Naturwissenschaften 78, 368- 370 (1991) © Springer-Verlag 1991 0028104291001225 Temporal Pheromonal and Kairomonal Secretion in the Brood of Honeybees J. Trouiller, G. Arnold, Y. Le Conte, and C. Masson Laboratoire de Neurobiologie Compar6e des Invert6br6s, INRA-CNRS, F-91440 Bures-sur-Yvette B. Chappe Institut de Chimie des Substances Naturelles, CNRS, F-91198 Gif-sur-Yvette In a honeybee colony, brood care is en- sured by adult bee behavior adapted to the different ages and castes of larvae. The brood is incubated, the larvae are fed, and their cells are capped [1- 3]. In order to adopt the appropriate behav- ior, adult workers must be able to re- cognize the age and the caste of a larva. Like most social Hymenoptera, chem- ical signals play an essential role in brood recognition for the honeybee [4]. When the worker larvae are 9 days old (from the time of egg-laying), adult bees close the top of the brood cell with a wax cap [5]. This behavior can be ar- tificially triggered by four of the ten methyl and ethyl fatty acid esters pres- ent on the surface of larvae: methyl palmitate, methyl oleate, methyl li- noleate, and methyl linolenate (MP, MO, ML, MLN) [6]. The six other fatty acid esters identified in the larval cu- ticle are: ethyl palmitate (EP), methyl stearate (MS), ethyl stearate (ES), ethyl oleate (EO), ethyl linoleate (EL), and ethyl linolenate (ELN). The parasitic mite, Varroa jacobsoni, reproduces on the brood of honeybees. It is attracted by worker and drone lar- vae just before their cells are capped, and reproduces in the sealed cell until the adult bee emerges at 21 days of age. Among the compounds present in cuticular larval extract, only three es- ters - methyl and ethyl palmitates, and methyl linolenate (MP, EP, MLN) - are attractive to Varroa [7]. Thus, two of these esters act both as pheromones and as kairomones. It seems likely that the presence of these esters in the larval cuticle on the 9th day is responsible for both the entry of Var- roa females into the cells and for the capping behavior of workers. This hy- pothesis is now supported by evidence 368 that these compounds are present in the cuticle only a few hours before the cell is capped, and by the fact that they are released by the larvae themselves. We have quantified esters in the cuticle of workers from one honeybee colony (Apis mellifera ligustica) at different ages in their development. According to most authors and to our observa- tions, cell capping takes place on the 9th day of the preimaginal period [5]. From our data, 8.5-day-old larvae have almost reached their definitive weight and the cell has no evidence of capping. When the larvae are 8.75 days old workers begin to close the cells, and by 9 days they are totally capped and with- out any trace of the larval cocoon. At 9.5 days of age the larvae have begun to spin cocoons. We therefore analyzed larvae at 6, 7, 8, 8.5, 8.75, 9, and 9.5 days of age; ll- day-old pre-pupae; and 12.5-day-old pupae at the white-eye stage. A queen was allowed to lay eggs during 1 day, thereafter, she was removed from the comb. Because the capping of the differ- ent cells of the comb occurred within 2-3 days, we backdated our samples by comparing them with similar control larvae (size, aspect of the cap, pig- mentation) from the completion of capping. Two or three replicates of each age group were analyzed, each sample consisting of about 15 g of lar- vae or pupae. The insects were taken in- tact from their cells and their cuticles extracted twice with n-pentane and there- after once with dichloromethane. For each sample, the combined extracts were fractionated by chromatography on a column of silica gel (Merck, 230- 400 mesh, diameter 0.5 cm, length 20cm). We first passed 6 ml of pentane through the column; thereafter Naturwissenschaften 78 (1991) a second fraction containing the esters was obtained using 6ml of dichloro- methane. This second fraction of larval or pupal extract was purified by HPLC with a si- lica column (Interchrom 250, diameter 4.6ram, intersphere 5/zm). The mobile phase was a mixture of n-heptane and ethyl acetate (95:5) and the flow rate was 1 ml/min. The pump (Waters 6000) was equipped with a Waters U6K in- jector and an Erma ERC 7510 refrac- tometry detector. The quantification was performed by GC on a capillary Carbowax column (SGE, length 50m, internal diameter 0.32mm, film thickness 0.5/zm). Hy- drogen was used as carrier gas (inlet pressure 40kPa). The temperature of the oven increased from 55 to 230 °C at a rate of 7 °C/rain and the temperature was kept at 230 °C over 20 min. The oven was a Carlo Erba 6000 equipped with an on-column injector and flame ionization detector. Compounds were identified by their re- tention times and molecular fragmenta- tion by GC-MS, and we compared the mass spectra of the compounds present in the extract with those of reference esters (Sigma). The GC-MS analysis was performed with a Varian 3400 oven, with the same column and tem- perature program, coupled with a mass spectrometer INCOS 50 (Finnigan). In 6- and 7-day-old larvae, there is a small amount of the total esters present in the cuticle (extreme values: 3- 30ng/larva); at 8 days, this amount is 56 - 59 ng/larva. During the 9th day, the amount of total esters increases up to 500- 590 ng/larva, while the amount of esters responsible for triggering capping (CE) is up to 230 - 360 rig/larva, and the amount of esters attractive to Varroa (VE) is up to 200 - 250 ng/larva (Fig. 1). For the five esters (MS, ES, EO, EL, ELN), which do not trigger these biological activities (NAE), the increase is six to eight times smaller. The variability observed for 8.5-day- old larvae indicates that the increase in secretion begins around that time. After capping, the amount of esters de- creases: Twelve hours after capping, when the larvae are 9.5 days old, CE is reduced to 90-165 and VE to 60 - 100 ng/larva. In 11-day-old pupae, these amounts drop to 20-30 and 4- 7 ng/pupa, respectively, and the level of esters is lower than at the 8th © Springer-Verlag 1991