Introduction

Insect parasitoids rely mainly on semiochemicals for location and recognition of hosts, and several studies have been published on the role played by various chemical cues in the host selection behavior of hymenopteran parasitoids (Godfray, 1994; Quicke, 1997; Vinson, 1998, 2010; Fatouros et al., 2008). Parasitoids, at the initial steps of the host location process, exploit mainly volatile cues (Group I cues sensu Vinson, 1998, 2010) since they can be perceived at long range distance, whereas slightly volatile cues or contact cues become progressively more important when parasitoids are in the close vicinity (Group II cues) or in contact (Group III cues) with the host (Vinson, 1998, 2010). Experience is particularly important in affecting the initial phases of host selection behavior, especially for generalist species, as experienced females have been reported to respond more efficiently to the host-plant complex than naive females (Drost et al., 1986; McAuslane et al., 1991). However, the role of experience may be important at all levels of the host selection process (i.e., time required to recognize and parasitize a host decreases as the parasitoid handles successive hosts) (Wajnberg, 1989; Segoli et al., 2009).

Wasps in the genus Melittobia are gregariously developing ectoparasitoids that frequently attack solitary wasps and bees, but also attack a wide range of hosts that represent several insect orders (González et al., 2004; Matthews et al., 2009). Melittobia wasps are sexually dimorphic, and females are polymorphic (short winged and long winged) (González and Matthews, 2005, 2008; Matthews et al., 2009). Long-winged females, also called macropterous females, are the dispersal stage in all Melittobia species, while some non-dispersing short-winged females, also called brachypterous females, may be produced on high quality hosts. While several aspects of Melittobia’s behavioral ecology have been studied (reviewed in Matthews et al., 2009; Tanner et al., 2011), the chemical ecology of Melittobia species is poorly understood. The only such study (Cônsoli et al., 2002) elucidated the role of α- and β-trans-bergamotenes as the male-produced sex pheromone in M. digitata Dahms. Nonetheless, chemical cues are likely to be important in allelochemical communication of Melittobia too, as, for example, during the host location process. The antennae of Melittobia females possess numerous mechano- and chemoreceptors (Dahms, 1984; Imandeh et al., 2001), suggesting that chemical cues are important in Melittobia biology.

Macropterous females of Melittobia wasps appear to disperse randomly once out of their natal nest, but when close to host nests, little is known about cues that might direct them to parasitize a healthy host (González and Terán, 2001; Matthews et al., 2009). However, some progress has been made in understanding how Melittobia wasps locate their hosts (Silva-Torres et al., 2005a,b; Cusumano et al., 2010). Cusumano et al. (2010), working with M. digitata, suggest that once the mud dauber wasp host Trypoxylon politum Say (Hymenoptera: Crabronidae), is located the Melittobia female adopts a “sit and wait” strategy relying mainly on indirect host-related volatile cues. Females search for T. politum nests under construction, and once located, they enter the newly built cell and wait inside until the host becomes a prepupa, the stage attacked by the parasitoid. Before such transformation, the mature larva scrapes grains from the cell wall and glues them with saliva and the first released meconial fluids to build its cocoon, and once it is built, the larva releases the remainder of the meconium (Cooper, 1957; Cross et al., 1975; R.W. Matthews pers. comm.). The presence of the cocoon and the final release of meconium are indicators associated with host suitability (Cusumano et al., 2010).

Contact cues are important in host location, host recognition, and also in mate finding of parasitic wasps (Godfray, 1994; Fauvergue et al., 1995; Vinson, 1998; Colazza et al., 1999). The parasitoid response to contact cues usually consists of flight inhibition, an intense tapping of the substrate with its antennae, and a modification of the walking pattern characterized by a reduction in linear speed and an increase in turning rate (arrestment response) (Waage, 1978, 1979). Several host-produced stimuli act as contact cues, such as frass (Loke and Ashley, 1984; Takabayashi et al., 1985; Meiners et al., 1997), honeydew (Vinson et al., 1978; Gardner and Dixon, 1985; Mehrnejad and Copland, 2006), wing scales (Laing, 1937; Noldus and van Lenteren, 1985; Gardner and van Lenteren, 1986; Schmidt and Carter, 1992; Chabi-Olaye et al., 2001), footprints (Colazza et al., 1999; Borges et al., 2003; Conti et al., 2003; Rostás and Wölfling, 2009), and mandibular gland secretions (Waage, 1978).

Chemical investigations have revealed that both polar and non-polar compounds can be the biologically active substances responsible for eliciting wasps’ arrestment behavior. For example, active compounds in honeydew are often water or polar solvent soluble (Vinson et al., 1978; Bouchard and Cloutier, 1984); water-soluble propriety can be important for parasitoids because of rain that can wash away honeydew from area where hosts are no longer available. In contrast, in the case of parasitoids that exploit contact cues produced by the walking traces of the host, kairomones often are hexane soluble, suggesting that cuticular lipids may be involved (Howard and Flinn, 1990; Colazza et al., 2007; Rostás and Wölfling, 2009). Due to similar solubility properties, such compounds can be adsorbed and retained on the wax surface of leaves, and thus, can be easily perceived by the searching parasitoid. The parasitoid’s response to contact cues also can be affected by experience which modifies the innate response to contact stimuli (Peri et al., 2006).

In order to better clarify the role of contact chemical cues in the host-parasitoid system Trypoxylon politum Say - Melittobia digitata Dahms, a series of bioassays were developed to investigate: (1) if M. digitata exploits contact kairomones from the host prepupa and/or host by-products, the cocoon, and the meconium; (2) the solubility and the chemistry of the kairomones; and (3) if experience affects the parasitoid’s response to contact cues.

Methods and Materials

Insect Cultures

Parasitoid wasps of the species Melittobia digitata were reared on naked prepupae of T. politum from stock cultures maintained in the Entomology Research Laboratory of Texas A&M University’s Department of Entomology. Wasp cultures were kept in an incubator at 25°C and 70% RH in total darkness. Only macropterous (long-winged) females less than 5 d-old were used in the experiments. Because of their age, and since they were taken directly from cultures previous to their use in the bioassays, the wasps were presumably mated. About 1 hr before bioassays, wasps were transferred individually to size 1 gelatine capsules (vol: 0.5 ml; diam: 6.63 mm) and tested only once.

Hosts

Prepupae of Trypoxylon politum Say were obtained from nests collected in 2009 (HW 47, West of Bryan, Texas N 30° 36′, W 96° 24′) and kept in a refrigerator (<12°C) to maintain the prepupae in diapause for eventual use as hosts of Melittobia wasps. Previous to use in these bioassays, every cocoon was opened gently at the subtruncate cap (Cross et al., 1975) making a small hole in order to check if each host prepupa was healthy and not parasitized.

Open Arena Preparation

Bioassays with M. digitata females were conducted in an open arena consisting of a rectangular filter paper sheet (120 × 140 mm). At the center of the arena, a circle with 58 mm diam. (2,642 mm2, about 15% of the entire arena) was treated with the stimuli, while the rest of the filter paper was left untreated. We performed bioassays by using a flat surface in order to avoid any possible bias due to physical cues, since shape has been also shown to have a role in the host acceptance of M. digitata (Cooperband and Vinson, 2000). Each arena was used for testing five female wasps; it was then discarded and replaced with a new one. For each different experiment, the following preparation procedure used for bioassays was performed:

  1. (1)

    In vivo prepupa, cocoon, and meconium. Trypoxylon politum prepupae are found protected by cocoons inside cells of mud nests. Each cell chamber contains a single cocoon/host (Fig. 1). Collected hosts were divided into three parts and tested separately: 1) prepupa; 2) empty cocoon; 3) dried meconium originally released by the mature T. politum larvae, and easily scraped from the bottom and inside the cocoon. The portion of cocoon to which the meconium was in direct contact was discarded and not used in the experiments to avoid any possible contamination.

  2. (2)

    Hexane and acetone host-by product extracts. Hexane and acetone extracts were prepared for the following host-by products and tested separately: a) five cocoons with meconium but without prepupae inside (cocoon + meconium); b) five cocoons with no prepupae inside and meconium scraped out. The bottom part of the cocoon that was directly in contact with the meconium was not used in order to avoid any possible contamination with chemicals located on the meconium surface (cocoon only); c) five dried meconium pieces scraped from cocoons (meconium only).

    Since in vivo bioassays (see Results) indicated that the prepupa of T. politum did not elicit a response to M. digitata wasps, we did not test the prepupa extract.

    The cocoon + meconium, cocoon only or meconium only were placed in a 15 ml-glass vial and extracted with 10 ml of hexane or acetone (Sigma-Aldrich) at 28°C for 2 hr. After removal of host by-products, the resulting extracts were evaporated under a gentle nitrogen stream and redissolved in 200 μl of hexane or acetone. After each extract was chemically analyzed, it was stored at −18°C until bioassayed.

  3. (3)

    Reconstructed Blend of Straight-chain Carboxylic Acids. The wasps responded strongly to apolar extracts (see Results). Thus, we investigated only the chemical composition of hexane extracts. Additionally, since the parasitoid’s responses to cocoon and meconium hexane extracts were similar, and because of the high chemical similarities among identified compounds of the cocoon and meconium (see results and Table 1), we decided to use a solution containing an average quantity of each compound present in the cocoon crude extracts. This was done by reconstructing the hexane extract with a mixture of standard carboxylic acids, normalizing the concentration of the carboxylic acids found in crude extract to oleic acid, the most abundant compound. The stock of chemicals was prepared as a hexane solution containing (in mg/l): C18:1(1000), C18:2(618), C18(369), C16(294), C8(262), C16:1(125), C14(124), C7(94), C6(90), C9(30), C12(17), (Sigma-Aldrich) (numbers given in brackets: see Table 1). The working solution was obtained by using 10 ml stock solution and diluting it to 100 ml. The reconstructed blend was stored at −18°C until bioassayed.

Fig. 1
figure 1

Line drawing of Trypoxylon politum wasp mud nest showing the typical position of nest cells and the wasp cocoons inside. The dried meconium is shown inside the cocoons (in broken lines)

Full size image
Table 1 Mean amounts (± SE) of compounds identified from the cocoon and meconium of Trypoxylon politum. and their relative abundance (%)
Full size table

Open Arena Bioassay

A single parasitoid female was gently introduced inside the lid of a small Petri dish (diam = 35 mm) that was placed with the open end over the center of the treated area of the arena, thereby allowing the wasp to acclimatize. After 1 min and verification that the wasp was in direct contact with the arena, the lid was removed and observations of the unrestrained wasps’ behavior were recorded. Recording stopped when the wasp walked off of the arena or after 10 min. For each experiment, the following bioassay procedure was used.

  1. (1)

    In vivo prepupa, cocoon, and meconium. Prepupae, empty cocoons, or only meconium (5 of each) were gently rubbed in the center of the filter paper for about 5 min in order to contaminate the area homogenously with chemicals. Plain filter paper arenas served as controls.

  2. (2)

    Hexane and acetone host-by product extracts. Hexane and acetone extracts of the cocoon + meconium, cocoon only, meconium only were tested at doses of 5 insect equivalent (IE) by pipetting 200 μl extract onto the central area of the arena. The 5 IE dosage was applied by repetitive pipetting of about 50 μl solution, and allowing the solvent to evaporate between repetitions. Filter paper in which the central area was treated with 200 μl of hexane or acetone was used as a control. The solvent was left to evaporate completely (∼5 min.) before observation started.

  3. (3)

    Reconstructed blend of straight-chain carboxylic acids. Five IE of the reconstructed carboxylic acid blend of cocoon were tested by pipetting 20 μl of a hexane solution containing 0.1 mg/ml of the most abundant component (C18:1) and proportional amounts of the minor components onto the treatment areas. Filter paper with the central area treated with 20 μl hexane served as control.

The walking pattern of M. digitata was observed using a video camera (EyeCom, PRO-9000) connected to a screen (Symphonic, 19-inch) in order to monitor the wasp away from the experimental arena and to avoid any possible disturbance effect of the observer on the behavior of the wasp. Meanwhile, data was digitalized and processed by using XBug, a video tracking and motion analysis system. The parameters computed by the system that best defined and described the parasitoid’s behaviors were: (1) total residence time (sec); (2) average linear speed (mm/sec); (3) average angular speed (°/sec); (4) tortuosity index. This index ranges from 0 to 1, with 0 indicating a completely linear track, and 1 the maximum tortuosity. It was computed as follows: T = 1-mp/tl; where T, tortuosity index; mp, maximum projection of the track over the generic line in the plan; and tl the total length of the track. All parameters were processed only for the central area and not for the total arena because we assumed that data from the treated area better characterized the behavior of the searching wasps stimulated by contact chemicals. To minimize cues from the room, the walls were painted black, and the experimental chamber was equipped with red fluorescent light (provided by 8, 40-W fluorescent bulbs covered with red tube guards) in order to keep visibility in the room in accordance to parameters previously established by González et al. (1985) and Matthews et al. (1985). For experiments (1) and (2), 25 replicates were performed, while for experiment (3), 40 replicates were performed. Experiments were carried out in a completely randomized design performing five replicates per experiment/day. The room was maintained at a temperature of 20°C and a RH of 60%. All bioassays were conducted from 10:00 to 12:00 and from 14:00 to 16:00.

Gas Chromatography

GC analyses were performed on a Hewlett Packard 6890 system with 7683 injector. One μl of each sample was injected onto a 0.32 mm × 25 m SGE BP1 column in splitless mode. A temperature program of 50°C was set for 5 min., then 220°C at 10°/min. (5 min). Helium carrier gas at 20 psig was used. Ten samples in hexane of each type (cocoon and meconium) obtained as previously described in section (2) of the “open arena preparation procedure” were used for analysis. Mass spectral data were acquired on a Thermoelectron Trace GC2000 quadrupole GC-MS instrument. The attached GC unit was a Trace GC Ultra (Finnegan) equipped with a 0.25 mm × 25 m SGE BP1 column operated at constant flow of 1 ml/min, using the same temperature gradient. Spectra were acquired in total ion current mode over a range of 40–500 amu with 70 eV electron impact ionization. Peak analysis was by comparison of acquired mass spectral data with that in the NIST 2001 libraries. Samples of all tentatively identified acids and of their methyl esters were purchased from Sigma-Aldrich chemical company, and retention time and mass spectra of these compounds were used for confirmation. Samples of acids then were prepared for bioassay. Due to poor peak symmetry of carboxylic acids, methyl ester derivatives were used for quantification. To obtain the methyl esters, 100 μl of crude insect derived extract for analysis, or 100 μl of a fatty acid mixture containing 1–10 μg fatty acids for standards were placed in a 2 ml HP screw cap vial, and 100 μl at 10% of BF3/MeOH were added to the vial, which was placed in heater block at 60° for 30 min., removing to shake well at 10 min. intervals. Then, 1 ml of water and 800 μl of hexane were added, shaking well and allowing layers to separate. The upper layer was removed carefully with a disposable pipette and filtered through anhydrous sodium sulfate, around 2 cm in a disposable pipette held in place by a small plug of Kimwipes, collecting the filtrate in a 2 ml crimp cap vial.

Detection Limits

The limit of detection (LOD) of 1 ng/μl was determined by measuring progressively more dilute concentrations of the identified methyl esters until a signal-to-noise ratio of 3:1 was reached. Means and standard error for amounts of each compound in each sample type were calculated based on ten replicates (Table 1).

Effect of Experience

To test if there was any influence of the experience elicited by exposing wasps to host by-products (cocoon or meconium) or to the host itself (prepupa) on the parasitoid’s arrestment response, wasps less than 5 d-old were isolated together with the source of the stimulus into small glass vials (1 dram) with tight cotton stoppers and kept for 24 hr prior to bioassays; in the case of exposure to prepupa, M. digitata initiated host feeding and oviposition after 2–6 hr; such parasitoids were experienced with respect to oviposition. The response of cocoon-experienced wasps, meconium-experienced wasps, or prepupa-experienced wasps were tested subsequently on filter paper arena contaminated, as in vivo bioassays, with contact chemicals of the cocoon, meconium, or prepupa, respectively. As controls, we used naïve females of the same age that were tested in arenas contaminated with chemicals of the cocoon, meconium, or prepupa, respectively, in order to make paired comparison between experienced and naïve wasps. For each set of bioassays 25 replicates were performed.

Statistical Analysis

Data were not significantly different from a normal distribution (Shapiro-Wilk test P = 0.05). Data from the in vivo bioassays, from host by-product extracts, and from reconstructed blends of carboxylic acids were analyzed by one-way ANOVA followed by Tukey’s HSD post hoc test for comparisons between means. Comparisons between naïve and experienced female wasps were done by using a t-test. The amounts of each carboxylic acid extracted from cocoon and meconium also were compared by using a t-test. All statistical analyses were performed by using Statistica 6.0 for Windows (StatSoft Inc., Tulsa, OK, USA).

Results

For all experiments with control conditions, wasp residence time in the arena was short, and the tortuosity index was low. Control wasps spent little time in the central area, exploring at most only small sections of the arena’s surface (Fig. 2a). With the different treatment conditions, in those experiments that elicited a behavioral response of wasps, there was increased residence time in the central treated area compared to controls. In general, wasps spent more time within the contaminated (treated) area and performed a more convoluted locomotion pattern, thus resulting in a more complete exploration of the arena (=arrestment response) characterized by a high tortuosity index. Females often were observed turning back to reenter the treated area after exceeding its boundaries (Fig. 2b).

Fig. 2
figure 2

Examples of two recorded paths (A+B) of Melittobia digitata in an open arena consisting of a rectangular filter paper sheet (120 × 140 mm) (central circle area 58 mm diam.). a Untreated central area. b Central area rubbed with Trypoxylon politum cocoons. S start; E end

Full size image

In vivo Prepupa, Cocoon, and Meconium

Residence time in the central area was higher in the arena contaminated with host by-products compared to those contaminated with either the host prepupa or the control (F = 10.59; df = 3,96; P < 0.001) (Fig. 3a). In fact, no significant difference was found between time (mean ± SE) spent by wasps on the prepupa-contaminated arena (59.91 ± 22.11 sec) and time spent on the control (27.21 ± 6.40 sec). Linear speed was lower on patches contaminated by cocoon and by meconium compared to patches contaminated by prepupa or the control (F = 46.15; df = 3,96; P < 0.001) (Fig. 3b). Higher values of angular speed of parasitoid were found in the arena contaminated by prepupa compared to those contaminated by host by-products or control (F = 21.23; df = 3,96; P < 0.001) (Fig. 3c). Tortuosity was higher on the patches contaminated by host by-products compared with the control, and intermediate in those contaminated by prepupa (F = 30.12; df = 3,96; P < 0.001) (Fig. 2d).

Fig. 3
figure 3

Various measures (mean±SE) of Melittobia digitata responses to contact chemicals from rubbed cocoon, rubbed meconium, and rubbed prepupa of Trypoxylon politum. Untreated arena served as control. Columns with the same letter are not significantly different at P < 0.05 (ANOVA, Tukey tests)

Full size image

Hexane Host By-Product Extracts

Residence time was higher in the arenas contaminated with cocoon + meconium extract, cocoon extract, and meconium extract compared to the control (F = 4.06; df = 3,96; P = 0.009) (Fig. 4a). Linear speed was lower on patches contaminated with cocoon + meconium extract compared to the other tested conditions or the control (F = 6.67; df = 3,96; P < 0.001) (Fig. 4b). Higher values of angular speed of parasitoid were found in the arena contaminated by cocoon extract compared to the other tested conditions or control (F = 9.49; df = 3,96; P < 0.001) (Fig. 4c). Tortuosity was higher within the patches contaminated by extract of cocoon + meconium, cocoon, and meconium compared to the control (F = 7.30; df = 3,96; P < 0.001) (Fig. 4d).

Fig. 4
figure 4

Various measures of Melittobia digitata responses to hexane extracts (above) and acetone extracts (below) of cocoon + meconium, cocoon only, and meconium only of Trypoxylon politum. Arena treated with hexane or acetone served as control. Columns with the same letter are not significantly different at P < 0.05 (ANOVA, Tukey tests)

Full size image

Acetone Host By-Product Extracts

Central area residence time was higher for arenas contaminated with cocoon + meconium extract compared to the other host by-product extracts or control (F = 5.96; df = 3,96; P < 0.001) (Fig. 4a 1). There were no significant differences in the linear speed of the parasitoid on host by products contaminated patches over the control (F = 2.05; df = 3,96; P = 0.112) (Fig. 4b 1). Higher values of the angular speed were found for arenas treated with extract of cocoon + meconium or cocoon compared to those contaminated by meconium extract or controls (F = 14.95; df = 3,96; P < 0.001) (Fig. 4c 1). Tortuosity was higher on the patches contaminated by extract of cocoon + meconium or cocoon compared to control (F = 8.92; df = 3,96; P < 0.001) (Fig. 4d 1). However, no statistically significant differences were found between cocoon extract and meconium extract nor between meconium extract and control (Fig. 4d).

Reconstructed Blend of Straight-chain Carboxylic Acids

The mix of linear carboxylic acids (test) elicited a weak but significant response in terms of arena residence time compared to the control (F = 9.32; df = 1,78; P = 0.003). In fact, time (mean ± SE) spent by the wasps on the test arenas was 76.72 ± 13.77 sec, while the time spent on the control arenas was 32.29 ± 4.81 sec.

Chemical Analyses

Twelve detected compounds, all linear carboxylic acids, were identified in the host by-product hexane extracts of T. politum. The five most abundant compounds were octanoic (cocoon[c] = 8.7%; meconium[m] = 12.2%), palmitic (9.7%c; 8.5%m), linoleic (20.5%c; 22.5%m), oleic (33.1%c; 30.5%m),and stearic acid (12.2%c; 14.4%m) representing 84.2% and 88.1% of the total carboxylic extract of cocoon and meconium, respectively. Qualitative differences were found, since decanoic acid was only detected in small quantities from meconium extract, and quantitative differences also were found because, for all carboxylic acids, excluding myristic acid, the relative amount was significantly higher for meconium extract compared to the cocoon extract; however, the compounds were present in similar ratios (Table 1) (Fig. 5).

Fig. 5
figure 5

Representative gas chromatograms of compounds extracted from cocoon and meconium of Trypoxylon politum with hexane. For peak identities, see Table 1; “Methods and Materials” describe analysis conditions

Full size image

Effect of Experience

Central arena residence time was not significantly influenced by exposure of wasps to host by-products (cocoon: t = 0.677; df = 48; P = 0.501; meconium: t = 0.174; df = 48; P = 0.862) nor to exposure to the host itself (prepupa: t = −0.878; df = 48; P = 0.384); even though experienced parasitoids increased the time (mean ± SE) spent on an arena contaminated by prepupa chemicals (naïve = 59.91 ± 22.11 sec; experienced = 93.85 ± 31.70 sec), no statistical differences were found between values of naïve vs. experienced wasps; similar values were found for the time spent by naïve and experienced wasps on patches contaminated with host by-products (cocoon: naïve = 143.86 ± 22.50 sec; experienced = 120.87 ± 25.40 sec; meconium: naïve = 132.49 ± 12.88 sec; experienced = 129.38 ± 12.27 sec).

Discussion

The results of the in vivo bioassays showed clearly that M. digitata was arrested by contact kairomones from the host by-products (cocoon, meconium) providing the first evidence that contact cues are exploited by Melittobia wasps. Surprisingly, the host stage attacked, T. politum prepupa, did not induce the parasitoid to spend significantly more time in an arena treated with chemicals in their exocuticle when compared to the control. Contact chemicals located in the host cuticle are cues always associated with the host, but M. digitata seems not to exploit them during the host selection process. Such behavior may be explained by considering that the host selection process is divided into a series of hierarchical steps in which each step is influenced by the previous one: the parasitoid, under natural condition, is expected to find a prepupa after having found a cocoon and contacted the meconium; hence, the lack of occurrence of the host by-product cues may explain the non-response to contact chemicals from the prepupa since cocoon and meconium are always associated with the presence of a suitable host. In addition, it also is possible that physical or tactile cues play an important role when parasitoids first contact a potential host as shown by Cooperband and Vinson (2000); chemical information could be secondary or redundant to physical cues.

The bioassays conducted with polar and non-polar extracts revealed that M. digitata responded strongly to hexane extract of cocoon + meconium, cocoon only, and meconium only, suggesting that the nature of the kairomones is mainly non-polar soluble. Since the response of the parasitoid was similar when the hexane extract of the cocoon or the hexane extract of the meconium was tested, we assume that contact kairomones of these host-by products share similarity. Indeed, the cocoon and the meconium have the same origin, and they are both produced by the mature larva just prior to its transformation to the prepupa (which is the suitable stage for parasitoid offspring development). GC and GC/MS analysis of cocoon and meconium extracts showed the presence of linear carboxylic acids (C6–C18). Cocoon and meconium extracts contained almost identical compounds in similar proportion (the most abundant compounds were: octanoic acid; hexadecanoic acid; (Z,Z)-9,12-octadecadienoic acid; (Z)-9-octadecenoic acid; octadecanoic acid).

Fatty acids are common compounds in nature that play roles in various ecological contexts, including the host location process of some parasitoids (Renou et al., 1992; Horikoshi et al., 1997). However, when the reconstructed blend of fatty acids was tested, M. digitata showed a weak response compared with the control; hence, even if carboxylic acids play a role, probably some other unidentified compounds present in small quantities are also involved. Many studies show that synthetic blends have only low activity when compared to the activity of crude extracts, thus supporting the presumed high complexity of contact kairomones. For example, the egg parasitoid Trissolcus basalis (Wollaston) weakly responded to a reconstructed blend of straight-chain hydrocarbons compared to hexane extracts of other host cuticular lipids (Colazza et al., 2007); the larval parasitoid Cotesia marginiventris Cresson weakly responded to reconstructed blends of linear n-alkenes compared with hexane extracts of the host larvae footprint and of the larval ventral cuticle (Rostás and Wölfling, 2009).

Melittobia digitata can successfully parasitize a large variety of hosts that belong to different insect orders. In general, the less specialized a parasitoid is, the more likely that experience can affect its behavior (Vet and Dicke, 1992). In our experiment, no difference was found between experienced and naïve wasps regardless if they were experienced on a cocoon, on meconium, or on prepupa. In the latter case, M. digitata was able to host feed and even oviposit on the host, resulting in rewards experienced by the wasps, which could act as strong reinforcing stimuli. We expected that prepupa-experienced wasps would spend significantly more time compared with naïve wasps when tested on filter papers arena contaminated with contact cues of the prepupae. However, this did not occur. Unlike many other parasitoids that attack and parasitize multiple hosts, thereby gaining repeated oviposition experience, Melittobia females usually locate and parasitize (and/or superparasitize) only a single host; thus, prior host experience would not be selectively favored. Silva-Torres et al. (2005b) reared M. digitata on 3 different hosts: T. politum, Sarcophaga bullata (Parker), and Megachile rotundata (Fabricius), and found that the female’s natal host did not influence host choice. This suggests that pre-imaginal experience is irrelevant for M. digitata wasps as well.

Cusumano et al. (2010) stressed the importance of volatile cues emitted by host by-products of T. politum in the host location process of M. digitata macropterous females. The results of our study indicate that once M. digitata is within the host cell, host acceptance is mediated by chemicals of low volatility that arise following host meconium release and cocoon construction.