1983] 
Nuw'icki & Eisner — Bombardier beetles 
121 
enclosures. The beetles all stemmed from the larval collecting site, 
where they were taken at the very places on the edge of the pond 
where the larvae were also abundant. In three encounters we were 
fortunate to witness the beginnings of the attack. The events pro- 
ceeded quickly and were the same in each case. No sooner had a 
beetle brought one of its tarsi to rest upon the mouthparts of the 
larva, than it was grasped by that tarsus and caused to spray. There 
were sometimes several discharges, audible at times and visible as 
misty puffs, but the larva, which had withdrawn below the surface 
the moment it hooked on to the beetle’s leg, was already out of 
reach. The beetle struggled as it was gradually pulled into the mud, 
but the larva never released its hold. Partly submerged, the beetle 
eventually died (Fig. 1 B), and when retrieved next day was found to 
be largely eaten out. Five additional encounters that were not wit- 
nessed from the outset were equally fatal to the beetles. We assume 
that the death of the beetles was hastened by the salivary toxins that 
tabanid larvae are said to inject into their insect prey (Schmidt, 
1982). 
Given the ecological co-occurrence of T. punetifer larvae and 
bombardier beetles, we feel that encounters between the two must 
inevitably occur also in nature, with the same outcome as in the 
laboratory. Moreover, the larvae must have access also to diverse 
other insects that discharge noxious secretions, includings ants, tiger 
beetles, and additional Carabidae. Species of Chlaenius, for exam- 
ple, whose odor was unmistakably suggestive of the phenolic output 
that characterizes other beetles of the genus (Eisner et ai, 1963), 
scurried about together with Braehinus at our collecting site at 
night. Against such chemically protected insects, the predatory tac- 
tic of lurking just beneath the surface, and of withdrawing into the 
mud for total cover the moment a victim is seized and caused to 
activate its defenses, doubtless serves the larvae well. Other mud- 
dwelling tabanid larvae of similar opportunistic feeding habits might 
equally profit from the tactic. 
Acknowledgements 
Study supported by Grant AI02908 from NIH; we thank Drs. 
John F. Burger and Rodolfo Ruibal for helpful information and for 
identifying the larva and toad respectively, and Maria Eisner for 
technical help. 
