Relationships between air temperature (Tair) and avian body temperature (Tb), resting metabolic rate (RMR) and evaporative water loss (EWL) during acute heat exposure can be quantified through respirometry using several approaches. One involves birds exposed to a stepped series of progressively increasing Tair setpoints for short periods (<20–30 min), whereas a second seeks to achieve steady-state conditions by exposing birds to a single Tair for longer periods (>1–2 h). To compare these two approaches, we measured Tb, RMR and EWL over Tair=28°C to 44°C in the dark-capped bulbul (Pycnonotus tricolor). The two protocols yielded indistinguishable values of Tb, RMR and EWL and related variables at most Tair values, revealing that both are appropriate for quantifying avian thermal physiology during heat exposure over the range of Tair in the present study. The stepped protocol, however, has several ethical and practical advantages.

Quantifying interspecific and intraspecific variation in the thermoregulatory performance of endotherms at environmental temperatures approaching or exceeding body temperature (Tb) is essential for understanding how animals persist in hot environments and how heat tolerance and evaporative cooling capacity have evolved in response to climate and with organismal traits (Czenze et al., 2020b; Dawson, 1954; Tieleman et al., 2002; Weathers, 1981). Among birds and mammals, hyperthermia tolerance and evaporative cooling capacity vary substantially among taxa. For instance, phylogenetic variation in the primary avenue of avian evaporative heat dissipation is associated with large differences in heat tolerance under conditions of acute heat exposure. Representatives of taxa that employ gular flutter (e.g. caprimulgids) or rapid cutaneous evaporation (e.g. columbids) often tolerate air temperatures (Tair) 5–10°C higher than taxa in which panting is a primary pathway (e.g. passerines), yet also maintain lower maximum Tb values (reviewed by McKechnie et al., 2021). Quantifying the upper limits of endotherms, heat tolerance has taken on new urgency in the face of rapid anthropogenic global heating (IPCC, 2021).

Relationships between resting evaporative heat loss, metabolic heat production and Tb at high Tair are typically quantified using open flow-through respirometry (e.g. Dawson, 1954; Lasiewski et al., 1966). Two approaches are commonly used for obtaining data over a sufficiently wide range of Tair to accurately estimate inflections, slopes and upper limits for Tb, evaporative water loss (EWL) and resting metabolic rate (RMR). One, which hereafter we refer to as the stepped approach, involves exposing an animal to incrementally increasing Tair setpoints for intervals of 10–30 min per setpoint value (e.g. Smith et al., 2015; Whitfield et al., 2015). The use of a stepped Tair profile to quantify limits to heat tolerance is conceptually analogous to the sliding cold-exposure protocol used to elicit maximum rates of resting metabolic thermogenesis, or summit metabolism, where animals are exposed to continuously decreasing environmental temperature in a Helox (79% He, 21% O2) atmosphere (Swanson et al., 1996). The stepped approach has been used in a number of recent studies relating to the upper limits of avian thermoregulation (e.g. Albright et al., 2017; Freeman et al., 2020; Smit et al., 2018).

A second approach (hereafter, the steady-state approach) involves exposure to each setpoint Tair value for longer periods. Widely employed by ecological and evolutionary physiologists, the steady-state approach has been used for measurements at Tair>Tb (1–3 h, Marder, 1973; 1 h, Weathers, 1981, 1997) and Tair just below Tb (5–9 h, Cooper et al., 2020) when evaluating thermoregulation in the heat. Steady-state measurements with measurement intervals of >6 h are necessary for accurately quantifying standard physiological variables such as basal metabolic rate at thermoneutral Tair, where shorter measurement periods result in overestimates of baseline physiological rates (Cooper and Withers, 2009; Jacobs and McKechnie, 2014; Page et al., 2011).

However, steady-state measurements for extended periods are potentially problematic when investigating thermoregulation at Tair>Tb. First, rates of EWL increase rapidly at Tair >Tb and create a risk of lethal dehydration; in small passerines, rates of EWL at Tair approaching 50°C may reach values equivalent to 7% of body mass (Mb) per hour (Wolf and Walsberg, 1996a). Second, with the exception of species inhabiting Earth's hottest deserts, exposure to Tair≥50°C under laboratory conditions is equivalent to thermal conditions birds are likely to experience only briefly in sunlit microsites during the heat of the day (Bakken, 1976; Robinson et al., 1976; Wolf and Walsberg, 1996b). Therefore, exposing birds to Tair≥50°C in a metabolic chamber simulates thermal conditions most birds experience over brief time scales of minutes under natural conditions, as the hottest part of the day is usually associated with curtailment of activity and retreat to shaded microsites (Austin, 1976; Pattinson et al., 2020; Ricklefs and Hainsworth, 1968).

The stepped protocol has, over the last decade, been used to quantify upper limits to heat tolerance and evaporative cooling in 56 arid-zone bird species (reviewed by McKechnie et al., 2021), ∼40 species from mesic environments (Freeman et al., 2022) and smaller numbers of bats (Czenze et al., 2020a, 2022) and rodents (van Jaarsveld et al., 2021). However, concerns have been expressed from time to time by a subset of reviewers about the validity of this approach, with these concerns centred on the argument that longer exposure to each Tair, and hence, steady-state measurements, is required to accurately quantify RMR, EWL and Tb during heat exposure. Here, we compared patterns of Tb, EWL and RMR at Tair approaching and exceeding normothermic Tb quantified using each approach in a southern African passerine. Our aim was to establish whether conclusions regarding avian thermoregulation during acute heat exposure depend on the respirometry protocol and whether the quantitative conclusions drawn from a dataset vary with the approach used.

Study site and species

We captured 20 dark-capped bulbuls [Pycnonotus tricolor (Hartlaub 1862); Mb=36.92±3.21 g; hereafter, bulbuls] using mist nets at the University of Pretoria's experimental farm (25°45′6″S, 28°15′10.0″E) in Pretoria, South Africa, during March and April 2021. This species is widespread and common in woodland, savanna and urban habitats in eastern and southern Africa, and feeds on berries, fruits, insects and nectar (Lloyd, 2005). Some authorities consider it a subspecies of common bulbul (P. barbatus tricolor). Following capture, birds were transported <200 m in cloth bags to the university's Small Animal Physiological Research Facility, where they were housed individually in indoor cages (60×40×40 cm) and provided with ad libitum food and water for the duration of the study. The bulbuls' diet consisted of apple, banana, Pronutro breakfast cereal mixed with water, and a sucrose solution mixed with a protein supplement (Ensure, Abbott, Abbott Park, IL, USA) following Lerch-Henning and Nicolson (2013).

Gas exchange measurements

Rates of gas exchange were measured using two open flow-through respirometry systems. Our experimental setup consisted of a single custom-built temperature-regulated ice chest (100 litres), with a Peltier device (AC-162 Thermoelectric Air Cooler, TE Technology, Traverse City, MI, USA) mounted through one wall and controlled by a digital controller (TC-36-25 RS485 Temperature Controller, TE Technology), which regulated experimental Tair. Birds were placed in individual 3-litre airtight plastic metabolic chambers (20×15×10 cm, height×length×width; previously shown not to adsorb water vapour; Whitfield et al., 2015) within the temperature-controlled ice chest. Each metabolic chamber was equipped with a wire mesh platform on which the bird rested 10 cm above a ∼1 cm layer of mineral oil to trap excreta. In order to ensure that birds remained calm during the assessments, the chambers were kept in darkness. An air inlet in the form of an upward-facing elbow joint was situated just below the chamber lid, with an outlet below the mesh platform, maximising air mixing within the chamber.

Atmospheric air was supplied by an oil-free compressor (MacAfric, Johannesburg, South Africa) and scrubbed of water vapour using a membrane dryer (Champion®CMD3 air dryer and filter, Champion Pneumatic, Princeton, IL, USA). The dried air was split into baseline and experimental channels, with the rate at which air flowed through these channels regulated by a needle valve (Swagelok, Solon, OH, USA) and digital mass flow controllers (Alicat Scientific, Tuscon, AZ, USA), respectively. Chamber flow rates between 6 and 12 l min−1 were used. Elevated humidity within metabolic chambers impedes evaporative heat dissipation (Lasiewski et al., 1966) and we increased flow rates at higher Tair values to maintain a chamber dewpoint <−10°C. On account of the high flow rates we used, 99% equilibrium times calculated following Lasiewski et al. (1966) averaged 1.34 min and never exceeded 1.71 min.

Excurrent air from each chamber and baseline channel was subsampled using a respirometry multiplexer (model MUX3-1101-18 M, Sable Systems, Las Vegas, NV, USA), and then pulled through a CO2/H2O analyser (model LI-840A, LI-COR, Lincoln, NE, USA) using a subsampler (SS-4, Sable Systems). The CO2/H2O analysers were regularly zeroed with nitrogen and spanned using gas with a known CO2 concentration of 1900 ppm (Afrox, Johannesburg, South Africa) and humidified air with a dewpoint of 15°C produced by a dewpoint generator (DG-4, Sable Systems).

Air and body temperature measurements

The Tb of the bulbuls was measured using a temperature-sensitive passive integrated transponder (PIT) tag (Biotherm 13, Biomark, Boise, ID, USA) injected into the peritoneal cavity of each bird. Temperature data were collected using a PIT tag transceiver (HPR+, Biomark) placed on top of each metabolic chamber prior to the commencement of measurements. Before injection, PIT tags were calibrated over temperatures of 35–50°C in a circulating water bath (model F34, Julabo, Seelbach, Germany) against a digital thermocouple reader (model RDXL12SD, Omega, Stamford, CT, USA) calibrated against a mercury-in-glass thermometer with NIST-traceable accuracy. During gas exchange measurements, Tair within each metabolic chamber was measured using a thermistor probe (TC-100, Sable Systems) inserted through a rubber grommet fitted to one wall of the chamber.

Experimental protocol

We measured RMR, EWL and Tb of the bulbuls using a stepped and steady-state respirometry protocol (10 individuals per protocol). For both protocols, measurements took place during daytime (approximately 05:30–18:30 h during the study period), the active phase for this diurnal species. Before measurements commenced, birds were kept in the chamber at Tair=28°C for at least 1 h, to ensure they were habituated to chamber conditions. Typically, birds had not had access to food for 2 h prior to measurements. Measurements took place at each of six Tair setpoints (28°C, 32°C, 36°C, 40°C, 42°C and 44°C), with 4°C increments for Tair≤40°C and 2°C increments for Tair>40°C following previous studies (e.g. Czenze et al., 2020b; Smit et al., 2018). Unlike these previous studies, however, we did not expose birds to Tair>44°C on account of the potential dehydration risk involved (see below). During all measurements, the behaviour of the bulbuls was monitored using a small surveillance camera and infrared light source within the modified ice chest, with individuals removed from the chamber immediately if they exhibited signs of prolonged escape behaviour, loss of coordination, or rapid, unregulated increases in Tb or decreases in RMR or EWL, indicating the individual had reached its thermal endpoint (Whitfield et al., 2015). Two individuals reached thermal endpoints during the steady-state measurements at Tair=44°C and were immediately removed from the chamber. Upon removal from the chamber, heat loss was augmented by dabbing each bulbul's underparts (irrespective of whether thermal endpoints were reached) with 80% ethanol and exposing the bird to cooled air supplied by a portable air conditioner. Once Tb returned to normothermic values, birds were placed back in their cages with ad libitum food and water. After completion of the study, the birds were released at the site of capture.

All experimental procedures were approved by the Animal Ethics Committee of the University of Pretoria (protocol NAS114-2020) and the Research Ethics and Scientific Committee of the South African National Biodiversity Institute (protocol P2019-13).

Stepped protocol

Measurements conducted with this protocol commenced with baseline measurements for 5–15 min, before bulbuls were exposed to the lowest Tair setpoint (28°C) for at least 5–15 min, with the lowest, most stable 5-min trace of CO2 and H2O used to estimate rates of gas exchange. Thereafter, Tair was adjusted to the next setpoint and a second measurement of 5–15 min was obtained, whereafter individuals were exposed to the next Tair setpoint, with this cycle repeated until the end of measurements at Tair=44°C. Measurement intervals were extended in cases where the bird became agitated (no longer than ∼30 min), in order to allow the bird to settle and obtain stable traces of CO2 and H2O over at least 5 min.

Steady-state protocol

In contrast to the stepped protocol, birds were exposed to a single Tair setpoint for 4 h at Tair≤40°C and 2.5 h at Tair >40°C. These measurement intervals were estimated from EWL data previously collected for this species (M.T.F., unpublished data) and were calculated from the time required for cumulative EWL to reach values equivalent to 10% of Mb (Fig. 1). Measurements for each individual took place in a random sequence of Tair setpoints. With the exception of two individuals, birds were not used for measurements on successive days. In all cases, successive measurements were only taken when an individual's Mb was within 3 g (∼8%) of the value at the start of the first set of measurements. Similar to the stepped protocol, measurements commenced with 10–20 min baseline readings of CO2 and H2O, with another baseline taken once the measurement interval had come to an end (i.e. 2.5–4 h later).

Fig. 1.

Cumulative evaporative water loss (EWL) in dark-capped bulbuls (Pycnonotus tricolor) as a function of time spent in a respirometry chamber at a constant air temperature (Tair), expressed as a percentage of body mass (Mb). Estimates are based on data for 10 individuals not involved in the present study, collected using a stepped protocol in early 2020 (M.T.F., unpublished data). The horizontal dashed line shows cumulative EWL equivalent to 10% Mb, the threshold we used to estimate measurement durations in the present study.

Fig. 1.

Cumulative evaporative water loss (EWL) in dark-capped bulbuls (Pycnonotus tricolor) as a function of time spent in a respirometry chamber at a constant air temperature (Tair), expressed as a percentage of body mass (Mb). Estimates are based on data for 10 individuals not involved in the present study, collected using a stepped protocol in early 2020 (M.T.F., unpublished data). The horizontal dashed line shows cumulative EWL equivalent to 10% Mb, the threshold we used to estimate measurement durations in the present study.

Data analysis

We corrected traces of CO2 and H2O for analyser drift and lag using ExpeData software (Sable Systems). Carbon dioxide production (CO2) and EWL was calculated from the lowest stable 5-min periods of CO2 and H2O concentrations at each Tair setpoint using eqns 9.5 and 9.6 of Lighton (2008), assuming a vapour density of 0.803 mg ml−1. The allometrically predicted digesta retention time for a 37 g bird is 64 min (Karasov, 1990), but is likely to have been shorter in the bulbuls on account of their diet consisting predominantly of fruit (Downs, 2008; Karasov, 1990). As it is likely that all individuals were post-absorptive prior to measurements, resting metabolic rate at each Tair setpoint was estimated from CO2 assuming a respiratory exchange ratio (RER) of 0.71, and converted to Watts (W) using a conversion factor of 27.8 J ml−1 CO2 (Walsberg and Wolf, 1995). To calculate evaporative heat loss/metabolic heat production (EHL/MHP), we converted rates of EWL to W assuming a latent heat of vaporization of water of 2.406 J mg−1 at 40°C (Tracy et al., 2010).

All analyses were conducted in R 3.5.2. Relationships between Tair and RMR, EWL, Tb and EHL/MHP were analysed using general linear mixed-effects models, with inflection points identified in the R package segmented (Muggeo, 2008; version 1.3-4) and the upper slopes of thermoneutrality fitted using the R package nlme (https://CRAN.R-project.org/package=nlme; version 3.1-153). Individual identity was included as a random effect in all analyses to account for the repeated-measures design of our experiment. Variation in slopes of Tb, RMR, EWL and EHL/MHP as functions of Tair between the stepped and steady-state respirometry protocols were tested for using a general linear model analysis (ANCOVA), performed in the R package car (Fox and Weisberg, 2011; version 3.0-12). Differences were identified by the presence of significant interaction effects, with Tair setpoint (Tair≥36°C) and respirometry protocol serving as the categorical predictor variables and RMR, EWL, Tb and EHL/MHP as continuous response variables. Means of RMR, EWL, Tb and EHL/MHP at each Tair setpoint were compared in a pairwise manner, using a Tukey’s honest significance difference (HSD) test in R. Type I errors were accounted for by maintaining a family-wise error below α=0.05. Significance was assessed at α<0.05, with values presented as means±s.d.

Bulbuls exposed to the steady-state protocol exhibited no significant differences in minimum normothermic or maximum hyperthermic Tb when compared with individuals exposed to the stepped protocol (Table 1). However, the steady-state group had a significantly lower mean Tb (by ∼0.7°C) at Tair=28°C compared with the stepped group (Table 1, Fig. 2). Whereas similar inflection Tair values were observed for both protocols, the Tb of individuals experiencing the steady-state protocol increased more rapidly with increasing Tair compared with the stepped protocol (Table 1, Fig. 2), despite the lack of difference in maximum Tb reached.

Fig. 2.

Body temperature (Tb), resting metabolic rate, evaporative water loss and evaporative heat loss/metabolic heat production (EHL/MHP) in dark-capped bulbuls (Pycnonotus tricolor) measured using either a stepped protocol involving brief measurements at a series of progressively higher air temperatures (Tair) or a steady-state protocol involving a measurement duration of 4 h (Tair≤40°C) or 2.5 h (Tair>40°C). Significant differences between protocols at Tair=28°C are indicated by asterisks. In addition, the slope of Tb at Tair≥36°C differed significantly between protocols, with the solid and dashed lines indicating linear regression models fitted to the data.

Fig. 2.

Body temperature (Tb), resting metabolic rate, evaporative water loss and evaporative heat loss/metabolic heat production (EHL/MHP) in dark-capped bulbuls (Pycnonotus tricolor) measured using either a stepped protocol involving brief measurements at a series of progressively higher air temperatures (Tair) or a steady-state protocol involving a measurement duration of 4 h (Tair≤40°C) or 2.5 h (Tair>40°C). Significant differences between protocols at Tair=28°C are indicated by asterisks. In addition, the slope of Tb at Tair≥36°C differed significantly between protocols, with the solid and dashed lines indicating linear regression models fitted to the data.

Table 1.

Comparison of thermoregulatory variables in dark-capped bulbuls (Pycnonotus tricolor) measured using either a stepped protocol involving rapid exposure to each of a series of progressively increasing air temperature (Tair) setpoints or a steady-state protocol involving exposure for 2.5 or 4 h to a single Tair setpoint

Comparison of thermoregulatory variables in dark-capped bulbuls (Pycnonotus tricolor) measured using either a stepped protocol involving rapid exposure to each of a series of progressively increasing air temperature (Tair) setpoints or a steady-state protocol involving exposure for 2.5 or 4 h to a single Tair setpoint
Comparison of thermoregulatory variables in dark-capped bulbuls (Pycnonotus tricolor) measured using either a stepped protocol involving rapid exposure to each of a series of progressively increasing air temperature (Tair) setpoints or a steady-state protocol involving exposure for 2.5 or 4 h to a single Tair setpoint

No significant inflection Tair for RMR corresponding with an upper critical limit of thermoneutrality was evident in individuals exposed to either protocol (Table 1). Minimum and maximum RMR values did not differ significantly between the two protocols other than the steady-state group having a significantly lower RMR (by ∼25%) at Tair=28°C compared with the stepped group (Table 1, Fig. 2). To evaluate whether this difference potentially reflected a habituation effect, we compared RMR of the steady-state birds between 10 and 70 min following the end of the initial habituation period with values later in each run. The RMR measured during this initial 10–70 min period was significantly higher (0.47±0.07 W, P<0.001) than RMR measured later (90–150 min after the habituation period, 0.37±0.09 W; 170–230 min, 0.35±0.10 W), supporting the notion that the differences between the two treatments reflects a habituation effect during the early stages of measurements.

Neither minimum nor maximum EWL differed between the two protocols (Table 1). Moreover, individuals in the steady-state group had similar mean EWL values at each Tair setpoint compared with individuals in the stepped group (Tukey test, P>0.05; Table 1, Fig. 2). Similarly, the inflection point (38.34°C) and slope (0.11 g h−1 °C−1) for individuals exposed to the steady-state protocol did not differ significantly from the inflection point (37.90°C) and slope (0.09 g h−1 °C−1) obtained for individuals exposed to the stepped protocol (Table 1, Fig. 2). Neither minimum nor maximum EHL/MHP differed between the two protocols (Table 1), nor did mean EHL/MHP at each Tair setpoint (Table 1, Fig. 2). When plotted as a function of TairTb, neither inflection TairTb nor slope differed between individuals exposed to the steady-state protocol and those exposed to the stepped protocol (Table 1, Fig. 2).

Our data confirm that, with only a few exceptions (RMR at Tair=28°C and the slope of Tb as a function of Tair), patterns of thermoregulation in dark-capped bulbuls did not differ when measured using a stepped series of increasing Tair values or steady-state conditions involving longer exposure per Tair setpoint. One difference between the two protocols concerns the significantly steeper slope of Tb at Tair>36°C during steady-state measurements compared with individuals experiencing the stepped protocol. Whereas this finding could be interpreted as birds not reaching thermal equilibrium during their relatively brief exposure to each Tair setpoint in the stepped protocol, the lack of any corresponding differences in EWL, RMR or EHL/MHP argues against this interpretation, as do the similar maximum Tb values reached in both protocols. Instead, we argue the steeper slope for steady-state birds reflects an experimental artefact; longer measurement intervals at Tair>36°C reduced the periods over which birds could defend a setpoint Tb, reflecting limited endurance rather than thermoregulatory capacity during acute heat exposure. Furthermore, under natural conditions, the bulbuls in our study population would rarely be exposed to operative temperature >40°C (Te; Bakken, 1976; Robinson et al., 1976) for more than brief periods. The population from which we sampled birds occupies an area characterized by a mesic climate, with an average summer (October to March) maximum Tair of 28.0±1.4°C and only a single recorded instance of Tair=40.1°C between 1970 and 2000 (Fick and Hijmans, 2017). For dark-capped bulbuls in Pretoria, exposure to Tair≥40°C has historically been rare. Moreover, the species is predominantly frugivorous (Lloyd, 2005), with the majority of foraging occurring in shaded microsites. The only time these bulbuls are likely to experience operative temperatures of 44°C or above is when they are in sunlit microsites during the heat of the day, transient conditions likely experienced for only brief periods. Hence, the exposure to Tair setpoints of 40°C, 42°C and 44°C for 150 min each experienced by the birds in the steady-state protocol represents thermal conditions not likely encountered naturally.

The second significant difference between protocols involved Tb and RMR at Tair=28°C, with the steady-state protocol eliciting significantly lower mean values for both variables at this Tair compared with the stepped protocol. These findings are consistent with those reported by Page et al. (2011), who showed the RMR, EWL and Tb of budgerigars (Melopsittacus undulatus) measured at Tair=30°C decreased significantly with increasing measurement duration. The Page et al. (2011) study is one of several revealing that adequate habituation to metabolic chamber conditions is important for eliciting minimal values of standard physiological variables such as basal metabolic rate (Cooper and Withers, 2009; Page et al., 2011; Jacobs and McKechnie, 2014). In the present study, we interpret the lower Tb and RMR at Tair=28°C among birds experiencing the steady-state protocol as a habituation effect, reflecting the fact that the steady-state birds experienced Tair=28°C for 5 h (i.e. 1 h habituation+4 h measurements), compared with ∼100 min (i.e. 1 h habituation+10 min baseline+10 min measurements) in the stepped protocol. The significantly higher RMR in steady-state birds 10–70 min after habituation compared with 90–150 min or 170–230 min supports this interpretation, also suggesting that increasing habituation time by ∼30 min would likely have avoided this effect. A longer habituation period than we used here should thus be considered when making use of the stepped protocol. The lack of significant differences at any Tair value >28°C, however, argues against the possibility that habituation effects influenced overall patterns observed.

Our study has several limitations, most notably the fact that it was restricted to one species. In addition, our low sample size (10 individuals per protocol group) could have affected the reliability of our findings (Button et al., 2013). However, the small number of marginally insignificant P-values (Table 1) argues against low statistical power being a major influence on our conclusions (Chase and Tucker, 1976). Nevertheless, further studies with larger sample sizes for multiple species are needed to confirm our present findings.

The use of a stepped Tair protocol for quantifying avian thermal physiology at Te approaching and exceeding Tb has several advantages compared with a steady-state approach. One concerns the ethical aspects of investigating avian thermoregulation during acute heat exposure. Brief exposure to each Tair setpoint, combined with close monitoring of birds while in chambers, permits measurements at Tair far exceeding Tb (in some cases by∼20°C; (Czenze et al., 2021; McKechnie et al., 2016) and the quantification of maximum values of physiological variables just before the onset of thermoregulatory failure (e.g. O'Connor et al., 2017; Smit et al., 2018; Whitfield et al., 2015) without adverse effects on the birds.

In several instances, birds and bats subjected to the stepped protocol have been monitored and/or recaptured several weeks after release, with no apparent negative consequences (Czenze et al., 2022; Kemp and McKechnie, 2019). Attempting to collect the same data with a steady-state protocol in which birds are exposed to a single Tair for several hours creates a significant risk of lethal dehydration and, for many species, is probably not feasible for Tair much above Tb; the maximum Tair to which we exposed bulbuls in the present study (44°C) was associated with cumulative EWL equivalent to ∼10% of Mb over the 150-min duration of measurements.

The shorter measurement intervals and stepped sequence of Tair of the stepped protocol also have substantial practical advantages in terms of the time and resources required to collect data. To characterise thermal physiology for 10 individuals of a hypothetical species with a heat tolerance limit of 52°C starting at Tair=28°C, using 4°C increments in Tair≤40°C and 2°C increments in Tair>40°C (following Czenze et al., 2020b; Whitfield et al., 2015), would require 4 days (assuming a single respirometry system and three measurement sessions per day). Collecting the same data using a steady-state approach would require 10 separate sets of measurements per individual (i.e. one set per Tair value), equivalent to a total of 100 respirometry sessions over 25 days (assuming four respirometry sessions per day). In addition to taking ∼6-fold longer to collect data for one species, the steady-state approach would require either (1) 10 individuals held in captivity while data are collected, or (2) 100 individuals caught, measured at a single Tair each and then released. Thus, the stepped protocol allows for data collection for six species during the time required to collect data for a single species using the steady-state approach.

In conclusion, our data confirm that respirometric measurements of avian thermoregulation in the heat using rapid exposure to progressively higher Tair setpoints yield similar results to measurements made over longer durations. In addition to facilitating much more rapid data collection compared with steady-state protocols, the stepped approach is preferable from an ethical perspective as it avoids exposing birds for long periods to operative temperatures most species would experience transiently in natural environments.

We thank Mathome Makola for his assistance during the course of this study, and an anonymous reviewer for constructive comments that improved the manuscript.

Author contributions

Conceptualization: M.T.F., A.E.M.; Methodology: J.C.S., M.T.F., A.E.M.; Formal analysis: J.C.S., M.T.F.; Investigation: J.C.S., M.T.F.; Resources: A.E.M.; Data curation: J.C.S.; Writing - original draft: J.C.S.; Writing - review & editing: M.T.F., A.E.M.; Supervision: M.T.F., A.E.M.; Funding acquisition: A.E.M.

Funding

This work was supported by the National Research Foundation of South Africa (grant 119754 to A.E.M.). Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Research Foundation of South Africa.

Data availability

Data are available from Dryad (McKechnie et al., 2022): https://doi.org/10.5061/dryad.b8gtht7fz.

Albright
,
T. P.
,
Mutiibwa
,
D.
,
Gerson
,
A. R.
,
Smith
,
E. K.
,
Talbot
,
W. A.
,
O'Neill
,
J. J.
,
McKechnie
,
A. E.
and
Wolf
,
B. O.
(
2017
).
Mapping evaporative water loss in desert passerines reveals an expanding threat of lethal dehydration
.
Proc. Natl. Acad. Sci. USA
114
,
2283
-
2288
.
Austin
,
G. T.
(
1976
).
Behavioral adaptations of the verdin to the desert
.
Auk
93
,
245
-
262
.
Bakken
,
G. S.
(
1976
).
A heat transfer analysis of animals: unifying concepts and the application of metabolism chamber data to field ecology
.
J. Theor. Biol.
60
,
337
-
384
.
Button
,
K. S.
,
Ioannidis
,
J. P.
,
Mokrysz
,
C.
,
Nosek
,
B. A.
,
Flint
,
J.
,
Robinson
,
E. S.
and
Munafò
,
M. R.
(
2013
).
Power failure: why small sample size undermines the reliability of neuroscience
.
Nat. Rev. Neurosci.
14
,
365
-
376
.
Chase
,
L. J.
and
Tucker
,
R. K.
(
1976
).
Statistical power: derivation, development, and data-analytic implications
.
Psychol. Rec.
26
,
473
-
486
.
Cooper
,
C. E.
and
Withers
,
P. C.
(
2009
).
Effects of measurement duration on the determination of basal metabolic rate and evaporative water loss of small marsupials: how long is long enough?
Physiol. Biochem. Zool.
82
,
438
-
446
.
Cooper
,
C.
,
Hurley
,
L.
and
Griffith
,
S.
(
2020
).
Effect of acute exposure to high ambient temperature on the thermal, metabolic and hygric physiology of a small desert bird
.
Comp. Biochem. Physiol. A Mol. Integr. Physiol.
244
,
110684
.
Czenze
,
Z.
,
Naidoo
,
S.
,
Kotze
,
A.
and
McKechnie
,
A.
(
2020a
).
Bat thermoregulation in the heat: limits to evaporative cooling capacity in three southern African bats
.
J. Therm. Biol.
89
,
102542
.
Czenze
,
Z. J.
,
Kemp
,
R.
,
Van Jaarsveld
,
B.
,
Freeman
,
M. T.
,
Smit
,
B.
,
Wolf
,
B. O.
and
McKechnie
,
A. E.
(
2020b
).
Regularly drinking desert birds have greater evaporative cooling capacity and higher heat tolerance limits than non-drinking species
.
Funct. Ecol.
34
,
1589
-
1600
.
Czenze
,
Z. J.
,
Freeman
,
M. T.
,
Kemp
,
R.
,
van Jaarsveld
,
B.
,
Wolf
,
B. O.
and
McKechnie
,
A. E.
(
2021
).
Efficient evaporative cooling and pronounced heat tolerance in an eagle-owl, a thick-knee and a sandgrouse
.
Front. Ecol. Evol.
9
,
799302
.
Czenze
,
Z. J.
,
Smit
,
B.
,
van Jaarsveld
,
B.
,
Freeman
,
M. T.
and
McKechnie
,
A. E.
(
2022
).
Caves, crevices, and cooling capacity: roost microclimate predicts heat tolerance in bats
.
Funct. Ecol.
36
,
38
-
50
.
Dawson
,
W. R.
(
1954
).
Temperature regulation and water requirements of the brown and Abert towhees, Pipilo fuscus and Pipilo aberti
. In
University of California Publications in Zoology
, Vol.
59
(ed.
G. A.
Bartholomew
,
F.
Crescitelli
,
T. H.
Bullock
,
W. H.
Furgason
and
A. M.
Schechtman
), pp.
81
-
123
.
Berkeley
:
University of California Press
.
Downs
,
C. T.
(
2008
).
Aspects of diet choice and digestion in the dark-capped bulbul Pycnonotus barbatus
.
Ostrich
79
,
73
-
78
.
Fick
,
S. E.
and
Hijmans
,
R. J.
(
2017
).
WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas
.
Int. J. Climatol.
37
,
4302
-
4315
.
Fox
,
J.
and
Weisberg
,
S.
(
2011
).
An {R} Companion to Applied Regression
, 2nd edn.
Thousand Oaks, CA
:
Sage
.
Freeman
,
M. T.
,
Czenze
,
Z. J.
,
Schoeman
,
K.
and
McKechnie
,
A. E.
(
2020
).
Extreme hyperthermia tolerance in the world's most abundant wild bird
.
Sci. Rep.
10
,
13098
.
Freeman
,
M. T.
,
Czenze
,
Z. J.
,
Schoeman
,
K.
and
McKechnie
,
A. E.
(
2022
).
Adaptive variation in the upper limits of avian body temperature
.
Proc. Natl. Acad. Sci. USA
119
,
e2116645119
.
IPCC
(
2021
).
Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (ed. V. Masson-Delmotte, P. Zhai, A. Pirani, S. L. Connors and others). Cambridge: IPCC.
Jacobs
,
P. J.
,
McKechnie
,
A. E.
(
2014
).
Experimental sources of variation in avian energetics: estimated basal metabolic rate decreases with successive measurements
.
Physiol. Biochem. Zool.
87
,
762
-
769
.
Karasov
,
W. H.
(
1990
).
Digestion in birds: chemical and physiological determinants and implications
. In
Studies in Avian Biology No. 13
(ed.
M. L.
Morrison
,
C. J.
Ralph
,
J.
Verner
and
J. R.
Jehl
), pp.
391
-
415
.
Los Angeles
:
Cooper Ornithological Society
.
Kemp
,
R.
and
McKechnie
,
A. E.
(
2019
).
Thermal physiology of a range-restricted desert lark
.
J. Comp. Physiol. B
189
,
131
-
141
.
Lasiewski
,
R. C.
,
Acosta
,
A. L.
and
Bernstein
,
M. H.
(
1966
).
Evaporative water loss in birds—I. Characteristics of the open flow method of determination, and their relation to estimates of thermoregulatory ability
.
Comp. Biochem. Physiol.
19
,
445
-
457
.
Lerch-Henning
,
S.
and
Nicolson
,
S. W.
(
2013
).
Bird pollinators differ in their tolerance of a nectar alkaloid
.
J. Avian Biol.
44
,
408
-
416
.
Lighton
,
J. R. B.
(
2008
).
Measuring Metabolic Rates: A Manual for Scientists
.
Oxford
:
Oxford University Press
.
Lloyd
,
P.
(
2005
).
Dark-capped bulbul
. In
Roberts Birds of Southern Africa
(ed.
P. A. R.
Hockey
,
W. R. J.
Dean
and
P. G.
Ryan
), pp.
766
-
767
.
Cape Town
:
The Trustees of the John Voelcker Bird Book Fund
.
Marder
,
J.
(
1973
).
Body temperature regulation in the brown-necked raven (Corvus corax ruficollis)—I. Metabolic rate, evaporative water loss and body temperature of the raven exposed to heat stress
.
Comp. Biochem. Physiol. A
45
,
421
-
430
.
McKechnie
,
A. E.
,
Whitfield
,
M. C.
,
Smit
,
B.
,
Gerson
,
A. R.
,
Smith
,
E. K.
,
Talbot
,
W. A.
,
McWhorter
,
T. J.
and
Wolf
,
B. O.
(
2016
).
Avian thermoregulation in the heat: evaporative cooling capacity in an archetypal desert specialist, Burchell's sandgrouse (Pterocles burchelli)
.
J. Exp. Biol.
219
,
2409
.
McKechnie
,
A. E.
,
Gerson
,
A. R.
and
Wolf
,
B. O.
(
2021
).
Thermoregulation in desert birds: scaling and phylogenetic variation in heat tolerance and evaporative cooling
.
J. Exp. Biol.
224
,
jeb229211
.
McKechnie
,
A. E.
,
Short
,
J. C.
and
Freeman
,
M. T.
(
2022
).
Respirometry protocols for avian thermoregulation at high air temperatures: stepped and steady-state profiles yield similar results
,
Dryad, Dataset
.
Muggeo
,
V. M. R.
(
2008
).
Segmented: an R package to fit regression models with broken-line relationships
.
R News
8
,
20
-
25
.
O'Connor
,
R. S.
,
Wolf
,
B. O.
,
Brigham
,
R. M.
and
McKechnie
,
A. E.
(
2017
).
Avian thermoregulation in the heat: efficient evaporative cooling in two southern African nightjars
.
J. Comp. Physiol. B
187
,
477
-
491
.
Page
,
A. J.
,
Cooper
,
C. E.
and
Withers
,
P. C.
(
2011
).
Effects of experiment start time and duration on measurement of standard physiological variables
.
J. Comp. Physiol. B
181
,
657
-
665
.
Pattinson
,
N. B.
,
Thompson
,
M. L.
,
Griego
,
M.
,
Russell
,
G.
,
Mitchell
,
N. J.
,
Martin
,
R. O.
,
Wolf
,
B. O.
,
Smit
,
B.
,
Cunningham
,
S. J.
,
McKechnie
,
A. E.
et al. 
(
2020
).
Heat dissipation behaviour of birds in seasonally hot arid-zones: are there global patterns?
J. Avian Biol.
51
,
e02350
.
Ricklefs
,
R. E.
and
Hainsworth
,
F. R.
(
1968
).
Temperature dependent behavior of the cactus wren
.
Ecology
49
,
227
-
233
.
Robinson
,
D. E.
,
Campbell
,
G. S.
and
King
,
J. R.
(
1976
).
An evaluation of heat exchange in small birds
.
J. Comp. Physiol. B
105
,
153
-
166
.
Smit
,
B.
,
Whitfield
,
M. C.
,
Talbot
,
W. A.
,
Gerson
,
A. R.
,
McKechnie
,
A. E.
and
Wolf
,
B. O.
(
2018
).
Avian thermoregulation in the heat: phylogenetic variation among avian orders in evaporative cooling capacity and heat tolerance
.
J. Exp. Biol.
221
,
jeb174870
.
Smith
,
E. K.
,
O'Neill
,
J.
,
Gerson
,
A. R.
and
Wolf
,
B. O.
(
2015
).
Avian thermoregulation in the heat: resting metabolism, evaporative cooling and heat tolerance in Sonoran Desert doves and quail
.
J. Exp. Biol.
218
,
3636
-
3646
.
Swanson
,
D. L.
,
Drymalski
,
M. W.
and
Brown
,
J. R.
(
1996
).
Sliding vs static cold exposure and the measurement of summit metabolism in birds
.
J. Therm. Biol.
21
,
221
-
226
.
Tieleman
,
B. I.
,
Williams
,
J. B.
and
Buschur
,
M. E.
(
2002
).
Physiological adjustments to arid and mesic environments in larks (Alaudidae)
.
Physiol. Biochem. Zool.
75
,
305
-
313
.
Tracy
,
C. R.
,
Welch
,
W. R.
,
Pinshow
,
B.
and
Porter
,
W. P.
(
2010
).
Properties of Air: A Manual for Use in Biophysical Ecology
, 4th edn.
The University of Wisconsin Laboratory for Biophysical Ecology: Technical Report
.
van Jaarsveld
,
B.
,
Bennett
,
N. C.
,
Kemp
,
R.
,
Czenze
,
Z. J.
and
McKechnie
,
A. E.
(
2021
).
Heat tolerance in desert rodents is correlated with microclimate at inter- and intraspecific levels
.
J. Comp. Physiol. B
191
,
575
-
588
.
Walsberg
,
G. E.
and
Wolf
,
B. O.
(
1995
).
Variation in the respirometry quotient of birds and implications for indirect calorimetry using measurements of carbon dioxide production
.
J. Exp. Biol.
198
,
213
-
219
.
Weathers
,
W. W.
(
1981
).
Physiological thermoregulation in heat-stressed birds: consequences of body size
.
Physiol. Zool.
54
,
345
-
361
.
Weathers
,
W. W.
(
1997
).
Energetics and thermoregulation by small passerines of the humid, lowland tropics
.
Auk
114
,
341
-
353
.
Whitfield
,
M. C.
,
Smit
,
B.
,
McKechnie
,
A. E.
and
Wolf
,
B. O.
(
2015
).
Avian thermoregulation in the heat: scaling of heat tolerance and evaporative cooling capacity in three southern African arid-zone passerines
.
J. Exp. Biol.
218
,
1705
-
1714
.
Wolf
,
B. O.
and
Walsberg
,
G. E.
(
1996a
).
Respiratory and cutaneous evaporative water loss at high environmental temperatures in a small bird
.
J. Exp. Biol.
199
,
451
-
457
.
Wolf
,
B. O.
and
Walsberg
,
G. E.
(
1996b
).
Thermal effects of radiation and wind on a small bird and implications for microsite selection
.
Ecology
77
,
2228
-
2236
.

Competing interests

The authors declare no competing or financial interests.