Reliable reconstruction of cricket song from biophysical models and preserved specimens

Predicting the function of a biological structure solely from its morphology can be a very powerful tool in several fields of biology, but especially in evolutionary reconstruction. In the field of invertebrate acoustic communication, reconstructing the acoustic properties of sound-producing forewings in crickets has been based on two very divergent methods, finite element modelling (FEM) and vibrometric measurements from preserved specimens. Both methods, however, make strong simplifying assumptions that have not been tested, and the reliability of inferences made from either method remains in question. Here we rigorously test and refine both reconstruction methods using the well-known Teleogryllus oceanicus model system and determine the appropriate conditions required to reconstruct the vibroacoustic behaviour of male forewings. We find that when using FEM, it is not necessary to assume simplified boundary conditions if the appropriate parameters are found. When using preserved specimens, we find that the sample needs to be rehydrated for reliable reconstruction; however, it may be possible to accomplish rehydration in silico using FEM. Our findings provide a refined methodology for the reliable reconstruction of cricket songs, whether from fossils or preserved specimens from museums or field collections.

Methods
There are several datasets in this paper, using Laser Doppler Vibrometry, Optical Coherence Tomography-based imaging, and Finite Element modelling. All details of how data were collected and processed are presented in the materials and methods of the paper, including the supplementary materials.