Long‐term analysis of persistence and size of swallow and martin roosts in the US Great Lakes

Abstract

In this study, we combined a machine learning pipeline and human supervision to identify and label swallow and martin roost locations on data captured from 2000 to 2020 by 12 Weather Surveillance Radars in the Great Lakes region of the US. We employed radar theory to extract the number of birds in each roost detected by our technique. With these data, we set out to investigate whether roosts formed consistently in the same geographic area over two decades and whether consistency was also predictive of roost size. We used a clustering algorithm to group individual roost locations into 104 high‐density regions and extracted the number of years when each of these regions was used by birds to roost. In addition, we calculated the overall population size and analyzed the daily roost size distributions. Our results support the hypothesis that more persistent roosts are also gathering more birds, but we found that on average, most individuals congregate in roosts of smaller size. Given the concentrations and consistency of roosting of swallows and martins in specific areas throughout the Great Lakes, future changes in these patterns should be monitored because they may have important ecosystem and conservation implications.

Quantifying long‐term phenological patterns of aerial insectivores roosting in the Great Lakes region using weather surveillance radar

“…We explored phenological trends at three spatial levels: persistent local roost clusters, radar stations, and the Great Lakes region. Roost clusters were generated by a mean shift clustering algorithm (Pedregosa et al, 2011 ) using our roost track detections, which calculated density‐based clusters while maintaining a minimum shape (Belotti et al, in review ). First detection points included in the same cluster represent a region consistently used by swallows and martins to roost in a local region, thus defined as a roost cluster.…”

Monitoring aerial insect biodiversity: a radar perspective

“…Dual-pol data products improve estimates of shape, size and variety of animals aloft, and have been a key innovation for ecological applications [ 34 ], particularly by improving target differentiation [ 72 , 73 ]. Other rapidly moving fields such as machine learning or convolutional neural networks will probably expand and improve the (automated) identification of insect and other aerial organisms from radar signals [ 94 , 95 ].…”

DISCount: Counting in Large Image Collections with Detector-Based Importance Sampling

“…Previous work (Belotti et al 2023;Deng et al 2023) used a roost detector combined with manual screening of the detections to analyze more than 600,000 radar scans spanning a dozen stations in the Great Lakes region of the US to reveal patterns of bird migration over two decades. The vetting of nearly 64,000 detections was orders of magnitude faster than manual labeling, yet still required a substantial 184 hours of manual effort.…”

“…We use the exhaustively screened detections from the Great Lakes analysis in (Belotti et al 2023;Deng et al 2023) to systematically analyze the efficiency of sampling based counting. The data is organized into domains Ω sta,yr corresponding to 12 stations and 20 years (see Fig.…”