What bioacoustics researchers wish they knew before deploying (PART 2)

Each year, hundreds of scientists submit applications to our Scientific Grant program, and four winners are selected each quarter. These researchers provide regular progress updates and complete a final survey upon concluding their projects. A central question we pose to every participant is: What guidance would you offer others using bioacoustics to address similar research challenges?

Here, we share their collective insights for you to reap the benefits of 20/20 hindsight!
 

1. Noise is not an afterthought; it is part of the study design.

Urban sound, wind, water, aircraft, bridges, vegetation, and other environmental noise can reduce detection accuracy or contaminate recordings. Researchers recommended preliminary site surveys, strategic placement, and analysis tools that can help identify, classify, or filter unwanted sounds.

“First, I strongly advise conducting a preliminary survey of the study site, as the sensor deployment conditions are critical to the quality of the results. In my case, I was unable to [do this] to determine the optimal observation points because my study site is located in the Arctic. As a result, some of my recordings were contaminated by background noise.

Second, Kaleidoscope Pro is a highly powerful tool for detecting noise events. Using its cluster analysis function, it is possible to build a noise classifier that can then be trained to automatically detect various events. In my study, I used this function to classify background noise in order to identify the dominant contributors to signal contamination and to analyse the usage pattern of the bridge near the river. This tool holds great potential for both fluvial acoustic research and urban noise studies.”

“More generally, setting up any audio recorders well is an interesting experience because most humans tend to be quite visually-oriented species. It’s good to think about the range of detection of your target species and also the likely background noise. I would try to choose sampling locations in specific microhabitats that have lower background noise, which often end up being a little bit elevated.”

2. Bioacoustics produces a lot of data, so workflow matters.

A major takeaway is to plan storage, organisation, labeling, backup, and processing systems before data collection begins. Without a clear data management strategy, large acoustic datasets can quickly become overwhelming.

“A well-structured data processing workflow is crucial for effectively managing and analyzing large volumes of acoustic data. Developing an incremental approach to data processing helps ensure that information is reviewed systematically, allowing for timely insights.

Considering the software tools available, ensuring that team members have the necessary training, and establishing a realistic schedule for routine data analysis sessions will help prevent backlogs and ensure steady progress.”

“[Bioacoustics] is a specialized field. It's essential to develop a comprehensive understanding of acoustic principles, data management, and analysis techniques. Consider seeking guidance or collaboration with experienced bioacousticians. This can provide invaluable support in project design, data interpretation, and avoiding common pitfalls.

Dedicate time to becoming intimately familiar with the specific equipment and software you'll be using before deploying to the field. This includes understanding all functionalities, troubleshooting procedures, and optimal settings for your target species and environment. Don't underestimate the learning curve.”

“Executing a project similar to our avian ecology study entails careful planning and strategic considerations. One crucial piece of advice is to formulate a clear and detailed plan for data analysis before initiating the survey. This proactive approach ensures that data collection methods align seamlessly with the analytical goals, promoting efficiency and accuracy throughout the project.”

3. Autonomous recorders still need regular, documented checks.

Even when using field-ready autonomous recorders, researchers stressed the importance of routine maintenance: checking battery life, SD card capacity, schedules, settings, recorder status, and physical condition. Several also recommended carrying backup batteries, SD cards, permits, locks, keys, and waterproof bags.

“Though the recorders are meant for autonomous functionality, repetitive and documented checks on the units are essential. Check the unit's function, schedule, battery life, storage capacity, and heat.

Bring a ‘save-a-trip’ kit with batteries, SD cards, locks and keys, permits, and water-tight bags. Before you deploy, invest in a suite of backup SD cards and data storage to vault data during the project to avoid losses.”

“To maintain data quality and equipment functionality, a thorough process for in-field servicing of ARUs should be established.”
 
“Be mindful of daylight savings as this will shift when your ARUs are recording, and they may need to be reprogrammed. [We found that] the SD cards and batteries lasted longer than expected; however, it may have been more efficient to analyse the data continuously throughout the survey period.”

4. Validation and quality control are essential.

Automated tools can accelerate analysis, but researchers cautioned that acoustic data still requires oversight. Species identification may need to be verified with known recordings, local call libraries, visual confirmation, complementary methods, or expert review.

“If local call libraries are incomplete, consider checking if your acoustic data can be paired with other identification methods to strengthen species-level conclusions.”

“Develop a process for validating and verifying your acoustic data. This may involve visual confirmation of vocalisations or comparisons with known recordings.”

“Being able to bounce ideas off an experienced head was something we missed, and I think it would have been really valuable to have someone who knew the ins and outs of bats and bioacoustics. We also learned not to be overly ambitious and also make sure to vet your data and Auto-IDs with local species.”

“Before conducting passive sampling, I would recommend others to capture bats and collect reference calls. Our previous sampling in the area and reference call collection has been essential for the identification and confirmation of species in the area.”

“Invest time in training or familiarizing field personnel with the survey equipment and methodologies. Adequate training ensures consistency in data collection protocols and minimises variability introduced by different observers. This attention to detail enhances the reliability and comparability of the collected data.” 

5. Bioacoustics rewards patience, flexibility, and long-term thinking.

Many respondents framed acoustic monitoring as powerful but labor-intensive. Projects often require adaptation, persistence, and realistic expectations—especially when working in noisy, remote, urban, seasonal, or privately owned landscapes. The strongest results come from combining good equipment with thoughtful study design, consistent field protocols, and disciplined analysis.

“Do not underestimate how labor-intensive the processing of bioacoustic data is! Even if you are using advanced computational techniques, there is a level of oversight and quality control needed at just about every step of the way.”

“(1) Connect with different experts (taxa, bioacoustics, fieldwork, etc.) to get advice on the different parts of the project; (2) Plan ahead, test/optimize the protocol many times before jumping into the field, and adapt to tricky situations; (3) Do not underestimate the post-fieldwork data management process; (4) Enjoy running the project!”

“Remain adaptable and responsive to unexpected challenges or environmental conditions. Weather patterns, unforeseen logistical issues, or changes in the avian landscape may necessitate adjustments to the survey plan. Flexibility and the ability to adapt methodologies as needed are crucial for overcoming unforeseen obstacles.”

For more recommendations, head over to Part I. Special thanks to these grant recipients for providing sage advice:

• Ellen Connelly, Tikki Hywood Foundation, Zimbabwe.
• Dr. Vanessa Ferdinand, University of Melbourne, Australia.
• Luke Foster, Nature Connect, Cape Town.
• Dr. Jérémy Froidevaux, University of Stirling, Scotland.
• Chelsea Greer, Raincoast Conservation Foundation, Canada.
• Tara Hanf-Dressler, Leibniz Institute for Zoo and Wildlife Research, Germany.
• Charley Marlow Vance, Florida Gulf Coast University, United States.
• Ellen McArthur, Universiti Malaysia Sarawak, Malaysia.
• Alice Michel, University of California Davis, United States.
• Tomohiro Nakayama, Hokkaido University, Japan.
• Michael Sevigny, The Tulalip Tribes of Washington State, United States.
• Dr. Phakhawat Thaweepworadej, Mahidol University, Thailand.