Quantifying Habitat Use by Secretive Marsh Birds in Antebellum Rice Fields Using Drones and Autonomous Recording Units in Coastal South Carolina

Secretive marsh birds, such as rails, bitterns, and gallinules, are notoriously elusive and well-camouflaged, presenting significant challenges for study using traditional methods. These birds inhabit dense vegetation in wetland habitats, making observation and study difficult. Antebellum (pre-U.S. Civil War) rice field impoundments provide essential habitat for many species of secretive marsh birds in coastal South Carolina. These rice fields, developed between the 1600s and 1800s, were primarily managed by the labor of enslaved people during the colonial era. Rice culture declined post-Civil War, and recent mapping efforts show it spans about 95,000 hectares. Currently, many of these fields are primarily managed for wintering waterfowl. Recent studies suggest constant water level management creates higher sites within an impoundment and can develop patches of saltgrass, clump cordgrass, and salt meadow cordgrass (Distichlis spicata, Sporobolus bakeri, and Sporobolus pumilus), providing suitable cover and nesting habitat for secretive marsh birds.

​

Our research aims to 1) quantify habitat use by these birds within the antebellum rice fields of coastal South Carolina and 2) Design innovative techniques for monitoring them. We employed Autonomous Recording Units (ARUs) for Passive Acoustic Monitoring (PAM) to capture the call activities of secretive marsh birds. ARUs are less invasive and reduce time and labor costs. We used Audiomoth recorders to monitor king rail (Rallus elegans), clapper rail (Rallus crepitans), Virginia rail (Rallus limicola), sora (Porzana carolina), black rail (Laterallus jamaicensis), American bittern (Botaurus lentiginosus) and least bittern (Ixobrychus exilis) during their breeding season from March to mid-June when they are notably more vocal. Our recording schedule involved capturing audio for one minute every three minutes from 5:45 PM to 9 AM for three days, aligning with the peak activity periods of the focal species.

​

We developed a low-cost methodology using drones to deploy and retrieve ARUs mounted on lightweight, multi-terrain floating platforms from inaccessible areas. This approach aimed to overcome the limitations of traditional deployment methods, such as data bias from unapproachable large managed and unmanaged rice field impoundments. We maintain a 200-meter distance between each ARU, effectively covering entire habitats with less time, effort, and human resources. This method allowed us to deploy ARUs in the center of impoundments and far, inaccessible and less disturbed areas, which are crucial for monitoring potential habitats of species like the black rail.

​

For data analysis, we utilized the open-source Artificial Intelligence model called BirdNet, developed by Cornell Lab. Data recorded in .wav format was fed into the model, which analyzed each three-second spectrogram against eBird checklists and BirdNet algorithms, providing species identification with confidence intervals. Last season, we deployed 130 ARUs, surveyed 29 rice field impoundments, recorded approximately 119,000 files, and logged around

2,000 hours of recordings. Moving forward, we plan to expand our survey efforts, double the number of impoundments surveyed, and analyze the collected data to understand better the habitat use and preferences of secretive marsh birds in these historically significant landscapes.

​

This research underscores the importance of integrating advanced technologies, like drones and ARUs, in secretive marsh bird monitoring and conservation. The innovative methodologies developed here enhance our understanding of secretive marsh bird populations and offer scalable habitat monitoring solutions in challenging environments.

​

Image 1: Deploying and retrieving ARUs mounted on top of a designed multi-terrain lightweight floating platform in inaccessible areas using a drone.

Image 2: Successfully deployed ARUs with a multi-terrain lightweight floating platform, ready to record marsh birds.

Image 3: ARUs deployed using kayaking and on foot-based traditional methods in approachable areas.

Image 4: Typical federally endangered black rail habitat with high marsh and dense vegetation cover within antebellum rice fields.

Image 5: King rail crossing impoundments via managed rice field dikes.

Image 6: Least bittern spotted in a managed rice field canal.

Image 7: Virginia rail image captured by drone in the middle of a managed rice field impoundment.

Image 8: Data analysis process flowchart using the BirdNET AI model.