Abstract:

Aims: The selection of nocturnal roosting sites is a critical aspect of urban birds’ adaptation and survival within built environments. Existing studies have predominantly focused on macrohabitat or microhabitat characteristics, yet lack precise measurement methods and research on the perching branch information directly utilized by birds. LiDAR technology, with its high-precision data measurement capabilities, makes it possible to study the three-dimensional structure of avian perching branches. As a typical urban-adapted species, crows often form large-scale aggregated roosting groups in cities in winter, triggering human-wildlife conflict issues. To explore the key factors influencing their nocturnal roost selection and verify the feasibility of LiDAR in the refined research of bird habitats, this study took winter roosting crows in urban Beijing as the focal species and conducted surveys in the main crow aggregation areas within the 6th Ring Road of Beijing from 2023 to 2025.
Methods: Based on data collected on 18 habitat factors from 36 roosting quadrats and 36 control quadrats, we further employed terrestrial laser scanning to measure four variables for 1,361 roosting branches and 581 control branches. Among the control branches, 349 were from roosting trees and 232 from non-roosting trees. The Mann-Whitney U test was used to compare differences in the 18 habitat factors between roosting and control quadrats. The Kruskal-Wallis test was applied to examine pairwise differences among roosting branches, control branches on roosting trees, and control branches on non-roosting trees. Further, generalized linear mixed models (GLMMs) and generalized linear models (GLMs) were adopted to identify the key variables affecting crow nocturnal roost selection.
Results: The results indicated that tree height, canopy closure, crown width, noise, and illuminance were significantly higher in roosting quadrats compared to control quadrats. Furthermore, the height, diameter, and length of nocturnal roosting branches were significantly greater than those of both control branch groups, while the branch angle was significantly smaller. Between the two control groups, only branch height showed a significant difference. Analyses using GLMM revealed that tree height and noise were key habitat factors affecting nocturnal roost selection. GLM analyses revealed that branch angle, the quadratic term of branch height, and the interaction between branch diameter and branch length all had significant effects on the nocturnal roosting site selection by crows.
Conclusion: Crows exhibited a clear preference for nocturnal roosting sites characterized by taller trees and higher ambient noise levels. At the branch scale, they selectively perched on branches positioned at greater heights with shallower angles, which were typically either long and moderately thick or short and thick. This study elucidates the nocturnal roosting selection mechanism of crows in urban Beijing and offers a scientific basis for evidence-based urban bird management. Furthermore, it demonstrates the feasibility and distinct advantages of employing LiDAR technology in fine-scale habitat research, highlighting its potential to advance urban ornithological studies.

Key words: nocturnal roost selection, terrestrial laser scanner, crows, roosting branch architecture, urban birds