Appl. Entomol. Zool. 45 (2): 329–337 (2010) http://odokon.org/

Invasion pathway and potential risks of a bamboo-nesting , Xylocopa tranquebarorum (: ), and its micro-associated mite introduced into Japan

Kimiko OKABE,1,* Hayato MASUYA,1 Kazuhide KAWAZOE2 and Shun’ichi MAKINO1 1 Forestry and Forest Products Research Institute; Tsukuba, Ibaraki 305–8687, Japan 2 JR Tokai Takashimaya Co., Ltd.; Nagoya, Aichi 450–6001, Japan (Received 7 January 2010; Accepted 1 March 2010)

Abstract Despite plant quarantine systems, many organisms have been unintentionally introduced by the rapid expansion of in- ternational trade. Xylocopa tranquebarorum, a large species of carpenter bee that nests in bamboo shoots, was intro- duced into Japan approximately 5 years ago and, along with its specifically associated mite, has established popula- tions in Toyota and Nagoya in Aichi Prefecture and Anpachi in Gifu Prefecture. Because X. tranquebarorum only nests in dead bamboo shoots and because its associated mite differs from the mite species associated with X. tranque- barorum in Taiwan but is morphologically similar to the species described in India, Indonesia, and China, we suspect that X. tranquebarorum and its mite were introduced via processed bamboo shoots from continental Asia. Because the most probable pathway of associated microorganisms is host introduction, biological and ecological information, such as an inventory of symbionts, is crucial to avoid the risk of invasive symbiotic microorganisms. Although X. tranque- barorum may become a pest in the bamboo industry if its distribution is expanded by human transport, its associated mite is suspected to be of low risk as house dust but to have a high contamination risk for the Japanese endemic mite, Sennertia alfkeni, associated with native Japanese carpenter bees.

Key words: Cleptoparasite; fungivore; house dust; phoresy; symbiont

tine offices in Japan based on their pest status, INTRODUCTION small organisms have been unintentionally intro- International trade has rapidly expanded world- duced because searches of associated organisms wide, resulting in the concomitant transport of (usually called hosts) and assessments of their po- alien species (Meyerson and Mooney, 2007; tential pest risk have never been conducted as quar- Hulme, 2009). Although quarantine systems cover antine targets (Goka et al., 2001; Okabe and Goka, many traded goods, microorganisms are often un- 2008). Due to the extreme difficulty of controlling intentionally introduced, frequently causing devas- and eradicating such organisms, important preven- tating consequences within the novel environment. tative measures should include both potential risk Whereas plants and vertebrates tend to be inten- assessments in terms of crops, hygiene, and eco- tionally introduced across borders as crops and/or logical impacts and pathway analysis to develop ef- pets, invertebrates and microorganisms are typi- fective control measures, reexamine the current cally introduced because the quarantine system quarantine system, and determine effective eradica- failures or because of a lack of knowledge (e.g., tion measures (Pimentel, 2002; Perrings et al., horticultural plants: Reichard and White, 2001; mi- 2005). croorganisms and others in China: Xu et al., 2006; Large carpenter bees belonging to the genus Xy- forest herbivorous : Gandhi and Herms, locopa have diversified within tropical and sub- 2009). Although microorganisms such as fungi, ne- tropical regions and expanded their distributions to matodes, and mites are monitored by plant quaran- temperate regions (Hurd and Moure, 1963). Sev-

* To whom correspondence should be addressed at: E-mail: [email protected] DOI: 10.1303/aez.2010.329

329

AEZ45-2_329-337.pdf 1 10.5.26 1:52:45 PM 330 K. OKABE et al.

eral species are suspected to have invaded oceanic exact distributions of carpenter bees of Xylocopa islands, as only a few species exist within island subgenus Biluna and their associated mites are un- groups and the distances between the original and known. An additional bamboo-nesting carpenter close relative’s habitats are relatively short. For ex- bee, X. tranquebarorum, was recently introduced ample, Xylocopa sonorina Smith was allegedly into Japan; in 2006, this species was collected in transferred by humans from North America to Aichi and Gifu prefectures, where the native large tropical Pacific islands, including (Hurd, carpenter bee, Xylocopa appendiculata circum- 1958). Mites are associated with most species of volans, is distributed (Kamio, 2007; Yata, 2007). carpenter bees, often exhibiting host specificity or The exotic bee has established populations in To- at least narrow host ranges; these mite species be- yota, Aichi Prefecture (Okada and Takeda, 2009; long to the Mesostigmata (Dinogamasus and Hy- this study) and has likely expanded its distribution poaspis of Laelapidae), Prostigmata (Cheletophyes to Nagoya (Ohkusa, personal communication) (Fig. of Cheyletidae and Tarsonemus of Tarsonemidae), 1). Xylocopa tranquebarorum is associated with a and Astigmata (Sennertia of Caetodactylidae and mite belonging to Sennertia, the deutonymph of Horstia of Acaridae) (OConnor, 1993; Eickwort, which is morphologically similar to but genetically 1994; Klimov and OConnor, 2008). Mutualisms different from the Japanese large carpenter bee may exist between carpenter bees with acarinaria mite, S. alfkeni (Kawazoe et al., 2010). Hereafter, (external pouch-like structures that facilitate phoretic mite transfer with the bees) and their asso- ciated mites, but most interactions are considered cleptoparasitism, with mites feeding on the host loaf and feces (Abrahamovich and de Alzuet, 1990; Eickwort, 1994; Okabe et al., 2008). The mite life cycle is well synchronized with that of its host: mites leave a host to migrate into a nest during the initial stage to anytime afterward, molt to a feeding stage to consume provisioned pollen and host feces after the host larva has consumed the pollen loaf, molt to deutonymphs during host pupation, and attach to the host for dispersal (Abrahamovich and de Alzuet, 1990; Okabe et al., 2008). In Japan, five species large carpenter bees are associated with one mite species, Sennertia alfkeni (Oudemans) (Okabe and Makino, 2005; Okabe et al., 2008); however, molecular analyses have revealed five host-specific genetic lineages of the mite, suggesting long evolutionary associations between each pair (Kawazoe et al., 2008a, b). Sennertia dissimilis Zakhvatkin, associated with Xylocopa nasalis Westwood (from Japan and prob- ably originating from a Southeast or East Asian country occupied by Japan at the collection time), Sennertia horrida (Vitzthum) with X. nasalis (from India, Indonesia, and Japan at the time of collec- tion), and Sennertia potanini Zakhvatkin with Xy- Fig. 1. Collection sites of introduced Xylocopa tranque- locopa tranquebarorum (Swederus) (from western barorum in Japan. Top diagram shows the regions of sites in China) are all described as species of bamboo-nest- Japan, and the bottom diagram indicates each site with a cir- cle. Bee and mite samples were collected in Toyota (white cir- ing large carpenter bees of the subgenus Biluna cle), and bees were also reported in Nagoya and Anpachi (Zakhvatkin, 1941; Klimov and OConnor, 2008). (eastern and western black circles, respectively). Gray lines in Currently, the biology and ecology, including the the bottom diagram indicate large rivers.

AEZ45-2_329-337.pdf 2 10.5.26 1:52:45 PM Invasion of Xylocopa tranquebarorum with Mite 331

the introduced species is referred to as Sennertia nr Cut shoots with or without mites were stored sepa- alfkeni. Although the original habitat(s) is un- rately in a closed container and maintained out- known, this bee is unlikely to have originated from doors under a roof at ambient temperature until Taiwan (although the Japanese name of X. tranque- adult bee emergence. To use living mites in experi- barorum is directly translated to “Taiwanese bam- ments, adult carpenter bees were maintained with boo-nesting large carpenter bee”), where X. tran- phoretic mites in a cage (45 cm per side, con- quebarorum is associated with S. horrida; instead, structed of a wooden frame and synthetic fiber based on its associated mite species, it probably ar- mesh) and were provided diluted honey and sugar rived from Southeast and/or East Asia (Kawazoe et water soaked in cotton. al., 2010). We conducted three laboratory experiments to Although the introduced X. tranquebarorum at- examine the potential risks of the mite: compar- tacks commercial bamboo products for nesting isons of the developmental period of S. nr alfkeni purposes, the risks posed by the associated mite are to that of S. alfkeni, which was originally associ- not as direct because thus far no Sennertia mites ated with the Japanese carpenter bee X. a. circum- have been considered pests. However, potential volans to investigate the possibility of a host switch mite problems include the following: introduced by the invasive mite; fungal sampling in nests to mites may lead to the collapse of the interaction evaluate the potential house dust risk via a mite between endemic X. a. circumvolans (Smith) and vector; and assessments of the necessary condi- S. alfkeni, perhaps via a mite host switch, and S. nr tions for the deutonymphal molt to determine the alfkeni may become house dust by surviving with- risk of both a host switch and to contributing to out the host and vectoring fungi. Therefore, as a house dust. risk assessment of S. nr alfkeni, we investigated the To assess the mite developmental periods, pairs life cycle of the introduced mite to examine the of S. nr alfkeni were isolated from bee nests and possibility of an association between X. a. circum- placed in transparent plastic vials (1.6 cc, 35 mm in volans and S. nr alfkeni, and to determine if the length) with a small piece of wet filter paper and mite could survive without the host, with particular about 1 g fresh pollen loaf from a X. tranquebaro- focus on deutonymphal molt to tritonymphs in the rum cell. Ten vials were prepared and maintained absence of the host. We also investigated the poten- at 25°C. Each pair was transferred daily into an- tial of the mite as a vector of fungi. other vial with the same contents. Vials containing mite eggs were also maintained at 25°C, and mite development was checked once daily. The mite de- MATERIALS AND METHODS velopmental period was estimated from mass rear- Information related to carpenter bee invasion. ing, assuming that the first larva became the first To record the expansion of this species, we also protonymph within the population. collected information on observations of X. tran- To collect fungi associated with the carpenter quebarorum from the published literature and ento- bee and the mite, we collected 20 tritonymphal mologists near Toyota, Aichi Prefecture, where the mites from two nests in which fungal contamina- carpenter bee was first reported. tion had occurred but that lacked host juveniles Experiments. Nests of the exotic X. tranque- (likely because they had been abandoned before barorum were collected from several points along oviposition or because bee offspring had died at an the Yahagi River in Toyota on 29 May 2009 (Fig. early stage). The body surface of 20 mites from 1). A few nests were collected from natural bam- each nest was soaked in 70% ethanol for 30 min boo forests, but most nests were found in commer- and in a diluted Twin 80 solution (one drop in 10cc cial bamboo shoots at a bamboo timber yard. The distilled water) for 30 min to remove and/or kill nest portion between bamboo joints was removed fungal materials on mites. After cleaning in dis- in the field, and 2 days later in the laboratory, the tilled water for about 1 min, 10 mites were trans- nest was opened by cutting it in half lengthwise. ferred onto a malt agar (MA) plate (90 mm in di- The diameter of each nested shoot was recorded. ameter) to collect feces. Ten squashed mites were The number of cells in a nest and the numbers of then transferred onto another MA plate to cultivate juvenile bees and mites in each cell were counted. internal fungal materials. Plates were maintained at

AEZ45-2_329-337.pdf 3 10.5.26 1:52:46 PM 332 K. OKABE et al.

20°C for 1 month to identify living fungi inside of niles with and without mites was compared using mites. one-tailed Fisher’s exact tests. After testing for nor- To examine mite host specificity, we assessed mal distribution using Kolmogorov-Smirnov tests, whether deutonymphal mites could molt without mite developmental periods between sexes and be- hosts (some -associated mites need to ride on tween species [data for S. nr alfkeni originated specific phoretic hosts before molting; e.g., from this study and data for S. alfkeni were ob- Okamoto et al., 1991) and with fungi collected tained from Okabe et al. (2008)] were compared from bee nests. The experiment was also con- using Mann-Whitney U-tests. All statistical analy- ducted using deutonymphal molts with or without ses were conducted using the statistical package chilling treatment at 5°C (some astigmatid mites STATISTICA 06J (StatSoft Inc., 2005). overwinter as deutonymphs; e.g., Hayashi et al., 2009), but all deutonymphs maintained with or RESULTS without hosts died 1 month later. Deutonymphal mites originated from several adults in a cell and Information related to carpenter bee invasion were reared in a vial with a bee pollen loaf. Vials The bamboo-nesting carpenter bee has been re- were maintained for 30 days at ambient tempera- ported in Toyota, Aichi Prefecture (white circle in ture. Subsequently, 20 mites were transferred to an- Fig. 1) and in Anpachi, Gifu Prefecture (western- other vial with a piece of bamboo (5ϫ20 mm) and most black circle in Fig. 1). Since 2006, X. tran- an about 0.5 g pollen loaf; this procedure followed quebarorum has been observed every season in that used to molt deutonymphal S. alfkeni to Toyota (Yamagishi and Okada, personal communi- tritonymphs (Okabe et al., 2008). Twenty mites that cation). In January 2009, Okada and Takeda (2009) had molted to deutonymphs on a carpenter bee and reported overwintering of the carpenter bee in To- that had been maintained for 5, 30, or 60 days on yota. In 2009, X. tranquebarorum was also ob- the bees were transferred to a vial with the same served in Nagoya, located between Toyota and An- contents as described above. We collected 20 mites pachi (Ohkusa, personal communication; Fig. 2). from each of five different carpenter bees for a During bee collections in the field, we found one to total of 100 mites for the experiment. Vials were several nests per bamboo stand (ca. 100–500 m2) maintained at 25°C, and mites were checked once but easily found over 25 nests at a timber yard spe- per week to determine if they had molted to cializing in bamboo shoots; however, whether these tritonymphs, until all deutonymphs had died. commercial shoots were imported or domestically Statistics. The mortality of carpenter bee juve- transferred was unclear.

Fig. 2. The amount of bamboo (includes both processed and unprocessed) imported to Japan from each major export country (Taiwan and China). Minor export countries that are not shown include Thailand, South Korea, and Vietnam.

AEZ45-2_329-337.pdf 4 10.5.26 1:52:46 PM Invasion of Xylocopa tranquebarorum with Mite 333

Life history traits of invasive X. tranquebarorum that had already consumed the pollen loaf). Adult Bamboo with carpenter bee nests appeared fresh bees emerged between 20 and 28 July. Males and had probably been cut within a few months. emerged after all females from the same nest had We collected 33 nests at four different sites (three emerged. The sex ratio differed slightly between natural bamboo stands and one timber yard) within mite-laden and mite-free nests (Fisher’s exact test, 4 h (Table 1). At each site, multiple carpenter bees pϭ0.4929). were provisioning their nests, but most nests (25) were collected from the timber yard. The stages of Life history traits of associated organisms nests varied from having only an entrance (but not The mite life cycle was observed in maintained yet provisioned) to having at most seven provi- nests: mites migrated into a cell at the beginning of sioned cells. Because many of the nests contained host provisioning; quickly turned into tritonymphs eggs and 22 cells were unfinished, most collected and adults; laid an egg mass usually on a partition nests were incomplete. In four nests, we collected wall of a cell; fed on the pollen loaf and then on adult females that had not yet provisioned. The av- host feces, which perhaps contained fungal parti- erage inner diameter of nests was 14.5Ϯ1.22 cm cles; protonymphal mites aggregated on a pupal (ϮSD, nϭ19). When nests were opened, 26 off- host; and deutonymphal mites attached to a newly spring (eggs or early instars) were dead within emerged host (see Fig. 3 for mites on a pupa and mite-laden cells, and three were dead in mite-free emerged host). Perhaps because the typical number cells (Fisher’s exact test, pϭ0.2468). A total of 123 of mites had increased under artificial conditions, cells contained bee stages ranging from eggs to bees emerging late from a nest were heavily cov- final instar larvae (final instar larvae were those ered with deutonymphs (Fig. 3b). Under conditions

Table 1. Selected life history traits of Xylocopa tranquebarorum with and without the symbiotic mite, Sennertia nr alfkeni

No. of nestsa No. of provisioned cells No. of cells with bees No. of emerged beesb

Mite-laden 22 (1) 101 87 49 (33) Mite-free 10 (3) 22 20 5 (4)*

a Total number of nests (number of unprovisioned nests). b Total number of emerged adults (number of emerged females). * Sex ratio between mite-laden and mite-free offspring was tested using Fisher’s exact test; pϽ0.05.

Fig. 3. Mite aggregates on a host carpenter bee. (a) Protonymphal (whitish round particles) and deutonymphal (brown parti- cles) aggregates on the surface of a female pupa. (b) Deutonymphal aggregates attaching to a male for phoresy.

AEZ45-2_329-337.pdf 5 10.5.26 1:52:46 PM 334 K. OKABE et al.

Table 2. Developmental period of Sennertia nr alfkeni associated with exotic Xylocopa tranquebarorum (meanϮSE) reared at 25°C

Egg Larva Protonymph Tritonymph Egg to adult (/) Egg to adult (?)

n 36 24 19 21 22 26 Developmental 4.8Ϯ0.15 4.3Ϯ0.16** 5.8Ϯ0.29 4.1Ϯ0.21 19.0Ϯ0.29 19.6Ϯ0.30* period (d)

Each developmental period was compared to that of S. alfkeni in Okabe et al. (2008) and significant differences are indicated by *(pϽ0.05) and ** (pϽ0.001).

of both 25°C and ambient temperature, mites main- pieces, and/or processed into an ornament) was not tained separately from the host but with a host reported in the available statistics. Bamboo-export- pollen loaf and commercial pollen collected by ing countries have also changed: the main export honeybees turned into deutonymphs at nearly the country shifted from Taiwan to China in the 1990s, same time as mites maintained with the host. and in 1997, Vietnam, instead of Thailand, began The developmental period did not significantly to export bamboo to Japan. For over 10 years prior differ between female and male mites (Mann-Whit- to the first report of X. tranquebarorum invasion, ney U-test: pϭ1.00; Table 2). Due to the mass rear- over 90% of bamboo was exported from China, ex- ing method, juvenile (including egg) mortality was cluding Taiwan. Because bamboo was not a major not evident. Larval and male developmental peri- imported good in Japanese seaports, no specific in- ods of S. nr alfkeni mites significantly differed formation on the amount of imported bamboo was from those of S. alfkeni mites (Table 2). available to the public. Although we could not trace Two unidentified fungal species, Candida sp. imported bamboo within the country using the sta- and Cladosporium cladosporioides (Fres.) de tistical data, Chiba, Kyoto, and Aichi prefectures Vries, were collected from carpenter bee nests at house the most bamboo timber companies (Ma- the final stage (bees had became pupae or had tsuno, 2006). Two such companies in Aichi Prefec- emerged). In cells with fungi, we were able to de- ture primarily imported bamboo from China and tect the dark color of fungal materials within the transported it throughout almost the entire country. mite gut contents through the idiosomal outer skin. The invasive pathway of the carpenter bee and The effects of the fungi on bee development were its associated mite remains unclear, but these his- unclear, but fungal contamination (e.g., powdery torical data suggest that the two species likely orig- spores) was conspicuous in failed cells and nests. inated in bamboo material from an Asian country Regardless of whether they had attached to the or countries. We suspect that because the mite host, deutonymphs of S. nr alfkeni never molted to species associated with X. tranquebarorum both tritonymphs. genetically and morphologically differs from S. horrida, which is associated with the same bee species in Taiwan, but is morphologically similar DISCUSSION to S. potanini from continental Asia (Kawazoe et Japan consistently imported bamboo from 1981 al., 2010), and because there have been few recent to 2006, but the extent of imports has gradually imports from Taiwan, S. nr alfkeni likely originated declined since 1994 (Fig. 2; information was from the continental Asian region. Whether X. obtained by National Statistics Center (e-Stat), tranquebarorum expanded its distribution from http://www.e-stat.go.jp/SG1/estat/GL71050103. Anpachi and Toyota to Nagoya (midway between do;jsessionidϭYJYtKwcYLyWN2kThGCP93jz2R the two cities) or was distributed from Toyota to tRG2QrMdbD2LRYtdPdtQ7hQwm7J!46332371! Anpachi, including Nagoya, remains unclear. Fur- 1827439296?_toGL71050103_&listIDϭ00000 ther monitoring is needed to better clarify the dis- 1044766&forwardFromϭGL71050101 and the tribution of the species. Forestry Agency). Whether bamboo was processed In Japan, Phyllostachys bamboo is a popular ma- or unprocessed (“processed bamboo” means, for terial traditionally used for processed goods, con- example, that bamboo was well dried, cut into struction posts, and plant stakes in agriculture. The

AEZ45-2_329-337.pdf 6 10.5.26 1:52:48 PM Invasion of Xylocopa tranquebarorum with Mite 335

average inner diameter of the carpenter bee nest servlet/BinaryDownloaderServlet/133703_E.pd.pd (13–15 cm) is nearly equivalent to that of one of f?filenameϭ1240490152156_ISPM_15_Revised_ the most popular bamboo species, Phyllostachys 2009_E.pdf&refIDϭ133703). We conclude that bambusoides Sieb. et Zucc. The carpenter bee did processed wood represents an important pathway not appear to prefer very fresh bamboo, perhaps of exotic insect introduction, particularly those in- because mortality is higher in fresh bamboo; we habiting dried wood. Additionally, unintentionally observed that offspring failed to develop in a fresh introduced insects may serve as an invasive path- nest (nϭ1). Together, our results suggest that the way for symbiotic microorganisms by vectoring carpenter bee may have been introduced with them. processed (dried before being loaded for export) Although S. nr alfkeni is morphologically similar bamboo; the carpenter bee is a potential pest in the to the Japanese mite S. alfkeni, genetic analysis bamboo industry; and harm to humans via acciden- separated the two species, and the former is consid- tal stings can be expected in agricultural fields and ered to have been introduced with X. tranquebaro- perhaps in botanical gardens, where bamboo is rum (Kawazoe et al., 2010). The Japanese carpen- often planted. ter bee, X. a. circumvolans, sometimes damages The preferred nesting materials of X. tranque- wooden buildings through the construction of nests barorum also indicate that the bee was probably ac- (Sakagami and Maeta, 1986), but its associated cidentally introduced while nesting in relatively mite has only generated scientific, and not public, dry, processed bamboo. Beetles such as Purpuri- interest. However, mites, particularly astigmatids, cenus temminckii (Guerin-Meneville) (Cerambyci- are infamous house dust components and allergic dae) and Dinoderus minutus (Fabricius) agents (e.g., Hughes, 1976). The mite associated (Bostrychidae) are listed as domestic dried bamboo with X. tranquebarorum feeds on pollen and bee pests (Uchimura, 2004). Many insects bore into feces and, as a result, may accidentally ingest fun- deadwood in forests (e.g., Coulson and Witter, gal contaminants of the bee nest. Because mite 1984), and species ranging from termites, moths, numbers were very high under artificially con- sawflies, and beetles have been listed as timber and trolled conditions, as seen in pupal and emerged lumber pests (Nobuchi and Suzuki, 1993). In bees (Fig. 3), we suggest that the mite will achieve Japan, imported live plants and timber must be ex- high numbers if the carpenter bee constructs in- amined under the current plant quarantine system, door nests or if the mite spreads indoors. This pos- but such investigations are not required for im- sibility suggests that S. nr alfkeni may pose a house ported lumber and processed wood materials. dust risk but only under certain conditions. Be- Thus, xylophagous insects feeding on the core of cause the mite deutonymphs did not molt under ar- deadwood escape quarantine and gain the opportu- tificial conditions with or without hosts, we suspect nity to establish populations, often becoming pests that the mite requires a certain stimulus, perhaps of building materials. For example, a termite provided by the host, to become a deutonymph, as species, Incisitermes minor Hagen (Blattaria, Ter- seen in the house dust mite Lardoglyphus konoi mitidae), is suspected to have been introduced with (Sasa et Asanuma) and its phoretic host beetle imported woods and wooden furniture, and a (Okamoto et al., 1991) as well as in Hemisarcoptes wood-boring beetle, Lyctus planicollis LeConte mites and Chilocorus beetles (Houck, 1999). (Coleoptera, Lyctidae), may have been introduced Therefore, we conclude that S. nr alfkeni will not with dried lumber infected after the drying process become house dust when separated from its host, during storage for export (Goto, 2003; Ohmura and even if X. tranquebarorum (and the mite) becomes Tokoro, 2003). In North America, invasive wood- common in Japan. boring insects directly and indirectly affect many Because we did not exhaustively assess the life ecological processes (Gandhi and Herms, 2009). history traits and life cycle of S. nr alfkeni, we can Furthermore, wood packaging is recognized as a only suggest that this species is capable of switch- critical pathway of invasive wood-associated or- ing hosts from X. tranquebarorum to X. a. circum- ganisms and is now regulated by the International volans based on the similarity between S. nr alfkeni Standards for Phytosanitary Measures No.15 and S. alfkeni in the mite developmental period. (Reaser et al., 2008; IPPC https://www.ippc.int/ Other life history traits were also very similar be-

AEZ45-2_329-337.pdf 7 10.5.26 1:52:48 PM 336 K. OKABE et al.

tween the two species; for example, both mites left sociated mite during this initial invasive stage. the host during host provisioning and quickly ACKNOWLEDGMENTS changed feeding stages, and both became deu- tonymphs during host pupation and attached to We thank Drs. T. Mano and K. Goka for technical help and emerged hosts, but many of the deutonymphs as valuable comments during this study. Dr. K. Yamagishi, Mr. S. well as feeding stages remained in nests after hosts Ohkusa, and Mr. M. Okada provided information on the inva- sion of X. tranquebarorum. This study was supported by the had left and overwintered with the hosts (Okabe et Global Environment Research Fund (F-081). al., 2008; this study). Thus, if the life cycle of X. tranquebarorum is very similar to that of X. a. cir- REFERENCES cumvolans, mites that switch hosts could survive. Abrahamovich, A. H. and A. D. B. de Alzuet (1990) Tipos Because X. a. circumvolans overwinters in existing de asociacion entre acaros e hymenopteros. I: Sennertia cavities, abandoned nests of X. tranquebarorum (A.) splendidula Alzuet y Abrahamovich, 1988 (Acarina: provide the opportunity for a host switch of S. nr Chaetodactylidae) comensal y foretico de Xylocopa (S.) alfkeni (Okabe et al., 2008). Although a test cross splendidula splendidula Lepeletier, 1984 (Hymenoptera: Anthophoridae). Turrialba 40: 319–322. between mite species was not conducted (colonies Cardoza, Y. J., J. C. Moser, J. D. Klepzig and K. F. Raffa of S. alfkeni were not available), long-term con- (2008) Multiparite symbioses among fungi, mites, ne- cerns include genetic contamination if the two matodes, and the spruce beetle, Dendroctonus rufipennis. species can mate and/or the replacement of one Environ. Entomol. 37: 956–963. species via a host switch. Furthermore, the patho- Coulson, R. N. and J. A. Witter (1984) Forest Entomology. genicity of S. nr alfkeni or its associated microor- Wiley-Interscience, New York. 669pp. Eickwort, G. C. (1988) The origins of mites associated with ganisms to native Japanese carpenter bees, includ- honey bees. In Africanized Honey Bees and Bee Mites (G. ing X. a. circumvolans, remains unknown. R. Needham, R. E. Page Jr, M. Delfinado-Baker and C. E. The invasion of organisms associated with alien Bowman, eds.). Ellis Horwood Ltd., Chichester, pp. species poses both direct and indirect risks. For ex- 327–338. ample, Varroa mites, which are infamous parasites Eickwort, G. C. (1994) Evolution and life-history patterns of mites associated with bees. In Mites (M. A. Houck, ed.). of the European honeybee, are suspected to have Chapman and Hall, New York, pp. 218–251. been introduced from Asia into Europe and South Gandhi, K. J. K. and D. A. Herms (2010) Direct and indirect and North America during phorestic overwintering effects of alien insect herbivores on ecological processes (Eickwort, 1988). Fungi, nematodes, and mites are and interactions in forests of eastern North America. common associates of bark beetles as symbionts in Biol. Invas. 12: 389–405. beetle galleries and are dispersed by phoresy on the Gibbs, J. N. (1978) Intercontinental epidemiology of Dutch elm disease. Ann. Rev. Phytopathol. 16: 287–307. beetles (Cardoza et al., 2008). Because some mites Goka, K., K. Okabe, M. Yoneda and S. Niwas (2001) Bum- are predacious, parasitic, or hyper-phoretic hosts of blebee commercialization will cause worldwide migra- blue stain fungi (Klepzig et al., 2001), accidental tion of parasitic mites. Mol. Ecol. 10: 2095–2099. introductions of phoretics and possible future Goto, N. (2003) Invasive wood-boring beetles in Japan 2— phoretic host switches, as seen in the Varroa mites, dry wood pests and bark beetles—. Shinrin Kagaku 38: 17–20 (in Japanese). may impact beetle populations and forest dynam- Hayashi, K., T. Ichikawa and Y. Yasui (2010) Life history of ics. As observed with the well-known destructive the newly discovered Japanese tree sap mite, Hericia tree wilt pathogens, Dutch elm disease and the pine sanukiensis (Acari, Astigmata, Algophagidae). Exp. wilt nematode, vector switches often lead to the Appl. Acarol. 50: 35–49. collapse of endemic vegetation (Gibbs, 1978; Houck, M. A. (1999) Phoresy by Hemisarcoptes (Acari: Mamiya, 1988). The eradication of microorgan- Hemisarcoptidae) on Chilocorus (Coleoptera: Coccinelli- dae): influence of subelytral ultrastructure. Exp. Appl. isms, including mites, is extremely difficult. Once a Acarol. 23: 97–118. host switch occurs between X. a. circumvolans and Hughes, A. M. (1976) The Mite if Stored in Food and S. nr alfkeni, eradicating the mite will be nearly im- Houses. Her Majesty’s Stationary Office, London. possible without eradicating the native Japanese 400 pp. bee species. A similar situation may also occur on Hulme, P. E. (2009) Trade, transport and trouble: managing invasive species pathways in an era of globalization. J. other islands on which cryptic species of S. alfkeni Appl. Ecol. 46: 10–18. have been distributed; therefore, we highly recom- Hurd, P. D. (1958) The carpenter bees of the eastern pacific mend the control of X. tranquebarorum and its as- oceanic islands. J. Kansas Entomol. Soc. 31: 249–255.

AEZ45-2_329-337.pdf 8 10.5.26 1:52:48 PM Invasion of Xylocopa tranquebarorum with Mite 337

Hurd, P. D. and J. S. Moure (1963) A classification of the nese). large carpenter bees (Xylocopini) (Hymenoptera: Okabe, K. and K. Goka (2008) Potential impacts on Japa- Apoidea). Univ. Calif. Publ. Entomol. 29: 1–365. nese fauna of canestriniid mites (Acari: Astigmata) acci- Kamio, K. (2007) About an observation record of Xylocopa dentally introduced with pet lucanid beetles from South- tranquebarorum in Toyota, Aichi Prefecture. Tsuneki- east Asia. Biodiv. Conserv. 17: 71–81. bachi 12: 21–25 (in Japanese). Okabe, K. and S. Makino (2005) Mite faunas and morphol- Kawazoe, K., A. Kawakita, Y. Kameda and M. Kato (2008a) ogy of acarinaria on large carpenter bees (Hymenoptera: Redundant species, cryptic host-associated divergence, Apidae). J. Acarol. Soc. Jpn. 14: 105–115. and secondary shift in Sennertia mites (Acari: Chaeto- Okabe, K., S. Makino and T. Endo (2008) Polymorphism in dactylidae) associated with four large carpenter bees (Hy- the deutonymph and adult of Sennertia alfkeni (Acari: menoptera: Apidae: Xylocopa) in the Japanese island arc. Chaetodactylidae) associated with the large carpenter Mol. Phyl. Evol. 49: 503–513. bee, Xylocopa appendiculata circumvolans (Hy- Kawazoe, K., A. Kawakita, S. Sugiura and M. Kato (2008b) menoptera: Apidae). J. Nat. Hist. 42: 1361–1384. Phylogenetic position of the endemic large carpenter bee Okada, M. and K. Takeda (2009) Overwintering Xylocopa of the Ogasawara Islands, Xylocopa ogasawarensis (Ma- tranquebarorum in Toyota, Aichi Prefecture. Gekkan tsumura, 1912) (Hymenoptera: Apidae), inferred from Mushi 461: 59–60 (in Japanese). four genes. Zool. Sci. 25: 838–842. Okamoto, M., K. Matsumoto and R. Shirasaka (1991) Stud- Kawazoe, K., K. Okabe, A. Kawakita and M. Kato (2010) ies on the attaching behavior of Lardoglyphus konoi An alien Sennertia mite (Acari: Chaetodactylidae) associ- (Acari, Ladoglyphidae) hypopus and its molting into the ated with an introduced Oriental bamboo-nesting large tritonymph. Jpn. J. Saniy. Zool. 42: 219–228. carpenter bee (Hymenoptera: Apidae: Xylocopa) invad- Perrings, C., K. Dehnen-Schmutz, J. Touza and M. Williamson ing the central Honshu Island, Japan. Entomol. Sci. (in (2005) How to manage biological invasions under glob- press). alization. Trends Ecol. Evol. 20: 212–215. Klepzig, K. D., J. C. Moser, F. J. Lombardero and M. P. Ayres Pimentel, D. (2002) Biological Invasions. CRC Press, New (2001) Symbiosis and competition: complex interac- York. 369pp. tions among beetles, fungi and mites. Symbiosis 30: Reaser, J. K., L. A. Meyerson and B. Von Holle (2008) Sav- 83–96. ing camels from straws: how propagule pressure-based Klimov, P. B. and B. M. OConnor (2008) Morphology, Evo- prevention policies can reduce the risk of biological inva- lution, and Host Associations of Bee-Associated Mites of sion. Biol. Invas. 10: 1085–1098. the Family Chaetodactylidae (Acari: Astigmata).Mu- Reichard, S. H. and P. White (2001) Horticulture as a path- seum of Zoology, Ann Arbor. 243 pp. way of invasive plant introductions in the United States. Mamiya, Y. (1988) History of pine wilt disease in Japan. J. BioScience 51: 103–113. Nematol. 20: 219–226. Sakagami, S. F. and Y. Maeta (1986) From Loneliness to Un- Matsuno, K. (2006) Production and distribution of bamboo. equalness. Tokai Univ Press, Tokyo. 264 pp. In Study of Distribution of Forest By-products, Report of StatSoft Inc. (2005) STATISTICA Pro 0.6J. StatSoft Japan, the Forest Policy Research Institute. No. 69 (K. Matsuno, Tokyo. ed.). Forest Policy Research Institute, Tokyo, pp. Uchimura, E. (2004) Fascination and Uses of Bamboo. 236–277. Soshin-sha, Tokyo. 215 pp. Meyerson, L. A. and H. A. Mooney (2007) Invasive alien Xu, H., S. Qiang, Z. Han, J. Guo, Z. Huang, H. Sun, S. He, H. species in an era of globalization. Front. Ecol. Environ. Ding, H. Wu and F. Wan (2006) The status and causes 5: 199–208. of alien species invasions in China. Biodiv. Conserv. 15: Nobuchi, A and K. Suzuki (1993) Dry Wood Pests and In- 2893–2904. sects Collected Indoors. Forestry Science and Technol- Yata, N. (2007) Collection records of Xylocopa tranquebaro- ogy Institute, Tokyo. 96 pp. rum in Aichi and Gifu prefectures. Gekkan Mushi 439: OConnor, B. M. (1993) The mite community associated with 39–40 (in Japanese). Xylocopa latipes (Hymenoptera: Anthophoridae: Xylo- Zakhvatkin, A. A. (1941) Fauna of USSR Arachnoidea Vol. copinae) with a description of a new type acarinarium. VI, No.1, Tyroglyphoidea (Acari) (translated and edited Int. J. Acarol. 19: 159–166. by A. Ratcliffe and A. M. Hughes). American Institute Ohmura, W. and M. Tokoro (2003) Invasive termites from of Biological Science, Washington DC. 573 pp. overseas to Japan. Shinrin Kagaku 38: 7–9 (in Japa-

AEZ45-2_329-337.pdf 9 10.5.26 1:52:48 PM