Ecotoxicology, 12, 307±316, 2003 # 2003 Kluwer Academic Publishers. Manufactured in The Netherlands.

Neurotoxic Pesticides and Behavioural Effects Upon Birds

C. H. WALKER Cissbury, Hillhead, Colyton, Devon EX24 6NJ, UK

Accepted 1 October 2002

Abstract. Organochlorine, organophosphorus, carbamate, pyrethroid and neonicotinoid insecticides and organomercuryfungicides are all neurotoxic and therefore have the potential to cause behavioural disturbances in birds. A number of studies have described behavioural effects caused to captive birds byneurotoxic pesticides, but it is verydifficult to measure such effects in the field, which is a serous limitation given their potential to cause adverse effects at the population level. The mode of action, and the neurotoxic and behavioural effects of these compounds are brieflyreviewed before considering evidence for their effects in the laboratoryand field. Behavioural effects maycause adverse changes at the population level either directlyor indirectly.Direct effects upon avian populations maybe due to disturbances of reproduction, feeding, or avoidance of predation. Indirect effects on predators maybe the consequence of direct action upon the preypopulation leading to either (1) reduction of numbers of the preypopulation, or (2) selective predation bythe predator upon the most contaminated individuals within the preypopulation. Attention is given to the historic evidence for neurotoxic and behavioural effects of persistent organochlorine insecticides, raising the question of retrospective analysis of existing data for this once important and intensively studied class of compounds. Less persistent pesticides currentlyin use mayalso have neurotoxic effects upon birds in the field. Sometimes, as with some OPs, their effects mayoutlast the persistence of their residues, and the ecotoxicityand persistence of some maybe affected byinteractions with other environmental chemicals. The development of new mechanistic biomarker assays could improve understanding of behavioural effects and possible associated effects at the population level caused bysuch compounds in the field.

Keywords: neurotoxicity; pesticides; birds; behavioural effects; organochlorine insecticides; organophosphorus insecticides; organomercuryfungicides

Introduction use of organophosphorus, carbamate and pyrethroid insecticides and the organomercuryfungicides because Since the early1960s and the publication of `Silent of established or perceived harmful effects upon Spring' byRachel Carson in 1962 much attention has animals in the field. It is noteworthythat the four main been given to the possible effects of pesticides upon groups of insecticides which are, or have been, widely birds and other vertebrates in the field. Evidence of used are all neurotoxic (Walker, 2001); the recently harmful effects of organochlorine insecticides such as developed neonicotinoids are also neurotoxic. It DDT, aldrin and dieldrin upon wild populations of appears that receptors within the nervous system predatorybirds contributed to the decision to impose present especiallyvulnerable targets not onlyfor man wide ranging bans on the use of these compounds made pesticides, but also for plant toxins that have worldwide. Restrictions have also been placed on the evolved during the course of `plant-animal warfare' 308 Walker

(Harborne, 1993; Walker, 2001). Herbicides and varying considerably in their complexity and efficacy. fungicides, bycontrast, rarelyshow high toxicityto Unfortunatelysome of the tests which seemed animals. most relevant from an ecological point of viewÐfor The neurotoxic character of insecticides attracted example, on breeding behaviour, and capture of the attention and interest of behavioural biologists preyÐwere the most difficult to perform. In field working in the field of ecotoxicologyat an earlystage. studies it was particularlydifficult to identifybeha- Manykinds of physiological disturbances caused by vioural effects of environmental chemicals. Finally, exposure to pollutants could, in theory, lead to beha- he concluded that, although behavioural effects vioural changes but neurological disturbances seemed had been demonstrated down to doses one order of to be paramount importance. Alterations in the func- magnitude below those causing lethality, they were tion of the nervous system were very likely to be less sensitive than biochemical changesÐbiochem- expressed as changes in behaviour. Behavioural effects ical changes that could be incorporated into biomarker of pesticides were looked at from two related points of assays. view. On the one hand, toxicitytests working to beha- The present short review focuses on the possible vioural end points might present a sensitive means of ecological significance of behavioural effects caused testing for toxicityof environmental chemicals in an bypesticides. It draws on material from the above integrative way; that is, behavioural responses could review, and on discussions with David PeakallÐ represent an integrative measurement of lower level although interpretations are the responsibilityof the effects caused byone or several chemicals. The other author alone. point of view was that behavioural effects in the field might lead to population declines as a consequence of, for example, reduced abilityto find food or to repro- Ecological consequences of behavioural effects duce. The issue to emphasise here is that the useful- ness of the former approach is dependant on the Behaviour is here defined as the overt action an significance of the latter. The case for developing such organism takes to adjust to environmental circum- tests becomes a strong one if it can be shown that stance so as to ensure its survival (Odum, 1971). The pesticides cause population declines bybehavioural implementation of overt actions is a function of the mechanisms following realistic levels of exposure in nervous system, and the disruption of nervous function the field. byneurotoxic compounds maybe expected to cause An earlycontribution to this subject came from changes in behaviour. In the living environment, Warner et al. (1966), who developed a test system for pollutants maycause changes at the population level fish which utilised several behavioural parameters. becauseofDIRECT orINDIRECT behavioural effects. Theywere able to demonstrate changes in the beha- Direct effects maybe upon feeding, avoidance of pre- viour of goldfish caused bysublethal concentrations dation or reproduction or anycombination of these. of toxaphene (severe changes at 1.8 mg/l, more mild Adverse effects upon anyof these functions maylead effects at 0.44 mg/l). Toxaphene exerts neurotoxicity to population declines. Indirect effects upon predatory byacting upon GABA receptors. More recently, birds maybe due to pollutants having behavioural or Beauvais et al. (2000) have demonstrated behavioural other effects upon their prey. Thus, the decline of a changes in larval rainbow trout exposed to sublethal preyspecies caused bybehavioural effects maylead doses of the organophosphorus (OP) insecticides to a decline of the predator consequent upon a diazinon and malathion, even down to doses that reduction of food supply. An important consideration onlycaused 20% inhibition of brain cholinesterase. with putative behavioural changes caused bypollu- These and other studies have described assays for fish, tants in the field is the duration of the EFFECT. which can demonstrate behavioural changes at levels In theorypersistent neurotoxic compounds like of exposure well below lethal concentrations. A the organochlorine insecticides dieldrin and DDT critical and detailed review on the question of generallypresent a greater hazard than more biode- behavioural responses of birds to pesticides and gradable compounds such as organophosphorus, other environmental contaminants was published by carbamate or pyrethroid insecticides, because neuro- David Peakall (1985). He concluded that a wide toxic action is likelyto be much more prolonged. That varietyof behavioural tests had been tried in birds, said, there are some non-persistent compounds which Neurotoxicityin Birds 309 produce effects that outlast the compounds originally Mechanisms of neurotoxic action causing them, a point that will be returned to shortly. Considering direct effects upon feeding, these may Several different receptors within the nervous system be upon one or several different components of this are targets for the action of neurotoxic insecticides. activity(Atchison et al., 1996). Included here are Theyhave contrasting functions and locations within searching for, encountering, choosing, capturing and the peripheral and central nervous systems of verte- handling of prey. Learning and the functioning of the brates and invertebrates (Eldefrawi and Eldefrawi, sensorysystemare important in searching, encounter- 1990; Krieger et al., 2001). There are marked differ- ing and choosing. Capturing preymaysometimes be ences between species and groups in the form and verydifficult and involve highlyskilled hunting tech- distribution of these targets. In principle, the nature of niques and sophisticated coordination. Indeed, adverse behavioural effects caused byneurotoxic compounds effects of pollutants upon capturing and handling prey should depend on the type of receptor, and conse- have been demonstrated in a number of aquatic quentlythe mode(s) of action involved. The situation species (Atchison et al., 1996). Direct effects mayalso is complicated, however, bythe possibilityof differ- be upon the abilityof birds to avoid predation, as has ences between species/groups in their response to any been well demonstrated in a number of studies with particular insecticide. Some of the more important fish (Beitinger, 1990). Although it seems verylikely modes of action of insecticides will now be briefly that the same principle mayapplyin the case of birds, reviewed. hard evidence is very scanty. In one study, Buerger Organophosphorus and carbamate insecticides act et al. (1991) dosed bobwhite quail with methyl principallythrough inhibition of acetylcholinesterase parathion before releasing them and observing sub- which is located in the postsynaptic membrane of sequent events with the aid of radiotelemetry. Over a cholinergic synapses. Cholinergic synapses occur in studyperiod of 3 years, it was found that survival time both the central and peripheral nervous systems of was significantlyless in birds receiving 6 mg/kg of the vertebrates. Earlysymptoms of anticholinesterase OP than in control birds. This was attributed to greater poisoning in vertebrates are manyand varied, but death susceptibilityto predation. Direct behavioural effects is often due to respiratoryfailure. A few organopho- upon reproduction can, in theory, involve disruption sphorus compounds, for example, DFP, mipafox and of pairing or mating, failure to incubate eggs, desertion methamidophos can also cause delayed neuropathy, of nest, and failure to protect nest or to feed young. referred to as `organophosphate induced-delayed- One studywith four different species of ducks in the neuropathy(OPIDN)' (Johnson, 1992). Here, dete- field gave evidence of reduced survival of ducklings rioration of peripheral nerves occurs some time after following the application of methyl parathion (1.4 kg exposure and the disappearance of OP from the body. a.i./hectare) which mayhave been the consequence of The effect is the long term consequence of the brood abandonment (Brewer et al., 1988). combination of the OP with an esterase of the nervous Indirect effects maybe the outcome of behavioural system termed `neuropathy target esterase'; usually changes caused to preyspecies. Apart from the ques- the symptoms only appear with the ageing of the bound tion of reductions in preypopulations, behavioural OP moietyattached to the enzyme(Johnson, 1992). changes of preymayforeshadow selective predation. There is evidence that birds are particularlysensitive Neurotoxic compounds can impair the abilityof the to this type of neurotoxicity. Indeed, the domestic preyspecies to avoid predation; theymayalso cause fowl is the preferred laboratoryspecies when testing aberrant behaviour which draws the attention of the for OPIDN. Recentlyit has been shown that OPs can predator. For example, preyspecies mayshow unusual, bind verystronglyto sites other than cholinesterase or poorlycoordinated movements as well as tremors and neuropathytarget esterase in rat brain (Richards et al., twitches in the earlystages of poisoning byorgano- 1999). Evidence for the existence of OP targets other chlorine or organophosphorus insecticides. The than acetylcholinesterase has contributed to the potential significance of thisÐespeciallywith persis- ongoing controversyabout possible long term neuro- tent insecticidesÐis that predators will ingest rela- logical effects of OPs upon sheep farmers, and gulf tivelylarge amounts of toxic compounds, because war veterans. It remains a possibilitythat birds exposed theyfeed preferentiallyon the most contaminated to OPs in the field mayhave experienced long term members of the preypopulation. neurological and behavioural effects unrelated to the 310 Walker inhibition of acetylcholinesterase, although this has than its principal stable metabolite, p,p 0 DDE; it is not been demonstrated in field studies. some 3-fold more toxic over a range of species Pyrethroid insecticides and DDT act upon the (Environmental Health Criteria 9, 1979). In one study voltage-sensitive Na‡ channels of axonal membranes. with the feral pigeon (Columba livia), birds were Theyslow down the inactivation of Na ‡ channels dosed with feed containing p,p 0 DDE at the level of when an action potential is developed. Consequently 1,000 ppm over a period of 28 days (Bailey et al., theycan cause repetitive discharge from nerves, 1969a). Twelve out of thirty-six birds died during this therebydisrupting the transmission of the action poten- period, but none of them showed symptoms asso- tial. Typical symptoms of poisoning of vertebrates ciated with p,p 0 DDT poisoningÐin contrast with include tremors, twitches and prostration. Type II birds receiving the latter compound in the same experi- pyrethroids, which contain a cyano-group, can also ment, manyof which showed marked and sustained cause convulsionsÐa symptom that is characteristic tremoring (Baileyet al., 1969b). p,p 0 DDE has been of action upon gamma-amino butyric acid (GABA) shown to cause eggshell thinning in birds, probably receptors rather than Na‡ channels (see next para.). because of its abilityto inhibit the calcium pump oper- Cypermethrin and decamethrin are examples of ated bycalcium ATP-ase in the shell gland (Lundholm, Type-II pyrethroids. 1987). In sensitive species such as the American kestrel Cyclodiene insecticides such as aldrin, dieldrin, (Falco sparverius), p,p 0 DDE levels as low as 3 ppm heptachlor and endrin and gamma HCH (lindane) act in the feed can cause eggshell thinning (Wiemeyer as inhibitors of GABA receptors (Krieger, 2001). The and Porter, 1970). Thus, there is evidence not GABA receptor is the major type of inhibitory recep- onlyof differential toxicitybut also of a different tor in the vertebrate brain, and the cyclodienes act mode of action when comparing p,p 0 DDT with its upon the GABA A form of it. This interaction in the principal persistent metabolite. central nervous system leads to a number of neurolo- With these points in mind, it is a little surprising gical disturbances, including characteristic tonic that most of the behavioural studies on birds for these convulsions. Predatorybirds poisoned bycyclodienes compounds have been conducted with p,p 0 DDE often show clenched claws. rather than the markedlyneurotoxic parent compound Organomercuryfungicides are also neurotoxic (Peakall, 1985). One or two studies have demonstrated (Environmental Health Criteria 86, 1989; Wolfe behavioural effects of p,p 0 DDE, notablyin a studyof et al., 1998; Walker, 2001). Methyl mercury, for exam- ring doves (Streptopelia risoria) byHaegele and ple, can cause brain damage accompanied bya variety Hudson (1977). Courtship behaviour was shown to be of neurological and behavioural disturbances, includ- affected by10 ppm p,p 0 DDE in the diet administered ing difficulties with walking and flying (Borg et al., over a period of 63 days, which yielded a brain 1970). The toxicityof organic mercuryis related to its residue of 2.9 ppm. In another study, disturbance of tendencyto bind to the -SH groups of proteins within nocturnal activityof white-throated sparrows ( Zono- the brain. trichia albicollis) was caused byboth p,p 0 DDT and p,p 0 DDE, at dietarylevels of 5 ppm or 25 ppm (Mahoney, 1975). Effects of neurotoxic pesticides With DDT and other persistent neurotoxic com- pounds there maybe selective predation upon intoxi- DDT and metabolites cated individuals of the preypopulation. Such selective predation has been demonstrated experimentallyin The toxic effects of DDT have frequentlybeen related the case of newts (Triturus cristatus) preying upon to the levels of persistent residues of the parent tadpoles of the frog (Rana temporaria) (Cooke, 1971). compound and its stable metabolites. Thus, bythe Tadpoles dosed with DDT showed hyperactivity analysis of carcasses (found dead or shot) and of eggs, when theycontained 0.5±0.6 ppm of p,p 0 DDT, and it has sometimes been possible to attribute harmful became relativelyinactive (`resigned') at somewhat effects in the field to these compounds (Walker, 2001). higher tissue levels. The number of succesful attacks However, the issue is complicated bythe question of bythe newts was 8±9 fold greater upon dosed tadpoles the differential toxicityof the residues. p,p 0 DDT than upon controls. When DDT was widelyused there generallyhas higher acute toxicitytowards vertebrates were examples of birds being poisoned bythis Neurotoxicityin Birds 311 insecticide in the field (see for example, Stickel et al., or below blood concentrations 30% of those asso- 1966; Baileyet al., 1970). It is tempting to speculate ciated with lethal intoxication. that there mayhave been selective predationÐand Peakall (1985) cites a number of laboratorystudies therefore enhanced uptake of DDT residuesÐby where sublethal doses of dieldrin were shown to cause raptorial birds. behavioural effects in sharp-tailed grouse (Tympanu- chus phasianellus), bobwhite quail (Colinus virginia- The cyclodienes nus), Japanese quail (Coturnix japonica), pheasant (Phasianus colchicus) and mallard duck (Anas The following account refers mainlyto aldrin, dieldrin platyrhynchos). Effects reported included a decrease and heptachlor, the use of which was largelydis- of aggressive behaviour in mallard, effects on social continued bythe 1980s because of harmful effects and breeding behaviour in sharp-tailed grouse, in upon the environment and perceived human health operant behaviour of bobwhite quail, of avoidance hazards. Other cyclodienes include endrin, endosulfan, behaviour of Japanese quail and a varietyof effects in chlordane and telodrin. The toxicityof aldrin appears the pheasant. Behavioural disturbances were observed to be due to its stable metabolite dieldrin. in bobwhite quail with brain levels of 4 ppm dieldrin. With vertebrates, a characteristic symptom of In a studyconducted upon the mallard duck severe cyclodiene poisoning is convulsions, in keep- (Sharma et al., 1976), behavioural changes were ing with action upon GABA receptors associated with observed when birds received dietarylevels of 4, 10 or inhibitory synapses of the central nervous system. 30 ppm dieldrin in their feed, and these were related to However, with a continuing low level of exposure a reductions in the levels of the biogenic amines number of other milder symptoms are often shown serotonin, dopamine and noradrenaline. Increased before tissue concentrations are high enough to cause avoidance and decreased encounters were observed convulsions. This has been observed both in occupa- at the lowest dose (4 ppm), where the concentrations tionallyexposed workers and experimental animals of dieldrin in brain and liver were 0.12 and 2.3 ppm, dosed with cyclodienes in the laboratory (Jaeger, respectively. At a dose of 10 ppm, the effects on beha- 1970; Environmental Health Criteria 91, 1989). viour were more pronounced and the brain and blood Workers suffering from sublethal dieldrin intoxica- concentrations of dieldrin were 0.26 and 0.42 ppm, tion have experienced a varietyof symptoms includ- respectively. Heinz et al. (1980) observed decreases in ing headache, dizziness, drowsiness, hyperirritability, dopamine and norepinephrine in ring doves after nausea, anorexia and, in more severe cases, convul- receiving 4 or 6 ppm dieldrin in their food. sions. Broadlyspeaking, these symptoms were shown During the late 1950s and the 1960s there were many with blood levels of the order 200±600 mg/l dieldrin; reports of granivorous and predatorybirds dyingof above this came severe convulsions and death. In a cyclodiene poisoning in the field (see Walker, 2001). recent publication (Krieger, 2001), the effects of long In Great Britain long term studies were initiated to term chronic exposure to cyclodienes have been monitor residues in birds found dead in the field and in summarised. After exposure over a period of months eggs, and to relate these, where possible, to population (most frequently8 months or more), mild symptoms trends of certain raptors locallyand nationally. of poisoning began to be manifest, which progressed Particular attention was given to the sparrowhawk to more severe symptoms. Initially these symptoms (Accipiter nisus) (Newton, 1986; Newton and Wyllie, included mild, persistent headache, dizziness, insom- 1992) and the peregrine (Falco peregrinus) (Ratcliffe, nia, nystagmus, diplopia, tinnitus, slight involuntary 1993). With both species population crashes in movement and blurred vision. More severe symptoms Britain during the late 1950s/early1960s were related were myoclonic jerkings of one or more limbs, to the effects of cyclodienes. This interpretation was sometimes progressing to convulsions. strengthened bythe subsequent recoveryof both Experimental animals exposed to sublethal levels populations with declining residues of dieldrin and of cyclodienes have shown changes in electroence- heptachlor in the birds following largescale bans and phalogram patterns, disorientation, loss of muscular restrictions on the insecticides. One puzzling observa- coordination, vomiting and convulsions (Environ- tion was that sparrowhawk and kestrel (Falco mental Health Perspectives 91, 1989). In the human tinnunculus) populations began to recover in the worst studies overt symptoms of intoxication were shown at affected areas onlyafter dieldrin residues (geometric 312 Walker means) in the livers of birds found dead in the field sublethal effects were important not onlyin the initial fell to 1 ppm and below (Newton and Wyllie, 1992). decline of the sparrowhawk but also in the delayin Laboratorystudies with Japanese quail and domestic recoveryafter the critical bans on the use of cyclo- pigeons had suggested that individuals dying from dienes. It also seems highlyprobable that starvation acute dieldrin poisoning normallycontained 10 ppm consequent upon a loss of hunting skills contributed to or more of the chemical in liver (Robinson et al., these population changes. 1967). Thus, veryfew individuals (no more than 5%) of the birds found dead had high enough residues to Organophosphorus (OP) and carbamate suggest death bydieldrin poisoning. Was this suffi- insecticides cient to explain the failure of the raptors to recover? Residue data for sparrowhawks and kestrels for the With the withdrawal of organochlorine insecticides period 1963±86 were reexamined in an attempt to from the market, the attention of environmentalists throw further light on this question (Walker and moved increasinglyto the less persistent insecticides Newton, 1999). In the area most highlycontaminated such as OPs and carbamates which superceded them bydieldrin, both species showed a wide range of resi- and whose toxic effects were likelyto be more transi- dues, and frequencydiagrams showed peaks centring tory. This did not, however, resolve all concerns about on c 20 ppm which evidentlyrepresented individuals possible behavioural effects. Although the original dying from acute dieldrin toxicity. There was also, compounds and their active metabolites have short however, a substantial number of individuals contain- half lives, the biochemical lesions that theycause ing residues of 3±9 ppm, these amounting to some 20± maybe more prolonged. In the first place inhibited 35% of the entire samples collected when numbers of acetylcholinesterase can be slow to reactivate, raptors were verylow. It seems clear that most of the especiallywhere ageing has occurred, and restoration individuals with these residues would have experi- of activitymaydepend largelyupon the synthesis enced sublethal effects of the kind described earlier, of new enzyme. There is also the question of OPs and that effects such as loss of coordination, disori- operating through other sites of action. As described entation and convulsions would veryprobablyhave earlier, a few OPs have been shown to interact with impaired hunting skills. (As noted above, Sharma et al., neuropathytarget esterase (NTE) causing delayed 1976 demonstrated behavioural effects and depressed neuropathy(Johnson, 1992). There is also evidence of levels of biogenic amines caused bydieldrin in other sites in the central nervous system which are mallards at a concentration of 2.3 ppm in the liver.) more sensitive to attack byOPs than acetylcholines- Highlyspecialised predators such as the sparrowhawk terase itself (Richards et al., 1999). Thus, it is not safe which feed upon highlymobile prey(other birds in to assume that all neurotoxic effects produced in the this case) are dependant upon sophisticated hunting field will be transitory. Finally, even transitory skills (Newton, 1979). behavioural effects maybe of ecological significance, To explore these ideas further, data for the sparrow- notwithstanding the difficultyof detecting them in hawk was incorporated into population models (Sibly the field. et al., 2000). Two postulates were compared: In the investigation of OP effects upon birds and other vertebrates the inhibition of brain cholinesterase (1) that deaths attributable to dieldrin were limited to can be used as a biomarker assayfor both exposure individuals with 9 ppm or more of the insecticide and, more importantly, toxic effect. This topic has in liver (29% of total sample); been reviewed byMineau, 1991. Manystudies have (2) that deaths attributable to dieldrin included all shown that sublethal doses of OPs can cause beha- individuals with 3 ppm or more in the liver (65% vioural effects in birds, and frequentlythese effects of total sample). have been related to levels of cholinesterase inhibi- tion (Peakall, 1985; Mineau, 1991). In manycases Onlyin model 2 was the percentage mortality behavioural effects were apparent when depression attributed to dieldrin sufficientlyhigh to explain the of brain acetylcholinesterase exceeded 40% (Grue population decline of the sparrowhawk in Eastern et al. in Mineau, 1991). There was, however, England in terms of the adverse effects of this considerable variation between compounds and chemical. Overall, these results stronglysuggest that species, and in some cases such effects were not Neurotoxicityin Birds 313 apparent until there was a greater reduction of activity Organomercury fungicides than 40%. Grue et al. (1991) observed that `the general Like aldrin and dieldrin, cases of secondarypoisoning picture of acute anticholinesterase intoxication is byorganomercurycompounds were reported during mainlyone of ``depression'', as shown bya reduction the 1950s, and were particularlyassociated with the of a wide varietyof behavioural outputs, both innate use of methyl mercury as a fungicidal seed dressing and learned'. However, at an earlystage of intoxica- (Borg et al., 1970). Earlysymptoms of organomercury tion, reduced activitymayalternate with periods of poisoning in birds are similar to those observed in hyperactivity. Grue and Shipley (1981) dosed male mammals (Borg et al., 1970; Wolfe et al., 1998). starlings (Sturnus vulgaris) with dicrotophos to cause Reduced food consumption and weakness of the around 50% inhibition of brain acetylcholinesterase extremities leads to poor muscle coordination and activity. The dosed birds were less active than ataxia. In the later stages birds can neither walk nor controls, showing significantlymore time perching fly. Such effects in the field would certainly affect the and significantlyless time foraging, singing, and abilityof birds to feed; as with cyclodienes, early displaying. Hart (1993) dosed starlings with chlor- effects of organomercurypoisoning would be likelyto fenvinphos and observed that inhibition of brain affect the hunting skills of predatorybirds. Although cholinesterase byaround 40% reduced flyingand lethal and sublethal effects of organomercurywere singing and increased resting. Fryday et al. (1995) reported from the field in Scandinavia at the time that showed that a dose of 5 mg/kg chlorfenvinphos organomercuryseed dressings were in use, no studies reduced the flying activity of starlings over a 2 h from that time established effects at the population period. An important consequence of the depressive level. effect of OPs on birds is to reduce counting efficiency in field studies of pesticide effects, because affec- ted birds maybe overlooked on account of their Discussion immobility(Grue et al., 1991). Fenitrothion fed to black-capped chickadees (Parus atricapillus) caused The connection between the sublethal neurotoxic inhibition of brain cholinesterase by50±60% but did action of some pesticides and various behavioural not cause impairment of memoryin a food caching disturbances in birds and other vertebrates has been experiment (Mineau et al., 1994). demonstrated in a number experimental studies and It is much more difficult to establish behavioural this inevitablyraises questions about the significance effects in the field than in the laboratory(Peakall, or otherwise of such effects in the field. There have 1985). In one study, however (Busby et al., 1990), been manyreports of both lethal and sublethal effects evidence was produced of both behavioural effects, of such compounds from field studies, and it seems and associated effects at the population level after clear that behavioural disturbances must have been forest spraying with fenitrothion. Spraying of spruce caused as well (see, for example, Busbyet al., 1990). forest in New Brunswick, Canada, led to estimated However, behavioural effects are notoriouslydifficult reductions of mean brain cholinesterase activityof to quantifyin the field, and there is a lack of hard white-throated sparrows of 42% and 30%, respec- evidence linking them to population declines (Heinz, tivelyon two consecutive sprayoperations. There was 1989; Peakall, 1996). In the foregoing account a a wide range of individual responses, ranging from number of examples were given of the quantification deaths due to acute poisoning to a varietyof sublethal of behavioural effects in field studies. These and some effects. Reported behavioural effects included inability others are mentioned byPeakall (1996). However, he to defend territories, disruption of incubation patterns also points out that it can be verydifficult to prove that and clutch desertion. The reproductive successs of changes have actuallybeen caused bythe environ- birds in the sprayed area was about 1/4 of that in the mental chemicals with which theyare associated. In control area. In a field studyred-winged blackbirds concept, biomarkers for toxic effect can resolve this (Agelaius phoeniceus) exposed to preycontaminated problem byestablishing (1) that there is a dose± with parathion developed taste aversion to the prey response relationship between the level of an species in question even where uncontaminated bythe environmental chemical and a biomarker response chemical (Nicolaus and Lee, 1999). which provides clear evidence of causalityin the field 314 Walker and (2) that this biomarker response is causally Although the persistent neurotoxic pesticides are related to a quantifiable behavioural change (Walker, now little used, it should be remembered that we know 2001). This approach has alreadybeen successfully more about their environmental fate and long term followed when using brain cholinesterase inhibition effects than we do for most other pollutants and they as a biomarker for OP toxicityin the field, as now serve as valuable MODELS for other compounds discussed earlier in this review. There is, however, with similar properties. Our understanding of their the further question alluded to byPeakall (1996)Ð behaviour gives insight into the ecological risks even where behavioural changes can be related to the presented byother persistent neurotoxic pollutants. action of environmental chemicals, do these cause The marked persistence of some organochlorine population declines? There are two well documented insecticides and organomercurial fungicides, once so cases where the use of biomarker technologyhas troublesome to environmentalists, carried a hidden facilitated the identification of such causal relation- benefit; their residues could be determined in tissues ships. One is eggshell thinning caused byDDE of vertebrates recentlydyingin the field, and used as bringing the decline of the Peregrine falcon and evidence of toxic effect. The magnitude of residues other raptors in North America (Peakall, 1993); the found in the tissues of dead birds provides an index of other is imposex caused bytributyltin in the sublethal as well as lethal effects. A recent retro- dogwhelk (Nucella lapillus) leading to reproductive spective analysis of residue data for dieldrin in failure and population decline in the United Kingdom predatorybirds gave evidence suggesting that a (Gibbs and Bryan, 1986). Progress in this direction considerable proportion of birds found dead in the depends on utilising appropriate mechanistic biomar- field would have experienced sublethal neurotoxic ker assays in the context of properly designedÐ effects (Walker and Newton, 1999). When data from usuallylong termÐpopulation studies. Unfortunately this studywere incorporated into a population model such studies are expensive and time consumingÐ for the sparrowhawk, there was support for the view quick and easyfixes do not provide the answer! that sublethal as well as lethal effects of cyclodienes Understanding the long term ecological significance were implicated in the decline of this species in of behavioural effects of environmental chemicals Britain during the 1950s and early1960s (Siblyet al., will depend on long term commitments to funding of 2000). Sublethal neurotoxic effects would have environmental research programmes byagencies and included behavioural disturbances. It would be governments. interesting to carryout retrospective analysesof It maybe argued that, with the extensive bans that other population studies which relate decline to have been placed on the use of organochlorine cyclodiene residues to see whether they yield similar insecticides and methyl mercury fungicides, there is results. now little reason for concern about possible beha- Returning to less persistent neurotoxic compounds, vioural effects of neurotoxic pesticides. Whilst it is there is seldom residue data that can be used to true that these markedlypersistent neurotoxins have identifybehavioural effects which maybe relevant to presented a more obvious threat to birds than the more population declines. On the other hand, biomarker biodegradable organophosphorus, carbamate, pyre- assays can, in principle, provide a measure for throid or neonicotinoid insecticides, there are other neurotoxic and behavioural effects of non-persistent important considerations. In the first place, even the chemicals in the field (Peakall, 1992). Inhibition of readilybiodegradable compounds still in use may cholinesterase has alreadybeen used as a biomarker of cause transitorybehavioural disturbances which could neurotoxic effects which can be related to consequent be of ecological significance (Mineau, 1991; Grue and behavioural disturbances both in the laboratoryand Shipley, 1981). Also, it should be born in mind that the field (Peakall, 1985, 1992; Busbyet al., 1990; neurotoxic effects of degradable compounds maybe Mineau, 1991). With the widespread use of non- potentiated/prolonged due to interaction with other persistent neurotoxic insecticides, the development pesticides (see Walker et al., 2001, 153±62). Further, of biomarker assays for pyrethroids, carbamates and there is increasing evidence that organophosphorus neonicotinoids could provide the technologyneeded compounds can produce neurotoxic effects which to investigate possible behavioural effects in the field. persist for longer than the active form of the The need is for sensitive, specific, relativelyinexpen- insecticides (see discussion above). sive mechanistic biomarker assays that can be used by Neurotoxicityin Birds 315 non specialists, in the field as well as in the laboratory, Environmental Health Criteria 91 Aldrin and Dieldrin (1989). which can provide evidence of links between WHO Geneva. Fryday, S.L., Hart, A.D. and Marczylo, T.H. (1995). Effects of behavioural effects of neurotoxic pollutants and sublethal exposure to an organophosphate on the flying perfor- population changes in the field (Peakall and Shugart, mance of captive starlings. Bull. Environ. Contam. Toxicol. 55, 1993). 366±73. Gibbs, P.E. and Bryan, G.W. (1986). Reproductive failure of the dog whelk caused byimposex induced byTBT from antifouling Acknowledgements paints. J. Mar. Biologic. Ass. UK 66, 767±77. Grue, C.E. and Shipley, B.K. (1981). Interpreting population esti- The author is grateful to two reviewers for their mates of birds following pesticide applicationsÐbehaviour of valuable comments and suggestions. male starling exposed to an organophosphate pesticide. Stud. Avian Biol. 6, 292±6. Haegele, M.A. and Hudson, R.H. (1977). Reduction of courtship References behaviour induced byDDE in male ringed turtle doves. Wilson Bull. 89, 593±601. Atchison, G.J., Sandheinrich, M.B. and Bryan, M.D. (1996). Harborne, J.R. (1993). Introduction to Ecological Biochemistry, 4th Effects of environmental stressors on interspecific interactions edn. London: Academic Press. of aquatic animals. In M.C. Newman and C.H. Jagoe (eds) Hart, A.D.M. (1993). Relationships between behaviour and the Quantitative Ecotoxicology: A Hierarchical Approach, inhibition of acetylcholinesterase in birds exposed to organo- Chelsea: MI, Lewis. phosphorous pesticides. Environ. Toxicol. Chem. 12, 321±33. Bailey, S., Bunyan, P.J., Jennings, D.M. and Taylor, A. (1970). Heinz, G.H. (1989). How lethal are sublethal effects? Environ. Hazards to wildlife from the use of DDT in orchards. Pestic. Toxicol. Chem. 8, 463±4. Sci. 1, 66±9. Heinz, G.H., Hill, E.F. and Contrera, J.F. (1980). Dopamine and Bailey, S., Bunyan, P.J., Rennison, B.D. and Taylor, A. (1969a). norepinephrine depletion in ring doves fed DDE, dieldrin and The metabolism of DDE and DDMU in the pigeon. Toxicol. Arochlor 1254. Toxicol. Appl. Pharmacol. 53, 75±82. Appl. Pharmacol. 14, 23±32. Jaeger, K. (1970). Aldrin, Dieldrin, Endrin and Telodrin, Bailey, S., Bunyan, P.J., Rennison, B.D. and Taylor, A. (1969b). Amsterdam: Elsevier. The metabolism of DDT and DDD in the pigeon. Toxicol. Appl. Johnson, M.K. (1992). Molecular events in delayed neuropathy: Pharmacol. 14, 13±22. experimental aspects of neuropathytarget esterase. In Beauvais, S.L., Jones, S.B., Brewer, S.K. and Little, E.E. (2000). B. Ballantyne and T.C. Marrs (eds) Clinical and Experimental Physiological measures of neurotoxicity of diazinon and Toxicology of Organophosphates and Carbamates, pp. 90±113. malathion to larval rainbow trout and their correlation with Oxford: Butterworth/Heinemann. behavioural measures. Environ. Toxicol. Chem. 19, 1875±80. Krieger, R.I. (ed.) (2001). Handbook of Pesticide Toxicology Vol. 1 Beitinger, T.L. (1990). Behavioural reactions for the assessment of Principles, pp. 597±602. stress in fish. J. Gt. Lakes Res. 16, 495±528. Lundholm, E. (1987). Thinning of eggshells in birds byDDE: mode Borg, K., Erne, K., Hanko, E. and Wanntorp, H. (1970). Alkyl of action on the eggshell gland. Comp. Biochem. Physiol. 88C, mercurypoisoning in terrestrial Swedish wildlife. Environ. 1±22. Pollut. 1, 91±104. Mahoney, Jr., J.J. (1975). DDT and DDE effects on migratory Brewer, L.W., Driver, C.J., Kendall, R.J., Zenier, C. and condition of White-throated sparrows fed DDT for 5 and 11 Lacher, T.E. (1988). The effects of methyl parathion in ducks weeks. J. Wildl. Manag. 39, 520±7. and duck broods. Environ. Toxicol. Chem. 7, 375±9. Mineau, P. (ed.) (1991) Cholinesterase-inhibiting Insecticides: Buerger, T.T., Kendall, R.J., Mueller, B.S., De Vos, T. and Their Impact on Wildlife. Amsterdam: Elsevier. Williams, B.A. (1991). Effects of methyl parathion on northern Mineau, P., Boag, P.T. and Beninger, R.J. (1994). Effects of bobwhite survivability. Environ. Toxicol. Chem. 10, 527±32. fenitrothion on memoryfor cachesite locations in black-capped Busby, D.G., White, L.M. and Pearce, P.A. (1990). Effects of aerial chickadees. Environ. Toxicol. Chem. 13, 281±90. spraying of fenitrothion on breeding white-throated sparrows. Newton, I. (1979). Population Ecology of Raptors. Berkhamsted: J. Appl. Ecol. 27, 743±55. Poyser. Carson, R. (1962). Silent Spring. New York: Houghton Mifflin. Newton, I. (1986). The Sparrowhawk. Calton: Poyser. Cooke, A.S. (1971). Selective predation bynewts on frog tadpoles Newton, I. and Wyllie, I. (1992). Recovery of a sparrowhawk treated with DDT. Nature 229, 275±6. population in relation to declining pesticide contamination. Eldefrawi, M.E., Eldefrawi, A.T. (1990). NervousÐsystem-based J. Appl. Ecol. 29, 476±84. insecticides. In E. Hodgson and R.J. Kuhr (eds) Safer Insecti- Nicolaus, L.K. and Lee, H.S. (1999). Low acute exposure to cides: Development and Use, pp. 155±208. New York: Marcel organophosphate produces long term changes in bird feeding. Dekker. Ecol. Applic. 9, 1039±49. Environmental Health Criteria No 9 DDT and its Derivatives Odum, E.P. (1971). Fundamentals of Ecology, 3rd edn. (1979). WHO Geneva. Philadelphia: W.B. Saunders. Environmental Health Criteria 85 Mercury: Environmental Aspects Peakall, D.B. (1985). Behavioural responses of birds to pesticides (1989). WHO Geneva. and other contaminants. Residue Rev. 96, 45±77. 316 Walker

Peakall, D.B. (1992). Animal Biomarkers as Pollution Indicator. Stickel, L.F., Stickel, W.H. and Christensen, R. (1996). Residues of London: Chapman. DDT in brains and bodies of birds that died on dosage and in Peakall, D.B. (1996). Disrupted patterns of behaviour in natural survivors. Science N.Y. 151, 1549. populations as an index of toxicity. Environ. Health Persp. Walker, C.H. (2001). Organic Pollutants: An Ecotoxicological 104(Suppl. 2), 331±5. Perspective. London: Taylor and Francis. Ratcliffe, D.A. (1991). The Peregrine Falcon, 2nd edn. Calton: Walker, C.H., Hopkin, S.P., Sibly, R.M. and Peakall, D.B. (2000). Poyser. Principles of Ecotoxicology, 2dn edn. London: Taylor and Richards, P., Johnson, M., Ray, D. and Walker, C.H. (1999). Novel Francis. protein targets for organophosphorous compounds. In E. Reiner, Walker, C.H. and Newton, I. (1999). Effects of cyclodiene insecti- V. Simeon-Rudolf, B. Doctor and C.E. Furlong (eds). Esterases cides on the sparrowhawk in Britain-correction and updating Reacting with Organophosphorous Compounds. Chemico- of an earlier paper byWalker and Newton. Ecotoxicology 7, Biological Interactions 119±120 (Special Issue), 503±12. 185±9. Robinson, J., Brown, V.K.H., Richardson, A., Roberts, M. (1967). Warner, R.E., Petersen, K.E. and Borgman, L. (1966). Behavioural The residues of dieldrin in the tissues of experimentally pathologyin fish: a quantitative studyof sublethal pesticide poisoned birds. Life Sci. 6, 1207±20. toxication. In N.W. Moore (ed.). Journal of Applied Ecology, Sharma, R.P.D.S., Winn, D.S. and Low, J.B. (1976). Toxic, neuro- Vol. 3(Suppl.), pp. 261±70. chemical and behavioural effects of dieldrin exposure in juvenile Wiemeyer, S.N. and Porter, R.D. (1970). DDE thins eggshells of mallard ducks. Arch. Environ. Contam. Toxicol. 5, 43±53. captive American kestrels. Nature 227, 737±8. Sibly, R.M., Newton, I. and Walker, C.H. (2000). Effects of diel- Wolfe, M.F., Schwarzbach, S. and Sulaiman, R.A. (1998). The drin on population growth rates of UK sparrowhawks. J. Appl. effects of mercuryon wildlife: a comprehensive review. Ecol. 37, 540±6. Environ. Toxicol. Chem. 17, 146±60.