Courtship behavior and identification of a sex pheromone in Ibalia leucospoides (Hymenoptera: Ibaliidae), a larval parasitoid of Sirex noctilio (Hymenoptera: Siricidae)

Insect collection

Pinus sylvestris Linnaeus and P. resinosa Aiton (Pinaceae) naturally infested with S. noctilio were collected at two sites in New York State from January 2015 to April 2019. The locations were Chimney Bluffs State Park (43°16′50″N, 76°54′48″W) in Wolcott, NY, USA and the Pack Demonstration Forest (43°32′58″N, 73°49′12″W) in Chestertown, NY, USA. Fieldwork was approved by the New York State Office of Parks, Recreation, and Historic Preservation (application numbers: 2015-CB-001, 2017-CB-001, 2018-CHB-001). Trees were felled and cut into 1 meter bolts and transported to the laboratory at the SUNY-College of Environmental Science and Forestry in Syracuse, NY, USA. The bolt ends were coated with wax (Gulf Wax, Royal Oak Enterprises, LLC, Roswell, GA) to slow moisture loss, stored at 1−5 °C for two weeks to two months, and every two weeks a subset of logs was transferred to emergence cages at room temperature. After 6-8 weeks adult I. leucospoides began emerging and were collected daily and individually stored in 59 ml plastic mini-cups (Wal-Mart Stores, Inc., Bentonville, AR) at 5 °C. We did not observe mating in the emergence cages in the entire time of collections, so we assumed collected wasps were not mated (Fischbein et al., 2012). In addition, we collected and dissected newly emerged females and males and observed them under a compound microscope at 1000 × to assess the presence of spermatozoa in the spermathecae and testes (as a positive control). Spermatozoa were clearly visible in the testes but no spermatozoa were observed in the dissected spermathecae (n = 16 females, 10 males).

Mating behaviors

Courtship behavior of I. leucospoides adults was observed in a wire mesh observation cage (0.5 mm mesh, 75 cm length ×10 cm diameter). The observation cage was positioned horizontally on a table with a two-bulb white fluorescent strip light (40 W) centered 75 cm above it. Wasps were taken out of cold storage and held at room temperature (25 ± 3 °C) for at least one hour before starting observations. A male and a female were released at opposite ends and observed for 15 min. Any wasp that did not move or interact with other wasp within 15 min was removed and replaced. All courtship observations (n = 14) were conducted at 25 ± 3 °C between the hours of 09:00 and 12:00 and were recorded using an iPhone SE device. All wasps were presumably virgin, and each individual wasp was used only once. The ages of males and females were between 1 and 20 and 1 and 10 days post-emergence, respectively. At the end of each day, wasps were fed only with water (reverse osmosis [RO] water) (Pietrantuono et al., 2012), then refrigerated for other experiments, including olfactometer assays.

Cuticular hydrocarbons extracts

Ibalia leucospoides cuticular hydrocarbons were obtained individually from 1- to 5-day post-emergence virgin adults (males and females) by immersing the whole body of a live wasp into 2 ml of dichloromethane (DCM) (99.9%, OmniSolv, Billerica, USA) for 2 min. Each individual was immersed three times (Böröczky et al., 2009) and the extracts from each insect were combined, concentrated to 200 µl under a gentle stream of nitrogen, and stored at −60 °C until further analysis. A total of 24 extracts were prepared for each sex.

Body size measurement

The body size of Ibalia males is smaller than that of females, so the ratio of male/female body size was applied to the quantities of cuticular components in males as a correction factor for size when the quantities of cuticular components of males was compared to those of females.

To measure the body sizes, measurements were carried out with extracted individuals using a caliper accurate to 0.1 mm. Body surface area of both males (n = 32) and females (n = 24) was estimated based on the length, width, and height of the thorax (including propodeum but minus wings and legs) and abdomen. The surface areas of the thoraxes and abdomens were calculated as prolate ellipsoids (Wang & Messing, 2004; Macias-Ordonez & Draud, 2005; Wang et al., 2016).

Identification of EAD-active components

The cuticular hydrocarbon extracts were analyzed with a Gas Chromatograph (GC) (Hewlett-Packard, HP 5890 Series, Sunnyvale, CA, USA) with a flame ionization detector (FID) and a custom electroantennographic detection system. Analytical conditions were as follows: splitless mode, injector temperature 239 °C, detector temperature 240 °C, a non-polar HP5-MS column 30 m × 0.25 mm ID × 0.25 µm film thickness (Agilent Technologies, Santa Clara, CA, USA), and nitrogen carrier gas. The temperature program was 39 °C for 1 min, increased 10 °C/min to 240 °C, and held for 40 min. The GC was coupled with an electroantennographic detection (EAD) system and Clarity Lite integrator (Data Apex Ltd., Czech Republic). The column effluent was split 1:1 in the oven using a Y-splitter (Supelco, Bellafonte, PA, USA); the EAD side had nitrogen make-up gas added (8 ml/min) through a second Y-splitter. Both the FID and the heated EAD port were 280 °C. The EAD half of the effluent was introduced into a cooled humidified air stream (1 L/min) directed toward the antennal mount. Whole head preparations were made of individual insects by excising the head and positioning the antennae between two gold wire electrodes immersed in saline-filled (140 mmol NaCl, 10 mmol KCl, 6 mmol CaCl₂, 2 mmol MgCl₂) wells in an acrylic holder. The EAD signal was amplified (10×) with a custom-built high input impedance DC amplifier and signals from both the FID and the antenna were collected simultaneously and digitized with the Clarity Lite integrator and Clarity Lite software (version 7.1.0.151-2016; DataApex Ltd., Czech Republic). The electrophysiological activity of samples and the identified synthetic compounds were confirmed on GC–EAD for both males and females.

The cuticular hydrocarbon extracts were also analyzed on a gas chromatograph (Model 7890A; Agilent Technologies, Inc., Santa Clara, CA, USA) equipped with a mass spectrometric detector (Model 5975c; EI mode, 70 eV with a scanning range of 30.0–550.0 m/z; Agilent Technologies, Inc., Santa Clara, CA, USA) and a non-polar capillary column, HP5-MS (30 m × 0.25 mm ID × 0.25 µm film thickness; Agilent Technologies, Inc., Santa Clara, CA, USA). The carrier gas was helium at a constant flow rate of 1 ml/min. The EAD-active components were tentatively identified by comparison of mass spectra from National Institute of Standards and Technology (NIST) mass spectra (MS), Kovats indices (KI) were calculated relative to straight chain alkanes (Kováts, 1965), and confirmed by comparison with the KI, the retention times and electron impact mass spectra of available compounds. Female and male equivalents for each EAD-active compound were determined by GC-MS quantification using authentic samples of each EAD-active component as external standards to create calibration curves. Components 1 and 2 were quantified using the closest alkane standards (pentacosane and hexacosane) as external standards.

Synthesis of (6Z,9Z)-heptadeca-6,9-diene

Barton decarboxylation / bromination was used to synthesize this compound from linoleic acid (Loreau et al., 2000) analogous to (3Z,6Z,9Z)-heptadeca-3,6,9-triene synthesis from linolenic acid adapted from a previously published study of blueberry spanworm (Itame argillacearia, (Lepidoptera: Geometridae)). (De Silva et al., 2013) (See Scheme 1 and details in Supplemental Information).

Chemicals

Chemical standards of tricosane (99%) and pentacosane (99%) were purchased from Sigma-Aldrich Inc. (St. Louis, MO, USA) and BeanTown Chemical (Hudson, NH, USA), respectively. Hexacosane (99%) and heptacosane (99%) were obtained from Alfa Aesar (Ward Hill, MA, USA). (6Z,9Z)-heptadeca-6,9-diene was synthesized as described above.

Two-choice olfactometer assays

A Y-shaped glass olfactometer was used to determine male responses to pairs of stimulus blends (Table 1). The olfactometer consisted of a Y-shaped glass olfactometer (6 cm i.d.) with a 26 cm main stem and two 21 cm arms and a 70° angle between the arms. The olfactometer was positioned horizontally on a table with a 500 W halogen lamp centered 50 cm above the Y-tube. Charcoal-filtered air was passed through two Erlenmeyer flasks containing 100 ml water, connected to plastic tubing (1 m length), and attached to polytetrafluoroethylene (PTFE) tubing (10 cm length) at the ends of each of the choice arms. The air flow rate in each of the choice arms was 500 ml/min, and at the base the rate was 1000 ml/min. Air was removed from the base by vacuum. Cardboard (100 cm length × 60 cm width) surrounded the olfactometer to eliminate visual cues. A piece of filter paper (5 cm × 0.5 cm) was loaded with 5 µl of extract or solution and was placed at the end of a choice arm. For each trial, a male was released at the base of the main arm and was allowed 6 min to respond to an odor source by reaching the end of a choice arm. The positions of odor sources and the olfactometer arms were alternated after every trial to control for potential positional bias. The Y-tubes were washed between trials after every responsive insect with unscented detergent (Alconox, Jersey City, NJ, USA) and hot water, rinsed with RO water, and baked at 120 °C for at least 1 h. At the end of each day, the glassware was washed and baked overnight. All bioassays were carried out between 07:00 and 18:00 at 25 ± 3 °C, and each male was used no more than once per day. Males were re-used in different olfactometer experiments using different stimuli and none was reused for the same pair of stimuli, i.e., in the same experiment. A positive response was recorded when a male reached the end of the arm containing a stimulus, while a negative response was recorded for those that reached the end of the arm containing the control within 6 min. All males were 1–14 day old virgins, and were provided with RO water for one day before being used in assays (Pietrantuono et al., 2012). Females were not used in the olfacometer assays because only male attraction to females was observed in our study of mating behavior.

The following assays were conducted: (1) female body wash vs. DCM, (2) male body wash vs. DCM, (3) a blend of the five authentic compounds (Blend 1, commercially obtained or synthesized) (Table 1) vs. DCM, (4) Blend 1 vs. a blend of the four commercially available alkanes (Blend 2), (5) a series of subtractive assays in which one component was removed from Blend 1 at a time and the remaining blend was assayed against DCM, (6) Blend 2 vs. female body wash. The component ratios in the blends approximated those in extracts obtained from individual females and measured with GC-MS (Table 1).

Statistical analysis

The equality of variance and normality of body size measurements (surface area) were confirmed using the Kolmogorov–Smirnov test. The Student’s two-tailed t test was used to compare the surface area measurements for both males and females. The amount of each EAD-active component in individual females was multiplied by 0.75 to adjust for sexual dimorphism in surface area. The Student’s two-tailed t test was used to compare the quantity of each EAD-active component between males and females. P-values were compared with the Bonferroni corrected alpha level (0.05/7 = 0.0071). For each trial in Y-tube assays, the numbers of wasps that responded to stimuli were compared with a binomial test with expected 1:1 ratio at a probability of 5%. All statistical analyses were conducted in SAS 9.4.

Mating behaviors

A consistent sequence of courtship behaviors was observed. Once a male was released into an arena, it groomed its antennae with its mouthparts and front legs, and progressively walked toward the female. Once it encountered the female, the male occasionally bounced its abdomen vertically, approached from the rear or side, and initiated rhythmical lateral movements while holding its antennae upright. Meanwhile, the female slowly walked in the opposite direction and slightly vibrated its wings (wing fanning). The female then groomed its antennae with its front legs, occasionally bounced its abdomen vertically, and waited for the male to arrive. The male climbed onto the female (mounted) from the side while making antennal contact, and aligned its body with its head above that of the female. Once the male mounted the female, it initiated head-nodding behavior, starting with multiple fast strokes with its antennae against the female antennae, followed by slow upward sweeping movements. All strokes and sweeping movements occurred with one male antenna against one female antenna, alternating antennae at the end of each head-nodding cycle. The duration of each cycle was 3.5 ±0.4 s, and each male performed 17.7 ± 0.6 head-nodding cycles for a total of 62 ± 8.1 s (n = 14). Females held their antennae upright with no movement during the male head-nodding cycles. Two types of responses were observed in females: (1) in most cases the female ignored the male, particularly smaller and older males, and did not allow the male to mount, or (2) the male immediately mounted and initiated the head-nodding cycle, but the female was not sexually receptive, and kicked the male off with her back legs after multiple kicks. Meanwhile, males exhibited a short bout of wing fanning immediately following each rejection. This was the only instance in which males exhibited wing fanning. We observed no successful copulations (n = 14), although the majority of males tried to mount the female and then repeated the cycle after rejection by the female. Our observations showed that the majority of males, regardless of size or age, were attracted to females, although females only allowed males of similar or larger size to mount.

Body size measurements

There was no significant difference between thoracic surface area measurements between males and females (Student two-tailed t test, P > 0.05, t = 0.68) (Table 2). The abdominal surface area and the total surface area measurements of I. leucospoides females were significantly greater than those of males (Student two-tailed t test, P < 0.05, t = 3.75 for abdomen, and P < 0.05, t = 2.48) (Table 2).

Identification of EAD-active components

In GC-EAD analyses of female body washes, antennae of males repeatedly responded to seven components present in all extracts (Table 3, Fig. 1). In GC-MS analyses of female body washes, the retention time and mass spectra of five of the components matched those of (6Z,9Z)-heptadeca-6,9-diene (supplemental information) tricosane, pentacosane, hexacosane, and heptacosane (Fig. 2). Although the retention time, EI-mass spectra and KI of the synthetic diene support this structure, the identification of this compound remains tentative. Further microchemical analyses (e.g., epoxidation) followed by GC/MS analysis, were not successful in producing structural information because of high background. The mass spectra of two additional EAD-active components are indicative of the branch point; 2-methylalkanes give highly characteristic EI-mass spectra showing enhanced M-15 and M-43 (loss of an isopropyl radical leaving the positive charge on the primary fragment). Authentic compounds were not available. The Kovats indices of them matched with two methylbranched alkanes (2-methyltetracosane and 2-methylhexacosane; components 1 and 2, respectively, see supplemental information). All EAD-active components were present in all male and female body washes, except (6Z,9Z)-heptadeca-6,9-diene, which was present in all female samples, but found in low quantity in the body washes of only seven of forty males. The amount of each EAD-active component was compared between males and females after correcting for the sexual dimorphism in body surface area. The statistical analysis showed that the amounts of C₂₅, C₂₆, (6Z,9Z)-heptadeca-6,9-diene, and component 1 were significantly different between males and females (P < 0.05).

Two-choice olfactometer assays

In olfactometer assays, 30 I. leucospoides males were attracted to female body washes, compared to 15 males that responded to DCM alone (P < 0.05) (Fig. 3). I. leucospoides males were not attracted to male body washes. The five-component blend (Blend 1) was attractive to I. leucospoides males (P = 0.05). Similarly, 25 males responded to odors of the four-component Blend 2, compared to only 13 that responded to DCM (P = 0.03). Blend 2 was more attractive than Blend 1 to I. leucospoides males (P = 0.04). Male I. leucospoides were not attracted to other blends in which individual components were successively removed from Blend 2 (Fig. 3). When the female body wash was tested for male attractiveness against the four-component blend (Blend 2), Blend 2 was no more attractive to males than the female body wash (P = 0.5).