary bees are able to fly under cooler conditions than honeybees and become non-target casualties of pesticide application. It is difficult to integrate pesticide considerations into conservation because of the multitude of possible effects. Consequently, there is a general precautionary assumption made that pesticides applied for agricultural purposes are detrimental to wildlife and that sites valuable to conservation, for whatever reason, should be protected from pesticides [although note that pesticides are sometimes applied for conservation purposes, for example to control invasive weed species (http://www.english-nature.org.uk/pubs/Handbooks/default.asp)] It is possible that the role that pesticides might play in the reduction of animal and plant populations, through whatever routes, may come under closer scrutiny in the future for several reasons. First, the increased awareness of long-term genetic consequences of short-term poisoning events may focus more attention on the distribution of pesticides across farmland habitats. Another issue is global warming (Hulme and Viner 1998) resulting in weather pattern changes in many countries. It is quite possible that the growth rate of arable weeds could increase considerably in temperate regions as a result of warmer, and probably wetter, conditions at particular times of the year (Hossel 2001). Pest insects might also survive in higher numbers. For example, the winter of 1989/90 was exceptionally mild in the UK, which triggered a big increase in the number of aphids spreading disease to winter cereal crops. Crop yields fell by as much as one tonne per hectare (Hossel 2001). Global warming is likely to result in increased pesticide application. However, the general public have become more aware of the amount of toxic substances used in food production and an increase in the quantities applied is likely to meet opposition, so outcomes are difficult to predict. Another factor that may result in increased scrutiny of pesticides is governmental activity such as the biodiversity strategy for England entitled ‘Working with the Grain of Nature’ (http://www.defra.gov.uk/wildlifecountryside/ewd/biostrat/). The target is to ensure that species are part of ‘healthy functioning ecosystems’ and to ensure that ‘biodiversity considerations become embedded in all the main sectors of economic activity.’ With a focus on ecosystems comes an acceptance that species exist within complex food webs and communities (Giler and O’Donovan 2002). A change of practice affecting one component of such a community could have a cascade effect all the way up to a species of conservation concern. Agricultural practices are considered to be very important within the context of UK government’s overall vision for the country’s biodiversity. 3.3.3 Long-term effect data required by conservation community There is a paucity of data that unequivocally demonstrates a link between pesticide use and long-term population reductions in farmland environments. For example, most of the species of conservation concern in the UK are not associated with farmland (www.ukbap.org.uk) and it is generally accepted that, for the majority of species, habitat loss or alteration is clearly the most significant factor threatening species extinction at the moment. Consequently, effects of pesticides are rarely considered. Part of the reason for this is that following a poisoning event, long-term monitoring is infrequently carried out and the potential ways in which pesticides could alter community structure are not studied. The investigation of both of these issues is limited by the amount of available resource and with so many immediate problems, long-term or more complex considerations fall low on the agenda. Furthermore, with the political desire in many countries to see wildlife conservation largely directed by local communities, the scientific expertise to appreciate and investigate the more insidious, but nevertheless potentially damaging, changes is not usually available. This type of research is the realm of a few academics within universities or overstretched non-government organizations. Management of farmland to consider biodiversity is generally voluntary and often entails joining an incentive scheme, such as the Countryside Stewardship Scheme in the UK (http://www.defra.gov.uk/erdp/schemes/css/default.htm). The way that farmland within the scheme is managed is dependent on the conservation objectives of that part of the countryside. Within such schemes, particular pesticide application regimes need to be approved and here knowledge of the longer-term effects of pesticide application would be very useful. For example, spraying with a given active substance may result in the decline of a suite of butterfly species, but how long would it take for the butterfly community to recover? Armed with this information, regulatory authorities may be able to recommend a suitable period of time between subsequent sprays to maintain both the butterflies and to achieve the pest control objective. Alternatively, a pesticide may be particularly detrimental to a small number of species that could affect community structure and ultimately a species of conservation concern. For example, a proposed herbicide application could eliminate food for meadow grasshoppers, Chorthippus parallelus, in field margins in an area. This species is not rare, but its powers of dispersal are very limited (Chinery 1993) so it could take some years for the area to be recolonized. Meadow grasshoppers are the staple food source of growing cirl bunting (Emberiza cirlus) nestlings, so the proposed application regime could ultimately eradicate a rare, target species through an indirect route. In conclusion, there are very few examples known to conservationists in which pesticides have been shown to have direct or long-term damaging effects on wildlife. However, potential effects of pesticides are rarely considered in conservation and pesticide issues, for example in connection with nature reserves, are examined on a case-by-case basis. The conservation community would welcome a general reduction in the use of pesticides on farmland, often as a result of the perceived impact of their use rather than direct evidence. Reductions are most likely to be achieved through government or SID 5 (2/05) Page 9 of 30 consumer pressure, e.g., the move towards organically grown foods (de Boer 2003). An integration of pesticide application and effects into conservation strategy would be desirable, but can only be realized if appropriate data are collected. Currently, these data are generally unavailable. 3.4 Quantitative tools for population and community analysis 3.4.1 Population analyses Population effects can be quantified in a variety of ways. Application of demographic methods to chronic Daphnia test data is possible with only minimal changes to current methods (Newman and McCloskey 2002, Section 4.3 of ECOFRAM 1999). Nacci et al. (2002) applied a demographic approach to analyzing toxicant exposed field populations. Caswell (1996) describes a straightforward approach to performing demographic analyses for toxicant-exposed populations and, in his book (Caswell 2001), provides many details about the matrix approach to demographic analysis. ECOFRAM (1999, Section 4.4 of the ECOFRAM Aquatic Report) provides general details for implementing population models in studies of population effects. With knowledge of the relevant exposure duration, a stochastic projection of population dynamics over the exposure period can be used to estimate the risk of local extinction at different exposure concentrations, i.e., a statement of population risk can be made with these methods (Newman 2001). If spatially-explicit modelling is required, such an approach can be taken with a metapopulation model (O’Connor 1996, Newman 2001). Mackay et al. (2002) provide an example of a spatially explicit, population-based ERA. Several software packages are available for analysis of demographic information. Population Viability Analysis (PVA) has been available as a tool in wildlife conservation for over a decade (Soulé 1987), and involves the estimation of extinction probabilities through analyses that incorporate identifiable threats to population survival into models of the extinction process. Models, such as VORTEX, are available to carry out PVA (Lacy 1993) and there are many examples where PVA scenarios have been modelled and tested (e.g., Ball et al. 2003, Kaye and Pyke 2003). Studies have also been carried out using PVA to provide recommendations for the survival of rare vertebrates, such as the eastern barred bandicoot Perameles gunnii (Lacy and Clark 1990), and invertebrates, such as Fender’s blue butterfly, Icaicia icaroides fenderi (Schultz and Hammond 2003). While PVA has undoubtedly great potential use in conservation, it has yet to be widely adopted as a tool by practitioners. The RAMAS program (Ferson and Akhakaya 1990) is another inexpensive program useful for deterministic and stochastic modelling of populations and PVA, and is available in a version which explicitly considers ecotoxicological information. A further shareware example of a population modelling tool is the PopTools (http://www.cse.csiro.au/poptools) addin to ExcelTM which implements demographic methods to fit population data, and to make deterministic or stochastic population projections. 3.4.2 Community and assemblage analyses The conventional approach to effects characterization produces a series of relevant effect metrics such as LC50 values for test species for comparison with expected exposure concentrations. Recently, a probabilistic approach has emerged for many risk characterizations beyond Tier I. There are considerable differences in sensitivities among species and a focus on the most sensitive of a small number of tested species does not take full advantage of the available data. Species sensitivity distributions (SSDs) make fuller use of the available effects data. An SSD is a distribution of effect metrics for individual species thought to represent collectively the species of concern. Effect metrics for the test species are ranked from lowest to highest, and their ranks converted to approximate proportions. The paired proportions and effect metric concentrations are then fitted to one of several models (Posthuma et al. 2002).1 Some SSD models include all species; however, some require separate models for taxonomic subsets of species. For example, Solomon et al. (1996) performed a retrodictive ERA for atrazine in the North American cornbelt by exploring SSD models for logical species groupings (e.g., Figure 21 in Solomon et al. (1996)). In many recent “probabilistic” risk assessments, the creation of a SSD relies heavily on one of several statistical distributions such as the log logistic, lognormal, or triangular distribution. For aquatic risk assessments of pesticides, data points in these distributions are taken from endpoints in acute or chronic toxicity tests. In acute distributions, data points are normally taken from tests used to derive LC/EC50 values. For chronic toxicity distributions, no-observed-effectconcentrations (NOEC values) are commonly used. In acute toxicity tests, exposures to a contaminant are generally of short duration (e.g., 24 to 96 hours) and chronic toxicity tests are conducted over a full life cycle or an early life stage of an organism. For a pesticide ERA, it is important to identify a threshold hazard concentration above which ecological effects are likely to occur. With SSDs, this is often approximated by selecting a low centile of the distribution. The resulting metric is commonly called a hazardous concentration (HCp). Normally, the fifth or tenth centiles (HC5 and HC10) have been arbitrarily used in ERA (Aldenberg and Slob 1993, Wagner and Løkke 1991). These lower centiles of an effect concentration distribution have been applied historically in deriving US water quality standards and, in the case of the HC5, have more recently been recommended in the EU Technical Guidance Document. Software available for fitting SSDs and estimating HC5 values has been developed by Van Vlaardingen et al. (2003). As discussed by the Aquatic Dialogue Group (1994), a risk assessment that relies solely on the protection of a certain proportion of exposed species might not be protective if keystone, dominant, or legally protected species are ranked below the specified proportion on a SSD. In choosing a proportion from a distribution of acute or chronic effects, one makes the assumption that “protecting” a certain proportion of species will be protective of the structure and function of an 1 Grist et al. (2002) describe a nonparametric method that circumvents the need to identify a well-fitting distributional model. SID 5 (2/05) Page 10 of 30 ecosystem, and that the available single-species toxicity tests are representative of the ecosystem to be protected or the universe of species in the environment. In reality, it would be remarkably fortuitous if the issue of protecting crucial species were adequately addressed in laboratory testing of small numbers of conventional species. Also, the argument could readily be made that acute LC50 or chronic NOEC information is not adequate for predicting a concentration to protect a species population existing in a natural community (Hopkin 1993, Jagoe and Newman 1997, Newman et al. 2002, Newman and Unger 2003). Regardless, it is now a common risk assessment practice and Maund et al. (2001) introduced some supporting evidence for using the tenth centile of acute distributions based upon ecologically significant effects observed at higher concentrations in field studies. Effect quantification for population and species assemblages are less common than those for individuals yet, based on the materials discussed to this point, the need for such metrics is high in pesticide risk assessment. Fortunately, relevant methods are being applied more and more frequently, and higher tier methods such as mesocosm experiments, enclosure studies, and field surveys are amenable to their use. Community effects can also be extracted from laboratory, enclosure, mesocosm and field studies using conventional ecological methods. Newman (1995), Matthews et al. (1998), and Clements and Newman (2002) provide information specific to their application for the risk assessment of chemicals such as pesticides. The influence of simple community interactions such as predator-prey interactions can be quantified (e.g., Tagatz 19762) in laboratory assays but this is not often done to support ERA activities. More commonly, mesocosm community structure metrics are used in predictive pesticide risk assessment activities and field community structure metrics are used in retrodictive risk assessment activities. The most common indices are species richness, diversity and equitability indices. Communities or species assemblages might be compared for different exposures using distance metrics. Multivariate methods such as ordination or clustering methods can assess differences and similarities in species assemblages that have different exposure histories. All of these methods rely on community structure information such as species abundances or presence/absence data. Multimetric methods can incorporate structural and functional qualities of species assemblages during assessments of effect (Clarke 1999, Clements and Newman 2002). The most common multimetric index is the index of biological integrity (IBI) (Karr et al. 1986, Karr 1991, 1993). Karr’s IBI attempts to quantify the integrity of a system of concern relative to an undisturbed or intact system of the same type in the same geographical region. As such, the IBI score for a system has quantitative meaning only in comparison to that of an undisturbed system. The IBI concept has been applied successfully to summarize the integrity of numerous sets of mesocosm or field data. Numerous software packages implement conventional ecological metrics and multivariate methods. One example of such shareware is the BioDiversity package available from Neil McAleece (biodiversity@nhm.ac.uk) of The Natural History Museum and Scottish Association of Marine Science. Other frequently used software for multivariate analysis of ecological community data includes PRIMER (Plymouth Routines In Multivariate Research; see http://www.primer-e.com/in) and CANOCO (see: http://www.plant.dlo.nl/default.asp?section=products&page=/products/canoco/right.htm). Pastorok et al. (2002) provide a comprehensive review of available modelling tools and software for analysis of chemical effects on populations, ecosystems and landscapes. A development in wildlife conservation that followed the Bern Convention (1979) was the realization that, in order to conserve species, there is a need to address the state of the habitats in which they exist (http://www.englishnature.org.uk/baps/habitats/), because the two are inextricably linked. Attention to habitat and landscape issues has provided a platform for the use of Geographic Information System (GIS) software in wildlife conservation. Geographic Information Systems have been used in geographical studies for some time, but possible applications in risk assessment have only recently been appreciated. As with PVA, there are examples of potential uses of GIS in conservation and pesticide risk assessment (e.g. Markus et al. 2003, Holloway et al. 2003b), but its application has yet to be fully embraced by practitioners. 3.4 Discussion and conclusions It is difficult to envisage a risk assessment process capable of consistently predicting long-term pesticide effects founded primarily on individual-based effect metrics. A risk assessment process that includes more population or community effects metrics would substantially reduce uncertainty in predicting long-term ecological effects of pesticide use. The best illustration of this point is the current assumption that conservative calculations using individual-based effect metrics allow conservative expression of protection of communities from unacceptable risk. In many cases, a predictive risk assessment progressing to higher tiers would not include any information on indirect effects that might be common and important in ecological systems. Furthermore, laboratory test species tend to be r-strategy species. This condition creates high uncertainty about predictions of field population persistence for K-strategy species. Can the current ERA process form the foundation for estimating long-term ecological risks of pesticide use? Current tests do provide useful information about the direct effects of pesticides. It would be unwise to abandon these tests completely and to require only complex and expensive ecosystem studies during the predictive assessment of pesticides. However, more laboratory tests focused on population level effect metrics can be performed and would generate more insight than currently possible. For example, risk predictions from the species sensitivity distribution (SSD) approach would be greatly improved by using risk of local population extinction instead of risk of exceeding an LC50 value. It is currently impractical to attempt this with most published information from laboratory toxicity tests performed to date, either because information on time-dependent survival and fecundity is not collected, or because it is not reported in sufficient detail. 2 Table 2.6 in Clements and Newman (2002) gives details for several predator-prey experiments conducted to quantify the effects of toxicants on species interactions. Six involve pesticides as the stressor. SID 5 (2/05) Page 11 of 30 Another improvement would be if strict adherence to the requirements of legislation such as Directive 91/414, with listing of the results of sometimes inappropriate toxicity tests, were replaced by more formal adoption of the principles of ecological risk assessment. This would include formal construction of a conceptual model for each active substance of exposure sources and pathways, and identification of potentially sensitive receptors in Tier 1, rather than only at higher tiers, if at all. In drafting the risk assessment, as much effort should be spent on integrating known ecological relationships into the assessment as is presently being spent on the compiling of concentration and individual-based effects data. A rich literature on ecology remains grossly underexploited in pesticide risk assessment activities. The continuing use of mesocosm or enclosure studies might also result in more ecologically relevant information for conducting predictive risk assessments because of an increased likelihood of seeing an indirect effect of a pesticide before it is authorized and enters into wide use. However, most such studies currently performed in Europe omit vertebrate species because their extensive feeding can confound measurements of invertebrate and plant populations in mesocosms. This means that indirect effects on vertebrates remain unmeasured, and those that are observed in mesocosms may be due to the absence of top predators. The realism of mesocosms when predicting the effects of pesticides therefore remains uncertain. Finally, and of great importance, more post-authorization monitoring could serve as a safety net for the predictive risk assessment of pesticides, which cannot assess all plausible direct and indirect effects for all systems or pesticide mixtures. Pesticides are currently authorized for use in most countries by combining toxicity data with modeled predictions of exposure concentrations. There is some evidence that model scenarios combine conservative assumptions that do not occur widely in the environment (Hendley et al. 2001). Kapustka et al. (1996) describe the 'ecological disconnect' that results from the gulf between the current simple pesticide authorisation procedures and the complex ecological protection goals of these procedures, and suggest that, 'It is remarkable that with such imperfect information, so few reports of pesticide incidents occur. Alternatively, one could argue that the registration process is exceptionally conservative to the extent that some beneficial uses are excluded without cause.' Their answer to the problem of ecological disconnect is to treat risk estimates made during pesticide registration as working hypotheses. These would then require post-authorisation testing, through field monitoring of concentrations and biological effects in the environment during a probationary use phase. This seems to be a sensible and logical approach, and it is surprising that it has not been widely adopted by regulatory authorities. Field studies to examine the biological effects of pesticides must be able to distinguish between effects caused by natural stressors (e.g., extreme weather events), other anthropogenic stressors (e.g., habitat modification or exposure to nonpesticide chemical pollution), and natural covariates (e.g., the prevailing climate, geology and geography), as these may all influence the structure of organism assemblages (Wickham et al. 1997). In addition to this, the effects of stressors can only be determined in relation to a reference condition (Rykiel 1985), which is the condition that would occur in the absence of stress. Unfortunately, many reported field studies do not meet the underlying assumptions of hypothesis testing statistics because they are unreplicated or pseudoreplicated, and unable to assign treatments at random (Beyers et al. 1995). For example, comparison of upstream and downstream sites will likely include factors that covary with pesticide contamination, such as other agricultural impacts (e.g., nutrients, sediments, or physical disturbance (Sallenave and Day 1991)). Investigations performed in this way will almost inevitably attract criticism, especially when the politically charged subject of pesticide use and effects is under study. To address such defects, Suter (1993) emphasizes the need for rigorous logic when attempting to unravel the ecological epidemiology of pollution effects in field studies. Both he and researchers with experience in field studies of pesticides (e.g., Beyers et al. 1995) identify Koch's postulates and Hill's factors, both appropriated from medical epidemiology, as useful for focusing attention on the criteria needed to demonstrate causality in field studies. Koch's postulates, modified for use in environmental toxicology, are, 1. The effects of a toxicant must be regularly associated with exposure to the toxicant and any contributory causal factors. According to Suter (1993), such a regular association should normally consist of Kant’s criteria for causation (law of succession and concept of action). These are that cause and effect must always occur together, and that the effect must follow, not precede, the cause. 2. Indicators of exposure to the toxicant must be found in the affected organisms. This could be established by either measuring the toxicant in the organism or measuring a relevant biomarker induced by the toxicant. 3. The toxic effects must be observed when normal organisms or assemblages are exposed to the toxicant under controlled conditions, and any contributory factors should contribute in the same way during the controlled exposures. This criterion is best met through use of laboratory toxicity tests or mesocosms (e.g., Schulz et al. 2002) to confirm that organisms are affected to the same degree at measured toxicant concentrations. 4. The same indicators of exposure and effects must be identified in the controlled exposures as in the field. Again, this is best established through laboratory or mesocosm experiments to confirm that similar concentrations of toxicant within organisms lead to similar effects. Koch's modified postulates are augmented by Hill's criteria, which were used by Gilbertson (1997) and co-workers to establish that organochlorine pollutants had adversely affected fish, other wildlife and humans in the Great Lakes basin. The main elements of Hill's criteria are, 1. Specificity: Does only the potential cause lead to the effect, and does the potential cause lead only to the effect? Meeting this criterion would be fortunate in most environmental investigations, as it is often the case that multiple causes have multiple potential effects. However, a very high degree of cholinesterase inhibition, outside the normal range of natural variability, would be an example of a specific effect that is very likely to be caused only by exposure to organophosphorus or carbamate pesticides. 2. Strength of association: How precise is the relationship between the potential cause and the observed effect? 3. Time order: Does the effect follow the cause temporally? (Kant’s law of succession). SID 5 (2/05) Page 12 of 30 4. Consistency on replication: Is the association repeatedly observed at different times and places by different investigators? 5. Coherence: Does a cause-effect interpretation of the data seriously conflict with generally known facts? Are there plausible mechanisms of toxic action? Design of studies around Hill's and Koch's criteria is a logical approach to answering questions about the link between pesticides and effects in nature, as found by researchers studying organochlorine effects in the Great Lakes (Gilbertson 1997). We believe that this requires five main elements: 1. Studies should be extensive, rather than intensive, and include measurements at as many sites as possible, so that the consistency of association between pesticides and effects can be assessed. 2. Pesticide concentrations in surface waters should be measured rather than inferred through modeling, so that the presence of a causal agent can be demonstrated. 3. Ideally, researchers should possess detailed knowledge of pesticide application times, rates and locations near particular aquatic sampling sites so that the time order of possible pesticide cause and effect can be examined more accurately. This may not be necessary if organism exposure concentrations are well characterized in the field, although it would almost certainly be of great use when identifying potential study sites. 4. Concentrations of pesticides in organisms should also be measured as a direct indicator of exposure. 5. The plausibility of potential pesticide cause and effect relationships in the field should be demonstrated by laboratory studies to confirm that concentrations of pesticides found in the field can cause the observed effects on particular taxa. 4. ACCEPTABILITY OF PESTICIDE IMPACTS ON THE ENVIRONMENT: WHAT DO UNITED KINGDOM STAKEHOLDERS AND THE PUBLIC VALUE? 4.1 Methods Focus groups with specific stakeholders in the UK pesticides authorisation process were held to elicit a wide range of views and questions about the acceptability of pesticide effects. The outcomes from these focus group discussions were then used to formulate and prioritise questions for a public opinion survey. 4.1.1 Stakeholder focus groups The following specific stakeholders were identified as key decision-makers or opinion-formers in the UK pesticides authorisation process: • The Pesticides Safety Directorate, who are the competent authority for Directive 91/414/EEC in the UK. • Other government regulatory agencies with an interest in pesticides or nature conservation. • Pesticides manufacturers. • Environmental non-governmental organisations. • Farmers and their advisors. • Food distributors and retailers. • Academics with knowledge of the fate and effects of pesticides. • Environmental consultants with knowledge of the fate and effects of pesticides. Representatives from each of these stakeholder groups (Table 1) were invited to separate 3-h meetings held during December 2003 at which they were asked to discuss: 1. In general, what levels of effect (lethal or sublethal) that may be caused by normal agricultural pesticide use on crops are acceptable and what are unacceptable? 2. Does your response depend upon the particular species affected? 3. What environmental scale does your response relate to (e.g., a single field, a parish, a county, or the whole country)? 4. What frequency of occurrence does your response relate to (e.g., acceptable once a year, once every 10 years, etc.)? The views expressed by the different stakeholder groups were collated for each group after the meetings and sent to them for comment. Once group members had agreed on the summary, the eight individual summaries were consolidated into one report which was then sent to all the stakeholders for any further comment, and final agreement. Table 1 Stakeholder organisations represented at focus groups Organisation Type UK pesticide regulator Pesticides manufacturers Government SID 5 (2/05) Organisation Name Pesticides Safety Directorate Bayer Crop Science Ltd Dow AgroSciences Ltd BASF plc Dupont (UK) Ltd Syngenta Crop Protection UK Ltd Environment Agency of England and Wales Scottish Environmental Protection Agency Central Science Laboratory Veterinary Medicines Directorate Broads Authority Page 13 of 30 Organisation Type Environmental Non-Governmental Organisation Farmers & Advisors Food Distributors Academics Consultancy Organisation Name English Nature National Trust The Allerton Research & Educational Trust Pesticides Action Network-UK Wildlife & Countryside Link Farming and Wildlife Advisory Group Random Pulse Horticultural Development Council Home Grown Cereals Authority Farmer Farmer National Farmers Union Association of Independent Crop Consultants Velcourt Research & Development Ltd Farm Retail Association Fresh Produce Consortium (Farm Care) University of Warwick University of Sheffield Centre for Ecology & Hydrology Cranfield University University College London Ponds Conservation Trust Cambridge Environmental Assessments JSC International Ltd. Braddan Scientific Huntingdon Life Sciences TSG Europe Safepharm Envirocorp 4.1.2 Public opinion survey A representative quota sample of 2,049 adults aged 15+ in 201 sampling points across Great Britain was interviewed by MORI. The interviews were conducted face-to-face, in-home between 29 April – 4 May 2004 and the data were weighted to the known national population profile. Questions for the public opinion survey were formulated and prioritised by stakeholders and are shown in Table 2. Question 1 was asked separately, at the start of each interview, and Question 2 onwards was asked later in each interview. Questions 3 and Questions 4/5 were alternated, as were Questions 6 and 7, to avoid bias. At Question 8, the sample was split into two versions, with one half of the sample (Version One) asked about their concerns associated with the effects of human activities on the countryside, and the other half (Version Two) asked about their concerns associated with the effects of pesticides used on farm crops. At Question 11 a simple choice experiment was designed to derive willingness-to-pay estimates for different socioeconomic groups of the population. For this, respondents were each faced with one choice set consisting of two hypothetical food baskets: one labelled ‘standard’ and one labelled ‘no pesticides’. Prices of the latter varied between £14, £17 and £23 per week and were randomised across respondents, while the price for the former was fixed at £12 per week. As a substitute reminder, each respondent was in addition provided with one of two weekly budgets (£30 or £50 per week) to be allocated to both food and leisure. Choices in Question 11 were also alternated. Table 2 Public opinion survey questions Question Q1. Which of these products do you use at home or in the garden? Q2. Which of these are the two or three most important factors that society should take into account in deciding how to grow food? Q3. What possible causes of damage to wildlife and habitats in Britain have you heard about? SID 5 (2/05) Choices Rat/mouse poison Weedkiller for paths, drive or patio Weedkiller for lawn or other vegetated areas Ant powder/spray Wasp powder/spray Slug/snail pellets Sprays to protect plants from pests, fungus or disease Organic pesticides Other None of these Producing cheap food Protecting animal welfare Protecting jobs/Generating jobs Protecting the beauty of the countryside Protecting the health and variety of wildlife Protecting waterways Protecting human health Producing good quality food Other None of these Don’t know Industrial chemicals/Use of industrial chemicals GM crops Housing/Building houses in the countryside Page 14 of 30 Question Q4. Do you personally support or oppose the use of pesticides on farm crops? Q5. Why do you say that? (asked only of those who supported or opposed the use of pesticides in Q4) Choices Intensive farming Litter/People dropping litter Sewage discharges Livestock/ The way animals are kept Pesticides/Use of pesticides Roads/Road building Cars/traffic Salmon farming/Fisheries Other None of these Don’t know Strongly support Tend to support Neither support nor oppose Tend to oppose Strongly oppose Don’t know Positive Codes Kill unwanted pests Control diseases of crops Control plants that compete with crops Increase amount of food grown Provide cheaper food Q6. What are the good things about using pesticides on farm crops? Q7. What are the bad things about using pesticides on farm crops? Q8. How concerned are you personally about the effects of [human activities in the countryside] or [pesticides used on farm crops] on each of the following: A. earthworms B. butterflies C. insects D. songbirds (e.g. skylarks) E. carrion birds (e.g. crows) F. birds of prey (e.g., eagles) G. badgers H. rats I. frogs J. fish K. flowers L. weeds M. soil micro-organisms N. waterways (e.g., ponds & streams) O. hedgerows P. woodlands Q. the beauty of the countryside R. jobs and income in the countryside S. human health Q9. You said you are very concerned about [A to M, inserted from Q8]. How concerned would you be if only a small number of individual [A to M, inserted from Q8] died, but the overall population was unaffected? (asked only of those who were very concerned about effects on biota A – M in Q8). Q10. When thinking about buying food, which of these is the most SID 5 (2/05) Other Don’t know Cheaper food/Reduce cost of food Grow faster/Higher yield/productivity Improve quality of food Improve taste of food Improve healthiness of food Kill unwanted pests Control diseases of crops Control plants that compete with crops More profit Other None Don’t know Increases cost of food Kill insects that are not pests Kill plants that are not weeds Kill wildlife/Damages wildlife Contaminate waterways Possible health risks for humans Reduce quality of food Reduce taste of food Reduce healthiness of food Other None Don’t know Very concerned Fairly concerned Not very concerned Not at all concerned Don’t know Very concerned Fairly concerned Not very concerned Not at all concerned Don’t know Relative product price Page 15 of 30 Negative Codes Risk to/impact on human health Risk to/impact on farmers’ health Impact on/destruction of wildlife Impact on countryside/habitats Cost of using pesticides Question important and which is least important? Choices Food safety Environmental safety Q11 Assume you have a budget of A £30 per week B £50 per week You may allocate this budget between the purchase of a food basket – composed of bread, milk, fruit and vegetables - on one hand, and leisure items on the other. Clearly, if you spend more money on food, you will have less to spend on leisure items. Assume also you have made all your other regular purchases. While shopping for food, you are faced with the following choices – either: 1 ‘Standard food basket’, costing: A £12 (£18 left for leisure) B £12 (£38 left for leisure) C £12 (£18 left for leisure) D £12 (£38 left for leisure) E £12 (£18 left for leisure) F £12 (£38 left for leisure) 2 ‘Reduced pesticide use food basket’, costing: A £14 (£16 left for leisure) B £14 (£36 left for leisure) C £17 (£13 left for leisure) D £17 (£33 left for leisure) E £23 (£7 left for leisure) F £23 (£27 left for leisure) Which product would you choose? ‘Standard food basket’ 4.2 ‘Reduced pesticide use food basket’ Results and discussion 4.2.1 Stakeholder focus groups This section provides an overall summary of the stakeholder focus group discussions, as agreed by participants. Areas where consensus could not be reached are identified by asterisks. In general, what levels of effect (lethal or sublethal) that may be caused by normal agricultural pesticide use on crops are acceptable and what are unacceptable? For birds, there was a large degree of consensus amongst stakeholders that ideally there should be no adverse effects on individuals, including behavioural effects. However, one, or a very few, dead or otherwise affected animals may be acceptable*, while mortality of a large group (e.g., a flock) is not, irrespective of any population effects. Off-field effects in addition to on-field effects are more serious than on-field effects alone. Any effect that leads to population decline outside natural fluctuations at any scale is unacceptable, as are sublethal effects over a wide spatial or temporal scale, or product persistence that restricts recovery. However, changes in populations need to be understood in the context of other ecological and human factors that may influence population size. For example, the effects of pesticides should be considered in the context of other farming practices, such as whether manual weeding in organic systems is as disruptive to ground-nesting birds as the use of pesticides. There needs to be greater understanding of whether pesticides cause any additional effects beyond the effects of other agronomic practices. This implies that population modelling is required, although this is a major challenge given the limited toxicity data that are available. More post-authorisation monitoring is also required to check on population predictions and projections. Many other issues also need to be taken into account for birds, such as food chain uptake, bioaccumulation, the significance of sublethal effects (e.g., effects on behaviour that influence breeding success), and the duration, frequency, scale and spatial distribution of impacts. For mammals, the stakeholder consensus was similar to that for birds. However, there are potentially closer links between effects on mammals and human health effects, so stakeholders agreed that effects on mammals may be less acceptable to the public. Alternatively, public perception issues may be more important for mammals than for birds because some mammals are clearly regarded as pests and are controlled by rodenticides. What therefore is the ‘natural’ population abundance of rats and rabbits that should be the benchmark for population size? For fishes and amphibians, again the stakeholder consensus was similar to that for birds. However, the discrete nature of some water bodies (e.g., ponds), and differences in organism mobility mean that effects may be more severe in aquatic habitats. In addition, fishes are definitely considered to be off-field in the UK, so there should be greater concern over any predicted or measured effects on them. The visibility of fish kills may make them less acceptable to the public, irrespective of any population effects or the time to population recovery. Amphibians may be more susceptible to population impacts as they already appear to be in decline. For invertebrates and plants, loss of species richness (i.e., the number of different species, not the number of individuals) both in- and off-crop compared with what ‘should’ occur at a site is unacceptable, as are effects on population viability. Off-crop organisms should receive greater protection than in-crop organisms. In general, there should be an intolerance of off-field effects on either species richness or abundance of individuals within species. Any indirect effects on the food chain are unacceptable (e.g., lack of invertebrate food for fish fry or bird chicks). SID 5 (2/05) Page 16 of 30 For plants alone, 100% eradication in-field is unacceptable, especially if it includes rare arable plants. There should be tolerance in-field of weed species that are rare, do not compete with crops, or have no adverse effects on human health. For invertebrates alone there should be active encouragement of beneficial species and no adverse effects on their populations. There should be no effects on other organisms as a result of impacts on micro-organisms. However, the intrinsic value of microbes to humans is likely to be in their biogeochemical functions and not their species structure. Microbial function should therefore ideally be maintained, although a brief reduction in function is acceptable. This reduction may be extensive, e.g., for soil sterilants.* Maintenance of function needs to be established in the context of non-degraded microbial assemblages. When assessing acceptable recovery times for invertebrate, plant and microbial populations, there should be recognition of differences in acceptable effects between in-field and off-field non-targets. In-crop recovery should be within one year. Off-crop recovery period is probably of less importance than the spatial extent of any impacts (i.e., distance into field margin), and should be minimal. Transient off-field effects may be acceptable.* There should be no general impairment of population recovery potential either in-crop or off-crop. Acceptable and unacceptable recovery times will depend on particular species life-histories and ecology. For example, effects from application of an herbicide in autumn when aquatic macrophytes are already dying back may be acceptable, while similar effects in spring may be unacceptable. A range of species of differing mobility and recovery potential would need to be studied as some (e.g., springtails) are known to recover slowly. Does your response depend upon the particular species affected? For birds and mammals, in principle probably no, but it might depend on the conservation versus nuisance value of particular species, and desired population abundances of pest (e.g., rat) and non-pest (e.g., dormouse) species. Despite this distinction, vertebrate pest control should be deliberate and not an indirect artefact of insecticide, herbicide or fungicide use. Some birds and mammals may have important ecosystem functions, while others are particularly attractive to humans, and this may influence scientific or public concerns. Some consideration should also be given to adverse effects on the wider genetic pool of a species. Some apparently similar species have very different reactions to exposure to some pesticides (e.g., Canada versus Branta geese), so toxicity data need to be extrapolated between species with care. For fishes and amphibians, the stakeholder consensus was similar to that for birds. However, for fishes the level of concern may also depend upon the commercial value of particular species. For example, the angling community have significant views about the range and abundance of fish species that they expect at particular sites. For invertebrates and plants, in general all species should be treated equally, but the answer may at times depend on the conservation versus nuisance value of a particular species, and its overall role within the wider assemblage of organisms. It is also likely that there will be more public support for aesthetically appealing species, such as butterflies, dragonflies and poppies. For microbes, species composition is probably not as important as function. However, this issue is not well studied and considerations of microbial biodiversity may increase in importance as research and understanding increase. For example, more study is required to determine whether application of microbial insecticides (e.g., Bti) alters natural microbial structure in soils. Stakeholders considered that there should be no major difference in approach when considering aquatic and terrestrial species, except that aquatic species are always off-crop in the UK and may require more protection because of their small-scale habitat use and restricted opportunities for recolonisation (e.g., for ponds). For aquatic systems, there may also be a useful link between what is classified as ‘good quality’ under the Water Framework Directive, and what is acceptable to the public. What environmental scale does your response relate to (e.g., a single field, a parish, a county, or the whole country)? For birds, whatever scale is necessary to sustain a particular species population, given its range, territory and mobility. In general this translates into a field scale for kills, and county to regional scale for population effects. However, there may be more local issues such as the desire for songbirds in gardens, which will be influenced by the territoriality and life history strategies of certain species. This question also needs to be addressed in the context of other population stressors and the spatial structure of the agricultural landscape. The same treatment applied to a mixed farming landscape is likely to have a lower impact than when applied to a crop monoculture. For mammals, fishes and amphibians, mobility is likely to be more restricted than for birds, so protection should be at a more local scale, such as the field or individual water body, to sustain particular species populations. Most mammals, fishes and amphibians also do not move into cropped areas in the same way as birds, so spatial considerations may differ substantially. For invertebrates, plants and microbes, the field scale is important for loss of richness (this may be more important for organisms such as plants, with localised populations). Some stakeholder groups suggested county to regional scales as important for population effects. Other groups suggested that the scale should be related to particular species needs and the patchiness of impacts in their habitats, and there was a suggestion that an operational scale might be 25-50km2 grids. It was also pointed out that, in general and if possible, this issue should be looked at within a wider landscape use and metapopulation context. For example, putting an entire parish under oilseed rape will have ecological consequences irrespective of pesticide use. More diverse agriculture with consequent ‘patchy’ application of pesticides is likely to be less detrimental. SID 5 (2/05) Page 17 of 30 What frequency of occurrence does your response relate to (e.g., acceptable once a year, once every 10 years, etc.)? For vertebrates, stakeholders felt that an occasional local occurrence of effects on a small number of individual birds and mammals, but not fishes or amphibians, is probably acceptable, so long as they are not protected species.* Interactions with the timing of migration (birds) and hibernation (mammals) also need consideration. No frequency of occurrence leading to population decline over the long term (e.g., 20-30 years, and taking natural stressors into account) is acceptable. Responses to stressors depend upon the resilience and recovery rates of different populations, and will relate to interactions between generation time and fecundity and the direct and indirect effects of pesticides. It should be possible to make projections of these with population models to aid decision-making. However, appropriate benchmarks need first to be defined for what population sizes for different species ‘should’ occur in farmland (which is itself a human artefact), and efforts made to achieve this. Small effects every year in the same place or across widespread parts of the UK are less acceptable than larger effects at very infrequent intervals in particular places. A mechanism is required for rationally balancing the benefits and essential uses of a pesticide against any environmental effects. For invertebrates, plants and microbes, large repeated in-crop effects each year are acceptable so long as recovery occurs by the beginning of the following season. For the off-crop environment an occasional local occurrence of effects on a small number of individuals is acceptable, so long as they are not protected species. No frequency of occurrence leading to widespread population decline is acceptable. Again, it was suggested that this should be looked at within a landscape context. The key issue is what frequency of effects across the landscape is compatible with the maintenance of populations and food chains across the same landscape. Conclusions Stakeholders in focus groups, who constitute a cross-section of UK expertise on pesticide effects, were almost entirely concerned with the potential effects of pesticides on animal and plant population viability, and micro-organism function. Focus group discussions returned regularly to this theme, so that questions about acceptable geographical and temporal frequencies of effects and recovery times were answered in terms of organism life history strategies and the ways in which particular frequencies of impacts might affect population demographics. This is not to suggest that stakeholders were entirely unconcerned about individual organism welfare. However, all stakeholder groups recognised that a trade-off exists between the potential economic advantages of responsible pesticide use and the potential disadvantages of individual poisoning events. So long as individual off-field poisonings are infrequent and do not adversely affect population size and viability, stakeholders generally felt that the use of a pesticide leading to these effects should be considered as acceptable. This view is in keeping with guidance provided in support of Directive 91/414/EC (Campbell et al. 1999, EC 2002a&b, Giddings et al. 2002), which focuses on protecting populations rather than individual organisms. 4.2.2 Public opinion survey Products used at home or in the garden Most respondents (71%) had used at least one pesticide product in the home or garden. Important factors that society should take into account in deciding how to grow food Respondents were more concerned about the effects of food-growing methods on human health (65%) and food quality (52%) than on wildlife health and variety (32%), animal welfare (24%) or waterways (13%). Possible causes of damage to wildlife and habitats in Britain Pesticides were the single possible cause of damage to wildlife and habitats most frequently identified by respondents (26%), with the other possible causes mentioned on the questionnaire identified as such by 15% or less of respondents. Nineteen percent of respondents had not heard of any of the identified factors as possible causes of damage. Support for use of pesticides on farm crops More respondents opposed (31%) or strongly opposed (15%) the use of pesticides on farm crops than supported (19%) or strongly supported (1%) this use. Around a third of respondents had no strong views either way. When respondents who supported the use of pesticides (n=419) were asked why they held this view, a third or more agreed that it was because pests were killed or diseases were controlled by the use of pesticides. Twenty-five percent of these respondents also expressed the view that the amount of food grown could be increased through pesticide use. When respondents who opposed the use of pesticides (n=969) were asked why they held this view, most (65%) agreed that it was because of risks to human health, with 42% expressing the view that it was because of risks to wildlife. Similar views emerged when respondents were asked about the benefits and costs of pesticide use. Concern for particular environmental features Figures 1a and 1b show the concerns of respondents for the effects on particular environmental features of either i) non-specific human activities or ii) pesticides. In general, respondents ranked their concerns in a similar order, with effects on human health regarded as most important, and effects on highly regarded wildlife such as songbirds and badgers also ranking relatively highly. In contrast, effects on relatively unloved organisms such as rats and weeds were not ranked as important. However, there were some interesting differences in the views of respondents asked about generic human activities and those asked specifically about pesticides, which are summarised in Figure 1c. More respondents were very concerned about the effects of pesticides, when compared with the effects of general human activities, on waterways, human health, SID 5 (2/05) Page 18 of 30 fishes, soil micro-organisms, frogs, earthworms, carrion birds, butterflies, insects and badgers. In contrast, respondents tended to be less concerned about the effects of pesticides when compared with general human activities on woodlands, countryside jobs and income, and countryside beauty. Respondents who were ‘very concerned’ about the effects of human activities or pesticides on particular organisms were then asked what their view would be if only a small number of individuals died, but the overall population remained the same. In both cases, between 40% and 55% of respondents did not change their views and remained very concerned. However, the remaining respondents did revise their opinion and expressed less concern. A greater percentage of respondents remained very concerned about the effects of pesticides when compared with the effects of general human activities on insects, weeds, butterflies, songbirds, birds of prey, earthworms, carrion birds and badgers. Economic considerations When buying food, respondents ranked consideration of food safety as more important than product price, which was in turn considered more important than environmental safety. This ordering changed when respondents considered the least important considerations when buying food, with product price less important than environmental safety, which was less important than food safety. Although this suggests that price was consistently of low importance to respondents, there were clear trade-offs when the relative price of a ‘no-pesticide’ food basket and the available budget were varied (Figure 2). When the additional cost of buying non-pesticide food was small, half of the respondents chose a food basket produced without pesticides, and there was little difference in this choice when the total available budget was £30 or £50. As the additional cost of a ‘no pesticide’ food basket increased, fewer respondents were prepared to choose it, particularly when the overall budget was relatively low. A conditional logit model was fitted to these data and parameters for the linear indirect utility function were estimated, enabling calculation of willingness to pay (Table 3), conditioned on gender, age (below or above 45 years old), weekly food and leisure budget, and primary concern while buying food. Results broadly conformed to prior expectations. Respondents faced with a higher budget were willing to pay more, but the fact that this increase was not very large suggests that consumers do take into account substitutes (in this case leisure), and therefore the use of substitute reminders is warranted. Table 3 Willingness to pay estimates Gender Male Female Age Younger Older Younger Older £30 per week budget Relative Food price safety 2.6 14.9 5.4 17.2 7.4 26.2 10.4 27.5 Environmental safety 14.3 15.6 19.3 20.4 £50 per week budget Relative Food price safety 4.4 16.9 7.0 18.9 9.5 28.3 12.1 29.3 Environmental safety 15.4 16.6 20.4 21.4 Other factors The information presented in this report is a complete summary of the results from the stakeholder focus groups, but summarises only the headline results from the survey of public opinions. The survey opinions can also be stratified by gender, age, social class, household income, presence of children in the household, geographic region, rurality, and level of education. For example, female respondents were willing to pay more than male respondents for a no-pesticide food basket, as were older when compared to younger respondents. Detailed stratified results are not presented here, for reasons of space, but are freely available by email from Mark Crane (craneconsultants@aol.com). 4.2.2 Conclusions and Relevance for Pesticide Risk Management The survey of opinions from more than 2000 members of the public in the present study showed that although pesticides are widely used in homes and gardens, use on farm crops remains of concern to the public. Concerns are greatest on issues of human health and food quality, as noted in other surveys of public attitudes on pesticides (Anon 2000, Dunlap and Beus 1992). However, potential environmental effects are also an issue for a substantial number of people, particularly if attractive species could be affected. In contrast to the stakeholders in focus groups, about half of the public would remain very concerned about the effects of pesticides even if they affected only individual organisms and not populations. Despite these apparently widespread concerns over pesticide use on farm corps, the desire to purchase food that has not been produced with the use of pesticides depends on its price relative to the total available budget, with only one-fifth of the public, or fewer, prepared to pay a substantial additional amount for this. This identifies a divergence between the public’s perceptions of the potential risks associated with pesticides, and their practical willingness to pay for food production that involves none of these potential risks. Such findings are in agreement with Tait et al. (2001) who found from a review of the literature on chemicals and public values that there is usually only a rather weak correlation between public attitudes and values, on one hand, and actual behaviour on the other, often because of intervening variables, such as price. Female respondents were willing to pay more than male respondents, which agrees with Loureiro et al. (2002) who, in a similar context, estimated willingness to pay for eco-labelled apples, and to Veeman and Adamowicz (2000) in the case of willingness to pay for ‘safer’ skimmed milk. The latter suggest that women are more likely to purchase the majority of household food items and are therefore more aware of food safety issues. Also in agreement with Veeman and Adamowicz (2000) older respondents were willing to pay more than younger respondents. Finally, respondents with food safety and environmental safety as their main food shopping concern had a higher willingness to pay, while price-sensitive respondents expressed a lower willingness to pay. Again, these results agree with Loureiro et al (2002). SID 5 (2/05) Page 19 of 30 Previous studies have shown that experts and the public tend to rank the relative generic risks from pesticides rather consistently. For example, Slovic (2000) asked experts and lay people to rank the perceived risks of 30 potentially hazardous activities. In his survey, the lay people ranked ‘Pesticides’ ninth, while the experts ranked them eighth. This convergence is in marked contrast to activities associated with significant public dread (Perrow 1999), such as ‘nuclear power’ which was ranked as the most important hazard by lay people but was ranked only 20th by experts. In contrast to these findings, the current study suggests that when compared with experts in stakeholder groups the public believes that there are greater environmental risks from pesticide use, possibly because more ‘dreadful’ threats such as nuclear accidents were not included in the survey. Many scientists and industrialists believe that greater public understanding of science is the solution to public attitudes that seem to be irrational, or are at variance with expert views or the actual behaviour of the public. However, social science studies show that this is not the solution because once a person’s mind is made up about fundamental values they will use only the scientific information that supports their position, ignoring the science that does not (Tait 2001b) a Very concerned Fairly concerned Not very concerned Not at all concerned Don't know 60% Percentage of respondents 50% 40% 30% 20% 10% Ra ts W ee ds Fi sh es H ed Co ge ro un w try s sid Fl e o jo bs wer s & in co m e Bu tte rfl Ca ie s rri on bi r ds Ea rth w or m s In se c ts So il m Fr ic og ro -o s rg an ism s H um Co an un he try al th sid e be au ty So ng bi rd Bi s rd so fp re y W oo dl an ds W at er w ay s Ba dg er s 0% Environmental feature b Very concerned Fairly concerned Not very concerned Not at all concerned Don't know 60% Percentage of respondents 50% 40% 30% 20% 10% Environmental feature SID 5 (2/05) Page 20 of 30 R at s W ee ds Fr og So s il m I n ic ro sec ts -o rg an is m s C ou Fi nt sh ry es si de be au ty W oo dl an ds B ad ge rs Fl ow er s H ed ge ro w s B ut C te ou rf l nt ie ry Car s si de rion jo bi bs rd s & in co m e Ea rth w or m s H um an he al th W at er w ay s So ng bi rd B s ird so fp re y 0% 15% Very concerned Fairly concerned Not very concerned Not at all concerned Don't know 5% 0% Countryside beauty Countryside jobs & income Woodlands Hedgerows Rats Birds of prey Weeds Flowers Songbirds Badgers Insects Butterflies Carrion birds Earthworms Frogs Soil micro-organisms -10% Fishes Human health -5% Waterways Percentage difference in response 10% -15% Environmental feature Figure 1 Concerns of respondents over the effects of a) human activities in the countryside and b) pesticides on different environmental features; c) differences in the concerns of respondents over the effects of human activities versus the effects of pesticides in the countryside on different environmental features. Positive values mean that respondents were more concerned about pesticides. Difference in percentage of respondents choosing 'non-pesticide' food basket over 'pesticide' food basket 60% 51% 49% 50% 38% 40% 30% 27% 20% 20% 16% 10% 0% £2 (total budget available = £30) £2 (total budget available = £50) £5 (total budget available = £30) £5 (total budget available = £50) £11 (total budget available = £30) £11 (total budget available = £50) Additional cost of 'non-pesticide' food basket and total budget available (£) Figure 2 5. Effect of price and available budget on respondents’ choice of food basket produced with or without pesticides DECISION FRAMEWORK 5.1 Issues to consider when constructing a decision framework for pesticide authorisation Earlier outputs from this project show that the following ecological and socioeconomic issues should, where possible, be considered when constructing a decision framework for authorising pesticides: 1. Demographic data from laboratory studies should be applied more effectively during pre-authorisation pesticide risk assessment. 2. Production of a formal conceptual ecological risk assessment model for each product or active substance for which authorization is sought would provide an appropriate framework for integrating and applying ecological knowledge. 3. The current lack of targeted post-authorisation monitoring for the effects of pesticides in the natural environment leaves great uncertainty over the level of precaution of pre-authorisation risk assessment and justifies conservatism during this assessment. 4. Long-term adverse pesticide effects on populations of plants and animals, or on microbial function, are considered unacceptable by all stakeholders and the public, and should be assessed within the geographical and temporal context of individual species life histories. SID 5 (2/05) Page 21 of 30 5. Individual vertebrate deaths should be avoided, if at all possible, even if these deaths are unlikely to affect population size or viability, and particularly if the organisms at risk are those that the public care about. In addition to this, a useful decision framework must also be compliant with current regulatory requirements and be based upon scientific evidence and knowledge, except when decisions need to be made in the absence of such evidence or knowledge. 5.2 Proposed decision framework The proposed decision framework operates on the principle that scientific evidence, when available, takes precedence over the opinions of stakeholders and the public and that the trigger values stipulated by Directive 91/414 must be observed during the authorisation process. The framework therefore changes nothing during the earlier tiers of pesticide risk assessment. However, it does offer improvements at higher tiers when judgement must be applied in determining whether a predicted level of environmental effect is acceptable or not. It does this by weighting the importance of each result according to the value placed upon it by stakeholders and the public. The main contribution to this from the stakeholder groups is the consensus that the population viability of all non-target higher organisms, and the community function of microbes, must be maintained under all circumstances. This is entirely consistent with current practice in pesticide risk assessment and guidance. The results from the public opinion survey are then used to determine weightings if there are predicted effects on individual organisms, or if the results from higher tier mesocosm or field studies suggest that transient effects may occur. The weightings are based on survey respondents who were either ‘very’ or ‘fairly’ concerned about the effects of pesticides on a particular group of organisms, multiplied by the proportion of very concerned respondent who remained very concerned even when told that only a few individuals would be affected (Questions 8 and 9 in Table 2 above). The weightings derived from this are shown in Table 4. We should like to emphasise that if experts truly have scientific knowledge that can be applied consistently to interpret effects observed in such studies then this should be used in preference to use of public opinions. However, if there is no quantifiable expertise on the issue under discussion then we would argue that the public’s values should take precedence. This could be done as follows: 1. List the receptors in the environment of concern (in this case aquatic ecosystems). If the pathways or interactions between receptors are complex, then a conceptual model is likely to be the best way to capture and visualise all appropriate information, as discussed in Section 3 of this report. 2. Calculate a maximum achievable score that could be achieved for this system if all hazard quotients or TERs are 1, weighted by the overall weights in the final column of Table 4. 3. Calculate the actual score for the system by substituting the true TER, if it is less than 1. No concessions are made for receptors for which TERs are very high (i.e., very low risks) because this would counterbalance and therefore mask receptors for which TERs are low, so a fixed TER of 1 is applied if the measured TER is >1. 4. Calculate the ratio between the maximum and actual scores and compare with a pre-determined pass or fail threshold. For example, this could be that the sum of all the individual receptor scores must be no less than 90% of the total maximum score, and that no individual score for a receptor can be less than 50% of the maximum score for that receptor. These threshold values may be set more formally by using empirical information from past decisions made during the authorisation process. Expert judgement can be used to challenge the results from this formal decision making process on the basis of additional scientific evidence, but this would be the exception rather than the rule. The approach is illustrated in a short, simplified case study in the next section. Table 4 Weights for incorporating public values into the proposed decision framework. Weights for organism groups ‘in general’ are mean values. Organism Proportion of respondents very or fairly concerned about pesticide effects (A) Songbirds Carrion birds Birds of prey 0.82 0.63 0.79 Proportion of respondents remaining very concerned when effects on only a few individuals (B) 0.51 0.46 0.5 0.42 0.29 0.40 0.74 0.21 0.46 0.49 0.34 0.10 0.8 0.66 0.48 0.53 0.38 0.35 0.74 0.59 0.54 0.5 0.4 0.30 0.57 0.73 0.3 0.5 0.44 0.52 0.29 0.32 0.16 Overall weight (A*B) Birds in general Badgers Rats 0.37 Mammals in general Fishes Frogs 0.22 Aquatic vertebrates in general Butterflies Insects 0.37 Arthropods in general Earthworms Flowers Weeds 0.35 Plants in general SID 5 (2/05) 0.24 Page 22 of 30 5.3 Decision framework case study Table 5 shows some hypothetical data for Pesticide X, an organophosphate insecticide. Briefly, an assessment of single species data for this substance would find the following, on the basis of TER trigger values: 1. Negligible risks to mammals, fishes, algae, earthworms, plants and micro-organisms. 2. Marginal acute, but not chronic, risks to birds. 3. Considerable acute risks to aquatic and terrestrial arthropods. These single species data are likely to prompt an aquatic mesocosm study, the results of which are also presented, and which suggest that potential risks remain for aquatic arthropods at estimated exposure levels. The EC guidance document (EC 2002a) suggests that expert judgement should be used at this stage, although it is difficult to see what knowledge, rather than opinion, an expert could bring to bear on this issue. This is because personal values rather than any scientific knowledge are likely to determine whether an expert decides that the critical value from the mesocosm study is the community NOEC, the concentration that causes only temporary effects on arthropods, or the concentration at which more persistent effects are evident on a few arthropod species. Table 5. Data for a hypothetical organophosphate insecticide (Pesticide X) Test organism/system Mallard duck (Anas platyrhynchos) Mallard duck (Anas platyrhynchos) Rat (Rattus norvegicus) Rainbow trout (Oncorhynchus mykiss) Waterflea (Daphnia magna) Green alga (Pseudokirchneriella subcapitata) Duckweed (Lemna minor) Honeybee (Apis mellifera) Earthworm (Lumbricus terrestris) Soil micro-organisms Sewage micro-organisms Aquatic mesocosm Toxicity LD50 980 mg/kg Maximum Estimated Theoretical Exposure 120 mg/kg Toxicity:Exposure Ratio (TER) 8.2 Annex VI TER trigger and pass/fail result 10 (fail) Reproduction NOEC 640 mg/kg LD50 >1300 mg/kg LC50 115 µg/L 120 mg/kg 5.3 5 (pass) 120 mg/kg 0.9 µg/L >10.8 127.8 10 (pass) 100 (pass) EC50 0.01 µg/L EC50 1500 µg/L 0.9 µg/L 0.9 µg/L 0.01 1667 100 (fail) 10 (pass) EC50 700 µg/L LD50 0.2 µg/bee LC50 160 mg/kg soil 0.9 960 g/ha 2.5 mg/kg soil 778 4800 64 10 (pass) Hazard Quotient of 50 (fail) 10 (pass) NOEC for nitrogen cycling 32 mg/kg soil LD50 280 mg/L Community NOEC 0.5 µg/L. Effects on arthropod abundance for six weeks at 1.0 µg/L. More persistent effects (>8 weeks) on abundance of some arthropod species at 3.0 µg/L 2.5 mg/kg soil 12.8 10 (pass) 0.9 µg/L 0.7 µg/L 311111 Community NOEC: 0.7 Effects for six weeks: 1.4 Persistent effects: 4.3 10 (pass) No set trigger value. If PEC:PNEC ratio of 1 is acceptable threshold then Community NOEC falls below this value. If, instead, we follow the simple proposed decision framework for Pesticide X (Table 6), we begin by listing the main receptor groups in the system within which potential risks may occur (aquatic systems). We then assign maximum achievable TER scores of 1 to each receptor in this system, weighted by the extent to which the public is concerned about effects on individuals from each receptor group, or organisms that are similar to these receptors. The actual TER for each receptor is then input, and the two sets of scores are compared. In this very simple example, pesticide X would be authorised because the overall acceptability score is above the threshold of 90% and no single receptor has an acceptability score below the threshold of 50% (assuming that these thresholds are agreed as reasonable). It is clear from this example that the decision framework proposed here will only be useful when decisions are marginal, because a TER of <0.5 (i.e. predicted exposure is twice the no effect value) for any receptor will automatically produce an acceptability score of <50%, leading to rejection. However, it is precisely this marginal area that is important, as correct decisions will usually be self-evident to risk managers when TERs are either low or high for individual receptors or overall. Table 6 Example for pesticide X of a formal decision framework incorporating public values and TERs Receptor Waterfowl Aquatic mammals Fishes Arthropods Other invertebrates arthropods) Algae Aquatic macrophytes SID 5 (2/05) TERactual (=1 if TER > 1) Maximum achievable acceptability score (Public value x TER1) (non- 0.37 0.22 0.38 0.35 0.29 1 1 1 0.7 1 Actual acceptability score (Public value x TERactual 0.37 0.22 0.38 0.25 0.29 0.16 0.32 1 1 0.16 0.32 Page 23 of 30 Ratio actual:maximum acceptability score 1 1 1 0.71 1 1 1 Receptor Maximum achievable acceptability score (Public value x TER1) TOTAL TERactual (=1 if TER > 1) 2.09 Actual acceptability score (Public value x TERactual 1.99 Ratio actual:maximum acceptability score 0.952 The method described here is generally consistent in that, 1. Higher scores represent a utility that corresponds to the wishes expressed by the public. 2. The scores allow for discrimination between combinations of potential events. 3. There is a maximum achievable score that corresponds to a ‘perfect’ score. 4. All pesticides can be measured and assessed on the same scale. 6. 7. 8. MAIN FINDINGS AND IMPLICATIONS 1. There is currently little information available to determine whether normal use of pesticides poses long-term risks to wildlife. There is clearly the potential for risk, but the temporal and spatial realisations of this risk in the natural environment remain largely unknown. Reliable population projections of pesticide effects during pre-authorisation risk assessment require reporting of relevant demographic data (age-specific survival and reproduction), which is not currently the case. Post-authorisation monitoring for effects is also sparse, but is the only effective means of testing the effectiveness of the authorisation process. Pesticide risk assessments would also benefit from formal adoption of conceptual modelling approaches, which are widely used in other ecological risk assessment frameworks. These shortcomings in knowledge and procedure mean that current pesticide risk assessments should remain conservative in their assumptions. 2. Stakeholders in pesticide risk management from across the spectrum of views agree that maintenance of higher organism population viability and microbial function are the main aims of pesticide risk assessment and management. Substantial numbers of the general public are more conservative than these informed stakeholders, but a softening of their views is evident when only a few individual organisms, rather than entire populations, are potentially affected by pesticides,. Despite the public’s apparent concerns over pesticides, economic factors clearly influence the extent to which they are prepared to pay for food that has been grown without the use of pesticides. A balanced approach to incorporating public values into pesticide risk management would therefore take account of the public’s views on potential costs (i.e., environmental effects, or effects on human health) and the potential benefits of pesticides, although this is not currently possible under existing regulations. 3. A simple decision framework for pesticide risk management can be constructed which takes public values into account when interpretation of environmental information is value-laden rather than evidence-based. This framework complies with Directive 91/414 and its associated guidance and provides a consistent means for deciding on the acceptability of observed effects. REQUIREMENTS FOR FURTHER RESEARCH 1. There is a clear need for targeted monitoring studies in aquatic and terrestrial environments to determine the spatial and temporal patterns of any adverse effects caused by the approved use of pesticides. 2. The public should be surveyed at regular intervals to determine the values that they place upon different environmental receptors. This survey is most cost-effectively achieved as part of an omnibus survey performed by a professional polling organisation. The survey should attempt to obtain information on a broader range of organisms than in the current project so that there is less need to extrapolate public values between taxonomic groups. 3. The proposed decision framework should be applied to a wide variety of currently authorised substances to determine whether it is fit for purpose and provides improved consistency of decision-making. Three main areas should be explored: a. whether it is possible to construct generic conceptual models for aquatic and terrestrial environments which capture all of the main receptors that may be directly or indirectly affected by pesticide exposure, as suggested by Objective 1 of this project; b. whether the information supplied by notifiers in support of authorisation is sufficient to parameterise these generic conceptual models; and c. whether quantitative thresholds for acceptable ‘scores’ from the decision framework can be determined from the empirical data. REFERENCES Aldenberg T, Slob W (1993) Confidence limits for hazardous concentrations based on logistically distributed NOEC toxicity data. Ecotoxicol Environ Safety 25: 48-63. Anon. 2000. Termites and the public: attitude toward pesticides. Pest Cont Technol 28:50-52. Aquatic Dialogue Group (1994) Aquatic Dialogue Group: Pesticide Risk Assessment and Mitigation, SETAC Press, Pensacola, FL. SID 5 (2/05) Page 24 of 30 Bacietto JJ (1998) A framework for ecological risk assessment: Beyond the quotient method. In: Risk Assessment: Logic and Measurement, Newman MC, Strojan CL (eds), CRC/Lewis Press, Boca Raton, FL, pp 11-22. Ball SJ, Lindenmayer DB, Possingham HP (2003) The predictive accuracy of population viability analysis: a test using data from two small mammal species in a fragmented landscape. Biodiv Cons 12: 2393-2413. Beaumont P (1997) Where have all the birds gone? Pesticides News 30:3. Begon M., Harper JL, Townsend CR (1986) Ecology: Individuals, Populations and Communities. Blackwell Scientific Publications, London. Bellamy PE, Rothery P, Hinsley SA (2003) Synchrony of woodland bird populations: the effect of landscape structure. Ecography 26: 338-348. Benton TG, Bryant DM, Cole L, Crick HQP (2002) Linking agricultural practice to insect and bird populations: a historical study over three decades. J Appl Ecol 39: 673-687. Beyers DW, Farmer MS, Sikoski PJ (1995) Effects of rangeland aerial application of Sevin-4-Oil® on fish and aquatic invertebrate drift in the little Missouri River, North Dakota. Arch Environ Contam Toxicol 28:27-34. Bishop CA, Ng P, Mineau P, Quinn JS, Struger J (2000) Effects of pesticide spraying on chick growth, behaviour, and parental care in tree swallows (Tachycineta bicolor) nesting in an apple orchard in Ontario, Canada. Environ Toxicol Chem 19: 2286-2297. Bliss CI, Cattell M (1943) Biological assay. Ann Rev Physiol 5: 479-539. Blus LJ, Henny CJ (1997) Field studies of pesticides and birds: unexpected and unique relations. Ecol Appl 7: 1125-1132. Borg C, Toft S (2000) Importance of insect prey quality for grey partridge chicks Perdix perdix: A self-selection experiment. J Appl Ecol 37: 557-563. Broyer J (1994) The decline of the corncrake in France and the management of meadow habitats. Alauda. 62: 1-7. Cairns Jr J (1984) Are single species toxicity tests alone adequate for estimating environmental hazard? Environ Monit Assess 4: 259-273. Cairns Jr J, Niederlehner BR (1993) Ecological function and resiliency: neglected criteria for environmental impact assessment and ecological risk analysis. Environ Prof 15: 116-124. Cairns Jr J, Pratt JR, Niederlehner BR, McCormick PV (1986) A simple cost-effective multispecies toxicity test using organisms with a cosmopolitan distribution. Environ Monit Assess 6: 207-220. Campbell PJ, Arnold DJS, Brock TCM, Grandy NJ, Heger W, Heimbach F, Maund SJ and Streloke M. 1999. Guidance Document on Higher-tier Aquatic Risk Assessment for Pesticides (HARAP). SETAC Europe, Brussels, Belgium. Cardwell C, Hassall M, White P (1994) Effects of headland management on carabid beetle communities in Breckland cereal fields. Pedobiologia 38: 50-62. Caslin TM, Wolfe JO (1999) Individual and demographic responses of the gray-tailed vole to vinclozolin. Environ Toxicol Chem 18: 1529-1533. Caswell H (1996) Demography meets ecotoxicology: untangling the population level effects of toxic substances. In: Newman MC, Jagoe CH (eds), Ecotoxicology. A Hierarchical Treatment. CRC/Lewis Publishers, Boca Raton, FL, pp 255-292. Caswell H. (2001) Matrix Population Models, Second Edition. Sinauer Associates, Inc., Sunderland, MA. Chamberlain DE, Wilson AM, Browne SJ, Vickery JA (1999) Effects of habitat type and management on the abundance of skylarks in the breeding season. J Appl Ecol 36: 856-870. Chamberlain DE, Vickery JA, Gough S (2000) Spatial and temporal distribution of breeding skylarks Alauda arvensis in relation to crop type in periods of population increase and decrease. Ardea 88: 61-73. Chinery M (1993) Insects of Britain and Northern Europe. 3rd ed. Harper Collins, London. Chiverton PA (1999) The benefits of unsprayed cereal crop margins to grey partridges Perdix perdix and pheasants Phasianus colchicus in Sweden. Wild Biol 5: 83-92. Clarke KR (1999) Nonmetric multivariate analysis in community-level ecotoxicology. Environ Toxicol Chem 18: 118-127. Clements DK, Skidmore P (2002) The autecology of the hornet robberfly Asilus crabroniformis L. in Wales, 1997-99. Countryside Council for Wales Contract Science Report No. 525. Clements WH, Newman MC (2002) Community Ecotoxicology. John Wiley & Sons, Chichester, UK. Crain DA,Guillette Jr LJ, Pickford DB, Percival HF, Woodward AR (1998) Sex-steroid and thyroid hormone concentrations in juvenile alligators (Alligator mississippiensis) from contaminated and reference lakes in Florida, USA. Environ Toxicol Chem 17: 446-452. Cramp S (1988) Handbook of the Birds of Europe, the Middle East and North Africa: The Birds of the Western Palearctic. Vol. V – Tyrant Flycatchers to Thrushes. Oxford University Press, Oxford. Cramp S (1992) Handbook of the Birds of Europe, the Middle East and North Africa: The Birds of the Western Palearctic. Vol. VI – Warblers. Oxford University Press, Oxford. Cramp S, Perrins CM (1993) Handbook of the Birds of Europe, the Middle East and North Africa: The Birds of the Western Palearctic. Vol. VII – Flycatchers to Shrikes. Oxford University Press, Oxford. Cramp S, Perrins CM (1994a) Handbook of the Birds of Europe, the Middle East and North Africa: The Birds of the Western Palearctic. Vol. VIII – Crows to Finches. Oxford University Press, Oxford. Cramp S, Perrins CM (1994b) Handbook of the Birds of Europe, the Middle East and North Africa: The Birds of the Western Palearctic. Vol. IX – Bunting and New World Warblers. Oxford University Press, Oxford. Crane M (1997) Research needs for predictive multispecies tests in aquatic toxicology. Hydrobiologia 346:149-155. Crane M, Giddings JM (2004) ‘Ecologically Acceptable Concentrations' when assessing the environmental risks of pesticides under European Directive 91/414/EEC. Hum Ecol Risk Assess 10:1-15. Crowfoot JE and Wondolleck JM. 1990. Environmental Disputes: Community Involvement in Conflict Resolution. Island Press, Washington, D.C., US. De Boer IJM (2003) Environmental impact assessment of conventional and organic milk production. Livestock Prod Sc 80: 69-77. Dodson SI, Merritt CM, Shannahan J-P, Shults CM (1999) Low exposure concentrations of atrazine increase male production in Daphnia pulicaria. Environ Toxicol Chem 18: 1568-1573. Doebeli M, Killingback T. (2003) Metapopulation dynamics with quasi-local competition. Theor Pop Biol 64: 397-416. Douglas H. 2000. Inductive risk and values in science. Philos Sci 67:559-579. Dover J, Sotherton N, Gobbett K (1990) Reduced pesticide inputs on cereal field margins: the effects on butterfly abundance. Ecol Entomol 15: 17-24. Dunlap RE, Beus CE. 1992. Understanding public concerns about pesticides – an empirical examination. J Consumer Affairs 26:418-438. ECOFRAM (1999) Draft Guidance, Ecological Committee on FIFRA Risk Assessment Methods (ECOFRAM) Aquatic Report, http://www.epa.gov/oppefed1/ecorisk/. Epperson BK (2000) Spatial genetic structure and non-equilibrium demographics within plant populations. Plant Spec Biol 15: 269-279. Ernst SKM, Brown JH (2001) Delayed compensation for missing keystone species by colonization. Science 292: 101-102. EC (1993) Directive 91/414/EEC. European Commission, Brussels, Belgium. EC (2002a) Guidance Document on Terrestrial Ecotoxicology Under Council Directive 91/414/EEC. DRAFT Working Document, 17 October 2002, European Commission SANCO/10329/2002 rev 2 final. EC. (2002b) Guidance Document on Aquatic Ecotoxicology Working Document, 17 October 2002, European Commission SANCO/3268/2001, rev. 4 final. Ewald JA, Aebischer NJ (1999) Pesticide use, avian food resources and bird densities in Sussex. Joint Nature Conservation Committee Report No. 296, Peterborough, UK. Fagan WF, Meir E, Prendergast J, Folarin A, Karieva P (2001) Characterizing population vulnerability for 758 species. Ecol Letters 4: 132-138. Ferson S, Akhakaya HR (1990) Modeling Structure in Age-structured Populations. RAMAS/Age User Manual. Applied Biomathematics, Setauket, NY. Finney DJ (1947) Probit Analysis. A Statistical Treatment of the Sigmoidal Response Curve. Cambridge University Press, Cambridge. Forbes VE, Calow P (2003) Contaminant effects on population demographics. In: Newman MC, Unger MA. Fundamentals of Ecotoxicology, Second Edition, CRC/Lewis publishers, Boca Raton, FL, pp. 221-224. Frankham R, Ballou JD, Briscoe DA (2002) Introduction to Conservation Genetics. Cambridge University Press, Cambridge, UK. Frewer L. 2004. The public and effective risk communication. Toxicol Lett 149:391-397. Frewer L. 2003. Societal issues and public attitudes towards genetically modified foods. Trends Food Sci Technol 14:319-332. SID 5 (2/05) Page 25 of 30 Frewer L, Lassen J, Kettlitz B, Scholderer J, Beekman V, Berdal KG. 2004. Societal aspects of genetically modified foods. Food Chem Toxicol 42:11811193. Furse MT, Symes KL, Winder JM, Clarke RT, Blackburn JH, Gunn RJM, Grieve NJ, Hurley M (1995) The Faunal Richness of Headwater Streams: Stage 3 - Impact of Agricultural Activity. Volume 1 - Main Report. National Rivers Authority R&D Project 242, Bristol, UK. Giddings JM, Brock TCM, Heger W, Heimbach F, Maund SJ, Norman SM, Ratte HT, Schäfers C and Streloke M. 2002. Community Level Aquatic System Studies - Interpretation Criteria. SETAC Press, Pensacola, FL. Gilbertson M (1997) Advances in forensic toxicology for establishing causality between Great Lakes epizootics and specific persistent toxic chemicals. Environ Toxicol Chem 16:1771-1778. Giler, PS, O’Donovan G (2002) Biodiversity and ecosystem function: do species matter? Biol and Environ 102B: 129-139. Grant A (1998) Population consequences of chronic toxicity: incorporating density dependence into the analysis of life table response experiments. Ecol Modelling 105: 325-335. Gray Jr LE, Ostby J, Wolf C, Lambright C, Kelce W (1998) The value of mechanistic studies in laboratory animals for the prediction of reproductive effects in wildlife: endocrine effects on mammalian sexual differentiation. Environ Toxicol Chem 17: 109-118. Green RE (1984) The feeding ecology and survival of partridge chicks Alectoris rufa and Perdix perdix on arable farmland in East Anglia. J Appl Ecol 21: 817-830. Green RE (1996) Factors affecting the population density of the corncrake Crex crex in Britain and Ireland. J Appl Ecol 33: 237-248. Green RE, Gibbons DW (2000) The status of the corncrake Crex crex in Britain in 1998. Bird Study 47: 129-137. Green RE, Stowe TJ (1993) The decline of the corncrake Crex crex in Britain and Ireland in relation to habitat change. J Appl Ecol 30: 689-695. Grist EPM, Leung KMY, Wheeler JR, Crane M (2002) Better bootstrap estimation of hazardous concentration thresholds for aquatic assemblages. Environ Toxicol Chem 21: 1515-1524. Gruar D, Peach, W, Taylor R (2003) Summer diet and body condition of song thrushes Turdus philomelos in stable and declining farmland bird populations. Ibis 145: 637-649. Guillette Jr LJ, Pickford DB, Crain DA, Rooney AA, Percival HF (1996) Reduction in penis size and plasma testosterone concentrations in juvenile alligators living in a contaminated environment. Gen Compar Endocrin 101:32-42. Guillette Jr. LJ, Brock JW, Rooney AA, Woodward AR (1999) Serum concentrations of various environmental contaminants and their relationship to sex steroid concentrations and phallus size in juvenile American alligators. Arch Environ Contam Toxicol 36: 447-455. Hansen J, Holm L, Frewer L, Robinson P, Sandoe P. 2003. Beyond the knowledge deficit: recent research into lay and expert attitudes to food risks. Appetite 41:111-121. Hanski I (1999) Metapopulation Ecology. Oxford University Press, Oxford. Harrington LA, MacDonald DW (2002) A Review of the Effects of Pesticides on Wild Terrestrial Mammals in Britain. Wildlife Conservation Research Unit, University of Oxford, Oxford, UK. Hayes TB, Collins A, Lee M, Mendoza M, Noriega N, Stuart AA, Vonk A (2002a) Hermaphroditic, demasculinised frogs after exposure to the herbicide atrazine at low ecologically relevant doses. PNAS 99:5476-5480. Hayes TB, Haston K, Tsui M, Hoang A, Haeffele C, Vonk A. (2002b) Feminization of male frogs in the wild. Nature 419:895-896. Hayes TK, Tsui, HM, Hoang A, Haeffele C, Vonk A (2003) Atrazine-Induced hermaphroditism at 0.1 ppb in American leopard frogs (Rana pipiens): laboratory and field evidence. Environ Health Persp 111: 568-575. Henderson IG, Cooper J, Fuller RJ, Vickery J (2000) The relative abundance of birds on set-aside and neighbouring field in summer. J Appl Ecol 37: 335347. Hendley P, Holmes C, Kay S, Maund SJ, Travis KZ, Zhang M (2001) Probabilistic risk assessment of cotton pyrethroids: III. A spatial analysis of the Mississippi, USA, cotton landscape. Environ Toxicol Chem 20:669-678. Hickey JJ, Anderson DW (1968) Chlorinated hydrocarbons and eggshell changes in raptorial and fish-eating birds. Science 162: 271-273. Hill JK, Thomas CD, Lewis OT (1996) Effects of habitat patch size and isolation on dispersal by Hesperia comma butterflies: implications for metapopulation structure. J Anim Ecol 65: 725-735. Hill JK, Thomas CD, Lewis OT (1999) Flight morphology in fragmented populations of a rare British butterfly, Hesperia comma. Biol Cons 87: 277-283. Hoelzel AR (1999) Impact of population bottlenecks on genetic variation and the importance of life-history: a case study of the northern elephant seal. Biol J Linn Soc 68: 23-39. Hoelzel AR, Halley J, O’Brien SJ, Campagna C, Arnborn T, le Boeuf B, Ralls K, Dover GA (1993) Elephant seal genetic variation and use of simulation models to investigate historical population bottlenecks. J Hered 84: 443-449. Holloway GJ, Dickson JD, Harris PW, Smith J (2003a) Dynamics and foraging behaviour of adult hornet robberflies, Asilus crabroniformis: implications for conservation management. J Insect Cons 7: 127-135. Holloway GJ, Griffiths GH, Richardson P (2003b) Conservation strategy maps: a tool to facilitate biodiversity action planning illustrated using the heath fritillary butterfly. J Appl Ecol 40: 413-421. Hopkin SP (1993) Ecological implications of “95% protection levels” for metals in soil. Oikos 66: 137-141. Horne G, Fielding AH (2002) Recovery of the Peregrine Falcon Falco peregrinus in Cumbria, UK, 1966-99. Bird Study 49: 229-236. Hossel JE (2001) Climate change and UK farmland birds: a review of implications. RSPB Report No. 5060. Hulme M, Viner D (1998) A climate change scenario. Climate Change 39: 145-176. Jacquemyn H, van Rossum F, Brys R, Endels P, Hermy M, Triest L, de Blust G (2003) Effects of agricultural land use and fragmentation on genetics, demography and population persistence of the rare Primula vulgaris, and its implications for conservation. Belgian J Bot 136: 5-22. Jagoe R, Newman MC (1997) Bootstrap estimation of community NOEC values. Ecotoxicol 6: 293-306. Kadiri N, Lumaret J-P, Janati-Idrissi A (1999) Macrocyclic lactones: Impact on non-target fauna in pastures. Ann Soc Entomol France 35: 222-229. Kammenga JE, Busschers M, Van Straalen NM, Jepson PC, Baker J (1996) Stress induced fitness is not determined by the most sensitive life-cycle trait. Functional Ecol 10: 106-111. Kapustka LA, Williams BA, Fairbrother A (1996) Evaluating risk predictions at population and community levels in pesticide registration – hypothesis to be tested. Environ Toxicol Chem 15: 427-431. Karr JR (1991) Biological integrity: a long-neglected aspect of water resource management. Ecol Appl 1: 66-84. Karr JR (1993) Defining and assessing ecological integrity: beyond water quality. Environ Toxicol Chem 12: 1521-1531. Karr JR, Fausch KD, Angermeier PL, Yant PR, Schlosser IJ (1986) Assessing Biological Integrity in Running Waters. A Method and Its Rationale. Illinois Natural History Survey Publication 5. Illinois Natural History Survey, Champagne, IL. Kaye TN, Pyke DA (2003) The effect of stochastic technique on estimates of population viability from transition matrix models. Ecology 84: 1464-1476. Krebs JR, Wilson JD, Bradbury RB, Siriwardena GM (1999) The second Silent Spring? Nature 400:611-612. Kudoh H (2001) Gene flow among plant populations in an ecological landscape. Jap J Ecol 51: 193-201. Lacy RC (1993) VORTEX: a computer simulation model for Population Viability Analysis. Wild Res 20: 45-65. Lacy RC, Clarke TW (1990) Population viability assessment of the eastern barred bandicoot in Victoria. In: Clarke TW, Seebeck JH (eds) The Management and Conservation of Small Populations. Chicago Zoological Society, Chicago. Lane SJ, Alonso JC (2001) Status and extinction probabilities of great bustard (Otis tarda) leks in Andalucia, southern Spain. Biodiv Cons 10: 893-910. Liber K, Kaushik NK, Solomon KR, Carey JH (1992) Experimental designs for aquatic mesocosm studies: comparison of the ‘ANOVA’ and ‘regression’ design for assessing the impact of tetrachlorophenol in zooplankton populations in limnocorrals. Environ Toxicol Chem 11: 61-77. Liebig J (1840) Chemistry in Its Application to Agriculture and Physiology. Taylor and Walton, London. Loureiro, ML, McCluskey JJ, Mittelhammer RC. 2002. Will consumers pay a premium for eco-labeled apples? Journal of Consumer Affairs 36:203-219. Luoto M, Rekolainen S, Aakkula J, Pykala J (2003) Loss of plant species richness and habitat connectivity in grasslands associated with agricultural change in Finland. Ambio 32: 447-452. SID 5 (2/05) Page 26 of 30 Mackay CE, Colton JA, Bigham G (2002) Structuring population-based ecological risk assessments in a dynamic landscape. In: Newman MC, Roberts Jr MH, Hale RC (eds), Coastal and Estuarine Risk Assessment, CRC/Lewis Publishers, Boca Raton, FL, pp 273-296. Madsen T, Shine R, Olsson M, Wittzell H (1999) Restoration of an inbred adder population. Nature 402: 34-35. Markus P, Timo P, Juha T (2003) Habitat preferences of the skylark Alauda arvensis in Southern Sweden. Ornis Fennica 80: 97-110. Marshall J, Brown V, Boatman N, Lutman P, Squire G (2001) The impact of herbicides on weed abundance and biodiversity, PN0940. A Report for the UK Pesticides Safety Directorate. IACR-Long Ashton Research Station, 134 p. Martin PA, Johnson DL, Forsyth DJ, Hill BD (2000) Effects of two grasshopper control insecticides on food resources and reproductive success of two species of grassland songbirds. Environ Toxicol Chem 19: 2987-2996. Matthews RA, Matthews GB, Landis WG (1998) Application of community level toxicity testing to environmental risk assessment. In: Newman MC, Strojan CL (eds), Risk Assessment. Logic and Measurement, CRC/Lewis Press, Boca Raton, FL, pp 225-253. Matthews RA, Landis WG, Matthews GB (1996) The community conditioning hypothesis and its application to environmental toxicology. Environ Toxicol Chem 15: 597-603. Maund SJ, Travis KZ, Hendley P, Giddings JM, Solomon KR (2001) Probabilistic risk assessment of cotton pyrethroids: V. Combining landscape-level exposure and ecotoxicological effects data to characterize risks. Environ Toxicol Chem 20: 687-692. McCracken DI, Foster GN (1993) The effects of ivermectin on the invertebrate fauna associated with cow dung. Environ Toxicol Chem 12: 73-84. McGinnity P, Prodohl P, Ferguson A, Hynes R, Maoileidigh NO, Baker N, Cotter D, O’Hea B, Cooke D, Rogan G, Taggart J, Cross T (2003) Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. Proc Roy Soc Lond B 270: 2443-2450. Mills LS, Soule ME, Doak DF (1993) The keystone-species concept in ecology and conservation. BioScience 43: 219-223. Millsap BA, Kennedy PL, Mitchell A, Court G, Enderson JH, Rosenfield RN (1998) Review of the proposal to de-list the American peregrine falcon. Wildlife Soc Bull 36:522-538. Milnes MR, Woodward AR, Rooney AA, Guillette LJ (2002) Plasma steroid concentrations in relation to size and age in juvenile alligators from two Florida lakes. Compar Biochem Physiol (Part A) 131: 923-930. Mineau P (2002) Estimating the probability of bird mortality from pesticide sprays on the basis of the field study record. Environ Toxicol Chem 21: 14971506. Nacci DE, Gleason TR, Gutjahr-Gobell R, Huber M, Munns Jr WR (2002) Effects of chronic stress on wildlife populations: a population modeling approach and case study. In: Newman MC, Roberts Jr MH, Hale RC (eds), Coastal and Estuarine Risk Assessment, CRC/Lewis Publishers, Boca Raton, FL pp 247-272. Newman MC (1995) Quantitative Methods in Aquatic Ecotoxicology. CRC/Lewis Publishers, Boca Raton, FL. Newman MC (2001) Population Ecotoxicology. John Wiley & Sons, Chichester, UK. Newman MC, McCloskey JT (2000) The individual tolerance concept is not the sole explanation for the probit dose-effect model. Environ Toxicol Chem 19: 520-526. Newman MC, Unger MA (2003) Fundamentals of Ecotoxicology, Second Edition. CRC/Lewis Publishers, Boca Raton, FL. Newman MC, Ownby DR, Mezin LCA, Powell DC, Christensen TRL, Lerberg SB, Anderson BA, Padma TV (2002) Species sensitivity distributions in ecological risk assessment: distributional assumptions, alternate bootstrap techniques, and estimation of adequate number of species, In: Posthuma L, Suter II GW, Traas TP (eds), Species Sensitivity Distributions in Ecotoxicology, CRC/Lewis Publishers, Boca Raton, FL, pp 119-132. Newton I (1986) The Sparrowhawk. T & AD Poyser, Calton, Staffs, UK. O’Connor RJ (1996) Toward the incorporation of spatial temporal dynamics into ecotoxicology. In: Rhodes Jr OE, Chesser RK, Smith MH (eds), Population Dynamics in Ecological Space and Time, The University of Chicago Press, Chicago, IL. Pastorok RA, Bartell SM, Ferson S, Ginzburg LR (2002) Ecological Modeling in Risk Assessment: Chemical Effects on Populations, Ecosystems and Landscapes. Lewis Publishers, Boca Raton, FL. Pemberton JM, Smith RH (1985) Lack of biochemical polymorphism in British UK fallow deer Dama dama. Heredity 55: 199-208. Perrow C. 1999. Normal Accidents: Living with High-Risk Technologies. Princeton University Press, NJ, USA. Posthuma L, Suter II GW, Traas TP (eds) (2002) Species Sensitivity Distributions in Ecotoxicology. CRC/Lewis Publishers, Boca Raton, FL. Power ME, Tilman D, Estes JA, Menge BA, Bond WJ, Mills SJ, Daily G, Castilla, JC, Lubchenco J, Paine RT (1996) Challenges in the quest for keystones. BioScience 46: 609-620. Pratt JR, Cairns Jr J (1996) Ecotoxicology and the redundancy problem: understanding effects on community structure and function. In: Newman MC, Jagoe CH (eds), Ecotoxicology. A Hierarchical Treatment, CRC/Lewis Press, Boca Raton, FL, pp 347-370. Preston BL (2002) Indirect effects in aquatic ecotoxicology: implications for ecological risk assessment, Environ. Management 29: 311-323. Ralls K, Ballou J (1983) Extinction: lessons from zoos. In: Schonewald-Cox CM, Chambers SM, MacBryde B, Thomas L (eds) Genetics and Conservation: a Reference for Managing Wild Animal and Plant Populations. Benjamin/Cummings, Menlo Park, CA. Rands MRW (1985) Pesticide use on cereals and the survival of grey partridge Perdix perdix chicks in a field experiment. J Appl Ecol 22: 49-54. Rands MRW (1986) The survival of gamebird Galliformes chicks in relation to pesticide use in cereals. Ibis 128: 57-64. Ratcliffe D (1980) The Peregrine Falcon. T & AD Poyser, Calton, Staffs, UK. Rowe G, Beebee TJC (2003) Population on the verge of a mutational meltdown? Fitness costs of genetic load for an amphibian in the wild. Evolution 57: 177-181. Rykiel EJ Jr (1985) Toward a definition of ecological disturbance. Australian J Ecol 10:361-365. Ryttman H (2003) Breeding success of Wryneck Jynx torquiila during the last 40 years in Sweden. Ornis Svecica 13: 25-28. Saccheri I, Kuussaari M, Kankare M, Fortelius W, Hanski I (1998) Inbreeding and extinction in a butterfly metapopulation. Nature 392: 491-494. Sallenave RM, Day KE. (1991) Secondary production of benthic stream invertebrates in agricultural watersheds with different land management practices. Chemosphere 23:57-76. Savidge JA (1978) Wildlife in herbicide-treated Jeffery pine plantation in eastern California. J Forestry 76: 476-478. Sawchik J, Dufrene M, Schickzelle N, Baguette M (2002) Metapopulation dynamics of the bog fritillary butterfly: modelling the effect of habitat fragmentation. Acta Oecologia 23: 287-296. Schmiegelow FKA, Monkkonen M (2002) Habitat loss and fragmentation in dynamic landscapes: avian perspectives from the boreal forest. Ecol Appl 12: 375-389. Schroeder MH, Sturges DL (1975) The effect on the Brewer’s sparrow of spraying big sagebrush. J Range Management 28: 294-297. Schultz CB, Hammond PC (2003) Using population viability analysis to develop recovery criteria for endangered insects: case study of the Fender’s blue butterfly. Cons Biol 17: 1372-1385. Schulz R, Thiere G, Dabrowski JM (2002) A combined microcosm and field approach to evaluate the aquatic toxicity of azinphosmethyl to stream communities. Environ Toxicol Chem 21:2172-2178. Sheffield SR, Lochmiller RL (2001) Effects of field exposure to diazinon on small mammals inhabiting a semi enclosed prairie grassland ecosystem. I. ecological and reproductive effects. Environ Toxicol Chem 20: 284-296. Shelford VE (1911) Physiological animal geography. J Morphol 22: 551-618. Shelford VE (1913) Animal Communities in Temperate America. University of Chicago, Chicago, IL. Sherman PW, Runge MC (2002) Demography of a population collapse: the Northern Idaho ground squirrel (Spermophilus brunneus brunneus). Ecology 83: 2816-2831. Sherratt TN, Roberts G, Williams P, Whitfield M, Biggs J, Shillabeer N, Maund SJ (1999) A life-history approach to predicting the recovery of aquatic invertebrate populations after exposure to xenobiotic chemicals. Environ Toxicol Chem 18:2512-2518. Sibly RM, Calow P (1989) A life-cycle theory of responses to stress. Biol J Linn Soc 37: 101-116. Slovic P. 2000. The Perception of Risk. Earthscan Publications, London, UK. Smith MN (2000) The hornet robberfly Asilus crabroniformis: land use and livestock grazing regimes at sites in England. English Nature Report No. 387. SID 5 (2/05) Page 27 of 30 Smith MN (2001) The current distribution of the hornet robberfly Asilus crabroniformis Linnaeus (Diptera, Asilidae) in England and Wales. Dipterist’s Digest 8: 79-84. Snell TW, Serra M (2000) Using probability of extinction to evaluate the ecological significance of toxicant effects. Environ Toxicol Chem 19: 2357-2363. Solomon KR, Baker DB, Richards RP, Dixon KR, Klaine SJ, La Point TW, Kendall RJ, Weisskopf CP, Giddings JM, Giesy JP, Hall Jr LW, Williams WM (1996) Ecological Risk Assessment of atrazine in North American surface waters. Environ Toxicol Chem 15: 31-76. Sotherton N, Holland J (2003) Indirect effects of pesticides on farmland wildlife. In: Hoffman DJ, Rattner BA, Burton Jr GA, Cairns Jr J (eds), Handbook of Ecotoxicology. CTC/Lewis Press, Boca Raton, FL, pp 1173-1195. Soulé ME (1987) Viable Populations for Conservation. Cambridge University Press, Cambridge, UK. Southwood RE, DJ Cross (2002) Food requirements of grey partridge Perdix perdix chicks. Wild Biol 8: 175-183. Sparling DW, Fellers GM, McConnell LL (2001) Pesticides and amphibian population declines in California, USA. Environ Toxicol Chem 20: 1591-1595. Stearns SC (1992) The Evolution of Life Histories. Oxford University Press, Oxford. Stearns SC, Crandall RE (1984) Plasticity of age and size at sexual maturity: a life history response to unavoidable stress. In: Potts G. and R Wootton (eds) Fish Reproduction, Academic Press, London, pp 13-34. Stowe TJ, Newton AV, Green RE, Mayes E (1993) The decline of the corncrake Crex crex in Britain and Ireland in relation to habitat. J Appl Ecol 30: 5362. Suter GW II (1993) Ecological Risk Assessment. Lewis Publishers, Boca Raton, FL. Tagatz ME (1976) Effect of mirex on predator-prey interaction in an experimental estuarine ecosystem. Trans Am Fish Soc 105: 546-549. Tait J. 2001a. More Faust than Frankenstein: the European debate about the precautionary principle and risk regulation for genetically modified crops. J Risk Res 4:175-189. Tait J. 2001b. Pesticide regulation, product innovation and public attitudes. J Environ Monitor 3:64N-69N. Tait J, Bruce A, Lyall C. 2001. Studies on People's Values in Relation to Chemicals and their Effects on Humans and the Natural Environment: A Literature Review. Report to the Royal Commission on Environmental Pollution. SUPRA Paper 23, Edinburgh. Tanaka Y (2000) Extinction of populations by inbreeding depression under stochastic environments. Pop Ecol 42: 55-62. Tanaka Y (2003) Ecological risk assessment of pollutant chemicals: extinction risk based on population-level effects. Chemosphere 53: 421-425. Tew TE, MacDonald DW, Rands MRW (1992) Herbicide application affects microhabitat use by arable wood mice Apodemus sylvaticus. J Appl Ecol 29: 532-539. Van Apeldoorm RC, Celada C, Nieuwen huizen W (1994) Distribution and dynamics of the red squirrel (Sciurus vulgaris) in a landscape with fragmented habitat. Land Ecol 9: 227-235. Van den Brink PJ, Van Wijngaarden RPA, Lucassen WGH, Brock TCM, Leeuwangh P (1996) Effects of the insecticide Dursban® 4E (Active ingredient chlorpyrifos) in outdoor experimental ditches: II. invertebrate community responses and recovery. Environ Toxicol Chem 15:1143-1153. Van der Hoeven N, Gerritsen AA (1997) Effects of chlorpyrifos on individuals and populations of Daphnia pulex in the laboratory and field. Environ Toxicol Chem 16: 2438-2447. Van Vlaardingen P, Traas T, Aldenburg T (2003) ETX-2000: Normal Distribution Based Hazardous Concentration and Potentially Affected Fraction, RIVM, Bilthoven, The Netherlands. Veeman M, Adamowicz W. 2000. Consumers’ perceptions of environmental risks and the demand for food safety. Project Report 00-01, Department of Rural Economy, University of Alberta, Edmonton, 31 March 2000. Wagner C, Lokke H (1991) Estimation of ecotoxicological protection levels from NOEC toxicity data. Water Res. 25:1237-1242. Wake DB (1991) Declining amphibian populations. Science 253:860. Wang G, Edge WD, Wolff JO (2001) Demographic uncertainty in ecological risk assessments. Ecol Modelling 136: 95-102. Wardhaugh KG, Holter P, Longstaff B (2001) The development and survival of three species of coprophagous insect after feeding on the faeces of sheep treated with controlled-release formulations of ivermectin or albendazole. Aust Vet J 79: 125-132. Watson J, Watson A, Paull D, Freudenberger, D (2003) Woodland fragmentation is causing the decline of species and functional groups of birds in southeastern Australia. Pacific Cons Biol 8: 261-270. Wickham JD, Wu J, Bradford DF (1997) A conceptual framework for selecting and analyzing stressor data to study species richness at large spatial scales. Environ Management 21:247-257. Wilson AM, Vickery JA, Browne SJ (2001) Numbers and distribution of northern lapwings Vanellus vanellus breeding in England and Wales in 1998. Bird Study 48: 2-17. Wilson JD, Morris AJ, Arroyo BE, Clark SC, Bradbury RB (1999) A review of the abundance and diversity of invertebrate and plant foods of granivorous birds in northern Europe in relation to agricultural change. Agric Ecosyst Environ 75: 13-30. Wingfield Gibbons D, Reid JB, Chapmand RA (1993) The New Atlas of Breeding Birds in Britain and Ireland: 1988-1991. T & AD Poyser, London. Withgott J (2002) Amphibian Decline. Ubiquitous herbicide emasculates frogs. Science 296: 447-448. Woin P (1998) Short- and long-term effects of the pyrethroid insecticide fenvalerate on an invertebrate pond community. Ecotoxicol Environ Safety 41: 137-156. Zhao Y and Newman MC (2004) Shortcomings of the laboratory derived LC50 for predicting mortality in field populations: exposure duration and latent mortality. Environ Toxicol Chem 23: 2147-2153. References to published material 9. This section should be used to record links (hypertext links where possible) or references to other published material generated by, or relating to this project. SID 5 (2/05) Page 28 of 30 9.1 Refereed journal papers Crane M, Norton A, Leaman J, Chalak A, Bailey A, Yoxon M, Smith J, Fenlon J. Acceptability of pesticide impacts on the environment: what do United Kingdom stakeholders and the public value? Submitted to Pest Man Sci, April 2005. Newman MC, Holloway GJ, Crane M. Current pesticide risk assessment in Europe: does it account for long-term adverse effects on non-target organisms? Submitted to Rev Environ Contam Toxicol, June 2005. Crane M, Fenlon J, Smith J. A consistent environmental decision framework for pesticide risk management. In preparation. 9.2 Conference presentations Crane M, Norton A, Leaman J, Chalak A, Bailey A, Yoxon M, Smith J, Fenlon J. Public acceptability of pesticide effects. SETAC Europe 15th Annual Meeting, Lille, France, 22-26 May 2005. Newman MC, Holloway GJ, Crane M. Current pesticide risk assessment in Europe: does it account for long-term adverse effects on non-target organisms? SETAC Europe 15th Annual Meeting, Lille, France, 22-26 May 2005. SID 5 (2/05) Page 29 of 30 SID 5 (2/05) Page 30 of 30