Hydrocarbon Signatures of the Ectoparasitoid Sphecophaga vesparum Shows Wasp Host Dependency

Sphecophaga vesparum often parasitizes nests of vespid wasps such as Vespula vulgaris and Vespula germanica. Inside the colonies, the ectoparasitic larvae feed on the immature forms of the wasps. There are two adult forms of S. vesparum. The large, winged adults emerge from either rigid yellow cocoons or the orange cocoons used for overwintering. The small, brachypterous females emerge from soft, white cocoons. The species is facultative deuterotokous, producing mostly parthenogenic females and infrequently producing males. Here, we describe the production of chemical compounds related to the different developmental forms of the parasitoid S. vesparum (larvae, pupae and adults). We also compare the chemical profiles of the parasitoid wasp adults to those of their two main host species, Vespula vulgaris and Vespula germanica. The results show differences in hydrocarbon composition of larvae, pupae and adults of S. vesparum. Our results also suggest a partial mimicry of each of the two host species, mostly relating to linear alkanes present in both parasitoids and the host vespid wasp species. This matching is likely due to the recycling of the prey’s hydrocarbons, as has been found in other species of parasitoids.


Introduction
Social parasites often deceive their host species using different strategies: chemical cues, mimicry, camouflage, chemical insignificance, crypsis, usurpation and weaponry [1,2]. Although multiple strategies can be employed to mimic different classes of pheromones, most of the studied interactions between arthropod associations in social insect colonies have been based on hydrocarbon mimicking [3,4]. Cuticular hydrocarbons (CHCs) have a primary function to protect against desiccation, but have acquired a communicative function in social insects. This function is the most studied mechanism used in nestmate recognition [5]. These CHCs have also recently been shown to function as queen pheromones [6]. Obligate parasites have evolved several methods to avoid olfactory CHC detection by their hosts. Strategies used by parasites can include producing low concentrations of recognition cues, demonstrating chemical insignificance, or copying chemical profiles of their host either actively or passively (chemical mimicry). In order to avoid host detection, obligate parasites often express low concentrations of recognition cues, are chemically insignificant, or copy profiles of hosts from queens or workers by chemical mimicry [1,2,[7][8][9]. Whether the chemical mimicry strategy is active or passive is difficult to determine. It may be that the parasitoid is using active mimicry, where the parasite biosynthesizes the host hydrocarbon composition, or it may be that the mimicry is passive, where the parasite acquires CHC composition through contact with the host itself or nest material [2,10].
The arthropods associated with social wasps are the least studied group when compared to arthropods that live together with other social insects, such as ants or termites [4,9]. An example of the complex chemical ecology that can occur within the social insects and their visitors is the aphidiid wasp, Lysiphlebus cardui, that parasitizes the aphid, Aphis fabae cirsiiacanthoidis, and uses chemical cues to avoid aggressive behavior from the ants, Lasius niger, attending the aphids [11]. In honeybees, some work has been done using the ectoparasite Varroa destructor, showing that mites can adjust their chemical profiles depending on the host, either Apis mellifera or Apis cerana, to avoid detection [12]. Research has also shown colony-specificity in Apis mellifera [13]. In paper wasps, the social parasites Polistes atrimandibularis show lower concentrations of CHC's, enabling them to go undetected by the host species Polistes biglumis [7]. The parasitic beetle Metoecus paradoxus uses chemical mimicry to resemble some of the hydrocarbons that occur in the host species, and are frequently found in nests of Vespula vulgaris [8].
Like the beetle M. paradoxus, Sphecophaga vesparum vesparum (Curtis) (Hymenoptera: Ichneumonidae) is a social parasitoid that exploits vespid wasp nests [14][15][16]. The S. vesparum larvae feed as an ectoparasitoid on the newly pupated forms of the wasps. There are two adult forms of S. vesparum: winged adults, which emerge from either thin yellow cocoons or thick yellow cocoons for overwintering, and brachypterous females, which emerge from white cocoons [14]. The species is facultative deuterotokous, where females and males can be produced without sexual fertilization of the egg [14], although males are less frequently found. In wasp species, S. vesparum seem to be specific to the subfamily Vespinae [17], and can be especially abundant in nests of Vespula vulgaris and Vespula germanica. This parasitoid is even used as biological control of wasp populations in invasive ranges in Australia and New Zealand [17][18][19]. To date, there has been no published characterization of cuticular compounds of S. vesparum, which could provide information to assist in developing alternative strategies of wasp population control.
In this study, we characterized the hydrocarbon profiles of Sphecophaga vesparum, V. germanica and V. vulgaris to determine whether the parasitoids were chemically similar to their two most common host species. We also compared chemical profiles between the parasites from different host species of wasp that were collected, and investigated whether the presence of chemical compounds could be related to the different developmental forms of the parasitoid S. vesparum.

Collection of Specimen
We analyzed 47 samples in total (4 larvae of S. vesparum from V. germanica nest; 3 pupae of S. versparum from V. germanica nest; 5 big winged adults of S. vesparum from V. germanica nest; 5 small brachypterous adults of S. vesparum from V. germanica nest; 6 workers of V. germanica; 7 big winged adults of S. vesparum from V. vulgaris nest; 1 pupae of S. vesparum from V. vulgaris nest; 10 workers of V. vulgaris; 3 queens of V. germanica and 3 queens of Vespula vulgaris). The Sphecophaga vesparum specimens (n winged large adults = 7) from one Vespula vulgaris nest were collected in the United Kingdom in 2018. One pupae of S. vesparum (n pupae = 1) and five workers of V. vulgaris (n workers = 5) were collected in Belgium from another nest. Sphecophaga vesparum ( Figure 1) have two morpho-types of females, both of which were collected from one heavily infested Vespula germanica nest from Belgium in 2018 (n winged large adults = 5; n brachypterous small adults = 5; n larvae = 4; n pupae = 3 and only one worker of V. germanica, n worker = 1). The specimens of hosts were three queens and five workers of each species of wasp, Vespula vulgaris (n = 3 queens from three nests and n= 5 workers from one nest) and Vespula germanica (n = 3 queens from three nests and n = 5 workers from one nest), which were collected in the UK for the host comparisons.

Chemical and Statistical Analysis
Samples were extracted using 500 µl of pentane (Acros Organics, HPLC) for Sphecophaga vesparum and 1 ml of pentane for workers of Vespula vulgaris and Vespula germanica. After 1 min, the insects were removed from the glass vials and the extracts were evaporated under the fume hood at room temperature. The extracts were resuspended using 100 µl of hexane (HiPerSolv CHROMANORM, HPLC) for parasitoids. For wasps, we used 100 µl for workers of V. germanica, 150 µl for workers of V. vulgaris and 250 µl for both queen species. All samples were run using Gas Chromatography-Mass Spectrometry (GC-MS) (Thermo Fisher Scientific Trace 1300 connected to a Thermo Fisher Scientific ISQ mass spectrometer). The column was Restek MXT-5 (30 m, 0.25 mm and 0.25 µm film). 1 µl of each sample was injected using split-less injection at 320 °C. Initially, the temperature was held at 40 °C for 2 min, then increased to 120 °C with an increase of 20 °C/min. This was followed by an increase of 10 °C/min until 200 °C, then 7 °C/min to reach 250 °C, and a last increase of 5 °C to 350 °C/min, which was held for 4 min. The helium carrier gas had a constant flow rate of 0.9 mL/min. Alkane standards (C7 to C40 straight-chain alkanes (#49452-U, Supelco Inc., Bellefonte, PA, USA) were run as a series using the same program at three different concentrations (0.01 µg/µl, 0.005 µg/µl and 0.001 µg/µl). Peak integration was performed by integrating over total ion chromatograms using in-house developed software in R v.3.0.1. External alkane standards were used to calculate retention indices for all identified compounds based on the cubic spline method.
Peak areas of the cuticular compounds were converted to relative amounts and a principal component analysis (PCA) was performed with the prcomp function of the stats package. The distance matrix was obtained using the vegdist function with Bray-Curtis dissimilarity distance. The chemical difference between parasitoids and hosts were compared using multivariate analyses (PERMANOVA) to highlight possible variations between the groups tested (origin of the individuals or species groups) with the adonis function in the vegan package in default mode with 999 permutations. We then conducted a SIMPER analysis (distance measure: Bray-Curtis, permutations equal to 999) to investigate how much each component (or peak) contributed to the observed differences in the CHC composition among groups.

Results
CHC profiles of different types of individuals (n = 47) were analyzed using GC-MS analysis, in which we identified 69 different compounds (Table 1 for the parasitoids and Table 2 for the hosts), mostly consisting of hydrocarbons (Table A1 for all identifications of compounds, retention time, retention indexes and diagnostic ions). An example of chromatograms comparing adults of S. vesparum parasitoids and hosts of the wasps V. germanica and V. vulgaris is shown in Figure 2.

Chemical and Statistical Analysis
Samples were extracted using 500 µL of pentane (Acros Organics, HPLC) for Sphecophaga vesparum and 1 ml of pentane for workers of Vespula vulgaris and Vespula germanica. After 1 min, the insects were removed from the glass vials and the extracts were evaporated under the fume hood at room temperature. The extracts were resuspended using 100 µL of hexane (HiPerSolv CHROMANORM, HPLC) for parasitoids. For wasps, we used 100 µL for workers of V. germanica, 150 µL for workers of V. vulgaris and 250 µL for both queen species. All samples were run using Gas Chromatography-Mass Spectrometry (GC-MS) (Thermo Fisher Scientific Trace 1300 connected to a Thermo Fisher Scientific ISQ mass spectrometer). The column was Restek MXT-5 (30 m, 0.25 mm and 0.25 µm film). 1 µL of each sample was injected using split-less injection at 320 • C. Initially, the temperature was held at 40 • C for 2 min, then increased to 120 • C with an increase of 20 • C/min. This was followed by an increase of 10 • C/min until 200 • C, then 7 • C/min to reach 250 • C, and a last increase of 5 • C to 350 • C/min, which was held for 4 min. The helium carrier gas had a constant flow rate of 0.9 mL/min. Alkane standards (C7 to C40 straight-chain alkanes (#49452-U, Supelco Inc., Bellefonte, PA, USA) were run as a series using the same program at three different concentrations (0.01 µg/µL, 0.005 µg/µL and 0.001 µg/µL). Peak integration was performed by integrating over total ion chromatograms using in-house developed software in R v.3.0.1. External alkane standards were used to calculate retention indices for all identified compounds based on the cubic spline method.
Peak areas of the cuticular compounds were converted to relative amounts and a principal component analysis (PCA) was performed with the prcomp function of the stats package. The distance matrix was obtained using the vegdist function with Bray-Curtis dissimilarity distance. The chemical difference between parasitoids and hosts were compared using multivariate analyses (PERMANOVA) to highlight possible variations between the groups tested (origin of the individuals or species groups) with the adonis function in the vegan package in default mode with 999 permutations. We then conducted a SIMPER analysis (distance measure: Bray-Curtis, permutations equal to 999) to investigate how much each component (or peak) contributed to the observed differences in the CHC composition among groups.

Results
CHC profiles of different types of individuals (n = 47) were analyzed using GC-MS analysis, in which we identified 69 different compounds (Table 1 for the parasitoids and Table 2 for the hosts), mostly consisting of hydrocarbons (Table A1 in Appendix A for all identifications of compounds, retention time, retention indexes and diagnostic ions). An example of chromatograms comparing adults of S. vesparum parasitoids and hosts of the wasps V. germanica and V. vulgaris is shown in Figure 2.    The principal component analysis of relative abundance of all compounds explained 76.70% of the total variation, in which PC1 explained 47.75% and PC2 explained 28.95% (Figure 3). The principal component analysis of relative abundance of all compounds explained 76.70% of the total variation, in which PC1 explained 47.75% and PC2 explained 28.95% (Figure 3). There were significant differences in the chemical profiles of all individuals collected comparing the origin of nest species, Vv or Vg (PERMANOVA, F =11.853, R 2 = 0.208, p = 0.001 ***) (see also Figure  3). From SIMPER, the first five compounds responsible for the ordered cumulative contribution were n-C25 (0.197, p = 0.001 **), n-C27 (0.353, p = 0.34), 3-MeC37 and 11,17-diMeC27 (0.420, p = 0.22), n-C29 (0.485, p = 0.249) and 13-, 11-, 9-MeC27 (0.535, p = 0.002 **) (cumulative contribution for all compounds, Supplementary Table S1). When comparing between each group (Vg_Sv_larvae, Vg_Sv_pupae, Vv_Sv_pupae, Vg_Sv_Big, Vg_Sv_Small, Vv_Sv_Big, Vg_queen, Vv_queen, Vg_worker and Vv_worker), the difference was also significant (PERMANOVA, F = 10.271, R 2 = 0.714, p = 0.001***) (cumulative contribution for all compounds, Supplementary Table S2). We then pooled together the adults, using only wasp adults (queens and workers) and adult forms of S. vesparum (big and small) and the difference was also significant between the adult forms (PERMANOVA, F = 13.569, R 2 = 0.537, p = 0.001***) (cumulative contribution from the first 10 compounds are show in Table 3, all data available in Supplementary Table S3). Considering the linear alkanes, n-C25, n-C27 and n-C29, from adults of S. vesparum (Table 3A, B) and wasps, the SIMPER analysis showed a significant probability of getting a larger or equal average contribution in random permutation for the alkane n-C25. When comparing the hosts and its parasitoids in the SIMPER analysis (Table 3C, F), only n-C27 showed a significant probability of getting a larger or equal average contribution in random permutation in the V. vulgaris hosts and its parasitoids. From hosts and parasitoids collected in a different species nest (Table 3D, E), only the host V. germanica and parasitoids coming from V. vulgaris nests showed significant probability of getting a larger or equal average contribution in random permutation in the two alkanes, n-C25 and n-C27.  Table S1). When comparing between each group (Vg_Sv_larvae, Vg_Sv_pupae, Vv_Sv_pupae, Vg_Sv_Big, Vg_Sv_Small, Vv_Sv_Big, Vg_queen, Vv_queen, Vg_worker and Vv_worker), the difference was also significant (PERMANOVA, F = 10.271, R 2 = 0.714, p = 0.001 ***) (cumulative contribution for all compounds, Supplementary Table S2). We then pooled together the adults, using only wasp adults (queens and workers) and adult forms of S. vesparum (big and small) and the difference was also significant between the adult forms (PERMANOVA, F = 13.569, R 2 = 0.537, p = 0.001 ***) (cumulative contribution from the first 10 compounds are show in Table 3, all data available in Supplementary Table S3). Considering the linear alkanes, n-C 25 , n-C 27 and n-C 29 , from adults of S. vesparum (Table 3A,B) and wasps, the SIMPER analysis showed a significant probability of getting a larger or equal average contribution in random permutation for the alkane n-C 25 . When comparing the hosts and its parasitoids in the SIMPER analysis (Table 3C,F), only n-C 27 showed a significant probability of getting a larger or equal average contribution in random permutation in the V. vulgaris hosts and its parasitoids. From hosts and parasitoids collected in a different species nest (Table 3D,E), only the host V. germanica and parasitoids coming from V. vulgaris nests showed significant probability of getting a larger or equal average contribution in random permutation in the two alkanes, n-C 25 and n-C 27 . . The percentage of contribution for each chemical compound that explains the similarity between the compared groups is indicated. The compounds were classified from the highest to the lowest percentage of contribution, shown in the cumulative contribution (%). The p-values from SIMPER were obtained when permutations were calculated (Permutation p-value as the probability of getting a larger or equal average contribution in random permutation of the group factor).

Discussion
Our study shows that the commonly found parasitoid of Vespidae wasps, Sphecophaga vesparum, express different hydrocarbon compositions depending on whether they were found in V. vulgaris or V. germanica nests. Comparison of the relative proportions of all chemical compounds shows that there is a difference between the parasitoids and the wasps. The difference between adults of S. vesparum and Vespula wasp hosts was expressed by their different ordering of the most prevalent contribution of chemical compounds found in each. Interestingly, nest origin, V. germanica or V. vulgaris nest, separates the groups, and the alkane n-C 25 showed significant probability of getting a larger or equal average contribution in random permutation. Therefore, the alkane n-C 25 seem to be important in S. vesparum to differentiate the origin from V. vulgaris compared to those in V. germanica nests.
The host-specific hydrocarbons do not seem to be primarily acquired through contact with the adult host, since larvae and pupae have higher amounts of the alkane n-C 29 but are more likely acquired through contact with the pupal cell walls or through recycling hydrocarbons from consumed wasp pupae. We speculated that adults of the parasitoids may not be detected by the wasp host. During the pupal stage, the cocoons have a thick layer of silk that may be sufficient to protect the parasitoids during development into adults. The females can be seen on the nest walking fast and requesting trophalaxic food from the wasp larvae. Although hydrocarbon signatures of Sphecophaga vesparum seem to show wasp host dependency, we speculate that it is likely that chemical mimicry plays a role for S. vesparum to remain undetected in the colony. Partial mimicking by S. vesparum seems likely to be achieved via passive contact with the wasp hosts, similar to what happens for ant inquilines [20]. However, recycling of CHC by consuming the host is a mechanism that cannot be ruled out [21]. Another sphecophile, the beetle M. paradoxus, chemically mimics the wasp V. vulgaris by recycling CHC from the host [8]. The presence of S. vesparum has been described from nests of the vespine Vespa orientalis in the Middle East [16]. In this case, Sphecophaga vesparum would likely have (at least partially) adapted to match the chemical composition of the host due to the feeding of the wasp larvae. This is because CHC composition of the Vespa genus differs markedly from those of Vespula with a higher proportion of pentacosane and a lower proportion of heptacosane, for example [22]. Future chemical characterization of S. vesparum and subspecies collected from other wasp host species will provide more understanding about the chemical communication between hosts and parasitoids.
The mite, Varroa destructor, which parasitizes the honeybee Apis mellifera, can acquire methylalkane compounds which are present on pupae of honeybees, but the mites can also lose this chemical profile once they are in isolation, indicating that mites obtain the compounds by passive mimicry [13]. As an example from ants, myrmecophiles expressed lower amounts of CHC concentrations in comparison to their host [9]. Future studies can investigate if this is also the case for sphecophiles.
Overall, this is a first step towards understanding the chemical communication of sphecophiles of Vespidae host species. There is currently no knowledge of how parasites can locate suitable wasp nests or how they are able to infiltrate aggressive wasp colonies with usually efficient mechanisms of defense. As a next step, we suggest testing if Sphecophaga transplanted from one host species to another are able to change their CHC composition. Another interesting question is whether the nest invading S. vesparum are using chemical cues or visual cues to locate their hosts, or perhaps a combination of both. It would also be interesting to conduct bioassays to test whether the different ratio of the alkane n-C 25 in S. vesparum is an important characteristic to stay undetected in wasp host nests.

Conclusions
Hydrocarbon signatures of Sphecophaga vesparum seem to show Vespula wasp host dependency and it is likely that chemical mimicry plays a role in the parasite's ability to remain undetected in the colony. Partial mimicking by S. vesparum seems likely to be achieved via passive contact with the wasp hosts.
Supplementary Materials: The following are available online at http://www.mdpi.com/2075-4450/11/5/268/s1, Table S1: Contribution of compounds discriminating the origin of the samples, if they were collected from Vespula germanica (Vg) or Vespula vulgaris (Vv) using SIMPER Bray-Curtis dissimilarities (999 permutations). The percentage of contribution of each chemical compound that explains the similarity between the two groups is indicated. The compounds were classified from the highest to the lowest percentage of contribution, shown in the cumulative contribution (%). The p-values from SIMPER were obtained when permutations were calculated (Permutation p-value as the probability of getting a larger or equal average contribution in random permutation of the group factor). * p < 0.05, ** p < 0.01, *** p < 0.001.

Acknowledgments:
We thank An Vandoren for chemical analyses assistance and summer helpers that helped us to collect the wasp nests.

Conflicts of Interest:
The authors declare no conflict of interest.