Jpn. J. Infect. Dis., 63, 401-404, 2010

Original Article Laboratory Colonization of japonicus japonicus (Diptera: Culicidae) Collected in Narita, Japan and the Biological Properties of the Established Colony

Keita Hoshino, Haruhiko Isawa, Yoshio Tsuda*, and Mutsuo Kobayashi Department of Medical Entomology, National Institute of Infectious Diseases, Tokyo 162-8640, Japan (Received May 17, 2010. Accepted August 23, 2010)

SUMMARY: A laboratory colony of the Aedes japonicus japonicus, which has recently in- vaded the United States and is recognized as a highly competent vector of , was estab- lished from larvae collected in Narita, Japan. The mosquitoes were maintained with induced insemina- tion, blood-feeding on humans, and oviposition in water provided from the original collection site during the first few generations, then the colony was transferred to a large cage (40 × 40 × 100 cm in height) and adapted to conditions in which specimens were allowed to mate freely. White mice were provided as the blood source, and deionized water was available for oviposition. Approximately 185 eggs, most of which were tolerant to desiccation for at least 1 month, with some surviving for up to 2.5 months, were obtained per female following a single blood-feeding. The rate of successful emergence was nearly 90z, although this rate decreased significantly at high larval densities. The colony has been maintained for 5 years, and developmental profiles of the species have been obtained during that time.

tablishment of a laboratory colony of Ae. j. japonicus INTRODUCTION will therefore contribute to both basic and applied Aedes japonicus (Theobald) is a mosquito distributed research associated with the vector competence and the primarily in eastern Asia, including Japan, , reproductive ecology of this species. A colony of an in- , Russia, and (1). Aedes j. japonicus,the vasive population of Ae.j.japonicusfrom New Jersey original subspecies of this species, was first found to has been established recently (14), although no similar haveinvadedtheeasternUnitedStatesin1998(2)and colony of the original population of Ae.j.japonicusex- has since expanded to other areas including Canada, ists to date. Herein we report the establishment and France, and Belgium (3–6). Ae. j. japonicus has also maintenance of a laboratory colony of Ae. j. japonicus been confirmed to be an important vector of West Nile from Japan, together with descriptions of the coloniza- virus (WNV) in , based on experimental tion methodology and some basic biological properties analysis of its vector competence and field isolations of obtained from this colony. These findings may prove WNV (7,8). Additionally, it was reported that Ae. j. useful for the characterization of this species, including japonicus is also susceptible to invasive populations, and provide the information re- virus (9), St. Louis encephalitis virus (10), Eastern quired to design laboratory experiments using this equine encephalomyelitis virus (11), and La Crosse virus colony, such as infection experiments with mosquito- (12), indicating that this species is a potential vector for borne pathogens. these viruses. Ae. j. japonicus is a common inhabitant of wooded environments, including residential areas, in MATERIALS AND METHODS palearctic Japan, whereas two other subspecies, Ae. japonicus amamiensis (Tanaka, Mizusawa & Saugstad) Origin of mosquito colony: The various instar larvae and Ae. japonicus yaeyamaensis, (Tanaka, Mizusawa & of Ae. j. japonicus (approximately 500 individuals) were Saugstad), are distributed sporadically only in the collected from the stone basin of a shrine in Narita Amami and Ryukyu Archipelago, respectively, at the (35938?27!N, 140928?52!E), Chiba Prefecture, Japan, southernmost area of Japan (1,13). Despite this, in June and July 2004, and they were used to establish however, very few experimental data are available for the laboratory colony. Ae. j. japonicus in Japan due to the difficulties in estab- Maintenance of mosquitoes: Thelarvaewerereared lishing a laboratory colony, although the bionomics of at a density of approximately 10 larvae per 100 mL and this subspecies as regards distribution, host preferences, fedwithfishfoodTetraMin(fortheearlyinstarlarvae; and larval habitats have been sporadically reported. Es- Tetra, Melle, Germany) and SWIMMY BABY (for the late instar larvae; Nippon Pet Food, Tokyo, Japan) in a white plastic tray (W:L:D ＝ 30 × 20 × 5cm).Anap- *Corresponding author: Mailing address: Department of propriate amount of larval food (10–20 and 30–50 Medical Entomology, National Institute of Infectious mg/day during the 1st to 3rd and 4th larval stages, re- Diseases, Toyama 1-23-1, Shinjuku-ku, Tokyo 162-8640, spectively) was supplied on the basis of the number and Japan. Tel: ＋81-3-5285-1111, Fax: ＋81-3-5285-1147, developmental stage of the larvae. Additional food was E-mail: tsudayso＠nih.go.jp provided once the blocks of larval food had been con-

401 sumed and disappeared completely. Once the pupae ap- eggs to dry in an airtight container for between 0 and 11 peared, they were carefully transferred into a cup filled weeks after oviposition. The survival rate of eggs was with deionized water and placed inside a cage with stain- determined by counting the number of hatched larvae less supports covered with nylon-mesh net (20 × 30 × and confirming the completely collapsed non-hatching 20 cm) for eclosion. The emerging adults were supplied eggs by optical microscopy in November 2005. with cotton soaked in an approximately 3z sucrose so- To examine the developmental period of the colony, lution. The adults were reared in a 20 × 30 × 20 cm hatched larvae were reared individually in 20 mL of cage with frequent opportunities to feed, which in- pure water with 5 mg of larval food, which were volved introducing a human arm into the cage up to the changed daily, and their developmental stage recorded wrist as the first generation of mosquitoes was not at- twice a day at the beginning and end of the light period. tracted to white mice. Artificial mating was subsequent- The developmental and survival effects of the density ly performed as described previously (15,16). A glass of immature stages were determined by measuring the dish (110 mm in diameter, 60 mm in depth) was lined adult emergence rate and the mortality of immature with Advantech filter paper Grade 2 (Toyo Roshi stages at four different initial densities of hatched larvae Kaisha, Tokyo, Japan) positioned with the rough side with a fixed amount (25 mg per day) of larval food. A facing the inside for oviposition. This dish was filled to glass dish (110 mm in diameter, 60 mm in depth) filled the halfway mark with water and placed stably in the with 100 mL of water was used, and both the water and cage of approximately 50 gravid females for 1 week. food were changed daily. Water from the original larval habitat was used to Adult survivorship and longevity of the established stimulate oviposition. The eggs deposited on the filter colony were investigated by recording the number of paper were submerged in a tray filled with deionized deaths every day for three experimental cohorts of ap- water. A number of adults which had developed from proximately 20 pairs of newly emerged adults reared larvae in the original habitat were subsequently collect- with a 3z sucrose solution in a cage. ed and placed into the cage along with the second gener- ation in July 2004. The next five generations of the RESULTS AND DISCUSSION colony were maintained under the same conditions de- scribed above. Then approximately 3,000 adults from Colony maintenance: To improve free mating, the the following generation were transferred into 40 × large cage was designed to have a wide vertical space 40 × 100 cm cage, the top of which was covered with with no direct light other than a few beams of light, pro- aluminum foil containing a few small holes. The adults duced by covering the top of the cage with aluminium were allowed to mate freely for a period of 3–4 weeks foil containing several small holes, to simulate sun- after emergence. Adult females were allowed to take a beams passing through tree leaves in the mosquito's blood meal from anesthetized white mice (Japan SLC, natural wooded habitat. We observed active flight with Hamamatsu, Japan) for 2 h in the latter half of the light phototaxis toward the light beam from the top of the period in order to mimic the blood-feeding habit of this cage. Such environmental conditions have been suggest- species in its natural environment. Additionally, the ed to be essential for the successful mating of some blood-feeding activity on mice was enhanced for the Aedes and Toxorhynchites mosquitoes in their native first few generations by exhaling over the mice for a few habitat (woodland) (16,17). Deposited eggs were found minutes, since these mosquitoes exhibited a low prefer- on the filter paper 5 days after blood-feeding. These ence for white mice. Deionized water was used for eggs subsequently hatched into neonate larvae. The oviposition. All developmental stages of mosquitoes hatching process was observed within 2 days. The eggs were maintained in a controlled insectary at 50–70z rel- on the filter paper were submerged in the water and any ative humidity, a temperature of 259C, and a light/dark air bubbles attached to them were removed by pipetting photoperiod of 16/8 h, with no dawn or dusk light dim- the water. A number of eggs hatched, indicating that ming. they were fertilized and deposited by females which had Basic biological properties of the established colony: mated successfully under our laboratory conditions. We The numbers of eggs per female, the hatching rate, the confirmed that the hatched larvae grew and metamor- tolerance of eggs to desiccation, the larval development, phosed into adults, with the next generation being ob- the density effects of immature stages, and the adult tained in the same way. The colony was successfully survivorship were examined to describe the basic biolog- maintained under the same conditions thereafter. ical properties of the established colony. All data ob- Basic biological properties of the colony: The average tained are quoted as mean and standard deviation. number of deposited eggs per individual was 185.3 ± Statistical analyses were performed by one-way analysis 24.0 (range, 165–218) in June 2008 (n ＝ 4). Oviposition of variance (ANOVA) or Student's t test, followed by of fertilized eggs was rarely observed (1 out of 40 fe- Tukey's honestly significant difference (HSD) test for males) in the initial phase, immediately after the change multiple comparisons of means. to free mating, but took place frequently (4 out of 6 fe- The average number of eggs deposited by an individ- males) in June 2008, indicating that the rate of success- ual female after a single blood meal was estimated by ful mating increased with successive generations. using randomly selected gravid females. To evaluate the The successful hatching rate was only 2.0 ± 0.7z in hatching rate under free mating conditions, the number the initial colonization phase in June 2005, and it subse- of hatched larvae per 100 eggs of 7 days post-oviposi- quently increased to 9.6 ± 2.3z,32.8± 6.4z,46.6± tion were counted within 2 days after submersion in 6.2z,and51.4± 11.5z in November 2005, June water. The desiccation tolerance of colony laid eggs was 2006, June 2007, and June 2008, respectively. This in- measured by examining the hatch rate after allowing the crease in successful hatching rate with increasing gener-

402 ation number was also reflected in an increase in the free prior to desiccation. The rate did not change furthre for mating success rate. The successful mating of Ae. j. up to 5 weeks, but then started to decrease after 6 japonicas in our colony was also observed. Based on weeks; and no hatching was detected after desiccation these observations, Ae. j. japonicus appears to have for 9 weeks. However, a very low hatching rate (prob- a mating system comprised of swarming-like syn- ably º10－2) was observed up to a maximum of 11 chronized up-down flight and mating flight, similar to weeks when extremely large numbers of eggs (À104) that of other mosquitoes (16). The increase in the suc- were examined (data not shown). These results show cessful hatching rate may therefore be largely due to that eggs of Ae.j.japonicushave a similar tolerance to adaptation of the male mating activity to the laboratory desiccation as those of other Aedes mosquitoes (21). conditions, which was suggested to be important for This is consistent with the life history of this species, free-mating in Anopheles and Culex spp. in laboratory particularly as regards overwintering in the egg stage. colonies (18–20). The developmental periods of the mosquitoes Figure 1 shows the tolerance of eggs to desiccation. differed between males and females (Table 1). Thus, The hatching rate decreased slightly after desiccation following hatching, the shortest time necessary for the for 1 week with respect to the rate for eggs which had male and female in the colony to grow into not been desiccated, although this difference may be adults was 10.5 and 12.0 days, respectively. The females due to errors arising from the loss of larvae that hatched spent significantly longer (1.05 and 0.51 days; Student's t test, P º 0.001) in the 4th instar larval and pupal period, respectively. Interestingly, the laboratory colony of the New Jersey strain of Ae.j.japonicus(14) also spent a longer time in the 4th instar larval and pupal period, whereas the periods for other instar larvae were consistent with those of our laboratory strains. In addition, the deaths observed tended to be concentrated during the 4th instar larval and pupal stages; none were found in either the 1st or 2nd instar larval stages in our Narita strain, thus suggesting that any developmental fragility may occur during these specific stages. The adult emergence rates were affected by the rear- ing density of the larvae (Table 2). An extremely high Fig. 1. Hatching rate of Aedes japonicus japonicus eggs in vari- density of 500 larvae per 100 mL led to both develop- ous desiccation period. Bars represent means ± SE. Bars with different letters are significantly different (P º 0.01; mental delay and a significant decrease in the numbers ANOVA). of larvae. We also observed that these larvae began to

Table 1. Developmental periods of immature stage for the established colony of Aedes japonicus japonicus

Period (days)

Developmental 1) stage Male (n ＝ 30) Female (n ＝ 23) P value mean ± s.d. range mean ± s.d. range

Larva 1st instar 2.08 ± 0.19 2.0–2.5 2.04 ± 0.14 2.0–2.5 0.4054 2nd instar 1.65 ± 0.27 1.0–2.0 1.54 ± 0.21 1.0–2.0 0.1211 3rd instar 2.10 ± 0.20 2.0–2.5 2.35 ± 0.35 2.0–2.5 0.0022 4th instar 3.15 ± 0.48 2.5–4.5 4.20 ± 0.45 3.5–5.0 º0.0001

Pupa 2.51 ± 0.09 2.5–3.0 3.02 ± 0.10 3.0–3.5 º0.0001

Larva ＋ 11.50 ± 0.71 10.5–12.5 13.15 ± 0.80 12.0–14.5 º0.0001

1): The value was estimated by t tests between both sexes in the same developmental stage.

Table 2. Effect of rearing density of larvae on adult emergence and mortality in Aedes japonicus japonicus

Density Adult emergence rate (z) Mortality (larva/pupa) (z) Replication (Larvae/100 mL) mean ± s.d.1) mean ± s.d.

10 88.0 ± 8.7a 8.0 ± 4.5a/4.0 ± 5.5a 5 50 78.8 ± 6.1a 8.8 ± 2.3a/12.4 ± 3.8b 5 100 55.0 ± 10.2b N.D. 5 500 4.7 ± 1.5c N.D. 5

1): Values followed by the same letter are not significantly different in each vertical line (P º 0.05, Tukey's HSD test). N.D., not determined.

403 REFERENCES 1. Tanaka, K., Mizusawa, K. and Saugstad, E.S. (1979): A review of the adult and larval mosquitoes of Japan (including the Ryukyu archipelago and the Ogasawara Island) and Korea (Diptera: Cu- licidae). Contrib. Am. Entomol. Inst., 16, 1–987. 2. Peyton, E.L., Campbell, S.C., Candeletti, T.M., et al. (1999): Aedes (Finlaya) japonicus japonicus (Theobald), a new introduc- tion into the United States. J. Am. Mosq. Control Assoc., 15, 238–241. 3. Schaffner, F., Chouin, S. and Guilloteau, J. (2003): First record of Ochlerotatus (Finlaya) j. japonicus (Theobald, 1901) in metropolitan France. J. Am. Mosq. Control Assoc., 19, 1–5. 4. Morris, J.A., Lampman, R.L., Ballmes, G., et al. (2007): First Fig. 2. Survivorship curves for a cohort of adult Aedes japonicus record of Aedes japonicus japonicus in Illinois: defining its spa- japonicus. Adults of both sexes, male (n ＝ 56) and female (n tial distribution and associated mosquito species. J. Am. Mosq. ＝ 55), were examined. Control Assoc., 23, 243–251. 5. Hughes, T.H., Irwin, P.M., Kaufman, A., et al. (2008): First records of Aedes japonicus japonicus in Wisconsin. J. Am. Mosq. Control Assoc., 24, 583–584. bite solid foods, the surface of the tray, and even other 6. Versteirt, V., Schaffner, F., Garros, C., et al. (2009): Introduc- larvae, suggesting that the rapid decrease in the numbers tion and establishment of the exotic mosquito species Aedes of larvae observed at high rearing densities could be a japonicus japonicus (Diptera: Culicidae) in Belgium. J. Med. result of the increased likelihood of the larvae biting Entomol., 46, 1464–1467. 7. Turell, M.J., Dohm, D.J., Sardelis, M.R., et al. (2005): An up- each other. A rearing density of º50 larvae per 100 mL date on the potential of North American mosquitoes (Diptera: should therefore be sufficient to provide enough adult Culicidae) to transmit West Nile virus. J. Med. Entomol., 42, mosquitoes to maintain a stable colony. The increase in 57–62. mortality observed during the pupal stage in the same 8. Centers for Disease Control and Prevention (2008): Mosquito experiments was also due to this biting habit, although it species producing WNV positives by year. Online at ºwww.cdc. gov/ncidod/dvbid/westnile/mosquitospecies.htmÀ. could be avoided by transferring the pupae to a separate 9. Takashima, I. and Rosen, L. (1989): Horizontal and vertical cup, which is the usual method for maintaining a transmission of Japanese encephalitis virus by Aedes japonicus colony. (Diptera: Culicidae). J. Med. Entomol., 26, 454–458. The maximum lifespan of the adults in the colony was 10. Sardelis, M.R., Turell, M.J. and Andre, R.G. (2003): Experimen- tal transmission of St. Louis encephalitis virus by Ochlerotatus j. 101 days for females and 89 days for males under the japonicus. J. Am. Mosq. Control Assoc., 19, 159–162. conditions described in the Materials and Methods sec- 11. Sardelis, M.R., Dohm, D.J., Paga, B., et al. (2002a): Experi- tion. According to the data in the survivorship curve mental transmission of Eastern equine encephalitis virus by (Fig. 2), adults of both sexes in the colony lived for at Ochlerotatus j. japonicus (Diptere: Culicidae). J. Med. Entomol., least 2 months under these conditions. The mean lon- 39, 480–484. 12. Sardelis, M.R., Turell, M.J. and Andre, R.G. (2002b): Laborato- gevity of male and female mosquitoes was 76.9 ± 9.9 ry transmission of LaCrosse virus by Ochlerotatus j. japonicus. J. and 84.5 ± 10.8 days, respectively, which is sufficient Med. Entomol., 39, 635–639. to be able to carry out infection experiments with 13. Toma, T. and Miyagi, I. (1986): The mosquito fauna of the Ryu- mosquito-borne pathogens. kyu archipelago with identification keys, pupal description and notes on biology, medical importance and distribution. Mosq. In conclusion, we have succeeded in adapting Ae. j. Syst., 18, 1–109. japonicus to laboratory conditions, and have stably 14. Williges, E., Farajollahi, A., Scott, J.J., et al. (2008): Laboratory maintained a colony for approximately 35 full genera- colonization of Aedes japonicus japonicus. J. Am. Mosq. Con- tions over a 5-year period. In addition, we have con- trol Assoc., 24, 591–593. firmed that a sufficient number of eggs to maintain the 15. Ow-Yang, C.F., Sta-Maria, F.L. and Wharton, R.H. (1963): Maintenance of laboratory colony of Anopheles maculatus Theo- colony can be obtained by a single blood-meal, and that bald by artificial mating. Mosq. News, 23, 34–35. additional oviposition as a result of secondary or fur- 16. Gerberg, E.J., Barnard, D.R. and Ward, R.A. (1994): Manual ther blood-feedings can also be successful. This new for Mosquito Rearing and Experimental Techniques. Am. Mosq. colony of Ae. j. japonicus will be used to provide Control Assoc. Bull., 5. Allen Press, Lowrence. 17. Clements, A.N. (1992): The biology of mosquitoes. vol. 1. material for further studies concerning virus susceptibil- Development, Nutrition and Reproduction. Chapman & Hall, ity and many other basic characteristics of this organ- London. ism. 18. Horio, M. and Tsukamoto, M. (1985): Successful laboratory colonization of three Japanese species of Toxorhynchites mosqui- toes. Jpn. J. Sanit. Zool., 36, 87–93. Acknowledgments This work was partly supported by a Grant-in- 19. Knight, J.W. and Nayer, J.K. (1999): Colonization of Culex Aid for Scientific Research of Emerging and Reemerging Infectious nigripalpus Theobald (Diptera: Culicidae) by stimulation of mat- Diseases from the Ministry of Health, Labour and Welfare of the ing using males of other mosquito species. J. Am. Mosq. Control Japanese Government (H18-Shinko-009, H21-Shinko-005). Assoc., 15, 72–73. 20. Lima, J.B., Valle, D. and Peixoto, A.A. (2004): Adaptation of a Conflict of interest None to declare. South American malaria vector to laboratory colonization sug- gests faster-male evolution for mating ability. BMC Evol. Biol., 4, 12. 21. Sota, T. and Mogi, M. (1992): Interspecific variation in desicca- tion survival time of Aedes (Stegomyia) mosquito eggs in correlat- ed with habitat and egg size. Oecologia, 90, 353–358.

404