Many military installations manage species listed as threatened or endangered (TES) in accordance with the Endangered Species Act. Integrating endangered species management with the primary mission of military readiness presents challenges. In particular, population censuses are traditionally performed by biological observers that intensively sample the study area. This is incompatible with live fire training activity and inconvenient for many other training areas.

The main objective of this project was to develop an aerial acoustical survey system that would extend census inventories into areas of military installations that are inaccessible to biological survey personnel. Additional objectives in developing this system were to eliminate the potentially biasing effect of human observers and to increase the speed at which large areas are inventoried.

Technical Approach

This system integrated the capacity to detect and identify animal vocalizations at ranges of several hundred meters with a balloon system that drifted with the wind at a controlled altitude. Fort Hood was chosen as the location to test the aerial acoustical survey system. Fort Hood contains the largest populations of the endangered golden-cheeked warbler (Dendroica chrysoparia, GCWA) and black-capped vireo (Vireo atricapilla, BCVI) under a single management authority. The size of the limited access area at Fort Hood (29,000 ha) and the limited ranges of detection for GCWA and BCVI presented substantial challenges for the proposed system.

To create the aerial acoustical survey system, lightweight, low-power autonomous recording units that monitor wildlife acoustic activity for many weeks without operator attention were developed. To carry acoustical instruments at a controlled altitude, a balloon system was developed that was easily launched, tracked, and recovered. Finally, software tools were produced to rapidly process acoustic data recorded by the digital recording systems. These tools yielded logs of calling events, measurements that characterize the structure of each sound, and had the capacity to determine the direction of arrival or location of each sound recorded by an array of sensors. The logs produced by this software were designed to be readily imported into statistical processing packages to develop algorithms for species identification and also had the capacity to automatically produce maps and other graphical summaries of the data.


This project demonstrated the feasibility of aerial songbird surveys using a small, inexpensive balloon system. Recordings from these balloon flights yield information about the acoustical behavior of bird species whose sounds are regularly recorded. Enormous volumes of acoustical data to document BCVI singing activity were collected and processed. A byproduct of the balloon design effort was identification of lightweight, low-cost components for a sensitive, long-term recording system that has low power consumption. These units were utilized to record tens of thousands of hours of environmental sounds, which were processed to detect and identify nearly three quarters of a million BCVI songs. The typical song intervals for several species were measured using an autocorrelation analysis. The typical range of detection was inferred from plots of song bout durations in relation to maximum allowable intervals for song linkage. These methods generalize readily to other vertebrate species. Most other species will pose much less challenging problems for automatic classification than BCVI, because this species has a diverse repertoire and it occurs with other species that produce similar sounds. With a couple of improvements, the system would be a practical tool for surveys of acoustically active animals in a variety of environments.


The data collected by this system can be used to determine the presence or absence and the estimated population density of target species in areas where such data are presently unavailable.  These data can be used by Department of Defense resource managers to support the development of effective management plans for threatened and endangered species on many U.S. military installations.  In additional, the balloon platform could be readily adapted for other survey applications: visual or thermal infrared video surveys, bird migration monitoring, and possibly even LIDAR surveys.