Many microorganisms are symbionts to more than one insect host species, because they are physiologically versatile or pervasive in the environment, can provide various functions or benefits, or because they have simply radiated with their host (Bright and Bulgheresi Reference Bright and Bulgheresi2010; Woodbury Reference Woodbury2012). For example, the bacterium Enterobacter cloacae (Jordan 1890) Hormaeche and Edwards 1960 (Enterobacteriaceae) produces the aggregation semiochemical of the desert locust, Schistocerca gregaria (Forkskal) (Orthoptera: Acrididae) (Dillon et al. Reference Dillon, Vennard and Charnley2000, Reference Dillon, Vennard and Charnley2002; Dillon and Charnley Reference Dillon and Charnley2002), helps blowflies and fruit flies locate oviposition sites (Emmens and Murray Reference Emmens and Murray1983; Jang and Nishijima Reference Jang and Nishijima1990), and improves the development and survival of various other fruit fly hosts (Fitt and O'Brien Reference Fitt and O'Brien1985; Brummel et al. Reference Brummel, Ching, Seroude, Simon and Benzer2004). Firebrats, Thermobia domestica (Packard), giant silverfish, Ctenolepisma longicaudata Escherich, and common silverfish, Lepisma saccharina Linnaeus (all Thysanura: Lepismatidae), inhabit human dwellings with preference for a particular microclimate (firebrats: 32–44 °C, 50–98% relative humidity (Sweetman Reference Sweetman1938); giant silverfish: 24 °C, 55–82% relative humidity (Lindsay Reference Lindsay1940); common silverfish: 22–32 °C, 75–97% relative humidity (Lindsay Reference Lindsay1940)).
All three Thysanura species aggregate on substrates that have previously been visited by conspecifics and by some heterospecifics (Tremblay and Gries Reference Tremblay and Gries2003; Woodbury and Gries Reference Woodbury and Gries2007, Reference Woodbury and Gries2008). Specifically, substrate previously exposed to firebrats or giant silverfish induces arrestment behaviour of firebrats and giant-silverfish, but not common silverfish (Woodbury and Gries Reference Woodbury and Gries2007). The arrestment response to E. cloacae and Mycotypha microspora allows firebrats, which lack any form of long-distance communication, to locate conspecifics and prospective mates in suitable shelters (Woodbury and Gries Reference Woodbury and Gries2013).
We have recently shown that firebrats carry, and deposit with their faeces, the symbiotic bacterium E. cloacae and the fungus M. microspora Fenner, 1932 (Mycotyphaceae), and that these microbes induce arrestment behaviour and aggregation of firebrats (Woodbury and Gries Reference Woodbury and Gries2013; personal observations). Fluorescently labelled E. cloacae and M. microspora were present within digestive tracts, faeces, and shelters of firebrats, but not within firebrat ovarioles or eggs, indicating that firebrats consume these symbionts and transmit them to new habitats by defecation (Woodbury et al. Reference Woodbury, Moore and Gries2013). Like their firebrat host, both symbionts require sustained hot and humid environments (35 °C, >75% relative humidity; personal observation), and proliferate only in those conditions optimal for their host. Firebrats aggregate where their microbial symbionts thrive, responding to and consuming digested cellulose (glucose) produced by M. microspora (personal observation). Here, we tested the hypothesis that E. cloacae and M. microspora induce arrestment in firebrats and giant silverfish but not in common silverfish.
Insect specimens used for bioassays were obtained from three-week-old colonies maintained at the insectary of Simon Fraser University (SFU), Burnaby, British Columbia, Canada under an 8:16 light:dark light-cycle, at 23–35 °C and 70–85% relative humidity, as described for firebrats (Woodbury and Gries Reference Woodbury and Gries2007). Colonies comprised insects of mixed age and gender that were obtained from autoclave rooms and boilers around SFU.
To test the response of single adult female insects to E. cloacae and M. microspora, bioassays were run in custom-built still-air, three-chamber olfactometers consisting of two lateral Pyrex® glass Petri dishes (Fisher Scientific Company, Whitby, ON, Canada), connected to a central dish (all dishes 3 × 9 cm) via Pyrex® glass tubing (2.5 × 2 cm). This olfactometer design mimics the natural still-air shelters of firebrats. Before the start of bioassays, a folded cone of sterile filter paper (Whatman® No. 1, 125 mm diameter; Whatman International Ltd., Maidstone, United Kingdom) was placed (tip facing central chamber) into each lateral dish to serve as an artificial shelter. Treatment and control stimuli were randomly added to each lateral dish that was capped to prevent insect escape (Woodbury and Gries Reference Woodbury and Gries2007). The randomly assigned treatment chamber received a 0.5-mg (wet-weight) aliquot of E. cloacae (1 × 108 colony forming units (CFU) or M. microspora (1 × 106 CFU) from GlcNAc-agar cultures (originally isolated from firebrats) that was applied to the centre of a 1/8th piece of filter paper (Whatman® No. 1, 125 mm diameter), smeared into a single 0.5-cm2 spot, and dried in ambient air before bioassays. Control chambers received an aliquot of water absorbed from uninoculated GlcNAc agar, and applied onto a 0.5-cm2 spot of a 1/8th piece of clean filter paper that was air-dried before bioassays. As firebrats, giant silverfish, and common silverfish are nocturnal foragers that arrest in shelters during daylight, an individual insect was released into an olfactometer during the dark period, allowing it to explore the chambers during 10 hours of darkness, and to arrest in one of the shelters during 6 hours of light. Any insect not found in contact with either the treatment or control paper was recorded as a non-responder. All experimental replicates (n = 22–32 for each experiment) were run at 21 ± 2 °C and 30–60% relative humidity, which are conditions conducive to the response of all three Thysanura (Tremblay and Gries Reference Tremblay and Gries2003; Woodbury and Gries Reference Woodbury and Gries2007, Reference Woodbury and Gries2008). Olfactometers were washed with LiquiNox detergent (Alconox® Inc., White Plains, New York, United States of America), and were oven-dried (Pacific Combustion Engineering Co., Torrance, California, United States of America) between each bioassay. Choice data of responding insects in each experiment were analysed by χ 2 test, using JMP™ software tables (SAS®, Cary, North Carolina, United States of America). Insects not in contact with the treatment or control paper at the termination of experiments were considered non-responders and were not included in statistical analyses. Voucher specimens of all three species were deposited in the museum collection of Simon Fraser University.
Paper inoculated with E. cloacae induced arrestment by firebrats but not by giant silverfish or common silverfish (Fig. 1), and paper inoculated with M. microspora arrested firebrats and giant silverfish but not common silverfish (Fig. 1). These results confirm reports that firebrats and giant silverfish interchangeably respond to the arrestment cue(s) conspecifics or heterospecifics deposited on substrate (Tremblay and Gries Reference Tremblay and Gries2003; Woodbury and Gries Reference Woodbury and Gries2007). The data also suggest that the fungal symbiont M. microspora, which we have isolated from field-collected firebrats and giant silverfish (data not shown), may represent or contribute to a common arrestment cue.
Number of adult female firebrats, Thermobia domestica (A), giant silverfish, Ctenolepisma longicaudata (B), and common silverfish, Lepisma saccharina (C) in single-insect bioassays arresting on a piece of filter paper inoculated with the bacterium Enterobacter cloacae or the fungus Mycotypha microspora. Numbers within bars indicate the number of insects responding to the treatment stimulus (upper bar) or control stimulus (lower bar). An asterisk (*) indicates a significant preference for a particular test stimulus (χ 2 test; *P ≤ 0.05). Numbers in brackets indicate numbers of non-responding insects.
Firebrats and giant silverfish as close relatives (Mendes Reference Mendes1991) seek shelters with similar microclimates for reproduction and offspring development (Sweetman Reference Sweetman1938; Lindsay Reference Lindsay1940). Firebrats aggregate in the sustained hot and humid environments (35 °C; >75% relative humidity) where M. microspora proliferates and breaks down cellulose to glucose, which firebrats consume (personal observations). As giant silverfish have similar microclimate and nutrient requirements as firebrats (Sweetman Reference Sweetman1938; Lindsay Reference Lindsay1940), they may accrue the same benefits when they arrest in response to the presence of M. microspora (Fig. 1)
No arrestment response by common silverfish, which are more distant relatives to firebrats and giant-silverfish (Mendes Reference Mendes1991), to paper inoculated with M. microspora (Fig. 1), or to paper exposed to firebrats or giant silverfish (Tremblay and Gries Reference Tremblay and Gries2003; Woodbury and Gries Reference Woodbury and Gries2007), implies that common silverfish require a different as yet unknown aggregation cue or signal, possibly one that is indicative of the type of microclimate (room temperature; high humidity) they prefer.
Acknowledgements
The authors thank E. Zabek and S. Byrne at the Animal Health Centre, SFU for identification of microbial samples, M. Moore, C. Lowenberger, R. Britton and members of the Gries laboratory for invaluable input during this research, and two anonymous reviewers and G. Boiteau for constructive comments. The research was supported, in part, by a Natural Sciences and Engineering Research Council of Canada (NSERC) – PGS-D Fellowship to N.W., and by an NSERC-Industrial Research Chair to G.G. with Contech Enterprises and Global Forest Science as industrial sponsors.
