Imen Blibech ，Ikbal chaieb，Mohiedine Ksantini(．Unity of Plant Protection and Environment，Olive Tree Institute，3000，Sfax，Tunisia;．Entomological laboratory，Regional center for research on horticulture and organic agriculture(centre régional de recherche en horticulture et en agriculture biologique)404，Chott Mariem，Sousse Tunisia)

1 Introduction

Egg parasitoids have been used in the biological control of pests as essential part of pest management strategies in crop protection．

Endoparasitoidsofthe genus Trichogramma，which are egg-parasitoids in the family Trichogrammatidae have met with variable success in agricultural cropping systems and have been reported to be efficient in regulating populations of lepidopteran pests(Wright et al．，2002，Hoffmann et al 2006，Hegazi et al 2007)．

Among main lepidopteran insect pest of olive tree，the olive moth Prays oleae Bern(Lepidoptera，Hyponomeutidae)is the most abundant throughout the Mediterranean and the Black Sea，the Middle East and the Canary Islands(Tzanakakis，2003)．

Trichogramma spp．are known to be fairly polyphagous(Smith，1996)，however host acceptance and preferences have been demonstrated for several species(Monje et al，1999;MansWeld and Mills 2003;Reznik et al．，1997)．In addition，several studies suggest that effectiveness of species/strains to control the insect pest is a first key step toward managing biological control(Wajnberg and Hassan，1994;Bruce et al．，2009)focusing on its ability to give good proportions of parasitism(Silva and Stouthamer，R．，1999)．Trichogramma species could be mass reared in large numbers using the eggs of stored product Lepidoptera moths as alternative hosts．Eggs of the mediterranean flour moth，Ephestia kuehniella Zeller(Lepidoptera:Pyralidae)has contributed to the mass production and the widespread use of Trichogramma species in augmentative biological control．

In Tunisia，the damage caused by P．oleae can reduce olive production by 50%(Jardak，1980)．As a consequence，many efforts have been developed in biological control programs for P．oleae pest of olive using the release of Trichogramma eggs(Herz et al．，2006)．Although widely used，a recent review of these programs worldwide concluded that because of considerable variability in success of releases and little evidence of consistently successful application of Trichogramma，the usefulness of these parasitioids is currently being debated(Knutson，2005)．There are still questions about the effectiveness and quality control of Trichogramma in many crop production systems because still the number of caterpillar eggs destroyed by native populations of Trichogramma is not sufficient to prevent the pest from reaching damaging levels．Releases and integration issues that are considered important for the development of a successful biological control program include important factors such as host suitability and preference studies(Roriz et al．，2006)as well as the high potential of parasitism during releases(Klug and Meyhöfer，2008)．Furthermore，it is important to select the appropriate Trichogramma species that are most adapted to the local field conditions for biological control programs(Hegazi et al，2007)．

This paper deals with laboratory research on the host-selection process of three native Trichogramma species T．oleae， T． cacoeciae and T．bourarachae(Herz et al．，2006)．These wasps were obtained from parasitized eggs of P．oleae and reared in the laboratory on E．kuehniella eggs．We examined whether Trichogramma from field collections of parasitized P．oleae eggs and reared under laboratory conditions on E．kuehniella eggs were able to show efficiency in parasitizing P．oleae eggs．Our first aim in this article is to investigate the oviposition behavior and to select the most adequate parasitoid species which accept and prefer to parasitize P．oleae than E．kuehniella wasps．Fast and reliable methods were established to determine host acceptance and preference．Quality measures can be associated with the effectiveness of native Trichogramma species parasitizing moth eggs and appropriate for use for P．oleae pest on olive．

2 Materials and methods

2.1 Insect pest and parasitoids

The colonies of Trichogramma species used in these experiments was originally established from parasitized eggs of Prays oleae Bernard(Lepidoptera，Yponomeutidae)found at Siris(North East Tunisia at 30Km from Sfax，)and of Virachola livia Klug(Lepidoptera，Lycaenidae)found at Oasis de Gafsa(South West Tunisia at 170 km from Sfax)．The colonies of T．oleae，T．bourarachae and T．cacoeciae were obtained from eggs collected in the above-mentioned locality and reared under laboratory conditions during 5 generations on E．kuehniella eggs．

Glued cardboard(8 cm ×1．5 cm)，each with E．kuehniella eggs，were presented to recently emerged Trichogramma females in glass tubes(12 cm×1．5 cm)at 26 ±2 °C，L16∶D9 and 60%RH．

The cards with a parasitism rate close to 100%were then transferred to other glass tubes and maintained until pupation(192 hours)．

2.2 Rearing host E.kuehniella

The culture was reared at 20 ±1C°on semolina wheat flour．Laboratory population was established from eggs obtained from a culture maintained at the insectary in the Entomology Department of Olive tree Institute University(Blibech，2003)．

2.3 Experimental protocol

All experiments were carried out in an insectary room controlled at a temperature of 25°C，65+5%RH．

2．3．1 Collection of E．kuehniella eggs

Eggs were collected daily from the laboratory cul-ture to ensure that the live eggs used in the experiments were notmore than 24h old．Fresh E-．kuehniella eggs were cleaned under a laminar flow and irradiated 50min at UV-B before being used for experiments．

2．3．2 Collection of fresh eggs of P．oleae

A laboratory population of P．oleae from fieldcollected pupae was transferred in plastic box to a raring chamber at 25°C，65+5%RH and 16 L∶D 8 until emergence．

Fifth mixed sex pairs of newly emerged adults of P．oleae from pupae were transferred onto fresh plants in plastic cages and provided with 10%sucrose solution for P．oleae females to lay eggs．Leaves with freshly P．oleae laid eggs were collected and used for the acceptance and preference test．

2.4 Host acceptance test

A single newly emerged female was transferred into another glass tube and offered to freshly P．oleae laid eggs(1-3)on one tender fresh leave for 6 hours at 25°C，70%HR and 16 L∶D 8 and assessed for parasitism after 6 days．We used 15 replicate females for each specie．To isolate a single female，Trichogramma adults were scattered on a smooth surface and an individual female was captured by placing an open end of a tube(6．0 × 1．5 cm diameter)around one individual．

A host was considered accepted when the female，after drilling the chorion，stayed more than 1min with the ovipositor inside the host egg，displaying a series of abdominal movements characteristic of egg deposition．Direct observations were stopped after the first oviposition and determined by the number of oviposition acts within one hour of insect release．Up to five oviposition acts per replication were counted to judge host acceptance．

2.5 Preference test

The preference was determined by offering the choice and observing contacts with eggs of the target pest P．oleae and eggs of the host E．kuehniella．We examined the wasps'host-selection behavior in a petri dish arena using four fresh(＜24 h old)egg masses of the two different host species．Each egg mass was of 10 eggs presented against a green background．These eggs were offered simultaneously to a single female wasp(＜24 h old)without previous oviposition experience．The following combinations were tested:T．oleae/P．oleae， T．cacoeciae/P．oleae， T．bourarachae/P．oleae，T．oleae/E．kuehniella，T．cacoeciae/E．kuehniella and T．bourarachae/E．kuehniella，Parasitism was therefore assessed 6 days after the experiment under similar laboratory conditions．Combinations were calculated and were tested by using them to identify responses from the test set(n=60;15 females/Trichogramma specie，approximately 15 responses/test)．Host eggs were glued with an insignificant drop of distillate water on the surface of a small piece of paper(1．0 ×1．0 cm)and a drop of honey was placed at the center of the petri dish arena．Egg masses were arranged on the cross．Therefore，two egg masses of one host species and two of the other host species were observed simultaneously to a single female wasp．Therefore，the ovipositional behavior of individual Trichogramma female was monitored continuously which is easy to examine using a binocular．We used a petri dish per female repetition．To exclude the locationeffect of the egg mass，host position was changed in half of the total replicates．

Each parasitoid was observed two times for a maximum of 20 min within the hour and 10 min was elapsed between any two observations．Having had 20 min of experience with host eggs，an individual Trichogramma female was held for 10 min in a glass tube(6．0 ×1．5 cm)before gently being released again onto the cardboard patch．After a series of preliminary observations of the ovipositional behavior of the three species we defined the following set of parameters and behavioral events．A sequence of four responses(behavioral events)was used to monitor host use by the Trichogramma females:

Walking:when a female leaves a host egg and lasts until the female starts drumming on the next host egg encountered．

Drumming:Starts with contact and examine of a newly host egg．

Drilling:When the female pierces a hole in the host chorion by wriggling her ovipositor alternately left and right．

Oviposition:When the female penetrate the host chorion during drilling．The female inserts her ovipositor full length into the host egg，and after trembling movements of the abdomen(Suzuki et al．，1984)．Oviposition ends with removal of the ovipositor．

2.6 Data analyses

Samples were first described with regular average and standard errors．Then，proportional data were normalized through arcsine transformation;all the other data were normalized through square root transformation．Data sets were compared by an analysis of variance(ANOVA)，followed by Duncan's tests with P＜0．05．The recorded behavioral events for each female specie and the mean duration of each event were treated separately．The following patches were noted for each female:Patch moving time;Equal to the time of total walking between eggs and drumming，and Patch handling time;The total time spent encountering(drilling，ovipositing)host eggs．

The number of parasitized eggs(detected visually from the black coloration of host eggs produced at parasitoid pupation)and wasp's emergence rates for each host species was determined after 6 days．All tests were done under natural laboratory conditions and data sets were compared by an analysis of variance(ANOVA)，

Pearson correlation analyses were also performed between the mean number of contact per host egg and the mean number of parasitized host eggs．

3 Results

The numbers of accepted eggs of P．oleae per female after 6h of contact，as well as the resulting number of parasitized eggs after 6d are shown in Table 1．

Table 1 Number of parasitized eggs，Adult emergence rates(means±SD)

A significant reduction of acceptance rates occurred between Trichogramma species and number of parasitized eggs(F=19．05;P=0．004)，due to the reduced performance of T．bourarachae on patches P．oleae eggs(P ＜0．05)．In contrast，no difference was observed in performance between T．oleae and T．cacoeciae acceptance rates(P=0．068)(Table 1)．In addition，the emergency rate per female was not significant between all three Trichogramma species(F=6．46;P=0．102)．

T．cacoeciae performed better with P．oleae eggs(100% of emergency rate)．Pearson correlation is significant at 0．01 level between the number of accepted eggs(parasitized eggs detected visually from the black coloration)and emerged Trichogramma adults(r=0．833，N=45，P=0．00)．

The number of contacts per female on the eggs of P．oleae and E．kuehniella，the moving and the handling times after 2 observations of 20 min as well as the resulting number of parasitized eggs after 6d are shown in Table 2．T．cacoeciae showed best results in searching for，and parasitisation of P．oleae eggs(7．93 ±3．08 contacts/female(first observation)and13．13 ±2．44 contacts/female(second observation)total during time 2400 s，Table 2)．Also，the number of contacts/female on P．oleae eggs differed significantly among the species(P ＜0．05);direct observations on preferences in searching for P．oleae eggs were shown by T．oleae(6．46 ±3．66 first observation and 07．2 ±3．93 second observation，Table 2)and T．cacoeciae．Moreover，mean number of contacts/female of three candidate species are significantly different between P．oleae and E．kuehniella hosts(F=3．52，P ≤0．005)．The correlation is significant at 0．01 between the number of contacts/female of the tested species on both P．oleae and E．kuehniella eggs and the resulting number of parasitized eggs(r=0．745，N=45，P=0．00)．There was a significant interaction between Trichogramma species and the duration time spent in a host eggs patch(F=3．72，P≤0．001)．T．bourarachae was spending substantially less time using P．oleae egg patch(P ＜0，005)．The handling time in P．oleae egg patch also showed a significant interaction between Trichogramma species(F=7．23;P=0．013)

Table 2 Number of contacts per female on the eggs of P.oleae and E.kuehniella(means±SD)，the moving and the handling times after 2 observations(s)and percent parasitism and emergence(%)after 6 days(means±SD)

4 Discussion

Quality measures helping to design the best biological control strategy against P．oleae pest were associated with the effectiveness of three native Trichogramma species parasitizing moth eggs．The objective of our work consists on the determination of ideal Trichogramma species searching for P．oleae eggs and leading to its parasitization．Ideal Trichogramma species should than be active in searching for，and showing parasitization preference for P．oleae eggs．

Important attributes for the quality control program were defined as follow:acceptance of the host egg，good searching behavior，percent parasitism and percent emergence．These traits were attributed mainly to Trichogramma parasitism viability and adult emergency as well as the criterions of acceptance and performance on searching capacity for the host eggs(Roriz et al．，2006)．T．cacoeciae was confirmed to be one of the best species，as well as T．oleae．In contrast，T．bourarachae was less active in searching for P．oleae eggs．

For native Trichogramma species used in the present study，T．cacoeciae required a longer handling time than the two other species(648±21 and 821±15s)and had maximal oviposition rates(70．72%and 72．00%，Table 2)．Trichogramma females are known to feed the host egg contents during parasitization(Huang and Gordh，1998)．It is interesting to note that probably，eggs of P．oleae are of a best nutritional quality for T．cacoeciae specie．

On the other hand，the capacity of T．cacoeciae to parasitize eggs of E．kuehniella was significantly influenced by the rearing host，being higher for the wasps that emerged from P．oleae eggs．It appears that eggs of the rearing host E．kuehniella had far less influence on percent parasitism and on the relative handling time of all three Trichogramma species．These may vary according to adapted individuals to the chosen rearing host(Parra，2010)．

In acceptance test，we compared individual wasps that were allowed to parasitize P．oleae host based on first contact with the egg without previous experience．Accepted and not accepted eggs and subsequent oviposition were indicated at the end of experiment．Variation in percentage egg parasitism can be attributable to variation in Trichogramma species and sources of host(Grieshop，1995)．Native Trichogramma species collected from parasitized P．oleae eggs in the field were shown to be with innate preference for this host pest．Contrary to what reported by Roriz et al．，2006 who show that the option of the first host accepted by the wasp was random，we can assume that the wasp may exhibit favor of acceptance for the host．

In addition，we could discuss the preference and the oviposition per female for a host that may be influenced by the natural substrate(such as the olive tree leave in our study)where hosts can be found(Lukianchuk and Smith，1997)and not by female size for example(Maini et al．，1991)．

Although we are not aware of any other study that have made similar comparisons of the ovipositional behavior of different native Trichogramma species when confront P．oleae eggs，these results have important implications for using appropriate species for mass production and inundation programs for P．oleae control．

Based on their performance on host acceptance and preference of P．oleae eggs，the two tested species T．cacoeciae and T．oleae were identified as suitable for further investigations for the control of P．oleae in olive tree farms．Our results suggest that T．cacoeciae can be effective in the attack of P．oleae eggs under field conditions．

While host preference has often been an important criterion for selecting Trichogramma species(Hassan，1989;Pak，1988;Brotodjojo and Walter，2006)，we believe that much more attention is required for the behavioral and reproductive characteristics of native Trichogramma species for the target host pest control(Roriz et al．，2006;Gingras et al．，2003)．It is interesting to test the relative importance of genetic and environmental factorsin hostpreferencesofTrichogramma(Takasu et al．，1997;Kaiser et al．，1989)

It appears from the evaluation of the methods used in this study that acceptance test and the direct observation of the parasitization behavior is necessary to determine preference of the Trichogramma specie to be used against P．oleae under release programs．

Blibech I，2003． Production massive des parasitoids du genre Trichogramma(Hymenoptera，Trichogrammatidae)élevés sur un hôte de substitution Ephestia kuehniella(Lepidoptera，Pyralidae):Performance de production de I'hôte et des parasitoides．Master Ecologie générale．Faculté des sciences．Sfax．Tunisie．85．

Brotodjojo RR R，Walter G H，2006．Oviposition and reproductive performance of a generalist parasitoid(Trichogramma pretiosum)exposed to host species that differ in their physical characteristics．Biological Control，39:300-312．

Gingras D，Dutilleul A，Boivin PG，2003．Effect of plant structure on host finding capacity of lepidopterous pests of crucifers by two Trichogramma parasitoids．Biological Control，27:25-31．

Grieshop MA，1995，Evaluation of three species of Trichogramma egg parasitoids for the biological control of the Indian meal moth in retail stores and warehouses．Phd thesis Department of Entomology，University of California at Santa Cruz．125．

Hassan SA，1989．Selection of suitable Trichogramma strains to control the codling moth Cydia pomonella and the two summer fruit tortrix moths Adoxophyes orana，Pandemis heparana(Lep．:Tortricidae)．Entomophaga，34:19-27．

Hegazi E，Herz A，Hassan SA，EKhafagi W，Agamy E，Zaitun A，Gehan AE，Showeil S，ElSaid S，Khamis N，2007．Field efficiency of indigenous egg parasitoids(Hymenoptera，Trichogrammatidae)to control the olive moth Prays oleae(Lepidoptera，Yponomeutidae)and the jasmine moth Palpita unionalis(Lepidoptera，Pyralidae)in an olive plantation in Egypt，Biological Control，43:171-187．

Herz A，Hassan SA，Hegazi E，Khafagi WE，Nasr FN，Youssef AA，Agamy E，Blibech I，Ksentini I，Ksantini M，Jardak T，Bento A，Pereira JA，Torres L，Souliotis C，Moschos T，Milonas P，2007．Egg parasitoids of the genus Trichogramma(Hymenoptera，Trichogrammatidae)in olive groves of the Mediterranean region．Biological Control，40:48-56．

Hoffmann MP，Pitcher SA，Cheever SA，Gardner J，Losey JE，Kuhar TP，Curtis AL，Youngman RR，2006．Efficacy of inoculative releases of Trichogramma ostriniae(Hymenoptera:Trichogrammatidae)against European corn borer Ostrinia nubilalis(Lepidoptera:Crambidae)in field corn．Biological Control，36:345-349．

Huang K，Gordh G，1998．Does Trichogramma australicum Girault(Hymenoptera:Trichogrammatidae)use kairomones to recognize eggs of Helicoverpa armigera(Hübner)(Lepidoptera:Noctuidae)?Australian Journal Entomology，37:269-274．

Jardak T，1980．Etudes bioécologiques de Prays oleae BERN(Lepidoptera，Hyponomeutidae)et de ses parasites oophages du genre Trichogramma (Hymenoptera， Trichogrammatidae)． Essais d'utilisation en lutte biologique．Thèse de 3ème cycle，Univ．Aix-Marseille．5-42．

Kaiser L，Pham-Delegue MH，Masson C，1989．Behavioural study of plasticity in host preferences of Trichogramma maidis(Hym．:Trichogrammatidae)，Physiological Entomology，14:53-60

Klug T，Meyhöfer R，2008．Performance of two Trichogramma brassicae strains under greenhouse and field conditions for biocontrol of the silver Y moth in spinach cultures．Journal of Pest Science，82:73-79

Knutson A，2005．A guide to the use of Trichogramma for bilogical control with special reference to augmentative releases for control of bollworm and budworm in cotton．The Trichogramma manual，1-21．

Lukianchuk JL，Smith SM，1997．Influence of plant structural complexity on the foraging success of Trichogramma minutum:a comparison of search on artificial and foliage models．Entomologia Experimentalis et Applicata，84:221-228．

Maini S，Burgio G，Carrieri M，1991．Trichogramma maidis host searching in corn vs．pepper．Redia LXXIV，121-127．

MansW S，Mills NJ，2003．A comparison of methodologies for the assessment of host preferences of the gregarious egg parasitoid Trichogramma platneri．Biological Control，29:332-340．

Monje JC，Zebitz CPW，Ohnesorge．1999．Host and host age preference of Trichogramma galloi and T．pretiosum(Hymenoptera:Trichogrammatidae)reared on different hosts．Journal of economic entomology，92:97-103

Pak GA，1988．Selection of Trichogramma for Inundative Biological Control．GaWsch bedrijf Ponsen en Looijen，Wageningen，Nethrlands．

Parra JRP，2010．Mass rearing of egg parasitoids for biological control programs progress．Biological Control，9:267-292．

Reznik SYa，Umarova TYa，Voinovich ND，1997．The influence of previous host age on current host acceptance in Trichogramma．Entomologia Experimentalis et Applicata，82:153-157．

Roriz V，Oliveira L，Garcia P，2006．Host suitability and preference studies of Trichogramma cordubensis(Hymenoptera:Trichogrammatidae)．Biological Control，36:331-336

Silva IMM，Stouthamer R，1999．Do sympatric Trichogramma species parasitize the pest insect Helicoverpa armigera and the beneficial insect Chrysoperla carnea in different proportions?Entomologia Experimentalis et Applicata，92:101-107．

Smith SM，1996．Biological control with Trichogramma:advances，successes and potential for their use．Annual Review Entomology，41:375-406．

Suzuki Y，Tsuji H，Sasakawa M，1984．Sex allocation and effects of superparasitism on secondary sex ratios in the gregarious parasitoid，Trichogramma chilonis(Hym．，Trichogrammatidae)．Animal Behavior，32:478-484．

Takasu HCJ，Ode PJ，Antolin MF，Strand MR，1997．Environmental and genetic determinants of ovicide in the parasitic wasp Bracon hebetor．Behavioral Ecology，8:647-654．

Tzanakakis ME，2003．Seasonal development and dormancy of insects and mites feeding on olive:a review．Netherlands Journal of Zoology，52:87-224．

Wajnberg E，Hassan SA，1994．Strategies to select Trichogramma species for use in biological control．Biological Control with Egg Parasitoids．Oxon，U．K．:CAB International．55-71．

Wright MG，Kuhar TP，Hoffmann MP，Chenus SA，2002．Effect of inoculative releases of Trichogramma ostriniae on populations of Ostrinia nubilalis and damage to sweet corn and field corn．Biological Control，23:149-155．