Introduction

Insect parasitoids find their hosts via a series of behavioral steps, which are mainly mediated by chemical cues directly or indirectly related to the host (Vinson 1998). Chemical stimuli lead the wasps in a hierarchical process to the host habitat, to the host plant, and finally to the susceptible host stage (Vinson 1985; Quicke 1997). The fitness of parasitoids strongly depends upon their ability to find hosts quickly, before hosts become unsuitable as a consequence of their fast development (Vet and Dicke 1992; Godfray 1994). This is particularly true for egg parasitoids, which must oviposit into recently laid eggs before embryo development renders them unsuitable as hosts (Strand 1986; Vinson 1998). To find host eggs quickly, egg parasitoid females can exploit semiochemicals arising from the host eggs, from the interaction of the plant and the eggs, or from stages other than the one attacked (Powell 1999; Fatouros et al. 2008). This last group of semiochemicals, termed indirect host-related cues, does not provide information on the location or the age of host eggs, but brings female wasps in proximity to potential host eggs. The adaptative value of indirect host-related cues is strongly related to their temporal and/or spatial correlation to oviposition sites (Colazza et al. 1999, 2007). Chemical footprints left behind by true bugs (Hemiptera: Heteroptera) on a substrate are indirect host-related contact cues for some Trissolcus and Telenomus species (Hymenoptera: Scelionidae) (Colazza et al. 1999; Borges et al. 2003; Conti et al. 2004; Salerno et al. 2006). Once in contact with a contaminated area, female wasps show an initial prolonged motionless period with the antennae kept in contact with the surface (arrestment response), followed by an intense searching behavior characterized by variation in orthokinesis and klinokinesis, and increased turning frequency. If no host eggs are found, this behavioral syndrome slowly decays, and the normal walking pattern is gradually resumed (Colazza et al. 1999; Conti et al. 2004; Peri et al. 2006; Dauphin et al. 2009). Recently, investigation of the chemistry of the footprint kairomone demonstrated that females of Trissolcus basalis (Wollaston) are stimulated by compounds present in the cuticular lipids of Nezara viridula (L.) (Heteroptera: Pentatomidae). Furthermore, the presence or absence of nonadecane (nC19) allowed female wasps to distinguish between residues left by male or female bugs, respectively (Colazza et al. 2007). To date, the role of chemical footprints of true bugs as contact kairomones for scelionid egg parasitoids has been investigated using open arenas made from filter paper (Colazza et al. 1999; Borges et al. 2003; Conti et al. 2004; Peri et al. 2006; Salerno et al. 2006). This method avoided potentially confounding or masking effects from the presence of leaf morphological features such as trichomes and veins (Obrycki 1986; Andow and Prokrym 1990; Romeis et al. 2003), but inadequately mimicked the role of the intact plant cuticle as the substrate for parasitoid-host interactions in nature (Rostás et al. 2008). For example, experiments using artificial substrates with egg parasitoids exclude the potential capacity of plant surfaces to adsorb and release host contact kairomones and do not address the female wasp’s ability to distinguish the components of the contact kairomone from similar components present in the epicuticular waxes of foliage and other plant tissues (Colazza et al 2007).

Epicuticular waxes form the outermost boundary covering the cutin matrix of higher plants. The waxes are usually present as a film, often proliferated as microscopic aggregate protrusions, named crystals, with a variety of forms (Barthlott et al. 1998). Epicuticular waxes represent the true surface of the aerial plant organs and regulate many plant physiological processes, e.g., non-stomatal water loss, protection against UV light, and gas exchange (Baker 1982). In addition, the epicuticular waxes play an essential ecological function in mediating the interactions with other organisms, such as insects and pathogens (recently reviewed by Müller and Riederer 2005). In particular, several experiments have demonstrated the role of plant cuticular lipids in reducing attachment and mobility of both insect herbivores and natural enemies (Espelie et al. 1991; Mc Auslane et al. 2000; Eigenbrode and Jetter 2002; Gentry and Barbosa 2006). Recent studies have also demonstrated a role for plant epicuticular waxes in supporting or functioning as semiochemicals for insect predators and parasitoids (Noldus et al. 1991; Eigenbrode 2004, Nakashima et al. 2004; Müller and Riederer 2005; Rostás et al. 2008; Rostás and Wölfling 2009).

In the current study, we observed responses of T. basalis females to footprints of N. viridula females on leaves of broad bean plants, Vicia faba L., with an intact epicuticular wax layer, and on leaves that had been dewaxed. Then, untreated leaf samples and dewaxed leaf samples were observed by scanning electron microscopy (SEM).

Methods and materials

Insect cultures

A colony of N. viridula was established from locally collected insects, reared in wooden cages (50 × 30 × 35 cm) with mesh covered holes (5 cm in diameter) and fed with sunflower seeds and seasonal fresh vegetables. Separate cages were used for rearing nymphs and adults. At every 2–3 days, food was replaced, egg masses collected, and newly emerged N. viridula adults isolated in different rearing cages so that individuals of known age were constantly available for the experiments.

A colony of T. basalis was established with adults obtained from N. viridula parasitized eggs collected from fields around Palermo. Adult parasitoids were reared on N. viridula eggs in 16-ml glass tubes and were fed with a 10% honey–water solution streaked on the inside surface. Three or four female wasps were used to parasitize each newly laid egg mass of N. viridula; after 3 days, parasitoids were removed, and parasitized egg masses were held in a rearing room until adult emergence. Both colonies were maintained in a rearing room at 24 ± 1°C, 70 ± 5% relative humidity, and 16 L:8D photoperiod and were refreshed periodically with field-collected materials.

Plants

Broad bean plants, V. faba var. Lux de Otoño, were grown in greenhouse conditions from seeds planted individually in 14-cm diameter plastic pots, fertilized with commercial soil (Trflor—HOCHMOOR) and watered as needed. Plants, 4–5 weeks old with at least four fully expanded leaves (each leaf on average 12.5 ± 4.9 cm2) were then transferred into a climate-controlled chamber (24 ± 2°C, 70 ± 5% RH, 16 L:8D) and used for the experiments.

Plants contaminated by footprints of N. viridula females

A single broad bean plant was kept inside a wooden cage (50 × 30 cm) with mesh-covered holes (10 cm in diameter) and contaminated by two to three mated N. viridula females that were allowed to walk over the leaves for about 30 min. Then, bugs were removed, and the plants were transferred into a climate-controlled room for the bioassays. Plants with leaves contaminated by bug’s feces were not used for the experiments. Noncontaminated plants were used as a control.

Mechanical removal of epicuticular waxes

The epicuticular waxes were mechanically removed from broad bean leaves using an aqueous solution of gum arabic (Sigma-Aldrich) as described by Jetter and Schäffer (2001). Approximately 0.05 g of a 50% (w/w) aqueous solution of adhesive was applied per cm2 of adaxial leaf surface using a small paintbrush and left to air dry at room temperature for about 2 h. Then, the resulting polymeric film was carefully stripped off with forceps. Dewaxed plants were then transferred into a climate-controlled room for bioassays. Small pieces of the thin polymeric films, peeled off the adaxial surface of the leaves of both contaminated and noncontaminated broad bean plants, were stored in a refrigerator at 4–6°C until being used for bioassays.

Bioassay procedures

All experiments were carried out between 1000 and 1500 hours in a climate-controlled room at 25 ± 1°C and 55–70% RH with fluorescent lighting. For all experiments, <6-day-old, mated, naïve female T. basalis were used. Each female wasp was placed in a 2-ml glass vial about 24 h before being used in experiments. The day of the experiments, wasps were allowed to acclimatize in the bioassay room for about 1 h prior to the bioassays. At the start of each experiment, an individual female parasitoid was released onto the middle of an adaxial leaf surface by gently tapping the vial, and its behavior was observed manually. A wasp was scored as “responding” when it displayed a typical arrestment response, with the body remaining motionless, while rubbing the plant surface with the antennae immediately after contacting the leaf surface (see Colazza et al. 2007). Females that did not assume an arrestment posture after touching the treated area were recaptured with the same vial, and retested after ∼1 min. A wasp was scored as “not-responsive” after three unsuccessful trials. Each broad bean plant was used for testing five wasp females. In a first set of experiments, plants with intact wax layers were used (1) without host chemical contamination, or (2) contaminated by host female footprints. Thirteen replications were carried out for each treatment. In a second set of experiments, dewaxed plant leaves were tested (3) without host chemical contamination, or (4) dewaxed a few minutes after being contaminated, or (5) dewaxed about 30 min before being contaminated by host female footprints. Thirteen replications were carried out for each treatment. In a third experimental set, female wasps were tested on the side of the gum arabic film that had been in contact with (6) the adaxial leaf surface of plants without host chemical contamination, and (7) to leaves contaminated by host footprints. Pieces of the polymeric films were peeled and mounted together, with the side that retained the wax layer up, to cover the bottom of a 5-cm diameter Petri dish. Each Petri dish with the polymeric film was used for testing five female wasps. Twenty replications were made for each treatment.

Scanning electron microscopy

Samples of about 1 cm2 of broad bean leaf were mounted on aluminum holders by double-sided adhesive tabs (ProSciTech). Half of each sample was dewaxed with gum arabic as described above while the other half was left untreated. Specimens were slowly allowed to dry overnight at 20–25°C, so that the wax microstructure could be examined at high resolution (Pathan et al. 2008). Furthermore, small portions of the polymeric film peeled from leaf surfaces were mounted with the side that had been in contact with the leaf surface upwards on an aluminum holder. All specimens were then sputtered with gold (20 mÅ × 300 s, Edwards Sputter Coater, S150A) and observed by SEM (FEI Quanta 200F, Holland).

Statistical analysis

The total numbers of responding and non-responding T. basalis females for each treatment were compared with Pearson χ 2 tests, and Goodman’s post hoc procedure was used for internal contrast among different treatments (Siegel and Castellan 1988).

Results

Wasp behavior

The presence of epicuticular waxes on the adaxial leaves of broad bean plants significantly affected female wasp responses to footprint kairomones of female N. viridula (χ 2 = 62.82, df = 5, P < 0.0001; Fig. 1). In trials with an intact adaxial wax layer, about 77% of T. basalis females showed no arrestment response on plants without host female chemical contamination (χ 2 = 4.59, df = 1, P = 0.032), whereas on contaminated plants, about 70% of observed females displayed typical arrestment responses (χ 2 = 3.45, df = 1, P = 0.041). In trials with dewaxed leaves, most T. basalis females did not respond to plants contaminated before de-waxing or contaminated after the waxes had been removed (χ 2 = 9.32, df = 1, P = 0.002; and χ 2 = 4.54, df = 1, P = 0.032, respectively). The mechanical removal of epicuticular waxes per se did not affect wasps’ responses because the frequency of nonresponding wasps was similar to the frequency observed in the treatments with plants with intact waxes layer and without host chemicals (Z = 1.01, P > 0.05) (Fig. 1). In trials using the wax layer adhering to polymeric film peeled from uncontaminated plants, the wax layer induced arrestment in about 15% of the observed wasps. In contrast, the wax layer adhering to the polymeric film peeled from leaves of contaminated plants induced arrestment in about 95% of the tested wasps, with a frequency that was not statistically different from the frequency observed on plants with intact wax layer and contaminated by host chemicals (Z = −2.29, P > 0.05; Fig. 1).

Fig. 1
figure 1

Effects of adaxial leaf epicuticular waxes of V. faba on T. basalis response to N. viridula female footprint contact kairomone. Wasps were scored as “responding” when after being released displayed the arrestment response; those that did not show arrestment response were scored as “not-responding.” Plants of broad bean were contaminated by allowing two to three mated, gravid N. viridula to walk over them for 30 min. Leaves were mechanically dewaxed by an aqueous solution of gum arabic. Replicates 1 N = 30, 2 N = 20. Columns with the same letters were not significantly different at P < 0.01 (Pearson χ 2 test)

Full size image

Scanning electron microscopy

Epicuticular waxes occurred as a film, densely crystallized into irregularly shaped platelets with spherical granules randomly distributed among them (major and minor axis of the spherical surface mean µm ± SD = 3.56 ± 0.64 and 2.56 ± 0.44, N = 5, respectively; Fig. 2a). At higher magnifications, some of the crystals appeared hollow (Fig. 2b). Scanning electron micrographs showed a clear border between the untreated and dewaxed areas; therefore, a single application of gum arabic seems able to remove substantial quantities of the epicuticular wax film and all the wax crystals from the adaxial leaf surface (Fig. 2a). The dewaxed area appeared to be a smooth cuticular surface (Fig. 2a). The film peeled from the adaxial leaf surface had an undulating topography, probably as a result of the underlying epidermal cells (Fig. 2c).

Fig. 2
figure 2

Scanning electron micrographs. a The adaxial V. faba leaf surface with the epicuticular waxes partly removed after a single treatment with gum arabic revealing the smooth surface of the cutin matrix. The line between the dewaxed area (bottom half of the picture) from the neighboring untreated area is highlighted by arrows. b Detail of the thin film of the epicuticular waxes sparsely covered by spherical granules which some of them appear hollowed. c Underside surface of gum arabic film after it has been applied to the adaxial broad bean leaf surface and peeled off. Single granules present in the upper side of the film are evidenced with arrows

Full size image

Discussion

The influence of plant surface on the efficacy of insect parasitoids has been mostly evaluated in terms of interference with wasp attachment and mobility, and it has been proposed that biocontrol success could be enhanced by development of plants with reduced leaf waxes (Eigenbrode and Espelie 1995; Romeis et al. 2003; Eigenbrode and Kabalo 1999; Rutledge et al. 2003; Chang et al. 2004; Eigenbrode 2004). Nevertheless, normal and dewaxed plants can be different in the composition of epicuticular waxes, so that the role of waxes or wax components in functioning as semiochemicals for insect predators and parasitoids could be different (Eigenbrode et al. 1998). For example, the efficacy of insect parasitoids could be positively influenced by plants with specific physiochemical properties of leaf waxes that permit the adsorption and release of volatile and/or contact host kairomones (Müller and Riederer 2005). Such indications are corroborated by the results presented in this study, which demonstrated that chemical footprints left behind by N. viridula females are perceived by foraging T. basalis females as a result of their absorption onto epicuticular waxes of V. faba. Hence, shortly after landing on a plant contaminated by chemical cues associated with the presence of a host egg, parasitoid females engage in a characteristic sequence of foraging behaviors. The effect of broad bean epicuticular waxes on semiochemical communication between N. viridula and T. basalis resembles the effects recorded for other host–parasitoid systems previously investigated (Noldus et al. 1991, Rostás et al. 2008). For example, the diurnally foraging females of Trichogramma evanescens Westwood can detect the volatile pheromone produced by their nocturnal host Mamestra brassicae (L.), due to its adsorption and subsequent release by the plant waxes of Brussels sprouts for up to 24 h (Noldus et al. 1991). The larval parasitoid Cotesia marginiventris Cresson can detect contact chemical footprints produced by walking second-instar caterpillars of its nocturnal host Spodoptera frugiperda Smith, as they are adsorbed on the wax surface of host plant (Rostás et al. 2008).

The influences of V. faba cuticle in multitrophic interactions have been reported for other plant–insect–parasitoid associations. For example, adults of the black bean aphid, Aphis fabae Scopoli, use broad bean epicuticular waxes as cues in the early stage of host–plant selection (Powell et al. 1999). Females of the parasitoid Aphidius ervi Haliday avoid broad bean leaves previously visited by adults of the predator Coccinella septempunctata L. (Nakashima et al. 2004). Micromorphological and chemical analyses of epicuticular waxes are needed to better clarify the interactions between V. faba cuticle and exogenous cues. The role played by plant surface waxes in the effectiveness of entomophagous insects has been investigated mostly using plants with reduced waxes caused by genetic mutations (White and Eigenbrode 2000; Chang et al. 2004; Eigenbrode 2004; Rostás et al. 2008, Rutledge et al. 2008). The epicuticular waxes can be removed from the plant using a range of adhesive methods while keeping the plant cutin intact. These methods have been used for chemical and micromorphological investigations of plant surfaces (Koch and Ensikat 2008; Pathan et al. 2008), and also to investigate the role of cuticular waxes in affecting host–plant selection by various herbivorous insects (Müller and Hilker 2001; Powell et al. 1999). However, to our knowledge, this was the first time that the removal of epicuticular waxes by gum arabic was used to investigate the absorption or dilution of semiochemicals into plant waxes. Normally, to exhaustively remove all epicuticular material, repeated applications of gum arabic are required (Wen et al. 2006). Normally, to exhaustively remove all epicuticular material, repeated applications of gum arabic are required (Wen et al. 2006). In our experimental conditions, SEM investigations showed that a single gum arabic treatment removed all the epicuticular wax crystals from the adaxial leaf surfaces of V. faba, and this removal was sufficient in preventing the adsorption of the chemical footprints left by N. viridula females. Therefore, these results suggest that host chemical footprints are retained mainly by wax crystals and not by the part of the underlying wax film that most likely remained after the removal of the gum arabic coating. To corroborate this evidence, further experiments to investigate the composition of V. faba wax crystals separated from the wax film are necessary. The chemical composition of broad bean epicuticular waxes has been investigated by dissolution of the wax layer with organic solvents (Kolattukudy and Walton 1972; Powell et al. 1999; Griffiths et al. 1999). It was found that V. faba had a characteristic wax composition rich in hydrocarbons and alcohols (Powell et al. 1999). Interestingly, the linear alkane nonadecane was not found; thus, it is conceivable that female wasps can detect the presence of nC19, a sex-specific component of the cuticular hydrocarbons of male N. viridula (Colazza et al. 2007), to efficiently discriminate between residues left on V. faba leaves by host males or host females. The interpretation of broad bean epicuticular waxes in terms of the polarity of wax constituents could allow investigation of the mechanisms by which T. basalis can discern host chemical footprints from the waxy background of the plant cuticle. Recent research found that cuticular lipids of N. viridula females contain long chain and high molecular weight glycerides (Lo Giudice 2009). Moreover, the most active fraction of N. viridula body extracts in inducing the arrestment posture in T. basalis females included compounds with polar functional groups such as free alcohols and free fatty acids (Lo Giudice 2009). This shows that female wasps may use both polar and apolar components present in host footprint and that both of them can be absorbed by V. faba waxy components.

In conclusion, broad bean plants might represent an ideal plant species for foraging T. basalis females because their wax film may have minimal effect on wasp locomotion, but may serve as an excellent substrate for the host footprint kairomone. However, large differences in chemical composition and microstructure of the waxes and the cutin network have been found in different plant species. More, N. viridula is a highly polyphagous herbivore distributed worldwide (Todd and Herzog 1980), and so future work should examine the role of epicuticular waxes of other plants attacked by N. viridula to determine the extent of plant waxes in general on foraging females of T. basalis.