Journal of Chemical Ecology, Vol. 10, No. 10, 1984 IDENTIFICATION OF TRAIL PHEROMONE OF THE ANT Tetramorium caespitum L. (HYMENOPTERA: MYRMICINAE) A T H U L A B. A T T Y G A L L E and E. D A V I D M O R G A N Department of Chemistry, University of Keele Staffordshire, ST5 5BG, U.K. (Received December 1, 1983 revised February 16, 1984) Abstract--The trail pheromone of the ant Tetramorium caespitum L. is a 70:30 mixture of 2,5-dimethylpyrazine and 3-ethyl-2, 5-dimethylpyrazine. The average total amount of the two pyrazines present in the poison vesicle was found to be 3.9 ng per ant, of which 2.7 _+ 0.4 ng is 2,5-dimethylpyrazine and 1.15 + 0.25 ng is 3-ethyl-2,5-dimethylpyrazine. The pyrazines con- stitute only 0.03% of the volume of the poison vesicle but account for the whole of the trail-following activity. A 70:30 mixture of the respective pyrazines evoked the highest activity in artificial trail-following tests. Key Words--Ant, Tetramorium caespitum, trail pheromone, 2,5-dimeth- ylpyrazine, poison gland, venom, Hymenoptera, Myrmicinae, synergism. INTRODUCTION A large n u m b e r of ant species are known to e m p l o y trail p h e r o m o n e s as a means of c o m m u n i c a t i o n , but only in a very few cases has the p h e r o m o n e been chemically identified. The first trail substance to be identified was methyl 4 - m e t h y l p y r r o l e - 2 - c a r b o x y l a t e from A t t a texana Buckley (Tumlinson et al., 1972). The same c o m p o u n d was subsequently d e m o n s t r a t e d to be active in evoking trail following in A. cephalotes L. (Riley et al., 1974) and Aero- m y r m e x octospinosus Reich ( R o b i n s o n et al., 1974). 3-Ethyl-2,5-dimethylpyrazine has been shown to be the m a j o r c o m - p o n e n t of the trail p h e r o m o n e o f A t t a sexdens rubropilosa Forel (Cross et al., 1979), and the same c o m p o u n d has since been identified as the single c o m - p o n e n t of the trail p h e r o m o n e of eight species of M y r m i e a (Evershed et al., 1981, 1982). In the beginning of the present study, only the a f o r e m e n t i o n e d 1453 0098-0331/84/1000-1453503.50/0 1984 Plenum Publishing Corporation

1454 ATTYGALLE AND MORGAN two compounds were known as trail substances that originate from the poison glands. Faranal, a terpenoid that originates from the Dufour gland, has been identified as the maj or trail pheromone of Monomorium pharaonis L. (Ritter et al., 1977). There is a controversy about the composition of the trail phero- mone of Solenopsis invicta Buren. Williams et al. (1981) have reported it to be (Z, Z, Z)-allofarnesene, while Vander Meer et al. (1981) describe it as a mixture of (Z, E)- and (E, E)-a-farnesenes and (Z, Z)- and (Z, E)-homofarnesenes. A mixture of C6-C12 and C14-C20 fatty acids are reported to be the active trail following mixtures for Lasiusfuliginosus Latrielte (Huwyler et al., 1975) and Pristomyrmex pungens Mayr (Hayashi and Komae, 1977), respectively. In Iridomyrmex humilis Mayr (Z)-9-hexadecenal has been identified as one of the components of its trail pheromone (Cavill et al., 1979 Van Vorhis Key and Baker, 1982). The above summary illustrates that the information available about the chemistry of trail pheromones is very limited. Some of the artificial trails laid with the above-mentioned single substances were not species specific, al- though the natural trails showed a much higher degree of species specificity, Although many trail pheromones had been recognized as multicomponent mixtures, the true quantitative and qualitative compositions of none of them were known at the beginning of this study. We have recently shown that the trail pheromone of T. caespitum L. contains two pyrazine compounds (At- tygalle and Morgan, 1983) and give here the full identification of these substances and show how the synergistic action of these two compounds together completely accounts for the activity of the natural pheromone. METHODS AND MATERIALS Insect Rearing. Colonies of T. caespitum were collected from Heartland moor in Dorset. The ants were maintained in the laboratory at room tempera- ture in a wooden box filled with moist soil and peat. The ants were fed on a diet of desiccated coconut, meal worm larvae (Tenebrio molitor), and sugar solution (10% w/v). Preparation of Glandular Extracts for Bioassay. The ants were killed by exposing them to the cold vapor from liquid nitrogen. The poison glands and Dufour glands were separated by dissecting the ants in water. The glands were macerated with a solvent such as hexane or acetone (100/zl) and kept ice-cold for further use. Bioassay of Trail-Following Behavior. The method of Pasteels and Verhaeghe (1974) was employed to measure the trail-following behavior of ants towards the test solutions. A circle of 5 cm radius was drawn with a lead pencil on a piece of white paper (13 X 13 cm). The circumference of the circle was marked with arcs (1 cm). The solution under investigation (usually 25-t00

TRAIL PHEROMONE OF T. caespitum 1455 #1) was injected into a Standardgraph (Blundel Harling, Dorset) "S" funnel pen (0.8 ram) and a continuous streak was drawn on the circle. The solvent was allowed to evaporate for 2 min, and the paper was placed in the foraging area of the ant nest. The number of arcs run along the trail by each individual worker ant was recorded for 20 min. The median of the values thus obtained was used as a measure of activity. The activities of extracts of two poison glands and two Dufour glands were tested separately. Median values were ob- tained by repeating the tests three times. A blank bioassay using solvent only was always performed before a test to ensure no residual activity was present in the pen. Thin-Layer Chromatography. An extract of two cleanly dissected poison glands was made in distilled acetone (50 #1). The extract was applied to the origin of a silica gel layer (20 cm X 5 cm X 0.3 ram) on a glass plate and developed with hexane-acetone (60: 40). The solvent front was allowed to run 15 era. The plate was air dried, and the silica was cut into ten bands (1.5 cm each). The bands were scraped separately into Pasteur pipets plugged with glass wool. Each fraction of silica was extracted with acetone (100 #1) directly into a Standardgraph pen. The trail-following activity evoked by each fraction was tested by bioassay. Blank bioassays using solvent only were performed before and in between each test to ensure no activity was present by contami- nation. The test was repeated in the same manner, except only the region between 4.5 and 9 cm was scraped with a small spatula and the width of each band was narrowed to 2 ram. All determinations were made in duplicate. Similar experiments were performed to test for functional groups. Two poison glands were extracted separately in HC1 in acetone (1%) and Br2 in hexane (1%, v/v) respectively. The reaction mixtures were separated by TLC, and the bioassays were carried out as before. Samples of synthetic 2,5-dimeth- ylpyrazine and 3-ethyl-2,5-dimethylpyrazine were chromatographed under the same conditions as above. The spots were visualized under a UV lamp and the the R I values were calculated. Gas Chromatography. Gas chromatography (GC) was performed with a Pye 104 gas chromatograph with a flame ionization detector using a packed column of 2.75 ��� 4 mm, 10% PEG 20 M on Chromosorb W (100-120 mesh) at 130~ The nitrogen carrier gas flow rate was 50 ml/min. The ionization amplifier was used at attenuation ��� Three poison glands were cleanly dissected without the Dufour gland and the sting, blotted dry, mounted on a small piece of glass and sealed in a glass tube (25 X 1.8 ram). The contents of the tube were chromatographed via a solid injection method (Morgan and Wadhams, 1972). A Dufour gland was chromatographed under the same conditions in order to distinguish any peaks that may arise as contaminants in the poison gland GC traces. Trapping of GC Effluent. Two poison glands were injected onto the PEG