Last updated: January 20, 2025
Article
Why Years of Bat Population Data Got a High-Tech Upgrade
Acadia National Park has one of the longest-running bat monitoring programs in the agency. A team of researchers modernized how the park processed its bat data. Now there are apps for that.
By Timothy Divoll, Bik Wheeler, and Molly Donlan
Image credit: NPS
Poop and wing swabs. Locations of animals from tiny radio transmitter signals.
Tens of thousands of acoustic echolocation recordings. Demographic data and countless biological measurements. National Park Service biologists have collected these and many other kinds of information on wildlife every year for decades. These long-term monitoring data help parks assess how wildlife populations are doing. But if the information isn’t well organized, it becomes harder to analyze with each additional year of data collection.
A team of researchers at Acadia National Park wanted to streamline data processing and availability from field collection all the way to management actions. We transformed a paper-based, internal, 16-year bat-monitoring dataset into a dynamic, digital process. This enables new information to flow in and results to flow out with regularity, just like the tides in this coastal park. Our new workflow makes it easier for park staff to collect and analyze information and make quick, scientifically sound decisions. It’s reproducible and can be scaled up to work in a wide range of settings and circumstances.
Then and Now
Though painters may reproduce the breathtaking views of Acadia’s Somes Sound, its coastal mountains and vertical cliffs hide plenty of bats that never make it onto paper or canvas. The Wabanaki and their ancestors have undoubtedly told stories about Acadia’s bats for thousands of years. Naturalists from Harvard collected and studied these species in the 19th Century.
Image credit: NPS
Fast forward to 2008, and the park participated in a regional survey of mercury in wildlife, including bats. The following year, I (Divoll) started researching Acadia’s bats for my master’s thesis. In retrospect, the 2008–2011 data represented the glory days of bat monitoring at Acadia. At that time, little brown, northern long-eared, and eastern small-footed bats dominated the nighttime skies. But by 2012, capture rates were declining so steeply they were reminiscent of the park’s infamously sheer Precipice Loop Trail. White-nose syndrome, a deadly fungal disease of bats, had come to Acadia.
We went from excitement about rediscovering booming bat populations to a determination to help them survive.
The mood of the park’s bat research team shifted. We went from excitement about rediscovering booming bat populations to a determination to help them survive. Funding for studying this devastating disease was sporadic. But as former park biologist Bruce Connery said, “When funding is limited or uncertain, you start small and collect what data you can. You have to start somewhere, otherwise you get nowhere. Even small bits of data are crucial in our ability to help species…it’s an opportunity to hone in on the most important questions.”
Sixteen years later, Acadia had become one of the longest running bat monitoring sites in North America. But most of the data were recorded on paper and filed away in cabinets, unavailable to scientists outside the park. I (Divoll) was now a data scientist at Brown University’s Center for Computation and Visualization. My colleagues in the National Park Service and I worked to modernize the collection, cleaning, and retrieval of these valuable long-term monitoring data.
The Digital Transformation
Conventional workflows use paper sheets to collect field data. But paper can get snatched up by the wind or become sopping wet after a summer rainstorm. Workers manually transcribing muddy, ripped, crumpled, or bug-stained data sheets into a database are likely to make errors. Instead, we used ArcGIS Survey123, which builds dynamic digital forms, to develop mobile apps for park staff to use when collecting and storing field data. The apps’ modern, accessible, user-friendly format saves time and leads to fewer errors compared to writing data on paper and then transcribing it.
Image credit: NPS
Image credit: NPS
The apps also improve the quality of the data by providing pre-selected options during data entry, limiting potential errors. We publish the data annually in the NPS DataStore for researchers and resource managers, first passing it through a template with project-specific information that conforms to the Ecological Metadata Language standard. In this way, we only add new data from the current year, recycling any information that hasn’t changed. This makes it easier to maintain data consistency and accuracy and reduces the effort it takes to publish a new dataset each year.
Once we had developed the apps, we built an accessible, interactive workflow for analyzing the cleaned, published data (for internal use only due to sensitive species). We used digital tools like R-Shiny and R-Notebooks to create the workflow. Our coding scripts can be recycled each year to generate outputs that incorporate another year’s worth of data.
One analysis showed, for example, that staff can sample bats effectively at fewer strategic locations annually.
With R-Shiny, we built an interactive app that enables park resource managers to quickly filter the 16 years of data on Acadia’s endangered bats and answer questions about the bat population’s vital signs. With R-Notebooks, we did some of our own analyses to help park managers generate consistent numbers and operate more efficiently.
One analysis showed, for example, that staff can sample bats effectively at fewer strategic locations annually. This would help minimize staff overtime. Before the analysis, researchers caught bats in mist nets at over 30 sites across the park. That provided some great geographic coverage. But we learned that only about six of those sites had bat activity consistently in the post-white-nose-syndrome landscape. So we designated those sites as “core sites” to focus on for future monitoring.
Image credit: NPS / Tim Divoll
Throughout the years, capture teams have opened mist nets to catch bats for four hours after sunset on average. But we learned from our analysis that most bats were captured in the first two hours. We found that if employees stopped surveys and returned to headquarters by midnight, they would still have accounted for about 86 percent of what they would have captured by staying out later.
Cheaper, Safer, and More Efficient
Shifting from manual data collection and curation to a digital workflow helps save money through reduced employee time, enabling them to address other tasks and projects. It’s also a major safety improvement. This is because injury risk increases as energy and attention fade the later researchers stay out chasing bats. In addition to cost, other major concerns when building specialized software are security and ease of use. We chose tools that conform to federal software security standards. After mobile field collection, the data are passed to a secure data portal before further processing.
Image credit: NPS
We expect the tools we used for data entry and hosting to remain relatively user friendly and supported, as they were created by the world’s largest geospatial software developer. For data cleaning and retrieval, the R language—and by extension R-Shiny and R-Notebooks—is popular in the field of biology. These tools are commonly part of a college curriculum in ecology, so staff biologists are likely to have some exposure to them, which is important for their long-term maintenance.
A Local Project with Global Value
“We’ve been at the right place and the wrong time,” said Rebecca Cole-Will, chief of Resource Management at Acadia National Park. Cole-Will was referring to watching and tracking the tragic decline of the park’s bats due to white-nose syndrome. “Now,” she added, “we may be seeing some glimmers of hope that Acadia will serve as a refugia for bat recovery. To tell this story, we need to access and communicate the baseline data.”
Improving the accessibility of long-term monitoring data is central to modern conservation.
Cole-Will’s comment underscores the importance of this project for the park. But our protocols for data collection can be adopted widely by other parks and protected areas for a wide range of resources. Improving the accessibility of long-term monitoring data, as we did, is central to modern conservation. Digital tools like these will enable scientists and land managers to make quick decisions that help wildlife and their habitats survive and adapt to rapidly changing conditions.
About the authors
Bik Wheeler is the lead wildlife biologist at Acadia National Park. He works to conserve the park’s wildlife populations through monitoring and research. Bik delights in the diversity of animal behavior, particularly when populations use unexpected life history strategies. Image © Friends of Acadia National Park / W. Greene
