Hummingbirds rarely use torpor when incubating eggs

Our study that started in 2017 has finally been published! Congratulations to Dr. Erich Eberts, who was project lead for this project while he was finishing his undergraduate degree at Loyola Marymount University, and who stuck with the writing, analysis, and manuscript handling. It is rather apt that the study was accepted and In Press around about the same time that Dr. Eberts defended his PhD!

Here is the abstract:

Reproduction entails a trade-off between short-term energetic costs and long-term fitness benefits. This is especially apparent in small endotherms that exhibit high mass-specific metabolic rates and live in unpredictable environments. Many of these animals use torpor, substantially reducing their metabolic rate and often body temperature to cope with high energetic demands during non-foraging periods. In birds, when the incubating parent uses torpor, the lowered temperatures that thermally sensitive offspring experience could delay development or increase mortality risk. We used thermal imaging to noninvasively explore how nesting female hummingbirds sustain their own energy balance while effectively incubating their offspring. We located 67 active Allen’s hummingbird (Selasphorus sasin) nests in Los Angeles, California and recorded nightly time-lapse thermal images at 14 of these nests for 108 nights using thermal cameras. We found that nesting females usually avoided entering torpor, with one bird entering deep torpor on two nights (2% of nights), and two other birds possibly using shallow torpor on three nights (3% of nights). We also modeled nightly energetic requirements of a bird experiencing nest temperatures vs. ambient temperature and using torpor or remaining normothermic, using data from similarly-sized broad-billed hummingbirds. Overall, we suggest that the warm environment of the nest, and possibly shallow torpor, help brooding female hummingbirds reduce their own energy requirements while prioritizing the energetic demands of their offspring.

Thermal images of a normothermic hummingbird (A) and one in torpor (B). Right hand images are a 3D-rendering of the surface temperatures.
Digital and thermal images of eggs and hatchling hummingbirds.
Thermal video of a Ruby-throated hummingbird feeding from a feeding station. Video captured at Brock University in 2012, and has no association with the study.


Eberts, ER, Tattersall, GJ, Auger, PJ, Curley, M, Morado, MI, Strauss, EG, Powers, DR, Camacho, NM, Tobalske, BW, and Shankar, A. 2022. Free-living Allen’s hummingbirds (Selasphorus sasin) rarely use torpor while nesting. Journal of Thermal Biology. Available online 5 December 2022, 103391.


We thank the numerous undergraduate assistants who completed much of the nest searching, equipment maintenance, and data collection, CURes, the LMU grounds and facilities maintenance staff for assisting with the location of and access to nests. We also thank Susan Wethington for providing broad-bill hummingbird nests. We also thank Welch lab members (University of Toronto) for helpful discussions. We especially thank our crowdfunding campaign donors who participated in the crowd-source campaign and FLIR Systems for their support.

Thermal imaging of respiratory patterns during vocalisation

Our paper “Vocalization associated respiration patterns: thermography-based monitoring and detection of preparation for calling” was just published in the Journal of Experimental Biology! This was one of the most enjoyable research projects I have been part of lately, but also one of the more complex journeys for a research paper. Huge credit must go to Vlad Demartsev (Max Plank Institute of Animal Behaviour), lead author on the project! Congratulations, Vlad!

Here is the abstract of the study

Vocal emission requires coordination with the respiratory system. Monitoring the increase in laryngeal pressure, needed for vocal production, allows detection of transitions from quiet respiration to vocalization-supporting respiration. Characterization of these transitions could be used to identify preparation for vocal emission and to examine the probability of it manifesting into an actual vocal production event. Specifically, overlaying the subject’s respiration with conspecific calls can highlight events of call initiation and suppression, as a mean of signalling coordination and avoiding jamming. Here we present a thermal-imaging based methodology for synchronized respiration and vocalization monitoring of free ranging meerkats. The sensitivity of this methodology is sufficient for detecting transient changes in the subject’s respiration associated with the exertion of vocal production. The differences in respiration are apparent not only during the vocal output but also prior to it, marking the potential time frame of the respiratory preparation for calling. A correlation between conspecific calls with elongation of the focal subject’s respiration cycles could be related to fluctuations in attention levels or in the motivation to reply. This framework can be used for examining animals’ capability for enhanced respiration control during modulated and complex vocal sequences, detect “failed” vocalisation attempts and investigate the role of respiration cues in the regulation of vocal interactions.

Here is the supplementary video from the paper, demonstrating a thermal image video (taken with a FLIR T1030) of a basking and vocalisating meerkat. We also demonstrate the syncrhonisation procedure and show how the machine learning algorithm trained to identify the nostrils was used to obtain the coordinates from which we could go back to extract the median nostril temperature associated with inhalation and exhalation.

Technologically this was one of the most complicated studies I’ve worked on. It involved 3 weeks of some of most exciting field work in the Kalahari Desert (waiting for Meerkats to come out of their burrows in the morning), high resolution thermal imaging, high resolution audio recording, elaborate device synchronisation, ImageJ, R, machine learning, cigarette lighters, and more PERL code than I ever want to have to write again.

Long story short: we estimated the rhythmic pattern of inhalation and exhalation from the periodic changes in nostril temperature due to evaporative cooling (inhalation) and respiratory warming (exhalation). These breaths were tracked along with the morning “sunning calls” (i.e. vocalisations), and we were able to detect the subtle changes in their breathing patterns that emerge as a result of their calls.


Demartsev, V, Manser, MB, and Tattersall, GJ. 2022. Vocalization associated respiration patterns: thermography-based monitoring and detection of preparation for calling. Journal of Experimental Biology, In Press,


This work was done while VD was funded by Minerva Stiftung and Alexander von Humboldt-Stiftung post-doctoral fellowships. Additional funding included Internationalization Initiative Start Up funding, University of Konstanz and Aharon and Ephraim Katzir Study Grant, The Israel Academy of Sciences and Humanities. The Natural Sciences and Engineering Research Council of Canada supported GJT’s research and thermal imaging camera (RGPIN-05814). MBM was funded by the University of Zurich. This article has relied on records of individual identities and/or life histories maintained by the Kalahari Meerkat Project, which has been supported financially by the European Research Council (Grant No 742808 to Tim Clutton-Brock, University of Cambridge since 1 July 2018) and the University of Zurich, as well as logistically by the Mammal Research Institute of the University of Pretoria.