462 ANNTJAL REPORT SMITHSONIAN INSTITUTION, 1961 



disregarded. Many others moved across the arena at various angles 

 but without marked deviation toward the light. When one of these 

 appeared to be in line with the loudspeaker the camera shutter was 

 opened and the sound pulses turned on. 



Some of the tracks registered by the camera as the illuminated moths 

 moved against the night sky are shown in plate 6. Many insects, 

 including some moths, showed no change in flight pattern when they 

 encountered the sound. In others, the changes in flight path were 

 dramatic in their abruptness and bewildering in their variety. The 

 simplest, and also one of the commonest reactions was a sharp power 

 dive into the grass (pi. 6, A, B). Sometimes the dive was not com- 

 pleted and the insect flew off at high speed close to the ground. Almost 

 as frequently the dive was prefaced or combined with a series of tight 

 turns, climbs, and loops (pi. 6, C, D) . 



It is not known whether these maneuvers are selected in some random 

 manner from the repertoire of individual moths, or whether they are 

 characteristics of different species. However, Webster (in press) has 

 shown that bats soon learn to plot an interception course with food 

 propelled through the air in a simple ballistic trajectory. The random 

 behavior elicited by simulated bat cries in the natural moth population 

 seems to be a natural answer to this predictive ability in bats, m hile the 

 sharpness of the turns must certainly tax the maneuverability of the 

 heavier predator. 



The reacting moths shown in plate 6 were mostly within 25 feet of 

 the camera and sound source, and were exposed to an unknown but 

 probably high sound intensity. Under these circumstances, the evasive 

 behavior appeared to be completely unorientated relative to the sound 

 source, as might be predicted from the binaural tympanic nerve record- 

 ings. In some instances, moths flying at a greater distance or only on 

 the edge of the sound beam appeared to turn away from the area and 

 fly off at high speed. This must be checked in future experiments. 



THE SURVIVAL VALUE OF EVASION 



In spite of the evidence that the moth ear is an excellent bat detector, 

 and that acoustic stimulation releases erratic flight patterns, one may 

 well ask whether this behavior really protects moths from attack by 

 bats. 



This question has been answered (Eoeder and Treat, in press) by 

 observing with a floodlight 402 field encounters between moths and 

 feeding bats. In each encounter we recorded the presence or absence 

 of evasive maneuvers by the moth, and the outcome, that is, whether it 

 was captured by the bat or managed to escape. From the pooled data 

 we determined the ratio of the percentage of nonreactors surviving 

 attack to the ratio of reactors surviving attack. Thus computed, the 



