SCOPE 49 - Methods to Assess Adverse Effects of Pesticides on Non-target Organisms 

4.1 INTRODUCTION

Pesticide use has resulted in acute and chronic ecological damage either by direct injury to NTOs such as birds and fish or by indirect effects such as elimination of natural enemies; particularly long-lasting effects have included the depression or stimulation of reproduction in organisms. Methods exist to detect, prior to commercial application, the potential of a pesticide to damage NTOs such as wildlife species.

When designing methods to assess the ecological consequences of pesticides, several issues must be considered. Assessing the toxic effects of pesticides on ecosystems is difficult, because so many species and processes are interacting. When significant changes in important ecological parameters become apparent only after a long time period, they are frequently undetectable in short-term experiments. Natural variation in ecological parameters requires very sensitive measurements and sampling at high frequency. Furthermore, observations made in one location may not apply to other sites because of variation among ecosystems. Finally, highly managed ecosystems may be more or less sensitive to a pesticide than a more natural community in the same site.

1. Accidental pesticide contamination of the environment may provide clues about how populations might be affected at the lower recommended use levels.
2. Testing of sensitive species can indicate whether new biological or chemical pesticides will likely have side effects.
3. Relying on experimentally created plots, field studies can be designed to measure the effects of pesticides on ecological parameters such as species diversity, energy flow, decomposition, trophic structure on indicator or monitored species.
4. Laboratory experiments can be performed with individuals, populations, or model ecosystems constructed from combinations of species chosen to represent the numerous components of natural ecosystems.

Field experiments have detected pesticidal effects on interspecies relationships such as predation. Pest resurgence has been observed following applications of select pesticides that reduce natural enemy populations and subsequently increase pest populations. Such consequences may be anticipated from laboratory investigations that measure susceptibility of a pest of its known natural enemies. Methods that measure susceptibility and use various direct-dosing and self-dosing techniques have been developed, and are recommended against many public health and agricultural pests. Host resistance to pesticides decreases selectivity and effectiveness. Biochemical techniques are under development to rapidly detect and diagnose resistance in the field.

At the ecosystem level, the challenge is to determine whether pesticide use causes changes in viability. Field methods that exist for such ecological studies have been adapted to identify effects of pesticides, and have been applied most often to the soil biota portion of the ecosystem. Soil biota have been observed to be affected in a complex manner; that is, some species increase in numbers, while others are reduced by injury. Thus, long-term observation is required for such studies. Similar effects take place above ground and in a plant host. These findings imply that extensive sampling of the total ecosystem is necessary to obtain a thorough and balanced understanding of the nature and extent of pesticide alterations likely to be manifest.

To understand the effects of pesticides on ecosystems, it is necessary to examine a spectrum of effects from lethality to subtle changes on reproduction, behaviour, and vital organ physiology.

Responses may vary between closely related species or even `biotypes' of a species, so that careful identification of test specimens is critical to learning much in depth about the genetic characteristics of the species of interest. This objective can be accomplished rapidly with newly developed molecular genetic techniques (e.g., restriction fragment length polymorphisms), so that at least the population genetics of organisms in experimental ecosystems can be taken into account.

Pesticides affect ecosystems by disrupting natural equilibrium; their effects can be observed by measuring the stability of populations, nutrient cycling, species diversity, interspecies food chains, primary production in and energy flow through trophic levels, and pollination. Some pesticides exert their effects on particular components of an ecosystem; for instance, some herbicides affect primary production in plants, and persistent organochlorine insecticides (such as DDT) bioaccumulate in higher trophic levels such as predators. Broadspectrum OP and carbamate insecticides with high acute toxicity to many species may acutely alter energy flow as well as other ecological parameters.

Prediction and assessment of ecological impacts caused by pesticides alert humans to the dangers of some alterations to the environment. Ecosystems are integrated and stable systems; they include humans, all other species on this planet, and basic biotic and abiotic processes. Components of ecosystems cannot be changed or destroyed without directly or indirectly impacting the human condition. It is to humans' benefit to consider all uses and risks of pesticides to ensure preservation of critical systems in the environment.

4.2 TROUBLESOME PESTICIDES

With about 1000 pesticide formulations in use throughout the world today, the listing of all hazards to ecosystems would be enormous, particularly with most individual pesticides having different effects on various species. Clearly delineating these effects is complicated by the fact that there are 510 million species in the environment.

The United Nations offers a list of pesticides considered most hazardous to humans. While valuable, this list is limited, since it is restricted to acutely lethal doses in humans (based on studies in the rat and other laboratory species).

In general, insecticides generally are the most toxic pesticides to the environment, followed by fungicides and herbicides. Exceptions exist for certain herbicides which are highly toxic, and are far more hazardous to the environment than are insecticides. The most hazardous pesticides include those that can be distinguished on the basis of either water or fat solubility. Water soluble compounds are easily transported out of the target area into ground water and streams; fat soluble chemicals are readily absorbed in insects, fish, and other animals, often resulting in extended persistence in food chains.

Some of the most troublesome pesticides to the ecology are:

1. insecticides: DDT, dieldrin, diazinon, parathion, and aldicarb; 
2. herbicides: 2-4-D, atrazine, paraquat, and glyphosate, and
3. fungicides: benomyl, captan, mercury, copper, and pentachlorophenol. 

4.2.1 MAGNITUDE OF PESTICIDE APPLICATIONS

Approximately 5 million ton of pesticides are applied annually in the world, of which about 70 per cent is used for agriculture, and the remainder by public health agencies and government agencies for vector control and by home owners. Thus, agriculture and forestry are the primary source of pesticides in ecosystems. In many countries, agriculture and forestry occupy approximately 50 per cent of the land area. When croplands are treated, some impacts of pesticides occur on non-target terrestrial and aquatic ecosystems, as well as on adjoining agroecosystems.

Forests are important wildlife habitats. Two broad classes of pesticides are used in forests: insecticides to control insect pests, and herbicides used to suppress the growth of shrubs during the regeneration process. Over the last thirty-five years, two large-scale pest control programmes have been used in forests of eastern Canada and of the north-eastern United States in attempts to control spruce budworm; from 1980 to 1983 in Czechoslovakia, similar programmes were initiated to control the larch bud moth. In the latter programme, a mixture of pyriphosmethyl and permethrin eliminated many invertebrate species, but recovery was relatively rapid (Tonner et al., 1983). In eastern Canadian forests, DDT caused fish mortality (Kerswill and Edwards, 1967) as a result of bioaccumulation through the food chain.

Extensive mortality of canopy-dwelling song birds has been observed with applications of phosphamidon and to a lesser extent, with fenitrothion; ground-nesting birds and small mammals were unaffected (Pearce et al., 1976). Generally, the safety margin of OPs for canopy-dwelling birds is small. Indirect effects of decreasing the biomass of insect food would be expected to affect insectivores, but that has rarely been demonstrated.

The use of herbicides to control broad-leaf plant growth during regeneration alters habitat and food-availability, but is unlikely to have any direct effect except on plants.

Most of the wide variety of other natural ecosystemsranging from alpine meadows to desertsare rarely treated with pesticides. Exceptions are programmes to control specific pests. The largest of these are those to control quella and locusts. Quella are controlled largely by the use of OPs; some direct mortality of non-target bird species and some secondary mortality have been reported. Locust control has been carried out using dieldrin-containing bait and by spraying with OPs (i.e, diazinon and fenitrothion). Studies of effects on NTOs have rarely been made. Some direct mortality would be expected, but the residence time under desert conditions is likely to be short.

Around human habitation, pesticides are used heavily; but these areas are so altered that pesticides are unlikely to exert much additional effect. Exceptions are semi-natural areas such as parks and golf-courses. The use of insecticides on the latter can be heavy, and has resulted in mortality of birds and mammals. The use of DDT around human habitation for control of malaria is unlikely to have much impact on non-target organisms, although it adds to the overall burden of this persistent material.

4.2.2 APPLICATION TECHNOLOGIES

The manner in which pesticides are applied is influential in determining the nature and magnitude of injury that may occur in ecological species. Likewise, this process also greatly impacts the effectiveness of pesticides.

According to some estimates, less than 0.1 per cent of all pesticides applied reach the target pests. If so, then a large fraction of applied pesticides may be available to contaminate water, soil, and atmosphere and may disrupt non-target species.

One reason for the small amounts of pesticides that actually reach target pests is the requirement that plant surfaces be thoroughly covered with pesticides to control small arthropods (e.g., aphids and mites) and plant pathogens (fungi and bacteria). To achieve this thorough coverage of plant surfaces requires that sprayed particles be extremely small, a situation that favours dissemination by gentle winds for distances of several thousand meters.

Achieving coverage of crop plant surfaces with pesticides is part of the problem, as is placing the pesticides into the target area. Under ideal conditions using aircraft, for instance, less than 50 per cent of the pesticide applied by aircraft reaches the target crops. With ultra-low-volume sprays (ULV) by aircraft, drift is even more extensive, placing less pesticide in the target area. Under ideal weather conditions, less than 25 per cent of the ULV-applied pesticide reaches the target. Likewise, most drifts often contaminate untreated terrestrial and aquatic ecosystems. Large quantities of pesticides are applied to orchards, vineyards, and similar crops by employing air blast sprayers. Only about 65 per cent of the pesticide applied by use of this equipment reaches the target area, whereas 35 per cent goes elsewhere.

Ground application equipment under normal weather conditions often place 7080 per cent of the sprayed pesticide into the target area. If, at the same time, a `plastic blanket' were used to protect the crop and spray boom, perhaps 90 per cent or more of the pesticide could be placed in the target area.

The physical form of pesticide formulations (i.e., wick, dust, granule, and chemigration technologies) greatly influences the extent to which a pesticide will be placed on the target site and be distributed elsewhere. One of the best means of placing pesticides on the target pests is illustrated with herbicides using the `wick' technology. A wick that has been wetted with the herbicide is drawn over the vegetation in a field. No drift occurs, and the herbicide is placed directly on the weeds without drift problems.

Dusts are highly susceptible to drift because of their extremely low weight. Seldom does 50 per cent of the applied dust remain in the target area when using ground equipment, even under ideal wind conditions.

Granules have become a popular means of applying dry insecticide materials to the surface of croplands and pasture lands. However, granules pose major problems to non-target pests, particularly birds. Birds mistake the granules for gravel, and may consume them with disastrous results.

Some pesticides can also be effectively applied in irrigation water; this practice is referred to as `chemigration'. Although this is an effective, easy method of applying pesticides to agricultural crops, it is also a means of spreading pesticides in the environment and of poisoning non-target species. In particular, it can contaminate the drinking water of birds, mammals, invertebrates, and humans. Ground and surface waters are also readily contaminated by chemigration.

Clearly, the need exists to improve pesticide application technology to assure that more pesticide reaches the target points and to reduce hazards to human and ecosystem NTOs. Finding the means to place more pesticide on target pests should not be difficult.

4.2.3 SPECIES AT RISK

Invertebrate NTOs are often killed by pesticides used against insects and acarine target pests. The susceptibility of NTOs, such as crustaceans, is due to the similarity of their physiological sites of action among target insects and certain non-target invertebrates; for example, AChE enzymes, octopamine receptors, and moulting processes are all similar insecticide targets in both pest and invertebrate NTOs. Terbufos, chlorpyrifos, and carbofuran are important insecticides for controlling corn rootworm larvae in soil; however, both terbufos and carbofuran are highly toxic to earthworms. Foliar applications of broad spectrum insecticides produce nearly total depletion of arthropod populations in crops such as cotton. While selectivity for herbivorous pests can be gained through systemic insecticides, some of these insecticides, such as aldicarb and oxamyl, are extremely toxic; thus, secondary poisoning can result in predators of treated pests.

Some selectivity of insecticides can be achieved. For example, the insect development-inhibitor methoprene controls mosquito larvae at doses not toxic to most NTO aquatic invertebrate species. Substrains of Bacillus thuringiensis are selective for certain insects in a given taxonomic order: B.t. israeleasis for mosquito larvae and B. t. kurstaki for lepidopterous larvae. Some herbicides and fungicides also affect NTO invertebrates, although they are the exceptions. Bipyridilium herbicides have a non-selective cytotoxic action, as do dinitrophenol pesticides which uncouple oxidative phosphorylation.

For pesticide registration, data are required about toxicity to representative invertebrates such as Daphnia magna, earthworms, and decapod crutaceans. Test protocols are specified by various national and international organizations such as the US EPA, OECD, and FAO.

Microorganisms are most susceptible to fungicides and bactericides aimed at target plant pathogens in the field. Some herbicides are effective against both fungi and bacteria by producing the same molecular lesions in each. In general, microorganisms lack the physiological and molecular targets (e.g., AChE and photosynthetic pathways) that render them susceptible to insecticides and herbicides. Insecticides and herbicides can deplete the insect and plant hosts of microorganisms.

Microorganisms are very important in the environmental detoxification and decomposition of pesticides. Soil ecosystems can change because of the enhanced capacity of select soils to degrade pesticides, as observed in the loss of effectiveness developed in certain soils after 10 to 20 years of continuous pesticide use. Some fungicides are known to interfere with the detoxification of OP insecticides in the soil, adding significantly to the energy flow of ecosystems which may affect species composition at lower trophic levels.

4.2.4 EFFECT AT INDIVIDUAL AND POPULATION LEVELS 

Ecosystems, whether plant or animal, can suffer from accrued damage to individuals. As an illustration, the complexities of such interactions among plants is described.

4.2.4.1 Plants

Plants are exposed to pesticides, whether as target organisms or as NTOs such as pathogenic fungi. The routes of contact include the uptake from soil and water and deposition via atmospheric drift.

Toxic and mutagenic effects and changes in the metabolism of plants (including formation of metabolites and residues capable of producing adverse effects) can be observed in NTO plants, even with quantitative and qualitative differences resulting from variations in plant sensitivity, and with the numerous metabolic pathways of pesticide degradation. Differential sensitivity of plant species to toxic and genotoxic effects of pesticides has been shown to cause overall changes in species proportion among weeds in croplands and in natural plant communities, because of the reduced abundance of susceptible species and concurrent increases in naturally tolerant species. However, neither the eradication of a susceptible species nor the occurrence of a new plant species has been established in the experimental plots even after long-term (36 years) spraying with formulations such as the herbicide 2,4-D.

The emergence among susceptible plant species of forms resistant to herbicides has become increasingly important. Pesticides can play a dual role as agents favouring the selection of pre-existing resistant mutants and as potential inducers of genetic changes. Genotoxic pesticides are potentially able to increase the pool of mutations in various qualitative and quantitative traits that could cause genetic instabilities of natural plant populations and in crop varieties.

Elements related to possible ecological damage include the formation of stable mutagenic and toxic metabolites and the accumulation of residues in plants that can be harmful to both human and animal populations via exposure through the food chain. This problem needs to be addressed despite the fact that in plants bound residues formed and incorporated into lignin, hemicellulose, and other carbohydrate components of the cell wall are usually less hazardous to the biosphere.

Pesticides have produced changes in both plant metabolism and nutritional patterns which may have further detrimental effects on the ecology. Several mechanisms are known to lead to this consequence.

1. Some herbicides, when applied at recommended dosages, have increased the attacks of insect pests and plant pathogens on crops. For example, when corn-growing areas were treated with 2,4-D at the recommended dosage of 1 kg/ha, the numbers of corn leaf aphids increased three-fold; corn borers were 26 per cent more abundant, and were 33 per cent larger than those insects present on untreated corn. Larger corn borers produce one third more eggs, and thus contribute to the build-up of corn borers on corn.

2. The insecticides monocrotophos and phosphamidon increased concentrations of nitrogen and phosphorus in rice plants, and these changes were thought to contribute to a resurgence in the numbers of the rice blue leafhopper.

3. Herbicides may increase damage due to plant pathogens. For instance, when corn was treated with a recommended dosage of 2,4-D (1 kg/ha), black corn smut grew five-fold larger than on untreated corn. Also, corn that was resistant to southern corn leaf blight lost its resistance to the blight when treated with 2,4-D.

4. Most nutrients, especially C, K, N, P, and S, are taken up by plants that, in turn, may be eaten by animals. These nutrients are eventually returned to the soil or atmosphere via decomposition of dead organisms. The amounts and forms of nutrients in soils and plants may be changed by pesticides, thereby altering the dynamics of these animals in the ecosystem.

5. Pesticides can alter the chemical make-up of plants. The changes that occur appear to be specific for both the plants and pesticides involved. For example, certain organochlorine insecticides have increased the amounts of some macro and micro-elements (Al, B, Ca, Cu, Fe, K, Mg, Mn, N, P, Sr, and Zn) of corn and beans. DDT, aldrin, endrin, and lindane have stimulated synthesis of the essential amino acids arginine, histidine, leucine, lysine, proline, and tyrosine in corn, but have decreased the content of tryptophan. The herbicide simazine increased water and nitrate uptake in barley, rye, and oat seedlings, resulting in increased plant weight and total protein content.

Increasing insect pest and plant pathogen attacks on crops by using some herbicides may, in turn, lead to an increase in spraying of additional pesticides, like insecticides and fungicides. Thus, the environmental problem arising from the use of herbicides may involve more than just the herbicide itself.

Methods are available to detect both non-genotoxicity and genotoxicity of pesticides in plants, including their effects on plant reproduction. The toxic effects of pesticides have been measured in plants growing in separated experimental plots using endpoints such as frequency of species and individuals in each species, survival of plants, fresh and dry weight of plants at the harvest, and seed setting. Cytogenetic analyses of mitotic cells in roots and shoots, meiotic cells in pollen mother cells, and postmeiotic cells in pollen represent the most frequently used methods to detect genetic changes caused in natural plant populations and communities or in plants growing in the field. Special plant assay systems are also available to monitor genotoxicity in situ.

About 150160 pesticides have been tested on plant genotoxicity assay systems. About 90 per cent of the test agents produced some kinds of changes in the chromosome structure, about 60 per cent caused disturbances in the meiosis and seedset reduction, and about 70 per cent caused gene mutations. Compared with other assay systems (e.g., microbial cells, mammalian cell cultures, insects, and whole animals), plant assays appear to be the most efficient to detect pesticide-related genotoxicity. Seasonal variation in cytogenetic endpoints have been correlated with the application of herbicide mixtures.

4.2.4.2 Animals

The more obvious effects of pesticides (i.e., reproductive failure caused by DDT and mortality caused by OPs and carbamates) have, so far, been the basis of concern and regulatory action.

The major adverse effects of organochlorine (OC) pesticides have been manifested through effects on reproduction. DDT has caused eggshell thinning in several high trophic level avian species and sufficient impact on reproduction to result in population declines (Risebrough, 1986). Likewise, the effects on fish occurred largely during the reproductive cycle (i.e., at the time that the yolk sac was absorbed) (Burdick et al., 1967).

By contrast, the major effect of OPs and carbamates has been direct mortality. The mechanism of action of these insecticides is the inhibition of AChE activity which causes the disruption of nerve function. Acute inhibition of 80 per cent and chronic inhibition of 50 per cent has been associated with mortality (Ludke et al., 1975; Hill and Fleming, 1982). Other esterases are also inhibited, and the inhibition of brain neurotoxic esterase (NTE) has been related to delayed neuropathy (Johnson, 1975). Lotti and Johnson (1978) compared the toxicity of inhibition of AChE and NTE for a range of OPs. The ratio of inhibition of the two enzymes varies over several orders of magnitude, but the degree of inhibition of AChE correlates well with the LD₅₀ and that of NTE with delayed neurotoxic damage. For those pesticides causing neurotoxic effects, inhibition of NTE can be one to two orders of magnitude more sensitive than that of AChE.

The relationship between dosage of the pesticide and degree of inhibition of AChE has been used to assess the impact of insecticide spray programme on forest songbirds (Mineau and Peakall, 1987). The data show a severe collection bias in favour of birds with inhibition of the enzyme of less than 30 per cent. This manifestation is probably due to behavioural effects on birds having AChE inhibition greater than this value.

Since behaviour is the result of integration of many inputs, it has long been considered as a potentially sensitive indicator of pesticide toxicity (Warner et al., 1966). For OPs, behavioural alterations are demonstrated only at AChE inhibition of 4050 per cent (Grue et al., 1982; Rudolph et al., 1984). The conclusion that behavioural effects are not particularly sensitive has been confirmed by a review of the effects of toxic chemicals on birds (Peakall, 1985). While it can be stated that behavioural changes will reduce an organism's ability to survive in the wild, little direct evidence exists to confirm this hypothesis. In some cases, predation experiments have been conducted in the laboratory (Brown et al. 1985); there remains a major extrapolation to field conditions. Furthermore, in this experiment a significant increase in predation was not seen until the concentration equivalent to half the LD₅₀ was used.

4.2.5 MEASURES OF EFFECTS ON STRUCTURE AND FUNCTION OF ECOSYSTEMS

Terrestrial and aquatic ecosystems are generally self-sufficient basic living units of nature that include a complex of species dependent on one another, and interacting with the physical-chemical environment. Pesticides may alter the structure (species richness, density, and biological diversity) and functional activities of an ecosystem. Pesticides may alter the self-sufficient nature of natural ecosystems that include plants (producers), herbivores, parasites and predators, and decomposers*.

Some pesticides are capable of destroying some species totally or of significantly reducing the populations of others. When the diversity of the ecosystem is reduced sufficiently, then food chains may be shortened or altered in diverse ways. When food chains are changed, the stability of ecosystems may be reduced, leading a susceptible ecosystem to extinction by a variety of mechanisms such as invasion by other complexes of species.

A critical component of all ecosystems is energy flow. Plants collect energy from the sun for growth, metabolism, and reproduction. Energy fixed by the plants eventually becomes available to herbivores and other species that make up the ecosystem. The more energy the plants collect, the more productive, diverse, and complex is the ecosystem. In aquatic ecosystems, eutrophicationa form of energy cyclingmay enhance or limit diversity. If a pesticide influences the growth of the plant populations, then the food/energy supply for the ecosystem is reduced, lowering its productivity and stability.

Another essential component of an ecosystem is the decomposition of organic matter. The basic elements (e.g., C, Ca, H, K, Mg, Mn, N, O, P) are vital to the proper functioning of all life, including ecosystems. Thus, the decomposers are essential to keep the vital nutrients in circulation for use and reuse by an ecosystem.

Some essential nutrients are in the atmosphere. Plants and other organisms may obtain some essential elements (C, H, O, N) from the atmosphere for use throughout an ecosystem. Other nutrients may be obtained directly from soil or water, and are cycled through the biota. Pesticides may be capable of reducing the variability of one or more organisms involved in the recycling process in an ecosystem. If this occurs to a large extent in an ecosystem, it may function at such a reduced rate as to threaten the entire web in the ecosystem.

*Decomposers are organisms (microbes, earthworms, insects, etc.) responsible for the decomposition of organic compounds other than their nutrients.

Methods to quantify ecological effects of pesticides on terrestrial and aquatic ecosystems are generally complex and costly, the magnitude and complexity being dependent on the nature of the pesticides and on the characteristics of the particular terrestrial and aquatic ecosystem in which pesticide is used.

Presently, ecosystem parameters are not among those data developed prior to registration of pesticides in most countries. The effects on two ecosystem components seem most urgently to require assessment prior to pesticide registration: these are the effects on the trophic structure (plants, herbivores, parasites, and predators), and on decomposer organisms, because of the unique vulnerability demonstrated by these components for ecosystem integrity and survival.

Any new pesticide, whether biological or chemical, should first be assessed for ecosystem effects in the particular agroecosystem of proposed primary use (e.g., an apple orchard or a cotton field). This approach requires the measurement of population dynamics for a representative number of resident species in addition to the target pests and their natural enemies. The number of species to be sampled should be in proportion to the complexity of the agroecosystem of interest. Effects on decomposers can be determined by cellulosic sampling by litterbag methods, for a minimum of three growing seasons.

4.3 RECOMMENDATIONS 

Non-chemical pest controls, such as biological, cultural, and environmental controls and host plant resistance, should be improved and used in order to reduce pesticide use.

4.4 REFERENCES

Brown, J. A., Johansen, R. H., Colgan, P. W. and Mathers, R. A. (1985) Changes in the predator-avoidance behaviour of juvenile guppies (Poecilia reticulata) exposed to pentachlorophenol. Can. J. Zool. 63, 2001-2005.

Burdick, G. E., Harris, E. J., Dean, H. J., Walker, T. M., Shea, J. and Colby, D. (1964) The accumulation of DDT in Lake Trout and the effect on reproduction. Trans. Amer. Fish Soc. 93, 127-136.

Grue, C. E., Powell, G. V. N. and McChesney, M. J. (1982) Care of nestlings by wild female Starlings exposed to an organosphosphate pesticide. J. Appl. Ecol. 19, 327-355. 

Hill, E. F. and Fleming, W. J. (1982) Anticholinesterase poisoning of birds: field monitoring and diagnostic of acute poisoning. Environ. Toxicol. Chem. 1, 27-38.

Johnson, M. K. (1975) The delayed neuropathy caused by some organophosphate esters: mechanism and challenge. Crit. Rev. Toxicol. 37, 113-115.

Kerswill, C. J. and Edwards, H. E. (1967) Fish losses after forest spraying with insecticides in New Brunswick, 1952-1962, as shown by caged specimens and other observations. J. Fish Res. Bd. Canada 24, 709-729.

Lotti, M. and Johnson, M. K. (1978) Neurotoxicity of organosphosphate pesticides: predictions can be based on in vitro studies with hen and human enzymes. Arch Toxicol 41, 215-221.

Ludke, J. L., Hill, E. F. and Dieter, M. P. (1975) Cholinesterase (ChE) response and related mortality among birds fed ChE inhibitors. Arch. Environ. Contamin. Toxicol. 3, 1-21.

Mineau, P. and Peakall, D. B. (1987) Review: an evaluation of avian impact assessment techniques following broadscale forest insecticide sprays. Environ. Toxicol. Chem. 6, 781-792.

Peakall, D. B. (1985) Behavioral responses of birds to pesticides and other contaminants. Residue Rev. 96, 45-77.

Pearce, P. A., Peakall, D. B. and Erskine, A. J. (1976) Impact on forest birds of the 1975 spruce budworm operation in New Brunswick. CWS Progress Note # 61.

Risebrough, R. W. (1986) Pesticides and bird population. Current Ornithology 3: 397-427.

Rudolph, S. G., Zinkl, J. G., Anderson, D. W. and Shea, P. J. (1984) Prey-capturing ability of American Kestrels fed DDE and acephate and acephate alone. Arch. Environ. Contamin. Toxicol. 13, 367-372.

Tonner, M., Vavra, V., Syrovatka, O. and Soldan, T. (1983) Einfluss der Luftbespritzung genen grauen Larchenwicker (Zeiraphera diniana) auf die Entomofauna des Rhithrons in Krkonose (Riesengebirge). Vest. Cs. Spol. Zool. 47, 293-303.

Warner, R. E., Peterson, K. K. and Borgman, L. (1966) Behavioural pathology in fish: a quantitative study of sublethal pesticide toxication. J. Appl. Ecol. 3, (Suppl), 223-247.