Using infrared thermography to detect night-roosting birds

William Francis Mitchell, Rohan Hartley Clarke

Research output: Contribution to journalArticleResearchpeer-review

3 Citations (Scopus)


Most birds sleep while roosting at night. Although a widespread behavior, few investigators have studied the nocturnal roosting behavior of birds. Studies conducted to date have either focused on species that roost communally or used radio-telemetry to locate sleeping individuals of a few focal species. Portable thermal cameras capable of detecting infrared (IR) heat signals may provide a more efficient and less invasive means of detecting nocturnal-roosting endotherms such as birds. Our objective was to assess the efficacy of using thermal cameras to detect roosting birds in a woodland bird community in southeastern Australia. To better understand the limitations of using thermography to detect roosting birds, paired bird surveys were conducted along 44 transects from May to September 2016 using both traditional survey techniques during the day and surveys with a thermal camera at night. We detected 195 birds representing 21 species at nocturnal roosts using IR thermography, with the detection rate of birds during nocturnal surveys approximately one-third (29.1%) that during diurnal surveys. Detection rates during nocturnal surveys declined more steeply with distance from observers than for diurnal surveys. Detection rates were significantly higher during diurnal surveys for 14 species of woodland birds, but did not differ between diurnal and nocturnal surveys for eight other species. Roost height, roost visibility, bird mass, and cluster size (i.e., two or more birds in physical contact) did not differ between species categorized as having high or low detectability during nocturnal surveys. Variability among species in nocturnal-detectability could not be attributed to roost-site visibility, roost height, or bird size. Positive detection biases associated with diurnal behavior, such as movement and vocalizations, and limitations of current IR technology, e.g., low resolution, likely contributed to overall lower detection rates during nocturnal surveys. However, our results suggest that infrared thermography can be an effective and useful technique for detecting roosting birds and studying roosting behavior, as well as for population monitoring under certain conditions.

Original languageEnglish
Pages (from-to)39-51
Number of pages13
JournalJournal of Field Ornithology
Issue number1
Publication statusPublished - 2019


  • bird sleep
  • distance sampling
  • nocturnal survey
  • survey bias
  • thermal camera
  • thermal imaging

Cite this