Science of the Total Environment 447 (2013) 274–279

Contents lists available at SciVerse ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv

The sublethal effects of deltamethrin on Trichogramma behaviors during theexploitation of host patches

Jean-Marie Delpuech ⁎, Maxime DelahayeUniversité de Lyon, CNRS, Université Claude Bernard Lyon 1, UMR 5558, Laboratoire de Biométrie et Biologie Evolutive, 43 Boulevard du 11 Novembre 1918, F-69622 Villeurbanne Cedex, France

H I G H L I G H T S

► Trichogramma species are parasitoids of numerous species of Lepidoptera.► Deltamethrin decreased the level of infestation of healthy host eggs.► Deltamethrin increased the level of infestation of unsuitable, previously infested hosts.► Deltamethrin increased antennal and ovipositor rejection of previously infested hosts.► These effects may impact the equilibrium of natural ecosystems.

⁎ Corresponding author at: Lab Biometrie Biologie EBoulevard du 11 Novembre 1918, 69622 Villeurbanne43 29 12; fax: +33 4 72 43 13 88.

E-mail address: jean-marie.delpuech@univ-lyon1.fr

0048-9697/$ – see front matter © 2013 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.scitotenv.2012.12.096

a b s t r a c t

a r t i c l e i n f o

Article history:Received 9 November 2012Received in revised form 18 December 2012Accepted 18 December 2012Available online 5 February 2013

Keywords:PyrethroidTrichogramma brassicaeInsecticideInfestationParasitoid

Trichogramma and parasitoids as a whole are key species because they regulate natural populations of otherinsects. As any non-target species, this parasitoid can be exposed to insecticides by environmental pollution.This study identified the effects of an LD 20 of deltamethrin (a pyrethroid) on the behavior of Trichogrammabrassicae females infesting a patch of host eggs. The study found that females that survived exposure to theinsecticide infested fewer host eggs; spent more time on unsuitable, previously infested host eggs; andinfested more previously infested host eggs than controls. The insecticide also induced an increase in antennaland ovipositor rejection of previously infested host eggs. These results are discussed in the light of the mode ofaction of pyrethroid insecticides. The findings of the study highlight sublethal effects that reduce the fitness ofparasitoids and that could consequently modify the equilibrium of natural ecosystems.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

When a Trichogramma female arrives at a host patch, she receivesinformation about the quality of this patch by perceiving differentstimuli that can emanate from chemical mediators, such as thekairomones of her host, which furnish information on the suitabilityof the host, and pheromones left by her congeners, which furnishinformation on the quality of the patch by informing her of the possi-bility that some host eggs are already infested. She also receives infor-mation from physical mediators that inform her of the size of the hosteggs and, therefore, of host characteristics, including the number ofher eggs that can develop in each host. From all this information,the female can evaluate the quality of the patch, decide to stay orleave, and select a strategy of infestation: how many eggs to lay, inhow many hosts, and to infest every host eggs or not. The acquisitionof information from the perceived stimuli, the transmission of this

volutive UMR CNRS 5558, 43Cedex, France. Tel.: +33 4 72

(J.-M. Delpuech).

rights reserved.

information to the central nervous system, the elaboration of anadapted response to the perceived information and, finally, the exe-cution of the behavior chosen according to the perceived informationall depend on the transmission of nerve impulses in the nervoussystem (Vosshall and Stocker, 2007). For this reason and because themajority of insecticides are neurotoxic (Casida, 2009) and interferewith the transmission of nerve impulses in insects, every behaviordescribed above may be disrupted by insecticides. Insecticides canproduce numerous sublethal effects on the life history traits of insects(see Desneux et al., 2007 for a review), particularly on those traitsinvolved in reproduction. Deltamethrin, the insecticide used in thisstudy, is a neurotoxic insecticide that interferes with the transmissionof action potentials along neurons (Soderlund, 2012). It belongs tothe pyrethroids, the second most widely used family of insecticides(organophosphates are the most-used family: EC, 2007; US, EPA, 2011).Deltamethrin is used on many crops (e.g., cereal, maize, crucifers, arti-choke, asparagus, beet, salad vegetables, tomato, pepper, potato, apple,pear, peach, grape, rice, peas, and onion, cf. Couteux and Lejeune, 2009).

Trichogramma parasitoids are key species because they regulatethe natural populations of other insects, including pests. Furthermore,they are the most widely used insect natural enemy in the world

275J.-M. Delpuech, M. Delahaye / Science of the Total Environment 447 (2013) 274–279

(Knutson, 1998). Nine species of Trichogramma are reared aroundthe world and released annually on an estimated 80 million acresof agricultural crops and forests in 30 countries (Knutson, 1998).Trichogramma are released to control 28 different caterpillar pestsattacking corn, rice, sugarcane, cotton, vegetables, sugar beets, fruittrees and pine and spruce trees (Smith, 1996; Yu and Byers, 1994).

Due to their intensive use in modern agriculture and urban protec-tion, insecticides produce environmental pollution, to which non-targetinsects, such as Trichogramma parasitoids, may be exposed. The impactof environmental pollution resulting from insecticides on this type ofinsect will therefore have ecological consequences because it canmodifythe equilibrium of natural ecosystems, but it will also have economicconsequences because of the pest populations that the insects naturallycontrol. In this study, we tested the effects of a 20% lethal dose (LD 20)of deltamethrin on the foraging behavior of Trichogramma brassicaeinfesting a patch of host eggs.

2. Materials and methods

2.1. Insects

A strain of T. brassicae Bezdenko (Hymenoptera, Trichogrammatidae)was used for the experiments. This strain was reared on Ephestiakuehniella eggs (Lepidoptera, Pyralidae) killed by UV radiation to preventnon-parasitized eggs from emerging. Host eggs were supplied in excess.Therefore, only one T. brassicae egg was laid per E. kuehniella egg. Therearing and experiments were conducted at 21 °C under a 12 L:12Dphotoperiod (light phase from 7:30 am to 7:30 pm).

2.2. Determination of lethal doses

E. kuehniella eggs parasitized by Trichogramma were individually(one egg per vial) isolated in glass vials (3 cm in length, 5 mm indiameter) containing a minute drop of honey to feed the insects uponemergence. Approximately 24 h after their emergence, the males andfemales were sexed. The females were individually exposed to the in-secticide. For this purpose, 3 μl of deltamethrin (99% certified purity;Cluzeau Info Labo, Sainte-Foy-La-Grande, France) diluted in acetonewas deposited on pieces of paper (2.2 cm×4 mm), which wereleft for 1 h on the lab bench to allow the total evaporation of acetoneto occur. The pieces of paper were then introduced into each vialcontaining a tested female. Papers on which pure acetone wasdeposited were used as controls. Mortality was determined after 24 hof exposure to the treated piece of paper at 21 °C, photoperiod 12:12(contamination occurred via tarsal contact). To calculate regressionlines for mortality, 5 groups of 50 individuals were exposed to a controlsolution and 4 solutions of increasing concentrations of insecticide. Themortality data were analyzed with a probit analysis (Finney, 1971), andthe LD 20 to be used for behavioral tests was then estimated with alinear regression using the log-probit program of Raymond (1985).The dilution corresponding to the LD 20 was stored at 4 °C betweenbehavioral experiments.

2.3. Protocol for the observation of the behavior of insects on host patches

Two days prior to emergence, E. kuehniella eggs infested byTrichogramma were individually isolated in glass vials (3 cm inlength, 5 mm in diameter) with a minute drop of honey. The malesand females were sexed 48 h after the first emergences, and eachfemale was then placed with one male for fertilization. After at least1.5 h, the males were removed, and the females were exposed to anLD 20 of deltamethrin as described in “Determination of lethaldoses” section. The females were left in their exposure vial untiltheir behavior on a patch of host eggs was observed. Accordingly,the exposure time to the insecticide ranged from 22 h to 27 h. Forall observations, a control female was always observed after a treated

female. The females were fertilized but naive (i.e., they had neverbeen in contact with host eggs and therefore had no ovipositionalexperience). The females were one to three days old at the momentof the observation of their behavior. A new host patch was used foreach female. Each host patch was made by gluing two groups of 9E. kuehniella eggs with water onto the center of a 10 cm×10 cmsquare sheet of paper. These host eggs had previously been killed byUV radiation. The first group of eggs consisted of 9 eggs infested byT. brassicae. The second group consisted of 9 healthy (not infested)eggs. The two groups were approximately 5 mm apart. Within eachgroup, the eggs were approximately 1 mm apart. The eggs of thegroup of infested eggs were infested 24 h before the experiment byleaving about 100 E. kuehniella eggs, killed by UV radiation, withabout 10 T. brassicae females in a closed Petri dish until the use ofthe eggs for the experiment. Only insects that were able to movenormally after their exposure to the insecticide were used for theexperiment (the discarded insects represented less than 10% ofthe total). For the experiment, one parasitoid female was placed onthe center of the group of infested eggs and was free to move insidethe patch, to leave it or to leave the sheet of paper. The female'sbehavior was observed with a stereomicroscope. The observationswere performed at 21 °C in a quiet room.

2.4. Behavior of the parasitoid in control conditions

After a female is released inside the group of infested eggs, shegenerally leaves it and enters inside the group of healthy eggs. Shethen walks over an egg and moves her antennae rapidly up anddown. This stage is defined as the drumming stage. The female thenpierces a hole into the egg by rotating her ovipositor alternately leftand right, the stage defined as the drilling stage. Once the chorionof the egg is pierced, the female can decide either to pull her ovipos-itor out of the egg and feed from the hemolymph that exudes or toinsert her ovipositor completely into the egg and lay an egg of herown. After infesting the first host egg several times, the female leavesit and continues to drum on the substrate with her antennae. Onceshe arrives at another egg, the female screens it by drumming on itwith her antennae. She can then decide either to drill it with herovipositor or to leave it. If she leaves without drilling, this event istermed an antennal refusal. If the female has drilled the egg, she canthen either infest it by completely inserting her ovipositor or pullher ovipositor out and not infest it (ovipositor refusal). Generally,an egg refusal (antennal or ovipositor refusal) occurs if the egg isalready infested. The female continues to perform these behaviorson several eggs, sometimes momentarily leaving the patch and thenreturning to it. Finally, after several ovipositions or rejections, sheconcludes by leaving the patch.

2.5. Observed behaviors

The following behaviors were counted and their duration recordedwith JWatcher software (Blumstein et al., 2006):

– Entry into the group of healthy (not infested) eggs– Entry into the group of infested eggs– Climbing onto an egg– Drumming on an egg– Drilling an egg– Antennal rejection (the egg is left after antennal drumming)– Ovipositor rejection (the egg is left after drillingwithout infesting it)– Egg rejection (sum of the last two behaviors)– Oviposition– Moving intra-patch (the parasitoid walks between eggs)– Exit from the group of healthy eggs– Exit from the group of infested eggs

Fig. 1.Meannumber of T. brassicae ovipositions as a function of host rank for control females(horizontal hatching) or females exposed to an LD 20 of deltamethrin (inclined hatching).Host rank is defined by the chronological order of infestation by the female parasitoid.Error bars: standard error. Wilcoxon rank sum test, ns: p>0.05.

276 J.-M. Delpuech, M. Delahaye / Science of the Total Environment 447 (2013) 274–279

– Moving extra-patch (the parasitoid walks on the square sheet ofpaper at a distance of at least 2 mm from the eggs)

– Resting by the parasitoid (immobility or feeding on the egg)– Revisiting the patch (the parasitoid comes back to the patch after

having left it).

During the observations, each of the 18 eggs was identifiedand the behaviors of the parasitoid monitored. These observationsallowed us to count the number of behaviors directed toward eachegg. The behavior of each female was observed until she left thepatch for more than 3 min, at which time the observation ended. Ifthe female remained on the patch for more than 1 h, the observationwas stopped at the end of the current behavior and the duration wastreated as censored data in the statistical analysis. After the observa-tion, the patches of host eggs were maintained at 25 °C until theinfested eggs became black to confirm infestation.

2.6. Data analysis

The variability of the duration of each behavior as a function ofthe rank of the repetition and treatment (insecticide or control) wasanalyzed with generalized estimating equations (GEE). GEE, a regres-sion method described by Liang and Zeger (1986) and Zeger andLiang (1986), allows us to test the influence of different factors on avariable that is non-normally distributed with repeated measuresfor the same individual (Ballinger, 2004). The GEE method is derivedfrom the generalized linear model, which allows the incorporationof correlations between measures. The data were compared using aWald test. The statistical software R (Ihaka and Gentleman, 1996,http://www.r-project.org/) with the package geepack (Halekoh etal., 2006) was used for the analysis.

3. Results

3.1. Lethal dose

The linear regression equation obtained from the observed mor-talities was Y (in probit)=3.274∗ log10(X)−1.108. The theoreticaldose inducing 20% of mortality (LD 20) was 40.60 ng (active ingredi-ent) per piece of paper (95% confidence interval: 28.83–50.00 ng).This dose was used for testing the effects of deltamethrin on thebehavior of Trichogramma.

3.2. Reliability of the observations of infestations

3.2.1. Group of healthy eggsA total of 85.2% (SE 1.70) of the host eggs identified as infested

became black during their development. Infested eggs become blackwhen the parasitoid has reached the 3rd larval stage (Voegelé,1978). Therefore, the emergence rate would be, at most, 85.2%. Thisresult corresponds to the observations of Pintureau (2009), whichestimated the emergence rate of T. brassicae reared at 25 °C inE. kuehniella eggs killed by UV radiation to range between 75 and95%. Based on these findings, our observations of infestation appearto be reliable. Our determination that an egg was infested was valid.

3.2.2. Group of infested eggsA total of 78.8% (SE 1.45) of the eggs used for the group of infested

eggs became black. The 6.4% difference between this result and the85.2% blackening of the infested eggs in the group of healthy eggs isstatistically significant (Wilcoxon rank sum test: W=173, pb0.05).Therefore, it is possible that the group of infested eggs was infestedat an approximate percentage of only 94%, even though the value of78.8% falls within the 75–95% interval of emergences estimated byPintureau (2009).

3.3. Effects of deltamethrin on the behavior of parasitoids

At the beginning of each observation, females were deposited onthe group of infested eggs. The majority of them rapidly left thisgroup and approached the healthy group. However, 28% of thefemales exposed to the insecticide stayed on the infested group andinfested their first host egg in that group, whereas only 11% of thecontrol females did so (χ2=1.60, df=1, NS).

Both the females exposed to the insecticide and the controlssuperparasitized the first egg, which was infested at least twice,but they rarely superparasitized the subsequent eggs. No significantdifferences between the exposed females and the controls wereobserved (Fig. 1).

The females exposed to deltamethrin spent significantly moretime on the group of infested eggs than the controls (Fig. 2). Theylaid significantly fewer eggs in the healthy group and more in theinfested group than the controls (Table 1). They also infested fewerhost eggs in the healthy group than the controls (74% and 94%,respectively, Table 2).

The female parasitoids exposed to deltamethrin rejected morehost eggs than the controls (Fig. 3). This increase in the rejectionrate was due to an increase in both antennal and ovipositor rejections,whereas there was no significant difference between the exposed andthe control females in the number of climbing behaviors. The expo-sure to the insecticide had no effect on the behaviors of drumming,drilling, oviposition and intra-patch moving (results not shown).

4. Discussion

The behavior of the females that survived exposure to an LD 20 ofdeltamethrin was modified. They examined as many host eggs as thecontrols, but they rejected more eggs because they spent more timeon the group of previously infested host eggs; therefore, they infestedfewer host eggs. Furthermore, they infested fewer healthy hostsbut more previously infested ones. This outcome corresponds to aloss of eggs because a previously infested E. kuehniella egg is toosmall to support the development of a second parasitoid (Pintureau,2009). T. brassicae is proovogenic. The females carry their total stock

Fig. 2. Mean percentage of time spent by T. brassicae females on the healthy group, on theinfested group of host eggs and onneither of these twogroups (extra-patch area) for controls(horizontal hatching) or females exposed to an LD 20 of deltamethrin (inclined hatching).Error bars: standard error. Wilcoxon rank sum test, *: pb0.05, ns: p>0.05.

Table 2Contingency table for the numbers of host eggs infested and not infested in the groupof healthy eggs according to the type of treatment received by parasitoid females(exposed or not exposed to the insecticide).

Healthy group Total

Infested eggs Non-infested eggs

Control 152 10 162Treated 120 42 162Total 272 52 324

Treated insects were exposed to an LD 20 of deltamethrin. Effect of the insecticide onthe number of hosts infested: χ2=23.46, df=1, pb0.001.

277J.-M. Delpuech, M. Delahaye / Science of the Total Environment 447 (2013) 274–279

of eggs (approximately 50 eggs) upon emergence and can matureonly 4–5 eggs after emergence (Volkoff and Daumal, 1994). There-fore, any loss of eggs directly reduces their reproductive capacityand, consequently, their fitness.

It is possible for the females to detect a previously infestedhost because females of Trichogramma species imprint on their hosttwice during oviposition, first externally and then internally uponinsertion of their ovipositor inside the host egg (Vinson, 1998). Theimprinting corresponds to the deposition of molecules that serve asolfactory marks, and it enables the females to distinguish betweenhealthy eggs and those previously infested (Klomp et al., 1980). Thefemales can perceive these olfactory marks with their antennaethrough antennal drumming (Salt, 1937), but the marks can alsobe perceived with the sensory receptors located at the end of theovipositor when the female inserts it in the host egg (Le Ralec andWajnberg, 1990). Both controls and females exposed to deltamethrinsuperparasitized the first host egg by infesting it approximately twice,but the subsequent eggs were infested only once, demonstrating thatthe exposed females were still able to discriminate between theeggs already infested and those not infested. Females exposed todeltamethrin spent more time on the infested group of hosts andlaid more eggs on that group than the controls, but the number ofeggs laid on the infested group remained low relative to the numberof eggs laid on the healthy group of hosts (10 vs. 139). This resultrepresents additional evidence that the females exposed to deltamethrinwere still able to recognize previously infested host eggs although thisability was probably somewhat decreased.

The rejection of host eggs generally occurs if the host egg hasalready been infested. The increase in the number of rejected hostsobserved when females were exposed to deltamethrin was most

Table 1Contingency table for the total numbers of eggs laid by T. brassicae females according tothe group of host eggs in which they were laid (healthy or infested eggs) and the typeof treatment received by parasitoid females (exposed or not exposed to the insecticide).

Group Total

Healthy Infested

Control 181 2 183Treated 139 10 149Total 320 12 332

Treated insects were exposed to an LD 20 of deltamethrin. Effect of the insecticide onthe choice of the group of host egg: χ2=7.44, df=1, pb0.01.

likely due to the tendency of the females to spend more time on theinfested group of host eggs than the controls and to reject theseeggs after examination. This increase corresponded to an increase inboth antennal rejection and ovipositor rejection. This effect differssomewhat from the effect observed with the organophosphorus(OP) insecticide chlorpyrifos. Indeed, T. brassicae exposed to an LD20 of chlorpyrifos showed a progressive decrease in the antennal per-ception of previously infested hosts. This progressive loss of antennalperception was compensated for by the perception by the ovipositorof the hosts that had already been parasited (Delpuech and Leger,2011). This phenomenon was interpreted as a result of the directeffect of the OP insecticide, which induces the prolongation of nerveimpulses in the nervous system by inhibiting the hydrolysis of acetyl-choline in the synapses (Bloomquist, 2009). Indeed, the prolongationof the nerve impulses generated in the pathways from the antennaeto the central nervous system after each odor perception inducedby the insecticide would ultimately prevent any new transmissionsof odor perceptions by these pathways, whereas the ovipositor path-ways, which were active to a lesser degree, would be able to compen-sate for the loss of antennal perceptions. The pyrethroid insecticidedeltamethrin also induces a prolongation of nerve impulses in thenervous system of the insect by prolonging the opening of sodiumchannels along the axons. However, this action is reversible within afew minutes (Bloomquist, 2009), whereas the action of chlorpyrifoslasts several hours, even days, because the bond between chlorpyrifosand acetylcholinesterase (the enzyme hydrolyzing acetylcholine inthe synapses) is permanent (Bloomquist, 2009). To recover the nor-mal function of its nervous system after an exposure to chlorpyrifos,the insect must then degrade the insecticide–enzyme associationand synthesize new acetylcholinesterase. Therefore, if the insectis exposed to deltamethrin, its nerve stimulation is amplified butis rapidly restored and is not inhibited. In contrast, the stimulatednerve pathways become rapidly and permanently excited if the insectis exposed to chlorpyrifos, and perceptions through these pathwaysbecome impossible. This result explains why antennal perceptions arestimulated and antennal rejections are increased with deltamethrin,whereas antennal perceptions are progressively inhibited and antennalrejections are decreased with chlorpyrifos.

The recommended dose of deltamethrin to protect crops (wheat,maize etc.) against their pests ranges from 6.3 g/ha to 12.5 g/ha(Couteux and Lejeune, 2009). The dose used in this study (LD 20) wasequal to 40.60 ng per piece of paper (2.2 cm×4 mm), which corre-sponds to 4.6 g/ha. It is therefore a high dose. Insects may be exposedto such a high dose in the immediate vicinity of cultivated areas sprayedwith deltamethrin. The exposed insects would then need not only tosurvive the exposure to the insecticide but also to overcome its possiblesublethal effects. We have seen some of these effects in this study,but others have already been demonstrated. Insecticides have beenshown to modify the pheromonal communications between malesand females (Delpuech et al., 1998a, 1998b, 1999, 2001) and the per-ception of kairomones by parasitoids (Delpuech et al., 2005; Komezaet al., 2001). They have been shown to modify the offspring sex ratioin Trichogramma (Delpuech and Meyet, 2003). They also have beenshown to induce confusion in the discrimination by males of female

Fig. 3. Cumulative numbers for each behavior in both groups of host eggs (healthy and infested) and each parasitoid female (faded symbols) versus time. Lines represent thepredictions obtained from the GEE model for parasitoids exposed to an LD 20 for deltamethrin (dashed) and control (solid) treatments. *: pb0.05, ns: p>0.05.

278 J.-M. Delpuech, M. Delahaye / Science of the Total Environment 447 (2013) 274–279

sexual pheromones emanating from two closely related Trichogrammaspecies (Delpuech et al., 2012; Dupont et al., 2010). This confusioncould favor interspecific matings and would thus strongly decrease thefecundity of females (Delpuech et al., 2010). However, it is known thatthe sublethal effects of insecticides are not always detrimental. Forexample, insecticides have been shown to stimulate the tendency tosearch for hosts and to increase the infestation efficiency of parasitoidsof Drosophila larvae (Rafalimanana et al., 2002).

5. Conclusions

This study found that T. brassicae females that had survived exposureto an LD 20 of deltamethrin showed decreased reproductive capacityand, as a consequence, reduced fitness. The exposed females spentmore time on the group of previously infested host eggs. As a result,they infested fewer healthy host eggs. Parasitoids such as Trichogrammaare key species because they control populations of other insects.Accordingly, these sublethal effects could have important consequencesfor natural ecosystems.

Acknowledgments

We are thankful to S. Janillon-Martinez and H. Henri for technicalassistance, to Marie-Christine Carpentier for assistance with statisticalanalysis and to R. Allemand for advice during the experiments.

References

(EC) European Commission. In: Nadin P, editor. The use of plant protection productsin the European Union: data 1992–2003. Luxembourg: Eurostat Statistical Books;2007.

(U.S. EPA) U.S. Environmental Protection Agency, Office of Chemical Safety and PollutionPrevention. Pesticides industry sales and usage, 2006 and 2007 market estimates.http://www.epa.gov/opp00001/pestsales/07pestsales/market_estimates2007.pdf 2011.

Ballinger GA. Using generalized estimating equations for longitudinal data analysis.Organ Res Methods 2004;7:127–50.

Bloomquist JR. Insecticides: chemistries and characteristics. In: Radcliffe EB, HutchisonWD, Cancelado RE, editors. Radcliffe's IPM world textbook. St Paul: University ofMinnesota; 2009URL: http://ipmworld.umn.edu [accessed 22 October 2012].

Blumstein DT, Daniel JC, Evans CS. JWatcher. URL http://www.jwatcher.ucla.edu 2006[accessed 8 October 2012].

Casida JE. Pest toxicology: the primary mechanisms of pesticide action. Chem Res Toxicol2009;22:609–19.

Couteux A, Lejeune V. Index phytosanitaire ACTA 2010. Paris, France: ACTA Publications;2009.

Delpuech JM, Leger L. Modification of Trichogramma behaviors during the exploita-tion of host patches induced by the insecticide chlorpyrifos. Ecohealth 2011;8:190–8.

Delpuech JM, Meyet J. Reduction in the sex ratio of the progeny of a parasitoid wasp(Trichogramma brassicae) surviving the insecticide chlorpyrifos. Arch EnvironContam Toxicol 2003;45:203–8.

Delpuech JM, Froment B, Fouillet P, Pompanon F, Janillon S, Boulétreau M. Inhibition ofsex pheromone communications of Trichogramma brassicae (Hymenoptera) by theinsecticide chlorpyrifos. Environ Toxicol Chem 1998a;17:1107–13.

Delpuech JM, Gareau E, Terrier O, Fouillet P. Sublethal effects of the insecticide chlorpyrifoson the sex pheromonal communication of Trichogramma brassicae. Chemosphere1998b;36:1775–85.

Delpuech JM, Legallet B, Terrier O, Fouillet P. Modifications of the sex pheromonalcommunication of Trichogramma brassicae by a sublethal dose of deltamethrin.Chemosphere 1999;38:729–39.

279J.-M. Delpuech, M. Delahaye / Science of the Total Environment 447 (2013) 274–279

Delpuech JM, Legallet B, Fouillet P. Partial compensation of the sublethal effect ofdelthamethrin on the sex pheromonal communication of Trichogramma brassicae.Chemosphere 2001;42:985–91.

Delpuech JM, Bardon C, Boulétreau M. Increase of the behavioral response tokairomones by the parasitoid wasp Leptopilina heterotoma surviving insecticides.Arch Environ Contam Toxicol 2005;49:186–91.

Delpuech JM, Dupont C, Allemand R. Decrease in fecundity induced by interspecific matingbetween two Trichogramma parasitoid species. J Econ Entomol 2010;103:308–13.

Delpuech JM, Dupont C, Allemand R. Effects of deltamethrin on the specific discriminationof sex pheromones in two sympatric Trichogramma species. Ecotoxicol Environ Saf2012;84:32–8.

Desneux N, Decourtye A, Delpuech JM. The sublethal effects of pesticides on beneficialarthropods. Annu Rev Entomol 2007;52:81-106.

Dupont C, Allemand R, Delpuech JM. Induction, by chlorpyrifos, of the confusion ofmales in discriminating female sexual pheromones used for mate finding by twosympatric Trichogramma species (Hymenoptera: Trichogrammatidae). EnvironEntomol 2010;39:535–44.

Finney DJ. Probit analysis. Cambridge: Cambridge University Press; 1971.Halekoh U, Hojsgaard S, Yan J. The R package geepack for generalized estimating equations.

J Stat Softw 2006;15:1–9.Ihaka R, Gentleman R. R: a language for data analysis and graphics. J Comput Graph Stat

1996;5:299–314.KlompH, Teerink BJ, MaWC. Discrimination between parasitized and unparasitized hosts

in the egg parasite Trichogramma embryophagum (Hym., Trichogrammatidae): amatter of learning and forgetting. Neth J Zool 1980;30:254–77.

Knutson A. The Trichogramma manual. Texas Agricultural Extension Service B-6071,Agricultural Communications. The Texas A&M University System; 1998.

Komeza N, Fouillet P, Boulétreau M, Delpuech JM. Modification, by the insecticidechlorpyrifos, of the behavioral response to kairomones of a Drosophila parasitoid,Leptopilina boulardi. Arch Environ Contam Toxicol 2001;41:436–42.

Le Ralec A, Wajnberg E. Sensory receptors of the ovipositor of Trichogramma maidis(Hym.: Trichogrammatidae). Entomophaga 1990;35:293–9.

Liang KY, Zeger SL. Longitudinal data analysis using generalized linearmodels. Biometrika1986;73:13–22.

Pintureau B. La lutte biologique et les Trichogrammes. Application au contrôle de lapyrale du maïs. Paris: Le Manuscrit; 2009.

Rafalimanana H, Kaiser L, Delpuech JM. Stimulating effects of the insecticide chlorpyrifoson host searching and infestation efficacy of a parasitoid wasp. Pest Manag Sci2002;58:321–8.

Raymond M. Présentation d'un programme d'analyse log-probit pour micro-ordinateur.Cah ORSTOM Ser Entomol Med Parasitol 1985;22:117–21.

Salt G. The sense used by Trichogramma to distinguish between parasitised andunparasitised hosts. Proc R Soc B-Biol Sci 1937;122:57–75.

Smith SM. Biological control with Trichogramma: advances, successes, and potential oftheir use. Annu Rev Entomol 1996;41:375–406.

Soderlund DM. Molecular mechanisms of pyrethroid insecticide neurotoxicity: recentadvances. Arch Toxicol 2012;86:165–81.

Vinson SB. The general host selection behavior of parasitoid hymenoptera and acomparison of initial strategies utilized by larvaphagous and oophagous species.Biol Control 1998;11:79–96.

Voegelé J. Utilisation des trichogrammes. Bull Tech Inf Minist Agric 1978;332–333:447–52.

Volkoff AN, Daumal J. Ovarian cycle in immature and adult stages of Trichogrammacacoeciae and Trichogramma brassicae (Hym. Trichogrammatidae). Entomophaga1994;39:303–12.

Vosshall LB, Stocker RF. Molecular architecture of smell and taste in Drosophila. AnnuRev Neurosci 2007;30:505–33.

Yu DS, Byers JR. Inundative release of Trichogramma brassicae Bezdenko (Hymenoptera:Trichogrammatidae) for control of European corn borer in sweet corn. Can Entomol1994;126:291–301.

Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes.Biometrics 1986;42:121–30.