RMR

Measurements of RMR provide insight into the minimum heat load animals experience as a function of T_a and how internally produced heat contributes to the total heat load during periods of heat stress. Our measurements of RMR are broadly similar to those reported previously, with differences that likely reflect variation in measurement conditions. Our measurements on poorwills and nighthawks were conducted on the night of capture during the active phase of their diel cycle. In our wild-caught poorwills, the lowest RMR was ∼65% greater than RMRs found in previous studies using captive poorwills, acclimated to laboratory conditions (Bartholomew et al., 1962; Lasiewski, 1969), but was comparable to values observed in wild-caught South African nightjars (O'Connor et al., 2017). The lowest mass-specific RMR observed in our wild-caught lesser nighthawks was 25% higher than that of the larger common nighthawk (Chordeiles minor) acclimated to laboratory conditions (Lasiewski and Dawson, 1964). In our study, the lowest RMR observed in the owlet-nightjar averaged 0.40±0.07 W compared with 0.32 W measured by Doucette and Geiser (2008). In the latter study, birds were captured before dawn and studied during the rest phase of their diel cycle.

As expected, our RMR values are considerably lower than those measured for passerines; RMRs in passerines of similar mass were approximately 1.5–2.0 times higher than those we observed in the nightjars. For example, two 40 g desert passerines, the white-browed sparrow-weaver of South Africa (Whitfield et al., 2015) and the chestnut-crowned babbler of Australia (McKechnie et al., 2017) showed average minimal RMRs of 0.54 and 0.59 W. In contrast, desert-nesting Columbiformes showed very similar RMRs to those of the Caprimulgiformes, with the 37 g Namaqua dove (Oena capensis) exhibiting an average RMR of 0.31 W (McKechnie et al., 2016b). The RMR values obtained in this study correspond well to the low metabolic rates noted in caprimulgiform birds in general relative to other taxa (Lane et al., 2004), a trait that importantly reduces the contribution of metabolic heat to the total heat load that the bird must dissipate via evaporation at high T_a.

Passerine birds have higher resting metabolic rates than caprimulgids. When T_a increases and active heat dissipation (panting) begins, there is a more rapid increase in RMR in passerines than in caprimulgids. Increases in RMR with increasing heat stress add to the total heat load that must be dissipated through evaporation. The high metabolic costs of respiratory evaporative heat dissipation via panting contrast sharply with the negligible energetic costs of gular flutter in the nightjars. Cowles and Dawson (1951) commented on the highly effective nature of evaporative cooling via the gular apparatus in the Texas nighthawk and urged further study. Bartholomew et al. (1962) noted that the large highly vascularized mouth and gular area in the common poorwill and other nightjars provides a large, effective surface for evaporation, and vigorous gular flutter increases evaporation with little observed increase in metabolism. This highly vascularized mouth of nightjars was clearly illustrated in Audubon's (1830) print of a whip-poor-will pursuing a moth. Bartholomew et al. (1962) found no increase in RMR at T_a as high as 44°C, concluding that an upper critical temperature (T_uc) must lie above 44°C. Work by Lasiewski and Dawson (1964) on common nighthawks also failed to identify a clearly discernible T_uc and found only gradual increases in RMR over T_a ranging from 35 to 45°C. Doucette and Geiser (2008), in contrast, estimated the T_uc for owlet nightjars at 34.8°C during the summer, but based this value on few data points. Most recently, O'Connor et al. (2017) found low stable RMRs across a wide range of high temperatures with no apparent T_uc in rufous-cheeked and freckled nightjars in South Africa. In the present study, we found gradual increases in RMR from the minimal observed values to a point ranging from 40.3 to 56.2°C (identified by segmented regressions) at which there is an abrupt increase in the slope of the RMR as a function of T_a. We do not define a thermoneutral zone in our birds because the criteria suggested by Scholander et al. (1950) are not met. Instead of a range of T_a at which RMR remains stable, our birds showed decreasing RMR from T_a=30°C to a minimal value, then a subtle increase in RMR until T_a neared the HTL. It is likely that birds in the wild, which are not experiencing additional stress from handling or the experimental conditions, might more closely reflect the findings of O'Connor et al. (2017) and not show significant metabolic responses to heat stress until reaching the limits of heat tolerance. Other researchers have noted the low metabolic costs of gular flutter and associated it with the small mass of the gular area and the apparent resonant frequency at which the gular area moves, irrespective of temperature (Lasiewski and Bartholomew, 1966; Bartholomew et al., 1968). They found that common poorwills did not change the frequency of gular flutter, but birds increased the amplitude and the area of the gular apparatus engaged with increasing heat stress.

To gain further insight into the effectiveness of the gular flutter process in our species, we quantified the costs (increases in RMR) versus the benefits (increases in EHL) where gular flutter occurred intermittently and other variables were stable but before increasing heat stress produced continuous gular flutter. Over a series of 5–10 bouts of gular flutter, we measured changes in metabolic rate and rates of evaporation immediately before gular flutter commenced, and then during the gular flutter bout that followed. Strikingly, common poorwills showed a 123% increase in EWL when gular flutter commenced, with no concomitant increase in RMR (Fig. 5). Lesser nighthawks showed a somewhat less effective pattern, with a 50% increase in EWL and a 5% increase in RMR. The cavity-nesting owlet-nightjar's responses were intermediate to those of the other species and showed a 57% increase in EHL and a 6% decrease in RMR. From a practical standpoint, the very low metabolic cost of the gular flutter apparatus (see discussion below) allows for increasing rates of EWL as heat stress increases, without additional metabolic heat contributing to the total heat load, thus saving water and allowing for high heat tolerance. These data indicate that the effectiveness of the gular flutter system used for evaporative heat dissipation by nightjars clearly rivals that found in Columbiformes (Smith et al., 2015; McKechnie et al., 2016b), where cutaneous evaporation is paramount, and is far more effective then respiratory evaporation driven by panting observed in most small passerine birds (Whitfield et al., 2015; McKechnie et al., 2017). The differences in evaporative mechanisms among orders clearly affect individual HTLs and the trajectory of T_b with increasing heat stress, as well as overall water balance during chronic heat stress.

EWL

As T_a increases, all birds become increasingly reliant on EWL for heat dissipation as the gradient between T_a and T_b decreases. In nightjars, as T_a increased, birds transitioned from modest increases in EWL from cutaneous and respiratory surfaces (Lasiewski et al., 1971) to activation of the gular flutter apparatus, which greatly increased rates of EWL. In our species, minimum values for EWL ranged from 0.20 to 0.21 g h⁻¹ at T_a=30°C, the lowest test T_a in our study. These rates are ∼50% higher than the minimal EWL values measured by Bartholomew et al. (1962) in captive poorwills, but are very similar to EWL of wild-caught South African nightjars studied by O'Connor et al. (2017). As T_a increased, EWL increased gradually to an inflection point, varying among our species over T_a≈43–48°C, at which EWL increased markedly and reached ∼12–16 times the minimal values at the birds' thermal limits (HTL). Above this inflection point, EWL increased at 0.11–0.21 g⁻¹ h⁻¹ °C⁻¹. O'Connor et al. (2017) reported a relatively low and stable EWL in South African nightjars that increased rapidly starting at T_a≈38–39°C at rates of 0.093–0.099 g⁻¹ h⁻¹ °C⁻¹. Other taxa studied at the same sites and times and by the same techniques show similar values but differences in the inflection point at which there was an abrupt uptick in the rate of water loss. McKechnie et al. (2017) in a study of Australian desert passerines found that the spiny-cheeked honeyeater (∼42 g) and the chestnut-crowned babbler (∼52 g) increased EWL at rates of 0.18 and 0.23 g⁻¹ h⁻¹ °C⁻¹ above inflection points at 38.1 and 42.2°C. The slope of EWL for the South African Namaqua dove (∼37 g) was 0.096 g⁻¹ h⁻¹ °C⁻¹ (McKechnie et al., 2016b). Overall, rates of EWL did not differ greatly among these arid-zone taxa; however, inflection points marking a significant increase in EWL vary.

Efficiency of evaporative processes: EHL/MHP

We can use the ratio of EHL/MHP to examine the relative efficiency of different evaporative pathways as individuals become reliant on evaporative pathways for heat dissipation when T_a approaches and exceeds T_b. Pathways of EWL vary greatly among different taxa and the relative efficiency of these pathways varies within and among groups. In general, EWL occurs by means of the respiratory surfaces or skin and the relative importance of these pathways in heat-stressed birds differs greatly (McKechnie et al., 2016a,b). In passerine birds, the dominant pathway for evaporative heat dissipation is through respiratory panting, where individuals ventilate the full respiratory surfaces to varying degrees, which results in significant increases in RMR with increasing heat stress and an increased demand for EHL (Whittow, 1976; Wolf and Walsberg, 1996).

EHL/MHP values in the thermoneutral zone are typically <1 and provide insight into the relative EHL compared with total heat loss as denoted by metabolic rate. When T_a>T_b, the maintenance of a stable T_b demands that EHL/MHP>1 and in birds that tolerate very high temperatures (many Columbiformes), these values can exceed 4.0 (Smith et al., 2015; McKechnie et al., 2016a,b). Columbiform birds studied to date rely primarily on cutaneous EWL during periods of heat stress, which appears to have little or no metabolic cost associated with evaporation from skin surfaces (McKechnie and Wolf, 2004). A number of researchers have commented on the efficiency of the gular flutter apparatus for evaporative cooling in caprimulgiform birds (Lasiewski and Seymour, 1972; Grant, 1982; Cowles and Dawson, 1951). The highest EHL/MHP (5.15) recorded for any bird was noted in the rufous-cheeked nightjar, which employs the gular flutter mechanism during heat exposure (O'Connor et al., 2017). The highly vascular and capacious oral cavity is a constant feature of nightjar anatomy and in the lesser nighthawk, for example, this area is equivalent to 15% of the total body surface area (Cowles and Dawson, 1951). Cowles and Dawson (1951) showed the efficacy of gular flutter by observing increases in T_b in Texas nighthawks when the bill was held closed, thus preventing gular flutter. We found that at T_a≈40°C, the onset of gular flutter resulted in increases in EHL/MHP ratios of 44–129% (Fig. 5). We speculate that increases in the depth of gular flutter and the surface area employed with increasing heat stress probably allow our species to increase EHL/MHP to maximal values comparable to those found in the South African nightjars (O'Connor et al., 2017). As a consequence, very low RMRs combined with high rates of EWL provide a highly effective system for heat dissipation, which we believe is responsible for the very high heat tolerance limits in our nightjars.

Hyperthermic responses and HTLs

When T_a approaches T_b, birds typically respond by increasing EWL and T_b. Hyperthermic responses to T_a>T_b can serve to minimize the heat gain from the environment and thus reduce the EWL necessary for thermoregulation. Early research by Lasiewski and Seymour (1972) showed that common poorwills exhibit modest hyperthermic responses to increasing heat stress, with T_b increasing by 0.13°C per 1°C change in T_a to 41.6°C at a T_a of 45°C. O'Connor et al. (2017) found increases in T_b of 0.12 and 0.14°C per 1°C change in T_a in rufous-cheeked and freckled nightjars, with T_b reaching 41.5 and 40.1°C, respectively, at T_a=52°C. The rate of T_b increase in our North American nightjars is similar to that found in the African nightjars studied by O'Connor et al. (2017), with increases of 0.09 and 0.11°C per 1°C change in T_a for the poorwill and nighthawk, respectively, and 0.22°C per 1°C change in T_a in the owlet-nightjar. In contrast to the nightjars, heat-stressed passerines showed increases in T_b of 0.23–0.58°C per 1°C change in T_a (Whitfield et al., 2015; McKechnie et al., 2017), while T_b of the similar-sized Namaqua dove (∼40 g) increased by 0.11°C per 1°C change in T_a (McKechnie et al., 2016a,b). The observed hyperthermic responses led to maximum T_b in individual birds of 44.0, 44.1 and 44.4°C for poorwills (T_a=62°C), nighthawks (T_a=60°C) and owlet nightjars (T_a=52°C), respectively. Even at these high T_b, poorwills and nighthawks were able to maintain a T_a–T_b gradient of 17.3 and 15.1°C, respectively, whereas the tree cavity-roosting owlet-nightjar maintained a maximal T_a–T_b gradient of 7.7°C. The maximal gradients observed in comparably sized passerines are in the 2.4–5.4°C range (Whitfield et al., 2015; Smith et al., 2017; McKechnie et al., 2017). The performance of nightjars in the heat is currently only rivaled by doves and pigeons; Smith et al. (2015), for example, found that mourning doves and white-winged doves can maintain a maximal T_a–T_b gradient of 14.1 and 15.3°C, respectively. These differences in performance among avian orders translate directly into overall heat tolerance.

During high-temperature events, birds may be severely challenged to evaporate sufficient water to maintain a stable T_b and thus be subject to overheating and death. We wanted to establish baseline performance values for heat tolerance in the Caprimulgiformes, as earlier researchers (Lasiewski, 1969; Grant, 1982) have suggested that this group shows exceptional heat tolerance compared with most other birds. We thus estimated the HTL of nightjars to acute exposure to high T_a. We defined the HTL as the T_a at which birds showed rapidly increasing T_b approaching or reaching a sub-lethal endpoint of T_b=45°C. We chose T_b=45°C because the limited data on lethal temperatures suggest that the avian lethal range is from 45.7 to 47.8°C (Arad and Marder, 1982; Brush, 1965; Randall, 1943). For the ground-nesting/roosting common poorwill and the lesser nighthawk, we observed HTLs of T_a≈62–64°C and 60–62°C, respectively. These values are comparable to the performance observed (58–64°C) in a variety of Columbiformes (McKechnie et al., 2016a,b; Smith et al., 2015). For the Australian owlet-nightjar, which nests/roosts in tree cavities and experiences less extreme heat exposure, the observed HTL was T_a≈52–54°C. Thus, we found that the HTLs of the owlet-nightjar were closer to the HTLs of passerines (HTL T_a≈48–52°C) tested under similar conditions (Whitfield et al., 2015; McKechnie et al., 2017; Smith et al., 2017).

Conclusions

This study builds on earlier seminal studies on nightjars, which recorded their remarkable ability to tolerate heat (Bartholomew et al., 1962; Lasiewski and Dawson, 1964). We worked on heat-acclimated, recently captured birds and found that the low resting metabolic rate observed in all species reduces the contribution of metabolic heat to the overall heat load. The use of gular flutter increases evaporative cooling with little metabolic cost compared with the large increases in metabolic heat production associated with panting used by passerines, thus increasing heat tolerance limits and conserving water. This mechanism rivals the performance found in doves and pigeons that primarily rely on cutaneous evaporation for cooling, which also has negligible metabolic costs. Although we present the data for North American nightjars with the Australian owlet nightjar together, their relative performances are in part a function of the differing measurement conditions (active versus rest phase measurements) among species, and differences in their ecologies (ground dwelling versus cavity roosting). All species, however, demonstrate HTLs and capacities for thermoregulation that greatly exceed those of sympatric passerines, potentially making them more resilient to the rapid warming that is occurring in their current distributions.