Potential vector switching in the evolution of Bursaphelenchus xylophilus group nematodes (Nematoda: Aphelenchoididae)

Abstract To show the importance of vector switching of nematodes in the evolution of the Bursaphelenchus xylophilus group, we tested a hypothesis that “Bursaphelenchus doui (or its ancestor) was transferred by Acalolepta fraudatrix, Acalolepta sejuncta, and/or Monochamus subfasciatus (or their ancestral species) from broad‐leaved trees to conifers, switched vectors from these cerambycid beetles to Monochamus beetles in conifers, and then evolved into the common ancestor of Bursaphelenchus mucronatus and B. xylophilus.” We used a simple nematode‐loading method to beetles and produced 20 binary combinations of five B. xylophilus group species and four cerambycid beetle species in the tribe Lamiini. The affinity of the nematodes for the beetles was examined based on phoretic stage formation of the nematodes. Phoretic stages of B. doui appeared in all beetle species examined, namely Acalolepta luxuriosa, Psacothea hilaris, A. fraudatrix, and Monochamus alternatus, although the affinity of the nematode for M. alternatus was weak. This finding indicates that B. doui could switch vectors to conifer‐using Monochamus beetles after transfer by A. fraudatrix from broad‐leaved trees to conifers. We conclude that vector switching of nematodes could have potentially happened during the evolutionary history of the B. xylophilus group.

The habitats for the above nematodes are determined by their vector beetles. Therefore, B. conicaudatus and B. luxuriosae are found in broad-leaved trees, and B. xylophilus, B. mucronatus, and B. firmae inhabit conifers. By contrast, B. doui is present in both broad-leaved trees (Han et al., 2009) and conifers (Kanzaki et al., 2008) because vectors for this species, A. fraudatrix, A. sejuncta, and M. subfasciatus, use both. Monochamus saltuarius is also a vector but inhabits only coniferous species. Kanzaki and Futai (2002a) proposed that the ancestral species of B. xylophilus group, which had originated in the Eurasian Continent, obtained the ability to use tree species of family Pinaceae instead of broad-leaved ones and expanded their distribution throughout the coniferous forests ranging widely in the ancient Eurasia-North America continent. Molecular phylogenetic analyses inferred from rRNA gene segments D2-D3 LSU in Figure 3 of Kanzaki et al. (2012) showed that nematodes in conifers evolved from nematodes in broad-leaved trees. The higher genetic diversity of B. mucronatus could be the result of an earlier origin in Eurasia, and B. xylophilus could evolved recently from a B. mucronatus population in North America through geographical or reproductive isolation (Pereira et al., 2013). For this evolutionary process, cerambycid beetles must have transferred nematodes from broad-leaved trees to conifers.
We hypothesized that "B. doui, or its ancestor, was transferred by A. fraudatrix, A. sejuncta, and/or M. subfasciatus (or ancestral species of these beetles) from broad-leaved trees to conifers, switched vectors from these beetles to Monochamus beetles, that is, M. saltuarius, in conifers, and later evolved into the common ancestor of B. mucronatus and B. xylophilus." The life cycle of B. xylophilus is divided into propagative and dispersal phases. The fourth-stage dispersal juvenile (dauer juvenile; J IV ) of B. xylophilus is vital in the nematode life cycle as the phoretic stage carried by beetles. Bursaphelenchus xylophilus J IV develops when late pupae and callow adults of Monochamus beetles are present (Maehara & Futai, 1996Morimoto & Iwasaki, 1973;Necibi & Linit, 1998;Ogura & Nakashima, 2002) and enters the tracheae of the beetles. Phoretic stages of B. mucronatus (Mamiya & Enda, 1979), B. conicaudatus (Kanzaki & Futai, 2001), and B. firmae (Kanzaki et al., 2012) are also J IV . Phoretic stages of B. luxuriosae (Ekino et al., 2017;Kanzaki et al., 2009) and B. acaloleptae  are the phoretic adults (PA) and B. doui  both J IV and PA. J IV of B. conicaudatus and PA of B. luxuriosae are also induced by their vector beetles . Moreover, J IV of B. xylophilus is induced not only by its primary vector M. alternatus Hope but also by nonvector P. hilaris, although the numbers and the percentages of J IV are markedly higher in the former than in the latter (Maehara & Futai, 2001). The third-stage dispersal juveniles (J III ) of B. xylophilus molt into J IV in response to long-chain C16 and C18 fatty acid ethyl esters that are secreted from the body surface of M. alternatus, specifically during adult eclosion (Zhao et al., 2013(Zhao et al., , 2014. Thus, J IV and PA are specific and essential to vector association. In the present study, our objective was to test the above hypothesis and demonstrate the importance of vector switching of nematodes in the evolution of the B. xylophilus group. We used a simple nematode-loading method to cerambycid beetles (Maehara & Kanzaki, 2016), which could be used to examine the affinity of nematodes for not only their vectors but also nonvectors, and produced 20 binary combinations of five B. xylophilus group species and four cerambycid beetle species in the tribe Lamiini. These nematode/ beetle combinations were examined for the effects of the vector and nonvector beetles on the formation of the nematode phoretic stages, that is, J IV and PA.  (Akutsu, 1985), and fresh hand-rolled leaves of Morus bombycis Koidzumi , respectively.

| Beetle cultures
Eggs of M. alternatus were harvested from the logs by a chisel, and those of A. luxuriosa and P. hilaris were collected by opening the hand-rolled leaves. These eggs were put on wet filter paper with distilled water at 25°C in the dark until they hatched. Artificial diets were modified from the diet for M. alternatus proposed by Kosaka and Ogura (1990) and Kosaka and Enda (1991). Diet for M. alternatus was composed of 8 g of the current and 1-year-old needles of P. densiflora dried at 90°C for 1 day and milled into powder, 26.8 g of artificial silkworm diet (Silkmate 2M powder, Nosan Corporation, Kanagawa, Japan), 3.2 g of dried yeast (EBIOS, Asahi Group Foods, Ltd., Tokyo, Japan), and 62 ml of distilled water. For A. luxuriosa , diet consisted of 8 g of leaves of A. elata dried at 70°C for 1 day and milled into powder, 26.8 g of Silkmate 2M powder, 3.2 g of dried yeast, and 62 ml of distilled water. For P. hilaris (Maehara & Kanzaki, 2016), diet contained 8 g of leaves of M. bombycis dried at 70°C for 1 day and milled into powder, 26.8 g of Silkmate 2M powder, 3.2 g of dried yeast, and 62 ml of distilled water. Approximately 20 g of each diet was placed into 50-ml Erlenmeyer flasks. Flasks were plugged with a silicone-rubber stopper (Silicosen, Shin-Etsu Polymer Co., Ltd., Tokyo, Japan) and autoclaved at 121°C for 20 min. A hatched larva of M. alternatus, A. luxuriosa, or P. hilaris was placed into each flask.
Larvae were reared at 25°C in the dark for 3-5 months. When mature, larvae were incubated at 10°C in the dark for 9 months.
Larvae were subsequently removed from the flasks, rinsed in distilled water, dipped in 70% ethanol for 5 s, and then rinsed again in distilled water. The larvae for use in the first experiment were placed on wet filter paper with distilled water at 25°C in the dark until they pupated. Beetles for the second experiment were reared for one more generation in the same manner.
Mature larvae of A. fraudatrix were collected in April and May 2010 from P. thunbergii logs in Fukaura, Aomori, Japan and kept at 10°C in the dark in 15-ml centrifuge tubes with wet filter paper.
Most of them pupated at 10°C. Remaining larvae pupated only after incubation at 25°C in the dark in the first experiment. For the second experiment, in summer 2010, some adults of A. fraudatrix reared from the larvae were allowed to oviposit on Larix kaempferi (Lamb.) Carrière logs that were cut about 1 month prior. After the frass of beetles was found on the logs, larvae were collected and placed into flasks with the artificial diet for M. alternatus. Larvae were reared at 25°C in the dark for 6-8 months and, when mature, were incubated at 10°C in the dark for 5-7 months. Larvae were subsequently treated in the same manner as larvae of M. alternatus, A. luxuriosa, and P. hilaris until they pupated.
Nematodes were reared on Botrytis cinerea Pers. grown on autoclaved barley grains at 20°C in the dark for 9-16 days in the first experiment and at 25°C in the dark for 15 days in the second experiment, and were isolated aseptically from the culture using the Baermann funnel technique (Hooper, 1986). A nematode inoculum was prepared with 500 nematodes/30 μl suspension.
Petri dishes. These dishes were incubated at 25°C in the dark for 20 days. A 30 μl nematode suspension (= 500 mixed-stage nematodes) was inoculated into each dish and incubated at 20°C in the dark for 11 days and subsequently at 25°C in the dark for 22 days in the first experiment, and at 25°C in the dark for 15-20 days in the second experiment. In both experiments, a final incubation at 10°C in the dark continued until larvae of cerambycid beetles pupated.
After the pupation, one pupa was placed onto each dish ( Figure 1).
Control dishes received no pupae. Dishes were wrapped in Parafilm M ® (Bemis Flexible Packaging, Wisconsin, USA) and incubated at 25°C in the dark.
The development of pupae was observed daily. Eight days after adult eclosion, adults of the beetles were removed from the dishes.
After removal, each beetle was rinsed with distilled water, ground for 10 s using a blender in 40 ml of distilled water, and placed in a Baermann funnel overnight to extract the nematodes in the body.
To determine the number of nematodes that were unable to enter beetle tracheae, rinse water from beetles and agar medium were placed in another Baermann funnel overnight. Harvested nematodes were then counted using a stereomicroscope, and J III , J IV , PA, and all other developmental stages (propagative juveniles and adults) were recorded for each sample. When nematodes were too abundant to count, the suspension was diluted, and the numbers of nematodes were estimated. In the first experiment, we used 16 combinations of four nematode and four beetle species along with four controls with only nematodes. In the second experiment, eight combinations of two nematode and four beetle species were used along with two controls with only nematodes.

| Statistical analyses
All analyses were conducted using JMP ® 11 (SAS Institute Inc., Cary, NC, USA). The total numbers of nematodes, J IV , and PA represent those carried internally by a beetle, and those on the surface of the beetle and remaining in the agar. Two-way analysis of variance (ANOVA) was used to analyse differences in the total numbers of nematodes, J IV , and J IV + PA; the numbers of J IV and J IV + PA carried by a beetle; and the percentages of total J IV and total J IV + PA to total nematodes among beetle treatments.
For ANOVA, the numbers of nematodes were log 10 -transformed, and the percentages of J IV and J IV + PA were arcsine transformed (Yonezawa et al., 1988).

| Affinity of nematodes for beetles
The affinity of five species of B. xylophilus group nematodes for four species of cerambycid beetles was based on phoretic stage formation of the nematodes (= the percentage of total J IV + PA to total nematodes) in both the first and the second experiments (

| D ISCUSS I ON
To be carried by vector beetles, nematodes need to develop into the phoretic stages, because propagative juveniles and adults cannot transfer to beetles even if they are around pupal chambers of TA B L E 1 Effects of four cerambycid beetle species in the tribe Lamiini on phoretic stage formation of four species of Bursaphelenchus xylophilus group nematodes, and transfer of the nematodes to the beetles in the first experiment  (Jikumaru & Togashi, 1995), M. nitens (Bates) (Kanzaki & Akiba, 2014), and M. rosenmuelleri (Cederhjelm) = M. urussovii (Fischer) (Togashi et al., 2008), although the primary vector beetle for B. m. mucronatus is M. alternatus (Mamiya & Enda, 1979). Several studies reported that J IV of B. xylophilus was induced by its vector beetles, M. alternatus (Maehara & Futai, 1996Ogura & Nakashima, 2002) and M. carolinensis (Necibi & Linit, 1998). J III of B. xylophilus molt into J IV in response to long-chain C16 and C18 fatty acid ethyl esters that are secreted from the body surface of M. alternatus specifically during adult eclosion (Zhao et al., 2013(Zhao et al., , 2014. J IV of B. conicaudatus and PA of B. luxuriosae are also induced by their vectors, P. hilaris and A. luxuriosa, respectively . Moreover, J IV of B. xylophilus is induced not only by its vector M. alternatus but also by nonvector P. hilaris that inhabits not conifers but broad-leaved trees, although the numbers and the percentages of J IV are higher in the former species (Maehara & Futai, 2001). In the present study, PA of B. luxuriosae and both J IV and PA of B. doui were equally induced by nonvectors, that is, A. fraudatrix, and both A. luxuriosa and P. hilaris, respectively (Table 1). In the other combinations of five nematode and four nonvector beetle species, the phoretic stages appeared to some extent with the exception of B. luxuriosae and B.
conicaudatus in M. alternatus (Tables 1 and 2). Few PA and J IV were recovered from these two combinations. Chemical signals were not identified for induction of the phoretic stages by vectors and nonvectors, except for B. xylophilus J IV induction by M. alternatus described above (Zhao et al., 2013(Zhao et al., , 2014 The evolution of the B. xylophilus group nematodes from broadleaved tree species to species in conifers is indicated by molecular phylogenetic analyses in Figure 3 of Kanzaki et al. (2012). This evolution required cerambycid beetles to transfer nematodes from broadleaved trees to conifers. Our hypothesis was "B. doui (or its ancestor) The four species of beetles used in the present study inhabit East Asia, including Japan (Iwata, 1992;Makihara, 1992;Ohbayashi, 1992). Before the Japanese archipelago was separated

CO N FLI C T O F I NTE R E S T
None declared.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used in this paper are deposited in Dryad (https://doi. org/10.5061/dryad.5qftt dz3g).