NSF IDBR: Developing a Behavioral Acoustic Biome Measurement System


Recording System Block Diagram

Project Summary

Objective. Communication networks are likely fundamental to how a diversity of vertebrate species convey meaningful information with conspecifics. Rather than communication being limited to signals exchanges between dyads, pertinent information is communicated along several dimensions of ongoing interactions between multiple individuals.    Despite the significance of these networks, there are notably few data on the topic.  A key limitation to progress on the topic is technology.  Until recently, technological advances had not reached a point where engineering a system capable of recording and quantifying communication networks in vertebrates was logistically feasible.  The primary aim of this proposal is to develop a wearable collar mounted measurement system capable of recording multiple dimensions of data pertinent to explicating communication networks in wild vertebrate populations. To complement the data collection, we will also develop novel offline data analysis software to quantify the diversity of vocal interactions that occur over varying temporal scales in a social group. This technology will provide a unique opportunity to elucidate the meaningful nuances of communication networks in vertebrates.

Method. Novel technology for both data collection and data analysis of behavioral and acoustic communication networks will be developed in this proposal. For data collection, we will take advantage of  recent advances in microSD flash and smartphone technology to engineer a wearable recording collar consisting of a microphone and GPS. Each collar will store data on a small high capacity flash card that will allow days of continuous recording time.  Collars will be placed on all individuals within a social group and the data will be synchronized in order to correlate the vocalizations and spatial position of all members of the group. To insure that the final product is usable under field conditions, these recording systems will be tested at various stages of development on both captive and wild population of common marmosets (Callithrix jacchus).  In order to analyze the expansive dataset that will result from implementing this technology, analysis software capable of addressing the plethora of idiosyncratic questions possible in the study of a real world communication network will be developed in parallel.  This analysis software will utilize transfer entropy to measure the interdependence between the multiple variables that occur in these networks. Though developed with marmosets as the model organism, both the recording collars and analysis software will be usable in a wide range of species.