The Department of Defense (DoD) needs new, innovative technologies for monitoring avian populations in inaccessible areas to comply with existing environmental regulations and as a means to enhance military testing and training. This demonstration project addresses this need by presenting the results from field validation tests for an autonomous aerial acoustic recording system (AAARS). This system is based on a helium-filled weather balloon that transports an instrument payload over inaccessible areas (e.g., firing ranges, ordnance impact areas) to record avian vocalizations. Specific objectives of the project were to 1) demonstrate and validate the ability of the AAARS to collect data on avian vocalizations for threatened, endangered, and at-risk species (TER-S) of birds over accessible areas; 2) train project staff and installation staff on use of this technology; 3) evaluate the ability of trained project staff and installation staff to deploy the technology; 4) compare the accuracy and precision of monitoring data collected by AAARS; and 5) compare costs of data collection by AAARS to costs associated with conventional, human-based ground-sampling methods.

The AAARS instrument payload is comprised of a microprocessor for command/control functions, a servo-controlled valve for venting helium, a global position system (GPS) module for real-time location and altitude, and a radio frequency module for communication with a ground-based monitoring station. Avian vocalizations are recorded on a digital recorder (Zoom H2) with an active directional microphone during the flight. GPS telemetry data are used to track the balloon during flight, and locate and recover the system once it is on the ground. Once the system is recovered, audio data recorded during the flight are transferred from the digital recorder to a personal computer for analysis, along with the recorded spatial data.

The field validation portion of the demonstration was conducted at Fort Riley, KS; Fort Bragg, NC; and at Big Oaks National Wildlife Refuge, IN (formerly Jefferson Proving Grounds). Validation trials were conducted in static (tethered) and dynamic modes to simulate standard avian point counts and line transect methods. AAARS performance was compared against human-observer point counts and line transects on real bird populations and also on simulated bird populations. Training workshops were held for DoD staff in Kansas, Indiana, and North Carolina as a means of evaluation of the ability to transfer the technology to environmental staff from numerous installations and as a means of generating interest in, and identifying.

Overall, 331 free flights were attempted during validation testing in 2011-2013. The AAARS was successfully flown on >85% of the attempted flights in 2013, the final year of testing in terms of successfully launching, hitting the target area, tracking and recovery of the payload, and collection of audio data. AAARS performance met the objective of documenting the occurrence of all target species evaluated. In addition, abundance estimates were similar in accuracy and precision to estimates generated from human-observer based methods (point counts and line transects) depending on the target species being monitored. Compared to human-based methods, the AAARS could monitor approximately 20 times more area per monitoring day (2000 hectare [ha] vs. 94 ha) based on a field crew of three individuals and also provided a permanent geo-referenced audio record of what species were vocalizing on a given site on that given day. This apparent economy of scale is seen as a huge advantage of the technology, especially in situations where human access is problematic or where large areas need to be covered such as for documenting the relative abundance of threatened and endangered species, or DoD Mission-sensitive Species.

Implementation of the technology faces three challenges. First, because the AAARS was designed to cover large inaccessible areas, requiring launch points and recovery zones typically in training areas, deployment of the technology requires daily coordination with military range control and training schedules. Deployment on Fort Bragg and Fort Riley was limited to about 10-12 days per breeding bird season in which access was possible, usually on weekends and especially on holidays. Second, implementation is limited by appropriate weather for flying the AAARS, typically with winds aloft <10 knots and without rain, although these conditions are often also used as constraints on human-based monitoring. Finally, the analysis of acoustic data remains a challenge for any acoustic-based monitoring technique. The research team has developed an analytic pathway to facilitate this process but acoustic data analysis still remains a significant cost of using the technology although this cost is not considered prohibitive for implementation of the technology. At the completion of the demonstration, the research team concluded that the AAARS technology was successful in meeting project objectives and was cost effective for meeting a range of DoD bird monitoring needs.