A larval Microdon, mimicking an ant larva, is grasped by an adult Formica 

 ant, to be carried away for safekeeping. The papery cocoon just under the 

 Microdon holds an ant pupa. 



Roger D. Akre 



ant pupae and molting into the next larval 

 stage. The discovery of pupa popping 

 proved invaluable to our work. It ex- 

 plained why newly hatched Microdon had 

 rarely been found before in the field and 

 provided us with an efficient method of lo- 

 cating them. Now, instead of searching for 

 the fly larvae as we had in the past, we 

 concentrate on collecting ant cocoons, 

 which can be carefully opened in the lab to 

 see if they contain Microdon. 



Microdon larvae, especially later 

 stages, also feed on ant larvae, moving 

 freely about the ant brood chamber as they 

 do so. One day, some of the Microdon lar- 

 vae that we had exposed in a tree stump 

 provided us with another surprise. We saw 

 these instars fold themselves lengthwise 

 until they were practically indistinguish- 

 able from ant cocoons. After this transfor- 

 mation, agitated worker ants arrived, 

 seized the impostor young, and carried 

 them to the safe depths of the nest. We had 

 discovered a most unusual case of aggres- 

 sive mimicry. The ants perceived the fly 

 larvae to be ant cocoons. The prey was 

 tricked into protecting the predator. 



How were the Microdon able to accom- 

 plish this feat? Chemical communication 

 is important in ants, so we thought that this 

 deception might be chemically based. 

 Tests carried out by U. S. Department of 



Agriculture entomologist Ralph Howard 

 showed that the chemistry of the outer, 

 hard cuticle of the larval flies and that of 

 the larval ants matched almost perfectly. 

 On the outside, the flies were chemical 

 mimics of the ant larvae. The ants merely 

 mistook the folded Microdon for their own 

 developing offspring and transported them 

 to safety. Subsequently, we watched for 

 and observed this subterfuge many times. 

 We also saw ants carrying aggregates — 

 whole clumps — of Microdon larvae, just 

 as they often grasp and transport aggre- 

 gates rather than single larvae of their own 

 species. 



We wanted to find out if Microdon ac- 

 quired these recognition chemicals from 

 eating ant larvae or if they synthesized the 

 chemicals within their own bodies. To an- 

 swer this question, we studied Microdon 

 albicomatus and one of its host ants, Myr- 

 mica incompleta. In the spring of 1989, we 

 collected 235 fly larvae; we washed some 

 in a solvent to extract the chemicals for 

 analysis and kept more than a hundred 

 others alive for radioisotope testing. The 

 chemical analyses confirmed that the 

 chemicals on the surface of the Microdon 

 matched those of its host, and radioisotope 

 labeling revealed that a larva did indeed 

 synthesize the chemicals to match those of 

 its host — a case of true chemical mimicry. 



This chemical defense is employed only 

 by Microdon larvae; adult flies are readily 

 attacked and killed by the ants. The adults' 

 defense is solely behavioral. They pupate 

 near the nest surface so that they can make 

 a quick getaway, and they tend to emerge 

 early in the morning when worker ants — 

 especially carpenter ants, which are 

 largely nocturnal — are least active. 



We know that many species of Mi- 

 crodon are host specific, that is, they reside 

 with just one type of ant, but some can be 

 found with two or even three different 

 hosts. Microdon albicomatus, for ex- 

 ample, has turned up in nests of several 

 species of Formica ants, as well as in 

 colonies of the unrelated genus Myrmica. 

 We are still hying to unravel the relation- 

 ships that occur with multihost Microdon 

 and to determine if these insects can 

 change their recognition chemicals in re- 

 sponse to a change in host. 



Microdon larvae have a topography of 

 odd structures covering the back of their 

 sluglike bodies. Most highly developed 

 and visible on mature, third-larval instars, 

 some of these structures look like toad- 

 stools, others like flowers, and stiU others 

 are beyond analogy. On the underside of 

 the larvae are other elaborate protuber- 

 ances, some of which remind us of the 

 "Schmoos" created by Al Capp in his 

 comic strip LiT Abner. Although these 

 structures have long been known, their 

 function has not. We now suspect that they 

 contain glands or glandular openings for 

 secreting the chemicals that the larvae use 

 to mimic their hosts. Since the surface is 

 so convoluted, it would also present an 

 enormous area for the dispersion of these 

 chemicals. The reticulations may also 

 physically deter attacks fi^om the host ants. 

 Yet another possible function is as a recep- 

 tor system for chemical signals from the 

 ant larvae or from the adult ants. 



For all our educated guesses as to the 

 secrets of these sdoictures, perhaps just as 

 appealing is the suggestion made by the 

 European entomologist E. Heckt in 1912 

 fliat they are "a result of an exuberance of 

 forms, which overrides with elan the bor- 

 ders of the purely necessary forms." That 

 exuberance and elan can be perceived in a 

 larva is hardly more surprising than the re- 

 centiy discovered chemical and behavioral 

 ploys displayed by Microdon. 



Gregory Paulson is an instructor in the 

 Program in Biology and Roger D. Akre is 

 a professor in the Department of Entomol- 

 ogy at Washington State University in 

 Pullman, Washington. 



58 Natural History 1/94 



