Our field survey and insect collection was conducted at Tsukuba Experimental Forest, University of Tsukuba (36°07'10"N; 140°05'50"E (DMS), ca. 25 m a.s.l.), Tsukuba, Ibaraki Prefecture, Honshu, Japan. We found four T. apicalis infested trees belonging to different families from 2016 to 2018 (Table 2). All T. apicalis, parasitoids and basidiocarps used for dissection, observation and behavioral experiments were collected from these four trees.

Twenty T. apicalis females were collected from the aforementioned trees (Table 2). T. apicalis females were killed using ethyl acetate. The dissection method was based on previous studies (Thomsen and Harding 2011; Li et al. 2015). The abdomen was removed using micro scissors under a stereomicroscope (Leica MZ12). The dorsal plates were removed from the abdomen. A female Tremex longicollis Konow, which is known to have the mycangia (Tabata and Abe 1995), was also dissected using the same method. The T. longicollis female was collected in Yokohama, Japan on November 12, 2017.

Collected basidiocarp surfaces were removed to avoid potential contamination. The samples were then ground into a fine powder using a pestle, mortar and liquid nitrogen. Fungal DNA was extracted from the powdered samples by using DNeasy Plant Mini Kit (QIAGEN) following the manufacturer’s instructions. Each extracted DNA sample was used as a PCR template to amplify an Internal Transcribed Spacer (ITS) region by using KOD FX Neo (Toyobo) following the manufacturer’s instructions. Reactions were performed with 25 μl mixture containing KOD FX Neo, 2×buffer for KOD FX Neo, 2 μM dNTP, 0.3 μM of each primer. Primers used to amplify fungal ITS region were ITS4 (5’-TCCTCCGCTTATTGATATGC-3’) and ITS5 (5’-GGAAGTAAAAGTCGTAACAAGG-3’) (White et al. 1990). PCR reactions were performed in the following order: 94 °C for 2 minutes, followed by 40 cycles of 98 °C for 10 seconds, 50 °C for 30 seconds, and 68 °C for 90 seconds. PCR products around 650 baseparis were purified using a QIAquick Gel Extraction Kit (QIAGEN).

Sequence reactions were performed with BigDye Terminator v3.1 (Thermo Fisher Scientific) followed by purification using BigDye Xterminator (Thermo Fisher Scientific). The Sanger method was applied to determine DNA sequence of ITS region using Applied Biosystems 3130 (Gene Research Center, University of Tsukuba) and commercial sequencing services (Macrogen Japan and Eurofins Genomics). Fungi species were identified using the UNITE database (http://unite.ut.ee) (Kõljalg et al. 2013). ITS sequences were deposited to Genbank (accession numbers are: Cerrena unicolor, MH645754; Daedaleopsis confragosa, MH645755; Trametes hirsuta, MH645756).

Potato Dextrose Agar (PDA) medium (Nissui) was prepared by following the manufacturer’s protocol. C. unicolor was obtained from the Genebank Project (National Agricultural Research Organization). For behavioral experiments, fungal cultures were inoculated with a PDA medium in a 9-cm petri dish for two weeks at 25 °C.

To obtain newly emerged T. apicalis and I. japonica, we cut down woodwasp infested E. polyandra (Table 2, tree no. 2) and M. liliiflora (Table 2, tree no. 4) on November 11 and October 4, 2017, respectively. The wood was stored outside until May 2018. Newly emerged T. apicalis and I. japonica were collected upon emergence. Live adults of each species was stored in plastic containers (16 × 28 × 17 cm). All insects were allowed to mate and feed on a solution of sugar and water (30% w/w) for 24–48 hours before behavioral experiments. The olfactory preference of T. apicalis and I. japonica was examined with a Y-tube bioassay (Fig. 1a) in the laboratory (25 °C ± 1.1 °C). Arms of a glass Y-tube olfactometer (common arm 20 cm, arms 20 cm, diameter 3 cm) were connected to glass jars (17 cm × 12.5 cm) respectively. Each glass jar was connected to an activated charcoal filter. The Y-tube olfactometer was connected to an electric vacuum pump (KNF, Germany) through a flow meter. The arm side of the Y-tube olfactometer was inclined upward at 5-degrees using a metal stand (Fig. 1b). The airflow from outside passed through the charcoal, the glass jar and into the olfactometer. The flow rate of the pump was set at 1.5 L/min.

Figure 1. 

Set-up of the Y-tube olfactometer used to test T. apicalis and I. japonica attraction to volatiles from a fungal disc of C. unicolor in a top view (a) and in a side view (b) of the Y-tube olfactometer.

A single male or female woodwasp or parasitoid was introduced into the starting point of the Y-tube olfactometer. Its behavior was observed for maximum of 15 minutes. We recorded the choice of the wasp if it reached the finish line. If it did not make a choice within 15 minutes, the trial was recorded as, “no choice.” A total 43 T. apicalis (20 females and 23 males) and 57 I. japonica (25 females and 32 males) were tested. To avoid bias in the experimental setup, the positions of the two odors sources were exchanged after testing five woodwasps or parasitoids. Odor sources were renewed after testing five woodwasps or parasitoids. Using a binomial test, we determined the preference of both T. apicalis and I. japonica between volatiles from fungal disks and control disks.

Abdomens of the female T. apicalis had ovaries that contained on average 96 eggs (Fig. 2a). A pair of mucus glands, which looked like large whitish sacs, was highly developed (Fig. 2a). A single mucus reservoir connected to the basement of the ovipositor was detected under the mucus glands (Fig. 2b). The reservoir contained a sticky transparent fluid.

We could not locate mycangia-like structures in the abdomen of the female T. apicalis while other anatomical features were nearly identical to other woodwasps. There were no visible sac-like structures located behind the base of the ovipositor where T. longicollis has clearly identifiable mycangia (Fig. 2c).

Figure 2. 

Abdominal organs of female woodwasps. a Ventral view of the dissected abdomen of T. apicalis. Typical internal organs, ovaries and mucus glands were easily detected. b Upper left diagonal view of the dissected abdomen of T. apicalis. A mucus reservoir was detected under the mucus glands. c Comparative view of dissected region of abdomens close to the ovipositor. In T. longicollis, the mycangia are recognized as greyish balls located close to the basement of ovipositor. There were no such sac-like structures in T. apicalis. Abbreviations; my: mycangia, mr: mucus reservoir, mg: mucus gland, ov: ovary. Scale bar: 2 mm.

We performed DNA barcoding for basidiocarps found on T. apicalis infested trees using ITS region. The results showed that the ITS region from basidiocarps on: (1) S. macrophylla had a 99.67% match with C. unicolor; (2) E. polyandra had a 99.81% match to Daedaleopsis confragosa (Bolton) J.Schröt.; and (3) M. liliiflora was identical to Trametes hirsuta (Wulfen) Lloyd. All the fungus species we identified belong to the family Polyporaceae.

In the Y-tube bioassay, 85.0% (n = 20) of T. apicalis females, 52.2% (n = 23) of its males, 84.0% (n = 25) of I. japonica females and 62.5 % (n = 32) of males chose between the volatiles from fungal and control disks. Females of T. apicalis woodwasps (76.5 %, n = 17, P < 0.05) and I. japonica (76.2%, n = 21, P < 0.05) preferred volatiles from the fungal disk to the control disks (Fig. 3). In contrast, males of both T. apicalis (41.7%, n = 12, P = 0.77) and I. japonica (55.0%, n = 20, P = 0.82) did not display a statistically significant preference for either fungal or control disks.

Figure 3. 

Percent choice by T. apicalis and I. japonica in arms of Y-tube olfactometer with the volatiles from PDA discs (control) vs. fungal culture discs.

Subfamily Siricinae has a close relationship with the genus Amylostereum (Schiff et al. 2012; Tabata et al. 2012). For subfamily Tremecinae, C. unicolor is only known symbiotic fungus for many Tremex woodwasps (Stillwell 1967; Tabata and Abe 1995; Pazoutova and Srutka 2007). Basidiocarps on T. apicalis infested large-leaf dogwood tree (Swida macrophylla) were C. unicolor. Notably, we could not identify any apparent mycangia-like structures inside the female T. apicalis (Fig. 2).

Basidiocarps on other T. apicalis infested trees were members of the family Polyporaceae, inclusive of C. unicolor. These species are white-rot fungi that cause sap wood decay (Enebak and Blanchette 1989; Stajić et al. 2017; Račko et al. 2018). Our observations suggest that T. apicalis, which lacks identifiable mycangia, inhabits host wood that is already infected by wood decaying fungi.

The female T. apicalis’ preference for C. unicolor suggests that it uses volatiles from the fungus to locate suitable host wood. This strategy would be similar to female Xeris spectrum (Hymenoptera: Siricidae), whose mycangia is also absent and who use the odor from fungi Amylostereum to locate host wood (Fukada and Hijii 1997; Matsumoto and Sato 2012).

While mycangia carrying woodwasps Sirex nitobei Matsumura and Urocerus japonicus Smith form specific relationships with a particular Amylostereum fungus, mycangia-less X. spectrum can utilize more than one particular species. For example, the Amylostereum fungi species has a well-documented relationship with Sirex nitobei and Urocerus japonicas respectively (Fukuda and Hijii 1997).

It is possible that T. apicalis employs similar strategies that take advantage of their lack of mycongia. This may provide woodwasps without mycongia with an evolutionary advantage that increases their overall chances for reproduction and survival.

The Siricinae parasitoid I. leucospoides locates its host using volatile cues from symbiotic fungi (Martínez et al. 2006; Pietrantuono et al. 2012). In our study, we demonstrated that I. japonica is attracted to volatile compounds from C. unicolor even though this fungus is not a symbiont carried in T. apicalis. In contrast, I. japonica’s potential host, T. longicollis, has symbiotic relationship with C. unicolor (Tabata and Abe 1995; Watanabe et al. 2018). In order to locate potential hosts, Ibalia parasitoids seem to exploit olfactory cues not only from symbiotic fungi but also fungi species that live outside of the host.

To the best of our knowledge, this is the first report that connects white rot wood decaying fungus, Tremecinae woodwasp species and its parasitoid via volatile compounds.