Body vibration and sound production in an insect (Atta sexdens) without specialized radiating structures

Abstract

1. The leaf cutting antAtta sexdens stridulates using a file-and-scraper device (Markl 1968). The temporal and spectral characteristics of the cuticular vibration and airborne sound are fairly typical of the stridulatory sounds of insects that lack specialized radiating structures (Masters 1980). We used laser-Doppler vibrometry (LDV) to measure the velocity of different points on the ant's body during stridulation to determine what sorts of body vibration are produced by the impact of the scraper on the file, both in air and with mass-loading of the cuticle underwater, and how the vibration relates to this radiated sound. 2. The gaster, on which the file is located, appears to be the principal sound-radiating part of the ant. From theoretical considerations, we would expect four types of gaster vibration to be excited by the scraper impact: (I) dipole swinging of the entire gaster, (II) quadrupole distortion of the gaster shape due to volume flow of the gaster contents, (III) monopole pulsation of the gaster and (IV) bending waves travelling in the gaster shell. Each type of vibration has associated modal resonances whose frequencies are predicted in a semi-quantitative fashion from a simple mechanical model of the gaster. 3. Vibrometer measurement shows that each tooth strike typically excites low frequency (∼1 kHz) and high frequency (∼10 kHz) damped vibrations that can be detected at most points on the gaster. Broadband spectral energy can be detected up to about 30 to 40 kHz (in rare cases to ∼60 kHz). The low frequency resonance appears to be due to type I rather than type II vibration, and the high frequency resonance due to type III rather than type IV vibration. 4. Underwater, both low and high frequency resonances were still present, but their frequencies were reduced by about 15% and 35% respectively. We expect the decrease to be even greater when the ant is buried underground. 5. The rate at which the scraper strikes the teeth of the file is close to the ∼1 kHz modal frequency of the gaster, but in contrast to our expectation does not exactly match this frequency (Fig. 2). The reason for this discrepancy is not clear. 6. Below about 30 kHz there is fairly good correspondence between the energy spectra of the airborne sound recorded with a microphone and the gaster vibration recorded by LDV. Above this frequency the sound usually shows a complicated power spectrum that we can relate to the vibration of the body only in a qualitative fashion.