Multiple glandular origins of queen pheromones in the fire ant Solenopsis invicta

The poison sac of the fire ant Solenopsis invicta is the only identified glandular source of pheromones produced by a functional ant queen. This structure, which contains the poison gland, has previously been shown to be the source of a releaser pheromone that mediates queen recognition and tending by workers. The poison sac has also been demonstrated to be the source of a primer pheromone that inhibits winged, virgin queens from shedding their wings (dealating) and developing their ovaries. To determine if the poison sac was the only source of these pheromones, we excised the poison sac from queens and observed whether operated queens retained their pheromonal effects. In a first experiment, the poison sac was removed from functional (egg-laying) queens which were then paired with unoperated nestmate queens in small colonies. Counts of the workers surrounding each queen two weeks after the operation showed that queens without poison sac were as effective as their unoperated nestmates in attracting worker retinues. In a second experiment, we removed the poison sacs of virgin queens which had not yet begun laying eggs and thus had not begun producing queen pheromone. After allowing them to develop their ovaries, these individuals produced amounts of queen recognition pheromone comparable to those secreted by unoperated or sham operated virgin queens as determined by bioassay. Testing the head, thorax and abdomens of functional queens separately revealed that the head was the most attractive region in relation to its relative surface area. Bioassays of extracts of two cephalic glands-the mandibular gland and postpharyngeal gland-showed that the postpharyngeal gland is a second source of the queen recognition pheromone. Finally, we found that virgin queens whose poison sacs were removed before they began producing queen pheromone initiated production of a primer pheromone that inhibits winged virgin queens from dealating, indicating that this pheromonal effect also has an additional but as yet undetermined source. These results parallel those on the honey bee in which several of the pheromonal effects of functional queens appear to have multiple glandular sources.