The behavioral functions of larval firefly lights have been studied, and several theories have been proposed. 1 Male flash patterns are thought to be affected by sexual selection and females have been shown to prefer certain light signal characteristics. 2 In addition to these behaviors, it is generally assumed that most non-luminous fireflies locate mates through the use of pheromones. The courtship patterns of Japanese fireflies seem to indicate numerous types of communication, which include

more »

... heromones as well as light signals. 3 Our analytical study of fireflies was based on the hypothesis that a certain chemical signal mediates a particular interaction between male and female fireflies. This hypothesis can be tested by collecting, analyzing and identifying these compounds. Using our previously reported method, pheromones were adsorbed by direct contact with glass surfaces. 4 We call this method direct contact extraction. Because this technique is nondestructive, it is possible to subject extracted hydrocarbons to gas chromatography-mass spectrometry (GC-MS) directly using a solvent. Here, we report on the GC-MS profiles of extracts from three Japanese firefly species (Luciola lateralis, Luciola cruciata, and Lucidina biplagiata) and three North American firefly species (Lucidota atra, Photuris lucicrescens, and Photuris cinctipennis). Several research aspects concerning present analytical methods are discussed: (1) distinguish firefly species using the GC-MS profile as finger prints, (2) the intraspecific differences of fireflies, which could be interpreted as a geographic variation, and (3) chemical-ecological strategy of luminous and non-luminous fireflies. Experimental Luciola lateralis were collected in Yokohama, Kanagawa prefecture and Chino, Nagano prefecture. Luciola cruciata were collected in Yokohama, Kanagawa prefecture. Lucidina biplagiata were collected in Hayama, Kanagawa prefecture. Lucidota atra, Photuris lucicrescens, and Photuris cinctipennis were collected in Meramec state park Missouri, USA. Glass vials (50 ml) were washed in hexane and dried at 200˚C in an electric oven. Three fireflies were collected; if this was not possible, one or two fireflies were collected and placed into one clean glass vial at the field site. Fireflies were free to crawl inside the glass vials for 2 -3 h during the mating period. After the fireflies were released, 1 ml of dichloromethane was added to the glass vials, which were then shaken several times to collect any remaining compounds from the firefly without further condensation. Extracts (1 µl) were analyzed using a mass spectrometer (MS) (JMS-SX102A; JEOL, Tokyo, Japan) equipped with a gas chromatograph (GC) (HP 5890A; Hewlett Packard, Palt Alto, USA). An HP-5 capillary column (30 m length, 0.32 mm i.d., 0.25 mm film thickness; Hewlett Packard, Palt Alto, USA) was used to separate the products into the volatile components. The initial temperature was 50˚C, the final temperature was 280˚C and a linear temperature program of 20˚C/min was applied. The injector temperature was 280˚C and the carrier gas was helium. Electron ionization (EI) was carried out using an ionizing voltage of 70 eV. The ion source temperature was 250˚C and the ion-accelerating voltage was 10 kV. Molecular formulas and compounds were assigned by high-resolution mass spectrometry and spectral data base SDBSWeb: http://www.aist.go.jp/RIODB/SDBS. Results and Discussion Insect pheromones are typically multicomponent mixtures on the order of picograms or nanograms. If the production site of a pheromone is known, it is possible to extract the active compound using a suitable solvent, or via solid-phase extraction. We previously reported a method for collecting 1729 We characterized three Japanese firefly species (Luciola lateralis, Luciola cruciata, and Lucidina biplagiata) and three North American firefly species (Lucidota atra, Photuris lucicrescens, and Photuris cinctipennis) based on their surface hydrocarbons. The analysis of firefly extracts by gas chromatography-mass spectrometry (GC-MS) revealed clear differences in the chromatographic profiles and mass spectra. Each firefly could be distinguished by its GC-MS profile. A major difference was observed between Japanese fireflies and North American fireflies. Among the North American fireflies, non-luminous fireflies, Lucidota atra, showed much more complicated GC-MS profile than those of luminous fireflies, Photuris lucicrescens and Photuris cinctipennis. pheromones by direct contact with glass surfaces. 4 This is nondestructive and allows the extracts to be subjected to GC-MS directly using a solvent. Analysis is a two-stage process of separation followed by detection. Separation is carried out by GC. At the field site, pheromones should work in the gas-phase. GC profiles showed the number, ratio and retention time of each chemical component in the extract. Detection is carried out using MS. In MS, the molecule is ionized and the instrument measures the mass of the resulting ions. High-resolution MS allows mass to be quantified to within six decimal places. The mass of the molecule allows us to determine the molecular formula. A series of hydrocarbons were identified by the characteristic mass spectra using a mass spectral database. GC-MS revealed obvious differences in the GC profiles and mass spectra; these GC-MS profiles are summarized in Table 1 , and a typical GC profile (TIC) is shown in Fig. 2 . Though direct contact extraction can not collect all surface hydrocarbons, three fireflies were sufficient to obtain clear GC profiles. Hydrocarbons are proving to be very important in the activities of numerous insect species. 5 The primary function of surface hydrocarbons is to protect the insects from desiccation. Many bioassays also have shown that surface hydrocarbons are used in recognition systems of both solitary and social insects. 6 Among the Japanese fireflies, tricosane was observed in Luciola lateralis and Luciola cruciata. These fireflies are known as "heike" and "genzi", respectively. The chemical structure and mass spectrum of tricosane are shown in Figs. 3(a) and 4(a) , respectively. Tricosane is reportedly a sex attractant in the firefly Pyrocoelia oshimana. 4 Slight differences in the ratios of the component compounds and presence/absence of henicosane were observed in geographically distinct populations of Luciola lateralis from Yokohama and Chino. Among the North American fireflies, the non-luminous firefly Lucidota atra 1730 ANALYTICAL SCIENCES DECEMBER 2004, VOL. 20 Fig. 1 Collecting surface hydrocarbons of fireflies by direct contact with glass surfaces. a) Non-luminous fireflies, Lucidota atra; b) luminous fireflies Photuris lucicrescens. Fireflies were free to crawl inside the glass vials for 2 -3 h during the mating period. After the fireflies were released, dichloromethane was added to the glass vials, which were then shaken several times to collect any remaining compounds from the fireflies. Fig. 2 Typical GC profile of firefly Lucidina biplagiata in Hayama Japan. Fig. 3 Chemical structures of (a) tricosane and (b) heptacosane. Tricosane was observed in Japanese fireflies Luciola lateralis and Luciola cruciata. Heptacosane was observed in North American fireflies Lucidota atra, Photuris lucicrescens and Photuris cinctipennis. Fig. 4 Mass spectra of (a) tricosane and (b) heptacosane. (a) (b) exhibited mating behavior in response to extracted hydrocarbons on the glass. These hydrocarbons were identified as a mixture of pentacosene, hexacosane, heptacosene heptacosane, octacosene, nonacosadiene, nonacosene and nonacosane. The chemical structure and mass spectrum of heptacosane were shown in Figs. 3(b) and 4(b), respectively. Interestingly, the hydrocarbons in non-luminous fireflies, Lucidota atra, showed a much more complicated GC-MS profile than those of luminous fireflies, Photuris lucicrescens and Photuris cinctipennis. The geometry of the double bonds of these molecules and their concentrations remains to be determined. We believe the hypothesis that non-luminous Lucidota atra strongly depends on chemical communication, using a blend of hydrocarbons in suitable concentrations. In conclusion, we applied direct-contact extraction and GC-MS to a chemical-ecological study of fireflies, and characterized six firefly species based on their hydrocarbons. We successfully conducted our first principal step of identification of chemical compounds. Though the chemical structures of insects pheromones were determined in nanogram quantities by gas-chromatographic methods 7 and liquidchromatographic methods, 8 limited work has been done with fireflies. The large body of studies on insect pheromones is a refrection of their potential value in pest management. 9 To the authors' knowledge, no chemical structures of firefly pheromones have been reported, except in our previous work. 4 We next intend to conduct a bioassay design project, using model compounds of the multicomponent hydrocarbons, which were characterized by present study. Our findings will extend further understandings on the behavioral functions of chemical communication, the same as their physical communications. 10