Dose–response relationship in lethal and behavioural effects of different insecticides on the paras

Dose–response relationship in lethal and behavioural e?ects of di?erent insecticides on the parasitic wasp Aphidius ervi

N.Desneux *,H.Rafalimanana,L.Kaiser

Laboratoire de Neurobiologie Compar e

e des Invert e br e s,INRA,BP 23,91440Bures-sur-Yvette,France Received 18February 2003;received in revised form 31July 2003;accepted 3September 2003

Abstract

Neurotoxic insecticides are widely used for crop protection and behavioural perturbations can be expected in sur-viving bene?cial insects,including parasitoids of pest insects.The present study aims to investigate the relationship

between the dose of insecticide parasitoids have been exposed to,and the subsequent ability of these parasitoids to respond to host-related cues.A four-armed olfactometer,a design widely used to observe orientation responses in various insects and parasitoids in particular,was chosen to investigate the dose–response relationship.The species studied was Aphidius ervi ,a relatively generalist parasitoid of aphids,and commercialised for biological control and integrated pest management.Active ingredients with similar and di?erent modes of action on the nervous system were compared:a pyrethroid (lambda-cyhalothrin),an organophosphate (chlorpyrifos),a carbamate (pirimicarb)and a carbamyltriazole (triazamate).Adult females were exposed to dry residues on glass for 24h.LD 50were calculated and predicted a high risk of mortality at the ?eld application rate.The e?ect of ?ve increasing residual doses of each active ingredient was tested on responses to plant-host odour in the olfactometer,from sublethal doses to LD 50,and up to LD 70for some products.It appeared that none of the doses of lambda-cyhalothrin,chlorpyriphos and pirimicarb had any e?ect on A.ervi responses to the odour from the aphid-infested plant (Myzus persicae on oilseed rape).But for triazamate,a signi?cant dose–behavioural response was quanti?ed and attraction to the odour was no longer signi?cant in females surviving the LD 50.The possible explanations for the presence or absence of e?ect,depending on the in-secticide are discussed.

ó2003Elsevier Ltd.All rights reserved.

Keywords:Aphid parasitoid;Host location;Olfactometer;Sublethal e?ect;Pesticide

1.Introduction

Many natural enemies of phytophagous insects are Hymenopterous parasitoids.They are characterised by a parasitic larval development causing the death of the host.By reducing the population of their hosts,they can help to limit insect pest damage and,in some instances,prevent the outbreak of pests (Van Driesche and Bel-lows,1996).They are important organisms for biologi-

cal control.Among the biotic factors of pest mortality,insect parasitoids cause the strongest mortality (mor-tality compiled for 78pest species,Hawkins et al.,1997).However,crop protection is mostly based on broad spectrum chemical insecticides that are noxious to ben-e?cial insects (Haskell and McEwen,1998).For in-stance,pest resurgence or increase in populations of secondary pests can occur as a result of death or per-turbation of bene?cial arthropods by pesticides (Hardin et al.,1995;Longley et al.,1997).

Parasitoids can be exposed to pesticides through di-rect exposure to spray droplets (Jepson,1989),or to residues on the crop foliage when foraging for hosts (Jepson,1989;Longley and Jepson,1996a,b)or through

*

Corresponding author.Tel.:+33-1-6929-8767;fax:+33-1-6907-5054.

E-mail address:desneux@jouy.inra.fr (N.Desneux).

0045-6535/$-see front matter ó2003Elsevier Ltd.All rights reserved.

doi:10.1016/j.chemosphere.2003.09.007

Chemosphere 54(2004)

619–627

https://www.360docs.net/doc/e217294925.html,/locate/chemosphere

dietary exposure through feeding on contaminated water droplets,nectar,or honeydews(Longley and Stark,1996).Indirect exposure during the development in the host can also occur(S€u ss,1983;Hsiech and Allen, 1986;Longley,1999).

Exposure to low doses of insecticide residues is highly probable due to the widespread use of pesticides (Brown,1989)and could induce sublethal e?ects in contaminated surviving insects.Sublethal e?ects are particularly expected on the behaviour of insects ex-posed to neurotoxic insecticides(Haynes,1998),i.e.the majority of insecticides.Parasitoids spend a signi?cant proportion of their adult life searching for hosts.Host searching behaviour involves orientation to host and host-plant odours(Vinson,1998).From the detection of odours to associated behaviours,olfaction is entirely dependent on nervous transmissions,which are targeted by neurotoxic insecticides through di?erent modes of action.The most common targets are(1)the voltage-sensitive sodium channel on neuron membranes,whose openings are prolonged by pyrethroids(Soderlund and Bloomquist,1989);(2)the acetylcholinesterase,which is inhibited by organophosphorous insecticides,carba-mates(Fukuto,1979)and carbamyltriazoles(Padilla, 1995);(3)the Gaba receptor,which is inhibited by organochlorous insecticides,and(4)the nicotinic acet-ylcholine receptors inhibited by the family of the neon-icotinoids(Masuda et al.,2001)or activated by the family of spinosyns(Salgado,1997).Therefore,insecti-cides may interfere with olfaction,depending on the mode of action and level of exposure.

Few data are available on sublethal e?ects of pesti-cides on the orientation behaviour of insects.For ex-ample in honeybees,sublethal doses of insecticides have been shown to disturb the homing-?ight(Vandame et al.,1995)and odour learning(Abramson et al.,1999; Decourtye and Pham-Del e gue,2002).Orientation to sex pheromone is disturbed by permethrin in pink bollworm moth,Pectinophora gossypiella(Haynes and Baker, 1985).Komeza et al.(2001)demonstrated that the behavioural response of the parasitoid,Leptopilina boulardi to host chemicals could be disturbed by a low dose of chlorpyrifos.This insecticide has also been shown to alter sex pheromone communication in Trichogramma (Delpuech et al.,1998).An increased response to host-habitat odour after exposure to chlorpyrifos was also described by Rafalimanana et al.(2002).Deltamethrin at a sublethal dose increased the arrestment behaviour of treated Trichogramma males responding to female pheromone(Delpuech et al.,1999).Trissolcus basalis females exposed to a low dose of this insecticide reduced their walking speed and the time spent on host-patches (Salerno et al.,2002).

The dose–response relationship has not been inves-tigated in these previous studies,but it is valuable to be able to predict the impact of a known pesticide residue on the foraging behaviour.The aim of the present work was thus to investigate the potential relationship be-tween the dose of insecticide to which a parasitic wasp is exposed and its subsequent ability to respond to host-related cues.The species studied was Aphidius ervi Haliday(Hymenoptera:Braconidae),a parasitoid com-mercially used to control several aphid species in greenhouses.Exposure to increasing doses of insecticide was used to establish the dose–mortality relationship and to produce surviving insects with di?erent levels of contamination.These insects were observed in a four-armed olfactometer delivering the host-plant odour. This device is commonly used to observe olfactory re-sponses in various insect species(e.g.,aphids:Petters-son,1970;parasitoids:Vet et al.,1983;De Jong and Kaiser,1991;honeybees:Sandoz et al.,2000).The in-secticides belonged to the most common families used for crop protection,and di?ered in their mode of action on the nervous system of insects:a pyrethroid,lambda-cyhalothrin;a carbamyltriazol,triazamate;a carbamate, pirimicarb;and an organophosphorous,chlorpyrifos.

2.Material and methods

2.1.Insects

A.ervi adults were provided by Biobest(Belgium). Prior to testing,in order to allow mating,parasitoids were introduced in a plexiglas box(length:18cm; weight:12cm;height:7cm)provided with honey solu-tion(80%)and placed in an environmental chamber at 15±1°C,65±5%relative humidity,and light regime 12L:12D.

2.2.Insecticide solutions

Active ingredients were used instead of commercial formulations,because the aim was to document e?ects of the neurotoxic molecules on the behaviour,without in-?uence of adjuvant added in commercial products.They were provided by Cluzeau InfoLabo(Sainte-Foy-la-Grande,France):lambda-cyhalothrin(certi?ed purity 89.3%),triazamate(certi?ed purity97.0%),pirimicarb (certi?ed purity99%)and chlorpyrifos(certi?ed purity 99.1%).Lambda-cyhalothrin was in the form of a gel, pirimicarb and chlopyrifos-ethyl in the form of crystals and triazamate in the form of a solution in cyclohexane.

A preliminary experiment was run to determine the range of insecticide doses,by exposing insects to doses equal to recommended?eld application rates(ACTA, 2002)until observing mortality rates lower than100%. To establish the dose–mortality relationship,insects were exposed to seven doses increasing by a2factor for lambda-cyhalothrin and triazamate,and to?ve doses increasing by a factor1.25for chlorpyrifos and pirimi-

620N.Desneux et al./Chemosphere54(2004)619–627

carb.The?rst?ve doses of each insecticide were then used to evaluate their impact on the orientation behav-iour(Table1).Doses were expressed as ng of active ingredient per cm2because we worked on the e?ects of residues.

2.3.Exposure of parasitoids to insecticide residues

Insecticide solutions(with acetone or cyclohexane) were applied in glass tubes(length:9.3cm;diameter:2.3 cm;internal surface:67.4cm2).Pure acetone was used as control for lambda-cyhalothrin,pirimicarb and chlor-pyrifos and cyclohexane for triazamate because it was already diluted in cyclohexane in commercial form.To get a homogeneous deposit,we introduced200l l of solution,which allowed a total coverage of the internal surface of the tube,and the tube was then manually rotated until no more droplets were seen on the glass wall.Tubes were left for1h at room temperature to ensure complete evaporation of the acetone(or cyclo-hexane)before introducing parasitoids.Both internal surface of the tubes and volume of solution being?xed, it was then possible to express the quantity of insecticide residue per unit of surface.At the time of exposure, A.ervi adults were four day old(Biobest,Personal com-munication).Ten females were placed per tube,with two drops of honey on a small plastic strip.Tubes were closed with a?ne nylon gauze to allow air circulation. Exposure was performed at15±1°C,65±5%relative humidity,and under a12L:12D photoperiod.The tem-perature was set at15°C to avoid mortality in the control tubes,so it was always lower than the10% mortality recommended to evaluate insecticides toxicity (Hassan,1998).We checked that parasitoids were active in the tubes.Four replicates of at least30wasps(3tubes of10females)were carried out for each dose of each insecticide.After24h of exposure,the number of dead parasitoids was counted to determine the regression lines of mortality and thus the LD50.The survivors were collected and placed individually in Petri dishes,until being observed in the olfactometer(within the2h fol-lowing the end of exposure).2.4.Behavioural test in the four-armed olfactometer

Oriented responses towards aphid-infested plant odour were observed in a four-armed olfactometer(Fig.

1)(from Vet et al.,1983).Pressurised and humidi?ed air ?ew into the central chamber through four arms(200ml/ min per arm)and was extracted from the centre of the chamber,so that four?elds of equal area were estab-lished.The device was placed on a light-table providing a homogeneous?uorescent light(800lx),in a room at25°C(temperature ensuring both active locomotion and high odour release rate),and70%RH.In our experi-ments,only one?eld was odorised.The odour source was constituted by oilseed rape stems(kept in water) (Brassica napus var Goeland)with a total of seven to eight leaves infested by Myzus persicae Sulzer(Ho-moptera:Aphididae)(400–500aphids after seven days of infestation).To deliver the odour into one of the four ?elds of the olfactometer,the corresponding arm was connected to an air-tight glass jar(height:25cm;di-ameter:11cm),containing the odour source.

Table1

Range of insecticide doses used to determine the LD50and used subsequently in the olfactometer tests

Active ingredients

Lambda-cyhalothrin Triazamate Pirimicarb Chlorpyrifos

Doses used to determine LD50 (ng/cm2)

0.29? 1.05?

0.58? 2.34?25.53?0.22?

1.17? 4.68?31.92?0.28?

2.34?9.37?39.90?0.34?4.69?18.75?49.87?0.43?9.3737.5062.34?0.54?18.7575.00

?Indicates doses used for the olfactometer tests.

N.Desneux et al./Chemosphere54(2004)619–627621

A female parasitoid was introduced into a vial con-nected to the centre of the four-armed olfactometer,of which it could walk out freely.Observations started when the female entered the chamber,and lasted for1 min,which was su?cient to observe a signi?cant at-traction to the odour(Desneux et al.,2000).The posi-tion of the female(?elds numbered clockwise from one to four)was recorded on a computer using an event recorder software‘‘the Observer’’(Noldus Information Technology,Wageningen,The Netherlands),allowing to compute the overall time spent in each?eld.After approximately every20individual observations,the ol-factometer was carefully washed with ethanol and the position of the odorised?eld was changed.

E?ect of the four pesticides was examined on the orientation responses of surviving females.On each day of experiment,all doses were tested alternately.Sample sizes are reported on the graphs.

2.5.Data analysis

Determination of LD50:Dose–mortality relationship was determined using the computer program WIN DL (CIRAD-CA/MABIS,Montpellier,France),based on probit analysis(Finney,1971).

To evaluate the impact of insecticide exposure on the orientation behaviour,two statistical analyses were car-ried out.First,to investigate the existence of a dose–response relationship,we carried out a logistic regression of the percentage of time spent in the odour as a function of the pesticide dose(S-Plus,Venables and Ripley,1999). In this regression,the deviation of the observed data is calculated relatively to a linear model under the as-sumption that there is no dose e?ect.Second,to detect any e?ect of a given dose,the Kolmogorov–Smirnov statistic was used to compare the time spent in the odorised?eld between the control group and each trea-ted group(S-Plus,Venables and Ripley,1999).

3.Results

3.1.Dose–mortality relationship(Table2)

Linear regression of the dose–mortality relationship was?tted to the observed data for all tested active in-gredients,as indicated by the absence of signi?cant de-viation between the observed and the expected data. Then LD50was considered as valid.Mortality in all control groups was always inferior to5%.The slope of the regression line indicates how fast mortality occurs with increasing doses.The highest value was observed for pirimicarb and the lowest for triazamate.All LD50were signi?cantly di?erent(no overlapping of the con?dence intervals)and allowed a ranking of the insecticides in order of increasing toxicity:pirimicarb,triazamate, lambda-cyhalothrin and chlorpyrifos.The recommended ?eld rate of these active ingredients was always higher than the LD50,from15to about8000times(ratio be-tween recommended?eld rate and LD50,Table2).

For the olfactometer tests,the doses used induced 0.9±2.6%to71.6±15.2%of corrected mortality(Ab-bott,1925)for lambda-cyhalothrin,5.7±2.5%to53.4±11.2%of corrected mortality for triazamate,11.0±6.1% to76.0±7.0%of corrected mortality for pirimicarb and 8.7±5.2%to63.3±8.8%of corrected mortality for chlorpyrifos(Fig.2A–D).

Table2

Dose–mortality relationship.LD50(with inferior and superior limits of the95%con?dence interval),statistical results of adjustment to the log-probit model,rate recommended in the?eld on oilseed rape against M.persicae and ratio between recommended?eld rate and LD50,for four pesticide residues applied in glass tubes for exposure of A.ervi adults(for chlorpyrifos,the rates recommended against Ostrinia nubilabis on maize were used to calculate ratio because this pesticide is not used in France on oilseed rape)

Active ingredients Inf lim

(ng/cm2)Equations of mortality

linear regression and

statistics

Recommended?eld

rates(g a.i./ha)

Ratio:recommended

?eld rates/LD50

Lambda-cyhalothrin4:48<4:97<5:52Y?à2:30t3:37X7.515.09

v2?7:63;P?0:267

(df6)

Triazamate13:46<17:16<20:94Y?à3:03t2:45X7040.79

v2?5:05;P?0:410

(df6)

Pirimicarb33:50<36:77<39:30Y?à9:27t5:92X25067.99

v2?5:92;P?0:115

(df4)

Chlorpyrifos0:43<0:47<0:51Y?1:44t4:41X3757978.72

v2?5:53;P?0:137

(df4)

622N.Desneux et al./Chemosphere54(2004)619–627

3.2.Dose–behaviour relationship

All control groups exhibited a signi?cant attraction towards the aphid-infested plant odour(Friedman analysis of time allocation to the four?elds:P<0:05).

The relative time spent in the odorised?eld by fe-males exposed to the di?erent doses of lambda-cyhal-othrin(Fig.2A)was not signi?cantly di?erent from that of control females(Kolmogorov–Smirnov:dose0.29ng/ cm2:Z?0:22,P?0:188;dose0.59ng/cm2:Z?0:14,

N.Desneux et al./Chemosphere54(2004)619–627623

P?0:697;dose1.17ng/cm2:Z?0:10,P?0:923;dose 2.34ng/cm2:Z?0:25,P?0:108;dose 4.69ng/cm2: Z?0:18,P?0:402).The logistic regression of the percentage of time spent in the odour as a function of the dose was not signi?cant(v2?1:29;df1;P?0:255) indicating the absence of signi?cant dose e?ect.

The relative time spent in the odorised?eld by fe-males exposed to the?rst four doses of triazamate(Fig. 2B)was not signi?cantly di?erent from that of control females(dose1.17ng/cm2:Z?0:17,P?0:579;2.34ng/ cm2:Z?0:26,P?0:134; 4.68ng/cm2:Z?0:28, P?0:091;9.37ng/cm2:Z?0:26,P?0:116).Never-theless,the relative time spent in the odorised?eld by females exposed to the highest dose of triazamate,18.75 ng/cm2,was signi?cantly di?erent from that of control females(Z?0:43,P?0:003).Additionally,we found a signi?cant logistic regression of the percentage of time spent in the odour as a function of the dose(v2?6:16; df1;P?0:013).

The relative time spent in the odorised?eld by fe-males exposed to di?erent doses of pirimicarb(Fig.2C) was not signi?cantly di?erent from that of control fe-males(dose25.53ng/cm2:Z?0:11,P?0:922;31.92 ng/cm2:Z?0:20,P?0:319;39.90ng/cm2:Z?0:12, P?0:830;49.87ng/cm2:Z?0:15,P?0:640;62.34ng/ cm2:Z?0:11,P?0:922).The logistic regression of the percentage of time spent in the odour as a function of the pesticide dose was not signi?cant(v2?0:35;df1; P?0:554).

The relative time spent in the odorised?eld by fe-males exposed to di?erent doses of chlorpyrifos(Fig. 2D)was not signi?cantly di?erent from that of con-trol females(0.22ng/cm2:Z?0:12,P?0:938;0.28 ng/cm2:Z?0:16,P?0:625;0.34ng/cm2:Z?0:22, P?0:171;0.43ng/cm2:Z?0:16,P?0:632;0.54ng/ cm2:Z?0:10,P?0:972).The logistic regression of the percentage of time spent in the odour as a function of the pesticide dose was not signi?cant(v2?0:21;df1; P?0:649).

4.Discussion

4.1.Dose–mortality relationship

For each active ingredient tested,there was a good?t between the observed dose–mortality relationship and the linear regression,so our experimental conditions gave reliable estimations of LD50.Values allowed to rank the insecticides in order of increasing toxicity: pirimicarb,triazamate,lambda-cyhalothrin and chlor-pyrifos,the latter being about100times more toxic than pirimicarb.However,it is not possible to use the LD50to compare the sensibility of A.ervi to other parasitoid species,because almost no data are available in the lit-erature.Indeed,the toxicity of insecticides to bene?cial arthropods has always been estimated using mortality induced by the?eld application rate.The estimation of LD50was only recently recommended for parasitoids (Candol?et al.,2001).Our data indicated that recom-mended?eld rates were always higher than LD50for the four studied active ingredients.When testing commer-cial products,the ratio between the?eld application rate and the LD50,i.e.Hazard Quotient(HQ),gives an indication of the risk.This quotient is already used for evaluating the toxicity of pesticides to honeybees (EPPO,1999).Although our LD50were calculated for active ingredients,the ratio to the recommended?eld rate does allow to compare the risk among the four tested https://www.360docs.net/doc/e217294925.html,ing the?eld recommended rates against aphids on oilseed rape(and against the Euro-pean Corn Borer on maize for chlorpyrifos),ratios were equal to15for lambda-cyhalothrin(which means that1/ 15of the?eld application rate will kill50%of the A.ervi population),41for triazamate,68for pirimicarb and 7979for chlorpyrifos,presenting the highest risk to A. ervi.Both toxicity and risk of chlopyrifos-ethyl to A.ervi were the highest compared to the other three insecti-cides.But for these three insecticides,comparison of LD50on the one hand,and of the ratio on the other hand,gave inverse classi?cations.Ranking the ratios pointed out that lambda-cyhalothrin presented the lowest risk to A.ervi.

4.2.Dose–behaviour relationship

For three of the four active ingredients tested:lamb-da-cyhalothrin,pirimicarb,chlopyrifos-ethyl,the orien-tation behaviour in the olfactometer was not changed in insects surviving to any doses of insecticides.Survival was not due to reduced exposure to residues,because the regular increase of mortality with increasing doses indi-cates a homogeneous exposure of the insects.Unaltered orientation behaviour in contaminated females can have two explanations:either the insecticide molecules did not alter the functions necessary for olfactory responses in the olfactometer,or surviving insects were less suscepti-ble.As regards the?rst explanation,considering the cellular and molecular targets of pyrethroids,organ-ophosphorous and carbamate,e?ects on olfactory re-sponses are expected.This has been veri?ed in a number of studies on pyrethroids and organophosphorous:dif-ferent pyrethroids modi?ed olfactory learning in the honeybee(Taylor et al.,1987;Mamood and Waller, 1990;Abramson et al.,1999),and responses to sex pheromone in the parasitoid Trichogramma(Delpuech et al.,1999)and in a moth(Floyd and Crowder,1981; Haynes and Baker,1985).Chlorpyriphos-ethyl modi?ed responses of parasitoids to host and host-habitat odours (Komeza et al.,2001;Rafalimanana et al.,2002)and to sex pheromone(Delpuech et al.,1998).

624N.Desneux et al./Chemosphere54(2004)619–627

With regard to the second explanation,it implies a heterogeneity in susceptibility within the A.ervi popu-lation to insecticides.According to Croft(1990),toler-ance to pesticides may occur as a direct result of the organism’s short-and long-term exposure to toxins in nature.Variable level of tolerance can re?ect di?erences in vigour or in genetically determined susceptibility to pesticides.The latter was reported in some other Hy-menopteran parasitoid species(Hoy,1990).Both sour-ces of variability are likely in A.ervi.Vigour can vary according to characteristics that were not controlled precisely in our experiment,like size,stress and general physiological state.Genetic variability of sensitivity to insecticides may have resulted from the yearly incorpo-ration of?eld-collected females to the mass-reared strain (Biobest,Personal communication).Indeed exposure to various pesticides residues has been occurring for years under?eld conditions and may select less susceptible A. ervi.In addition,the results obtained with A.ervi in a previous study supported the hypothesis of heteroge-neous sensitivity to insecticides in A.ervi species.Indeed, we observed that the orientation behaviour of A.ervi was disturbed by a sublethal dose(LD0:1)of lambda-cyhalothrin(Desneux et al.,2000),which was not found again in the present study.These contrasting results may be due to di?erences in insecticide sensitivity between the strains of A.ervi,which did not come from the same company.

As third explanation,the experimental conditions of the test in the olfactometer may not be optimal to observe sublethal behavioural e?ects.This study was the?rst at-tempt to use a four-armed olfactometer to characterise the e?ects of neurotoxic insecticides on olfactory re-sponses of insects.The in?uence of some experimental parameters remains to be tested.For instance,a longer observation time might have revealed e?ects on odour habituation,as already shown in L.boulardi responding to patches of host odours after exposure to chlorpyrifos (Komeza et al.,2001).E?ects could concern other behavioural items than the time spent in the odorised ?ow,such as the walking speed,which decreased in the parasitoid T.basalis exposed to the pyrethroid delta-methrin(Salerno et al.,2002).We can also hypothesise that the nervous functions used by the insect to walk into the odour?ow are relatively limited and were not a?ected by the insecticides tested,whereas the same insects ex-posed to the same doses of insecticide might not be able to perform in-?ight orientation to an odour source,a task which requires di?erent muscular and nervous input,and the ability to detect a lower odour concentration diluted into the insect’s environment.For instance,Haynes and Baker(1985)have demonstrated that a pyrethroid,per-methrin,decreased the attractiveness of sex pheromone in a moth tested in a wind tunnel,but this sublethal e?ect did not occur when moths were close to the odour source,i.e. when placed in a higher odour concentration.Therefore,although exposure to lambda-cyhalothrin,chlorpyrifos and pirimicarb did not have any e?ect on A.ervi orien-tation in the four-armed olfactometer,it might impair in-?ight orientation,or other types of behaviour involved in the host location and recognition(attack behaviour, etc...).Indeed,we observed an irreversible sublethal ef-fect of chlorpyrifos within a few hours after the start of the exposure:some females were bending permanently their abdomen forward(as if attacking aphids),although they exhibited a normal orientation behaviour in the olfactometer.

In contrast to lambda-cyhalothrin,chlorpyrifos and pirimicarb,A.ervi exposed to triazamate showed a re-duction of the time spent in the odour?ow proportional to the dose of residues,and the attraction to the odour was no longer signi?cant in insects exposed to the highest dose.The progressive e?ect of triazamate on the orientation behaviour is in accordance with the fact that mortality also occurred more progressively with in-creasing doses than with the other three insecticides tested.This suggests that triazamate would have a mode of action somehow di?erent from that of the other two inhibitors of acetylcholine esterase.Indeed,the action on a given target can di?er among active ingredients because of the various chemical functions which cha-racterise insecticide molecules(Fukuto,1979).More-over,the insecticides often have secondary targets of actions(Soderlund and Bloomquist,1989).Alterna-tively,a lack of tolerance to triazamate could explain that the orientation behaviour of A.ervi was more sen-sitive to triazamate than to the other three insecticides tested.Indeed,triazamate is considered as relatively re-cent,because it has only been commercialised in1991 (Delorme et al.,2002).So it is probable that there is not yet any tolerance of parasitoids to this insecticide.In general,when insecticides have been used for a longer time(older),tolerance is more likely to have appeared. In M.persicae,known to rapidly develop resistance to insecticides,the?rst case of resistant strain to triazamate was only recently reported(Foster et al.,2002),whereas resistant strains to organophosphorous,carbamates and pyrethroids appeared much before(Devonshire and Moore,1982).

In conclusion,as regards the behavioural e?ects of insecticide residues,this study was the?rst attempt to establish a dose–behaviour relationship,using the four-armed olfactometer,a device frequently used to observe olfactory orientation in various insects.It pointed out that among four active ingredients,only one induced a dose–response relationship whereas three had no e?ect on the orientation behaviour of surviving parasitoids. Further investigations are necessary to conclude on the possible reasons:observation parameters in the olfac-tometer,weak sensitivity of the responses developed in the olfactometer,or tolerance to older insecticides in surviving A.ervi.

N.Desneux et al./Chemosphere54(2004)619–627625

Acknowledgements

We wish to thank Minh-H a Pham-Del e gue for her scienti?c advices,Aude Couty for general discussions on the subject,Alfred Martin-Canadel for providing bio-logical material and Annick Lacombe for linguistic corrections.This work bene?ted from a PNETOX grant from the French Ministry of the Environment and from a CETIOM/Aventis Crop Sciences grant.

References

Abbott,W.S.,1925.A method for computing the e?ectiveness of an insecticide.J.Econ.Entomol.18,265–267. Abramson,C.I.,Aquino,I.S.,Ramalho,F.S.,Price,J.M.,1999.

The e?ect of insecticides on learning in the Africanized honey bee(Apis mellifera L.).Arch.Environ.Contam.

Toxicol.37,529–535.

ACTA,2002.Index Phytosanitaire,37th ed.ACTA,Paris. Brown,R.A.,1989.Pesticides and non-target terrestrial inver-tebrates:an industrial approach.In:Jepson,P.C.(Ed.), Pesticides and Non-target Invertebrates.Intercept,Wim-borne,pp.19–42.

Candol?,M.P.,Barrett,K.L.,Campbell,P.,Forster,R., Grandy,N.,Huet,M.C.,Lewis,G.,Oomen,P.A.,Sch-muck,R.,Vogt,H.,2001.Guidance document on regula-tory testing and risk assessment procedures for plant protection products with non-target arthropods.In:SE-TAC/ESCORT2Workshop report,21–23March2000, Wageningen.

Croft,B.A.,1990.Arthropod Biological Control Agents and Pesticides.Wiley and Sons,New York.

De Jong,R.,Kaiser,L.,1991.Odor learning by Leptopilina boulardi,a specialist parasitoid(Hymenoptera:Eucoilidae).

J.Insect.Behav.4(6),743–750.

Decourtye, A.,Pham-Del e gue,M.H.,2002.The proboscis extension response as a laboratory method to assess the sub-lethal e?ects of pesticides on the honey bee.In:Devillers,J., Pham-Del e gue,M.H.(Eds.),Estimating the Environmental Impact of Chemicals to Honey Bees.Taylor and Francis, London,pp.67–84.

Delorme,R.,Leroux,P.,Gaillardon,P.,2002.Evolution des produits phytosanitaires a usages agricoles.III––Les insec-ticides-acaricides.Phytoma548,7–13.

Delpuech,J.M.,Gareau,E.,Terrier,O.,Fouillet,P.,1998.

Sublethal e?ects of the insecticide chlorpyrifos on sex pheromonal communication of Trichogramma brassicae.

Chemosphere36(8),1775–1785.

Delpuech,J.M.,Legallet,B.,Terrier,O.,Fouillet,P.,1999.

Modi?cation of the sex pheromonal communication of Trichogramma brassicae by sub-lethal dose of deltamethrin.

Chemosphere38,729–739.

Desneux,N.,Noel,B.,Kaiser,L.,2000.Sublethal e?ect of a pyrethroid on orientation behaviour of the parasitic wasp Aphidius ervi(Hymenoptera:Aphididae)in response to odour from oilseed rape infested by the aphid Myzus persicae.Bull.IOBC/WPRS23(9),55–64.

Devonshire,A.L.,Moore,G.D.,1982.A carboxylesterase with broad substrate speci?city causes organophosphorus,car-

bamate andpyrethroid resistance in peach–potato aphids (Myzus persicae).Pestic.Biochem.Physiol.18,235–246. EPPO,1999.Decision-making scheme for the environmental risk assessment of plant protection products.Bull.OEPP/ EPPO23,59–65.

Finney, D.J.,1971.Probit Analysis.Cambridge University Press,Cambridge.

Floyd,J.P.,Crowder,L.A.,1981.Sublethal e?ects of permeth-rin on pheromone response and mating of male pink bollworm moths.J.Econ.Entomol.74(5),634–638. Foster,S.P.,Denholm,I.,Devonshire, A.L.,2002.Field-simulator studies of insecticide resistance to dimethylcarba-mates and pyrethroids conferred by metabolic-and target site-based mechanisms in peach–potato aphids,Myzus persicae(Hemiptera:Aphididae).Pest Manag.Sci.58, 811–816.

Fukuto,T.R.,1979.E?ect of structure on the interaction of organophosphorus and carbamate esters with acetylcho-linesterase.In:Narahashi,T.(Ed.),Neurotoxicology of Insecticides and Pheromones.Plenum Press,New York and London,pp.277–295.

Hardin,M.R.,Benrey,B.,Coll,M.,Lamp,W.O.,Roderick,

G.K.,Barbosa,P.,1995.Arthropod pest resurgence:an

overview of potential mechanisms.Crop Prot.14,3–18. Haskell,P.T.,McEwen,P.,1998.Ecotoxicology––Pesticides and bene?cial organisms.Kluwer Academic Publishers, London.

Hassan,S.,1998.Guideline for the evaluation of side e?ects of plant protection products on Trichogramma cacoeciae.Bull.

IOBC/WPRS21(6),118–128.

Hawkins,B.A.,Cornell,H.V.,Hochberg,M.E.,1997.Preda-tors,parasitoids and pathogens as mortality agents in phytophagous insect populations.Ecologist78,2145–2152. Haynes,H.F.,1998.Sublethal e?ects of neurotoxic insecticides on insect behavior.Annu.Rev.Entomol.33,149–168. Haynes,K.F.,Baker,T.C.,1985.Sublethal e?ects of permeth-rin on the chemical communication system of the pink bollworm moth,Pectinophora gossypiella.Arch.Insect Biochem.Physiol.2,283–293.

Hoy,M.A.,1990.Pesticide resistance in arthropod natural enemies variability and selection responses.In:Roush,R.T., Tabashnik, B.E.(Eds.),Pesticide Resistance in Arthro-pods.Kluwer Academic Publishers,Netherlands,pp.203–236.

Hsiech, C.Y.,Allen,W.W.,1986.E?ects of insecticides on emergence,survival,longevity,and fecundity of the parasi-toid Diaeretiella rapae(Hymenoptera:Aphididae)from mummi?ed Myzus persicae(Homoptera:Aphididae).J.

Econ.Entomol.79,1599–1602.

Jepson,P.C.,1989.The temporal and spatial dynamics pesti-cide side-e?ects on non-target invertebrates.In:Jepson,P.C.

(Ed.),Pesticides and Non-target Invertebrates.Intercept, Wimborne,pp.95–127.

Komeza,N.,Fouillet,P.,Boul e treau,M.,Delpuech,J.M., 2001.Modi?cation,by insecticide chlorpyrifos,of the behavioral responses to kairomones of a parasitoid wasp, Leptopilina boulardi.Arch.Environ.Contam.Toxicol.41, 436–442.

Longley,M.A.,1999.A review of pesticide e?ects upon immature aphid parasitoids within mummi?ed hosts.Int.

J.Pest Manage.45(2),139–145.

626N.Desneux et al./Chemosphere54(2004)619–627

Longley,M.,Jepson,P.C.,1996a.E?ects of honeydew and insecticide residues on the distribution of foraging aphid parasitoids under glasshouse and?eld conditions.Entomol.

Exp.Appl.81,189–198.

Longley,M.,Jepson,P.C.,1996b.The in?uence of insecticide residues on primary parasitoid and hyperparasitoid foraging behaviour in the laboratory.Entomol.Exp.Appl.81,259–269.

Longley,M.,Stark,1996.Analytical techniques for quantifying direct,residual,and oral exposure of an insect parasitoid to an organophosphate insecticide.Bull.Environ.Contam.

Toxicol.57(5),683–690.

Longley,M.,Jepson,P.C.,Izquierdo,J.,Sotherton,N.,1997.

Temporal and spatial changes in aphid and parasitoid populations following applications of deltamethrin in winter wheat.Entomol.Exp.Appl.83,41–52.

Mamood, A.N.,Waller,G.D.,1990.Recovery of learning responses by honeybees following a sublethal exposure to permethrin.Physiol.Entomol.15,55–60.

Masuda,K.,Ihara,M.,Nishimura,K.,Sattelle,D.B.,Komai, K.,2001.Insecticidal and neural activities of candidate photoa?nity probes for neonicotinoid binding sites.Biosci.

Biotechnol.Biochem.65(7),1534–1541.

Padilla,S.,1995.The neurotoxicity of cholinesterase inhibiting insecticides:past and present evidence demonstrating pers-itent e?ects.Inhal.Toxicol.7,903–907.

Pettersson,J.,1970.An aphid sex attractant.Ent.Scand.1, 63–73.

Rafalimanana,H.,Kaiser,L.,Delpuech,J.M.,2002.Stimulat-ing e?ects of the insecticide chlorpyrifos on host searching and infestation e?cacy of a parasitoid wasp.Pest Manag.

Sci.58,321–328.

Salerno,G.,Colazza,S.,Conti,E.,2002.Sub-lethal e?ects of deltamethrin on walking behaviour and response to host

kairomone of egg parasitoid Trissolcus basalis.Pest Manag.

Sci.58,663–668.

Salgado,V.L.,1997.The mode of action of spinosad and other insect control products.Down Earth52(1),35–44. Sandoz,J.C.,Laloi,D.,Odoux,J.F.,Pham-Del e gue,M.H., 2000.Olfactory information transfer in the honeybee: compared e?ciency of classical conditioning and early exposure.Anim.Behav.59,1025–1034.

Soderlund,D.M.,Bloomquist,J.R.,1989.Neurotoxic actions of pyrethroid insecticides.Annu.Rev.Entomol.34,77–96. S€u ss,L.,1983.Survival of pupal stage of Aphidius ervi Hal.in mummi?ed Sitobion avenae F.to pesticide treatment.In: Cavalloro,R.(Ed.),Aphid Antagonists.Balkema,Rotter-dam,pp.129–134.

Taylor,K.S.,Waller,G.D.,Crowder,L.A.,1987.Impairment of a classical conditioned response of the honey bee(Apis mellifera L.)by sublethal doses of synthetic pyrethroid insecticides.Apidologie18(3),243–252.

Van Driesche,R.G.,Bellows,T.S.,1996.Biological Control.

Chapman and Hall,New York.

Vandame,R.,Meled,M.,Colin,M.E.,Belzunces,L.P.,1995.

Alteration of the homing-?ight in the honey bee Apis mellifera L.exposed to sublethal dose of deltametrhin.

Environ.Toxicol.Chem.14(5),855–860.

Venables,W.N.,Ripley,B.D.,1999.Modern Applied Statistics with S-Plus.Springer-Verlag,New York.

Vet,L.E.M.,Van Lenteren,J.C.,Heymans,M.,Meelis,E., 1983.An air?ow olfactometer for measuring olfactory responses of hymenopterous parasitoids and other small insects.Physiol.Entomol.8,97–106.

Vinson,S.B.,1998.The general host selection behavior of parasitoid Hymenoptera and comparison of initial strategies utilized by larvaphagous and oophagous species.Biol.

Control11,79–96.

N.Desneux et al./Chemosphere54(2004)619–627627