Cold- and exercise-induced peak metabolic rates in tropical birds
Edited by Ewald R. Weibel, University of Bern, Bern, Switzerland, and approved October 30, 2007
Abstract
Compared with temperate birds, tropical birds have low reproductive rates, slow development as nestlings, and long lifespans. These “slow” life history traits are thought to be associated with reduced energy expenditure, or a slow “pace of life.” To test predictions from this hypothesis, we measured exercise-induced peak metabolic rates (PMRE) in 45 species of tropical lowland forest birds and compared these data with PMRE for three temperate species. We also compared cold-induced PMR (PMRC) with PMRE in the same individuals of 19 tropical species. Tropical birds had a 39% lower PMRE than did the temperate species. In tropical birds, PMRC and PMRE scaled similarly with body mass (Mb), but PMRE was 47% higher than PMRC. PMRE averaged 6.44 × basal metabolic rate (BMR) and PMRC averaged 4.52 × BMR. The slope of the equation relating PMRE to Mb exceeded the slope for the equation for BMR vs. Mb, whereas slopes for the equations of PMRC and BMR vs. Mb did not differ. Mb-adjusted residuals of PMRE were positively correlated with residual BMR, whereas residual PMRC and residual BMR were not correlated. PMRE and PMRC were not correlated after we corrected for Mb. Temperate birds maintained their body temperature at an 8.6°C lower average air temperature than did tropical species. The lower PMRE values in tropical species suggest that their suite of life history traits on the slow end of the life history continuum are associated with reduced metabolic rates.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
Ecological physiologists have devoted considerable effort to examining the upper limits of physiological performance, especially locomotor performance and metabolic power production during cold exposure. An implicit or explicit hypothesis in this work is that individual variation in performance is correlated with fitness: higher performance such as faster running or greater metabolic power output leads to increased survival and/or reproductive success. In the past few decades, a variety of performance indices have been measured in numerous species, and analyses have examined performance in mechanistic, phylogenetic, ecological, and evolutionary contexts.
In birds, sustained high aerobic metabolism is a salient feature of their physiological performance and is fundamental to their capacity for powered flight and their tolerance of extreme cold (1). The most-studied aerobic index in birds is basal metabolic rate (BMR), the lower limit to aerobic power production (see refs. 2–4). However, it is questionable whether BMR per se, rather than other physiological parameters that may correlate with BMR, is ever a “target” for direct selection except when the requirement for reducing heat load or water loss is critical to survival, as might be the case in hot deserts (5). Moreover, BMR is not directly responsible for the remarkable flight capacity or temperature tolerance of birds. Instead, it is the upper limit to aerobic power output (peak metabolic rate; PMR) that forms the metabolic foundation of the high-intensity avian way of life. Intuitively, PMR can be linked to fitness in numerous contexts, such as survival during migration, predator avoidance, or survival during extreme cold.
A complication in analyses of maximum aerobic limits is that several approaches have been used to elicit PMR. Some investigators have used maximal metabolism during exercise (PMRE) and others maximum rates of metabolism during cold-induced thermogenesis (PMRC; sometimes called “summit metabolism”). Typically, PMRC is measured during brief exposure to low ambient temperatures (Ta) (e.g., ref. 6), often in a helium–oxygen (heliox) atmosphere to increase conductance (7). Although measuring PMRE is a technical challenge, it has been quantified in birds trained to fly in wind tunnels (e.g., ref. 8) or run on treadmills (e.g., refs. 9 and 10) or by using related methodologies such as “flight wheels” (11–13). Most data on avian PMR have been obtained during cold challenge, even though exercise often elicits higher rates of energy expenditure (e.g., ref. 14). The fact that these two measurements differ presents a problem in interpretation. To understand the relationship between PMRE and PMRC, it would be of value to measure both parameters on the same individuals, but such data are currently unavailable. Another interpretive bias concerning PMR stems from the fact that most species studied are native to temperate climates. No data exist on PMR for tropical birds. Although the tropics comprise many different habitats, we use the term “tropical” here to refer specifically to tropical lowland forest.
Tropical and temperate birds lie at opposite ends of a life history continuum, with tropical species falling at the “slow” end; they show lower mortality, longer lifespan, and reduced reproductive effort (15–19). A reasonable deduction from these differences in climate and life history is that selection on the upper limits of physiological performance—especially aerobic metabolism—may be relaxed in tropical species. Because thermal conditions in their habitats are warm and stable, tropical birds do not experience selection for high PMRC. This is consistent with measurements of lower PMRC in tropical birds than in temperate birds (20). Predictions for differences between tropical and temperate birds in PMRE are less intuitive. Aerodynamic power requirements for flight are unaffected by latitude, so on that basis there is no reason to expect divergence of PMRE between temperate and tropical species. Nevertheless, the lack of long-distance migration in lowland tropical species, potentially shorter flight durations in dense forests, and lower brood-care requirements in most tropical birds make it plausible to expect reduced emphasis on high-endurance sustained flight and hence lower PMRE. Also, some data (e.g., refs. 21 and 22) suggest that PMR and BMR are functionally coupled and, therefore, should show positive correlations within and among species. On that basis, we predicted that tropical birds would have a low PMRE because we have previously shown that tropical birds have a reduced BMR (20).
In this study, we measured PMRE in tropical birds and compared it with BMR and PMRC. First, we tested whether PMRE of tropical species would be higher than PMRC, which seems to be a pattern among temperate endotherms (14). Second, we tested whether PMRE was lower in tropical birds than in temperate species, as would be predicted by a functional link with their low BMR (20) or whether PMRE was unaffected by climate, as predicted from aerodynamic considerations. Third, we searched for a positive correlation between BMR and PMR. Correlations between BMR and PMR have been found in some studies (e.g., refs. 21 and 22) but not others (e.g., refs. 23–25). Finally, we tested whether temperate birds have better tolerance for low temperatures than do tropical species.
Results
PMRs.
We measured PMRC for 77 individuals of 19 species [see supporting information (SI) Table 3]. Among species, PMRC scaled to Mb0.773 (Fig. 1 and Table 1). The relationship between PMRC and Mb remained significant when phylogenetic independent contrasts (PICs) (PMRCPIC; Table 1) were used. On the basis of BMR measurements of the same species, cold-induced factorial aerobic scope (fASC) equaled 4.52 ± 0.32 × BMR (Table 1). The allometric slope of logPMRC did not differ significantly from the slope of logBMR, which equaled 0.644 ± 0.034 for tropical birds (20) (F1,49 = 2.9, P = 0.09).
Fig. 1.
Table 1.
| Variable | Intercept | logMb | n |
|---|---|---|---|
| logPMRE | −0.720 (0.058)*** | 0.757 (0.045)*** | 45 |
| logPMREPIC | — | 0.734 (0.050)*** | 44 |
| logPMRC | −0.899 (0.149)*** | 0.773 (0.109)*** | 19 |
| logPMRCPIC | — | 0.615 (0.095)*** | 18 |
| logTCL | 2.495 (0.009)*** | −0.0323 (0.0063)*** | 19 |
| logTCLPIC | — | −0.0226 (0.0060)** | 18 |
| logfASE | 0.599 (0.076)*** | 0.148 (0.055)* | 31 |
| logfASC | 0.482 (0.166)** | 0.115 (0.121) | 19 |
| log(PMRE/PMRC) | 0.148 (0.146) | 0.006 (0.104) | 16 |
Standard errors are given in parentheses. Note that PIC regressions were forced through the intercept.
*, P < 0.05;
**, P < 0.01;
***, P < 0.001.
PMRE was measured in 139 individuals of 45 species (see SI Table 3). For 53 of these individuals, from 14 species, we also measured PMRC. For two additional species, we had both measurements but from different individuals. PMRE scaled to Mb0.757 (Fig. 1 and Table 1). The exponent for the relationship between PMRE and Mb was larger than that for BMR and Mb (F1,75 = 5.85, P = 0.018). The association between PMRE and Mb remained significant when we used PICs (PMREPIC; Table 1). For 31 tropical species for which we measured both BMR and PMRE, fASE averaged 6.44 ± 0.22 × BMR. fASE scaled positively with Mb0.148 (Table 1). The slope for logfASE did not differ from the slope for logfASC (F1,46 = 0.08, P = 0.8).
In tropical birds, PMRE was higher than PMRC (paired t test: t15 = 6.2, P < 0.001; Fig. 2). The average ratio of species means of PMRE divided by PMRC was 1.47 ± 0.09 (n = 16, range 0.85–2.61), and there was no association with Mb (Table 1). Slopes of the allometric relationships of PMRE and PMRC did not differ (F1,60 = 0.01, P = 0.9).
Fig. 2.
Residual logPMRE and residual logBMR were significantly positively correlated, which supports the view of a functional link between PMR and BMR (r = 0.39; Table 2). However, residual logPMRC and residual logBMR were not correlated. On the basis of averages for 16 species, residuals of logPMRE and logPMRC were not correlated (Table 2). Use of PICs showed the same results (Table 2).
Table 2.
| Residuals of | Residual logBMR | Residual logPMRC | Residual logBMRPIC | Residual logPMRCPIC |
|---|---|---|---|---|
| logPMRE | 0.387* (31) | 0.119 (16) | — | — |
| logPMRC | 0.132 (19) | — | — | — |
| logPMREPIC | — | — | 0.406* (30) | 0.00 (15) |
| logPMRCPIC | — | — | −0.174 (18) | — |
Standard errors are given in parentheses. Variables marked PIC were transformed to phylogenetic independent contrasts.
*, P < 0.05.
Temperate vs. Tropical Comparisons.
The three temperate species had 39 ± 12% higher PMRE than tropical species (F1,45 = 10.6, P < 0.01; Fig. 1). Tested separately (and omitting the satin bowerbird, Ptilonorhynchus violaceus, which had an Mb exceeding that of all tropical species), the PMRE of 23.0-g temperate house sparrows, Passer domesticus, was 3.35 W, a value 95% above predicted PMRE for tropical species (t79 = 11.2, P < 0.001). PMRE of 20.3-g temperate red-eyed vireos, Vireo flavoviridis, was 2.74 W, 62% above estimated PMRE for tropical species (t53 = 5.08, P < 0.001). For house sparrows and red-eyed vireos, fASE was 10.58 ± 0.35 and 10.42 ± 0.42, respectively. These values differ significantly from fASE of 6.44 ± 0.15 in tropical species (64% higher, t65 = 10.3, P < 0.001 and 62% higher, t39 = 11.3, P < 0.001, respectively). The temperate satin bowerbird had an fASE of 11.2, assuming a BMR of 1.26 W (3). This factorial scope is also considerably higher than our estimate for tropical species.
Lower Limit of Cold Tolerance.
We compared the temperature at cold limit (TCL) values of tropical birds with those of 21 summer-acclimatized temperate species (6), using general linear models (GLMs) on log10-transformed data with Mb and Climate as independent variables. Temperate birds could maintain their PMR at a lower TCL than tropical birds. In tropical species, TCL was 8.6°C higher than in temperate species (F1,37 = 53.1, P < 0.001). We found no interaction between Climate and Mb (F1,36 = 0.87, P = 0.36).
Discussion
We have presented measurements of PMRC and PMRE in birds from tropical lowland forests and compared these two variables in the same species. In combination with BMR data for the same species (20), our results enable us to examine several aspects of avian aerobic performance with relatively few confounding factors related to methodology, phylogeny, or seasonality.
One of our goals was to compare peak aerobic metabolism during exercise with that elicited by intense cold exposure in tropical birds. We found a 47% higher PMRE than PMRC in tropical birds (Fig. 2). In warm-acclimated small mammals, PMRE and PMRC are often similar (24, 26), but in many species, cold acclimation induces hypertrophy of brown adipose tissue, which augments shivering thermogenesis (27). As a result, PMRC is often considerably greater than PMRE in small mammals after seasonal or laboratory acclimation to low temperature (e.g., refs. 26 and 28). In large mammals, PMRE seems to consistently exceed PMRC (14, 29), but that may be a reflection of the difficulty in eliciting PMRC in large, well insulated endotherms.
Birds lack tissues specialized for heat production (30, 31). Therefore, regulated power production in both exercise and thermogenesis relies on skeletal muscle (32). Most studies of avian PMR have focused on PMRC in small species because that measurement is technically easier than eliciting maximal exercise metabolism and because of the challenge of attaining PMRC in large birds. In tropical species, forced exercise in our flight wheel elicited a higher PMR than cold exposure. There are no data for temperate birds where both PMRE and PMRC were measured in the same species. However, an assortment of interspecific measurements of cold-induced and exercise-induced PMR in temperate species shows the same trend: PMR is higher in exercise than in thermogenesis (14). On average, PMRC in temperate birds is 5–6 × BMR (1, 22), occasionally reaching 8 × BMR [e.g., in winter-acclimated black-capped chickadees, Poecile atricapillus (6)]. In contrast, PMRE obtained in flight wheels or treadmills is ≈7–12 × BMR (9–13).
Differences between PMRE and PMRC may be due to the way the flight musculature is used. In flapping flight, the pectoralis–supracoracoideus complex can operate at maximum intensity. In contrast, effective shivering requires isotonic, simultaneous antagonistic contractions, which limits force generation by the large downstroke muscles and thus constrains total power output (32). Among tropical species, the mangrove swallow, Tachycineta albilinea, had the highest PMRC. Because this species is an active aerial forager, this raises the question of whether rate of heat production can be a byproduct of adaptations to an active aerial lifestyle (see ref. 33).
There is an apparent difference between the PMRE of birds measured during long-duration steady flight vs. forced exercise in flight wheels or treadmills. We found a lower PMRE in flight wheel tests than the average value of 16 × BMR for birds in long-duration steady flights (e.g., refs. 8, 14, 34, and 35). We tentatively conclude that PMRE measured in a flight wheel is lower than PMR measured in flight.
Another goal of our study was to compare aerobic limits in temperate and tropical birds, and we found considerably lower PMRE in tropical species than in the small number of temperate birds tested with similar methods. The difference is puzzling, but analyses of performance traits associated with contrasting environmental regimes might provide useful insight into proximate and ultimate causes of physiological diversity (36). Several aspects of climate and life history suggest that temperate and tropical bird species might show metabolic divergence. First, tropical lowland forest environments are considerably warmer and have less daily and seasonal Ta variation than is typical of temperate habitats; in particular, tropical lowlands lack the prolonged periods of cold characteristic of temperate-zone winters. Therefore, unlike temperate residents, tropical birds are presumably not under strong selection for high thermogenic capacity and cold tolerance. This hypothesis is supported by data on 57 rodent species that show a negative correlation between PMRC and mean minimum annual Ta (37). Hence, one might predict a lower PMRC in tropical birds. Second, compared with temperate birds, tropical lowland birds generally lie at the slow end of the life history continuum (15, 16). In combination with low thermoregulatory costs, these life history differences might be expected to result in a lower average daily rate of energy expenditure (a syndrome of characteristics sometimes referred to as the “slow pace of life”) in tropical species (20). Both theory (38) and, for several taxa, data (39, 40) suggest negative correlations between the pace of life and both reproductive rate and survival.
Although it is intuitive to predict lower thermoregulatory capacity in tropical species, it is less clear how differences in life history and pace of life might affect the limits to aerobic exercise capacity. On the one hand, the highest power output in volant species presumably occurs during flapping flight. Other factors (altitude, temperature, etc.) being equal, air density and viscosity are quite similar in tropical and temperate regions. Accordingly, the aerodynamic forces required to support flapping flight are largely independent of latitude. Therefore, the metabolic power requirements for flight should be the same at all latitudes, so if the energy demands of flight drive the upper limit to aerobic power output in exercise, it is not evident that tropical and temperate birds would show divergence in PMRE.
Alternatively, aspects of life history and environment in tropical birds are consistent with relaxed selection on the upper limits of sustainable exercise. Warm temperatures reduce the energy costs of thermogenesis and hence the need for foraging to support those costs. Smaller brood sizes and slower offspring growth could ameliorate the need for extensive foraging to provision offspring; for example, tropical house wrens, Troglodytes aedon, lay fewer eggs, make less frequent feeding trips to the nest, and have lower daily energy expenditures than temperate house wrens (41). Additionally, the reduction or absence of long-distance movements and migration in many tropical species could reduce selection favoring high flight endurance. For both of these reasons, tropical species might be expected to have relatively low PMRE.
If PMR and basal metabolism are functionally linked (42), low PMRE and PMRC in tropical birds should also result in a reduced BMR. This prediction is supported by the 18% lower BMR found for tropical birds when compared with their close relatives from temperate habitats (20).
Our finding that PMRE is lower in tropical bird than in temperate birds (Fig. 1) supports the hypothesis that life history traits and a thermal environment are reflected in aerobic performance limits. However, our data cannot reveal which of several possible causal factors—high ambient temperatures, low rates of energy use, or life history traits—is responsible for the reduction in aerobic performance of tropical species.
What explains the ability of tropical birds to fly effectively if their maximal aerobic power output is substantially less than that of similar-sized temperate birds? Because of the densely forested habitats used by most of the Panamanian birds in our study, typical flight distances and durations are quite short. Therefore, most flights for these species are “burst” activities, and some of the necessary power production may come from anaerobic pathways, whereas our methods measured only aerobic power output over periods of a minute or more. Obviously, this argument is not applicable to some of the species that routinely fly for long periods, such as mangrove swallows and gray-breasted martins, Progne chalybea.
Because both PMR and BMR are reduced in tropical birds compared with temperate species, the fAS (PMR/BMR) of tropical birds is not dramatically different from that of temperate birds. fASE was positively correlated with Mb and, based on the allometric relationship (Table 1), averaged 5.3–7.6 g over the 7- to 79-g Mb range in our dataset. These fASE are at the low end of the range of ≈6.5–30 × BMR measured in other birds by using a variety of methods (11–14, 43, 44). The thermogenic aerobic scope of tropical birds, fASC, was not related to Mb, which may be an effect of the small range in Mb of 9.6–71, and averaged 4.5 × BMR or 15–34% lower than fASE (Table 1). fASC was 22% lower in tropical birds than in temperate birds (20): a significant difference despite extensive overlap with values reported for temperate bird species that fall within the range of 3–9 × BMR (6, 23, 45).
It has been suggested (e.g., refs. 21 and 46) that PMR and BMR are functionally coupled. The mechanistic basis of the linkage is usually postulated to derive from indirect effects on the “central” organs or organ systems (cardiopulmonary, digestive, etc.) that support the peripheral musculature responsible for PMR by providing fuel, oxygen, and waste removal. Selection favoring high PMR presumably requires correspondingly high capacity, and hence high metabolic rates, in the central support organs (47). Because BMR is often thought to be largely involved in the “maintenance” metabolic activity of central organs (48, 49), it follows that high PMR should be associated with high BMR, and vice versa.
Relationships between BMR and PMR have been examined in a variety of vertebrate endotherms, with variable and often ambiguous results. Some studies, including both interspecific comparisons and intraspecific tests of within-population variation found positive correlations in both mammals (24, 50–52) and birds (e.g., refs. 11, 21, and 22). However, other studies found no association between BMR and PMR (e.g., refs. 12, 23, 25, 53, and 54). Some of the inconsistency may be due to seasonal or acclimatory differences among the species. Cold or seasonal acclimation greatly increases PMRC in many small mammals and in some species may also affect PMRE (28). Similarly, in some high-latitude birds, acclimatization to winter temperatures increases PMRC (6, 55). Also, PMRC is higher in the migration season in some temperate species, presumably as a consequence of increased flight muscle mass associated with preparation for long-duration migratory flights (33). Another potentially confounding factor is the use of different methods to elicit maximum performance (e.g., PMRE vs. PMRC; with different measurement techniques used for each). As a case in point, in Belding's ground squirrels, Spermophilus beldingi, BMR is significantly correlated to PMRE but not to PMRC (24).
In our work with tropical birds, we also obtained mixed results: PMRE and BMR were significantly correlated, but PMRC and BMR were not (Table 2). Correlations between PMRE and PMRC were not significant. Because PMRE was substantially higher than PMRC, the presence of a significant correlation with PMRE but not PMRC is consistent with the concept of symmorphosis (56) and the “aerobic capacity model” of the evolution of endothermy (42, 46), both of which predict that BMR should be linked to the highest level of aerobic power production. In contrast, Dutenhoffer and Swanson (21) and Rezende et al. (22) found correlations between BMR and PMRC in birds. However, most of their species were of temperate origin and so presumably experienced stronger selection (or acclimatory responses) for high thermogenic capacity, with correspondingly higher PMRC than our tropical species.
Mass scaling for BMR, PMRE, and PMRC was similar in tropical birds (Table 1), with no significant difference in interspecific mass exponents except between BMR and PMRE. This finding contrasts with other reports that indicate that PMRE may scale with a higher exponent with respect to Mb than PMRC in both mammals and birds (6, 44, 57). Because in birds both exercise and thermogenesis share a common peripheral effector (skeletal muscle) and common central supply organs, it is unclear to what extent these functions can evolve or respond to acclimation or conditioning independently.
Among temperate birds, tolerance to cold is reduced in species with low PMRC (6). As expected, tropical species do have inferior cold tolerance, with an average TCL 8.6°C higher than in temperate species. Nevertheless, all of the tropical birds we tested can tolerate Ta far colder than they are likely to encounter. The highest estimated TCL, converted to a value for air instead of heliox, was −2°C for the 11.5-g plain xenops, Xenops minutes. Because minimum Ta in lowlands in Panama rarely falls below 15°C, even this small species has a comfortable reserve capacity for heat production, at least for dry, low-convection environments like those in our metabolic chambers.
Materials and Methods
Capture and Handling of Birds.
We mist-netted birds in secondary growth and rainforests around Gamboa, Panama (9°7′N, 79°42′W; elevation ≈36–100 m), from April 20 to May 25, 2006, the start of the rainy season during which many birds breed (58). Birds were kept in small cages and provided water and food ad libitum. The type of food depended on their natural diet: if insectivorous, rehydrated crickets and live mealworms; if granivorous, a commercial mix of seeds; if frugivorous, fresh fruit. We usually measured PMR on the day of capture; a few individuals that fed readily were housed for 1–3 days before measurement.
We weighed birds (Mb, g) with a Pesola scale at the start of each metabolic trial. Body temperature (Tb, °C) was measured at the start and end of each cold-exposure trial, using a 36-gauge thermocouple inserted into the cloaca. Thermocouples were read to ±0.1°C by a Bailey Bat-12 thermocouple reader.
All procedures were approved by the Institutional Animal Care and Use Committee of Ohio State University (protocol IACUC2004A0093). Catching of birds was permitted by Panamanian Autoridad Nacional del Ambiente (permit no. SE/A-36-06) and Autoridad del Canal de Panamá.
Measurements of PMRC and TCL.
The methods used for our measurements of PMRC in tropical birds are given in Wiersma et al. (20). Briefly, we used heliox in a flow-through respirometry system, with a bird placed in a metabolism chamber that was in a temperature-controlled freezer. Inlet and outlet O2 concentration and chamber Ta were recorded at 1-sec intervals. Instantaneous V̇O2 (V̇O2inst) was calculated by using equations from Bartholomew et al. (59) and equation 4 of Hill (60), based on 5-min running averages of O2 concentrations. The effective volume of the system was estimated as 5,397 ml from washout curves. We used 20.08 J/ml O2 to convert V̇O2inst to heat production in watts (61).
Measurements of PMRE.
We used a metabolic flight wheel to measure PMRE (11). Dry air was supplied under positive pressure from a cylinder of compressed air, regulated at 5 liters/min (at standard temperature and pressure) with a Tylan mass flow controller calibrated before and after the field season against a DTM-113 dry volume meter (Singer American Meter Division). Excurrent air was subsampled at ≈100 ml/min, dried with Drierite, scrubbed of CO2 by using soda lime, redried, and routed through a Sable Systems Oxilla dual-channel O2 analyzer. Reference air for the differential reading came from the compressed air cylinder. Flow rate, wheel rotation speed, and O2 concentration were recorded every 1.0 sec with a Macintosh laptop computer and Sable Systems UI-2 AD converter running Warthog LabHelper software (http://warthog.ucr.edu).
Measurements of PMRE were made during the day at Ta of 23 ± 0.5°C. After weighing birds, we placed them inside the wheel, sealed it, and obtained an initial baseline reading of O2 concentration. After 1–2 min, we began rotating the wheel slowly (≈0.3 m/sec at the rim), increasing speed as birds became oriented to the direction of movement. We exercised birds at increasing intensity until they showed signs of exhaustion (panting and gaping; refusal to run or fly despite wheel rotation) and O2 consumption reached a plateau, whereupon the wheel was stopped and a second baseline reading was obtained. Typical tests lasted <15 min. This method has been used to elicit maximum V̇O2 in birds (11, 12) with high repeatability (63). Some individuals refused to run or flap in the wheel chamber, and we excluded these from our analyses.
For wheel measurements, we also used calculations of V̇O2inst (59). The effective volume of the wheel, estimated from washout curves, was 8,300 ml. Calculations of effective volume, baseline adjustment, smoothing to eliminate electrical noise, and computation of V̇O2inst were performed with Warthog LabAnalyst. Because the flowmeter was upstream of the chamber and excurrent CO2 was absorbed, metabolism was calculated as V̇O2inst = V̇ · [FiO2 − FEO2]/[1 − FEO2], where V̇ is flow rate and FiO2 and FEO2 are incurrent and instantaneous excurrent O2 concentrations, respectively (FiO2 was 0.2095). PMRE was computed as the highest continuous 1-min average of V̇O2inst (11) and converted to watts, as described for PMRC.
Data for Tropical–Temperate Comparisons.
We compared our PMRE measurements with data from the literature for three temperate species that were measured by using a flight wheel. We used mean values for adults of house sparrows (n = 36; ref. 11) from Australia, red-eyed vireos (n = 10; ref. 13) from North America, and satin bowerbirds from northeast New South Wales, Australia (Mb = 216.2 ± 1.7 g, PMRE = 14.1 ± 0.23 W, n = 36; J. Savard, J. Siani, M.A.C., and G. Borgia, unpublished data). Satin bowerbirds and house sparrows were caught and measured during the breeding season: the same period during which we made measurements on tropical species. Red-eyed vireos were caught during fall migration and kept in small cages for 1–2 months before measuring. It is not known whether this management may have affected their PMR, but a decrease in maximum muscle output in the red-eyed vireos might be expected because they were constrained to small cages, inhibiting muscle usage. Species-specific averages of BMR from Wiersma et al. (20) were used to calculate fASs [fASC and fASE (= PMR/BMR)].
Statistical Analyses.
We tested for statistical significance by using t tests and GLMs. To compare PMRE of tropical and temperate species, we tested for a significant effect of climate (tropical or temperate) in a GLM. Because Mb of satin bowerbirds substantially exceeded Mb of our tropical species (see Fig. 1), we also analyzed PMRE of house sparrows and red-eyed vireos separately, using t tests. For these t tests, we first regressed log10-transformed PMRE against logMb for the tropical species and compared each average temperate species' PMRE with the predicted tropical PMRE. To obtain the correct standard error of the predicted tropical PMRE, we fitted the regression line with a constant equaling Mb of the focal temperate species, instead of x = 0.
We calculated Mb-independent residual metabolic rates from regressions of species-average logPMR or logBMR on logMb. Where variables could covary, such as PMR and BMR, we tested for associations by using Pearson's correlation coefficient. To test for differences between slopes of the allometric relationships, we rearranged data of the variables PMRC, PMRE, fASC, fASE, and BMR into a single variable, while adding a variable coding for the different metabolism measurement types. The Mbs associated with the different metabolism variables were likewise reorganized. This procedure allowed us to then test for differences in logMb slopes by testing the interaction term of logMb and the “measurement type variable” in a GLM, along with logMb and the measurement type variable.
Because species are phylogenetically related to varying degrees, associations between traits of species may be differentially affected by common ancestry. Accordingly, we transformed our measurements to PICs (64, 65) to reduce phylogenetic effects in analyses. The use of PICs relies on assumptions about trait evolution that are sometimes difficult or impossible to verify, so we used both conventional and PIC analyses when interpreting results (66–68). We constructed a phylogenetic tree (see SI Fig. 3) primarily based on Sibley and Ahlquist (69) and calculated PICs by using PDTREE (70). Details on PIC analysis and tree construction are given in Wiersma et al. (20).
Statistical tests were performed using SPSS version 14.0, with α = 0.05. Values are shown as mean ± 1 standard error.
ACKNOWLEDGMENTS.
We thank Thomas Dijkstra, Ana María Jiménez, and Jennifer Ro for assistance in the field and the laboratory and Lisa Miller, Brianne Addison, Rubi Zambrano, Betzi Pérez, and Ben Lascelles for support in Gamboa. Ed Hice (University of California, Riverside) built the flight wheel, and Jill Soha of the Borror Laboratory (Ohio State University) provided a microphone. Staff from the Smithsonian Tropical Research Institute, especially Raineldo Urriola and Orelis Arosemena, greatly facilitated our research. We also thank the anonymous reviewers for their helpful comments. This study was funded by National Science Foundation Grant IBN 0212587 and by the University of California, Riverside.
Supporting Information
07683Table3.xls
- Download
- 24.00 KB
07683Fig3.jpg
- Download
- 81.10 KB
References
1
WR Dawson, C Carey J Comp Physiol 112, 317–333 (1976).
2
PS Reynolds, RM Lee Am Nat 147, 735–759 (1996).
3
AE McKechnie, BO Wolf Physiol Biochem Zool 77, 502–521 (2004).
4
BI Tieleman, JB Williams Physiol Biochem Zool 73, 461–479 (2000).
5
BI Tieleman, JB Williams, ME Buschur Physiol Biochem Zool 75, 305–313 (2002).
6
DL Swanson, ET Liknes J Exp Biol 209, 466–474 (2006).
7
M Rosenmann, P Morrison Am J Physiol 226, 490–495 (1974).
8
VA Tucker J Exp Biol 48, 67–87 (1968).
9
JH Brackenbury, MS el-Sayed J Exp Biol 117, 349–355 (1985).
10
DJ Ellerby, M Cleary, RL Marsh, CI Buchanan Physiol Biochem Zool 76, 695–703 (2003).
11
MA Chappell, C Bech, WA Buttemer J Exp Biol 202, 2269–2279 (1999).
12
KA Hammond, MA Chappell, RA Cardullo, R-S Lin, TS Johnsen J Exp Biol 203, 2053–2064 (2000).
13
BJ Pierce, SR McWilliams, TP O'Connor, AR Place, CG Guglielmo J Exp Biol 208, 1277–1285 (2005).
14
DS Hinds, RV Baudinette, RE MacMillen, EA Halpern J Exp Biol 182, 41–56 (1993).
15
TE Martin J Avian Biol 27, 263–272 (1996).
16
RE Ricklefs Condor 102, 9–22 (2000).
17
EM Russell, Y Yom-Tov, E Geffen Behav Ecol 15, 831–838 (2004).
18
HC Schaefer, GW Eshiamwata, FB Munyekenye, K Böhning-Gaese Ibis 146, 427–437 (2004).
19
AD Chalfoun, TE Martin Anim Behav 73, 579–585 (2007).
20
P Wiersma, A Muñoz-García, A Walker, JB Williams Proc Natl Acad Sci USA 104, 9340–9345 (2007).
21
MS Dutenhoffer, DL Swanson Physiol Zool 69, 1232–1254 (1996).
22
EL Rezende, DL Swanson, FF Novoa, F Bozinovic J Exp Biol 205, 101–107 (2002).
23
F Vézina, KM Jalvingh, A Dekinga, T Piersma J Exp Biol 209, 3141–3154 (2006).
24
MA Chappell, GC Bachman Physiol Zool 68, 421–442 (1995).
25
TP O'Connor J Comp Physiol B 165, 298–305 (1995).
26
MA Chappell, KA Hammond J Comp Physiol B 174, 41–48 (2004).
27
B Cannon, J Nedergaard Physiol Rev 84, 277–359 (2004).
28
JP Hayes, MA Chappell Physiol Zool 59, 473–481 (1986).
29
CR Taylor, GMO Maloiy, ER Weibel, VA Langman, JMZ Kamau, HJ Seeherman, NC Heglund Respir Physiol 44, 25–37 (1981).
30
RM Brigham, P Trayhurn Condor 96, 208–211 (1994).
31
JE Bicudo, AC Bianco, CR Vianna J Exp Biol 205, 2267–2273 (2002).
32
RL Marsh, WR Dawson Advances in Environmental and Comparative Physiology: Animal Adaptation to Cold, ed L Wang (Springer, Berlin), pp. 205–253 (1989).
33
DL Swanson, KL Dean J Avian Biol 30, 245–254 (1999).
34
VA Tucker Am J Physiol 222, 237–245 (1972).
35
JR Torre-Bueno, J Larochelle J Exp Biol 75, 223–229 (1978).
36
RE Ricklefs, M Wikelski Trends Ecol Evol 17, 462–468 (2002).
37
EL Rezende, F Bozinovic, T Garland Evolution (Lawrence, Kans) 58, 1361–1374 (2004).
38
JR Speakman J Exp Biol 208, 1717–1730 (2005).
39
MRE Symonds J Zool 249, 315–337 (1999).
40
BG Lovegrove J Comp Physiol B 173, 87–112 (2003).
41
BI Tieleman, TH Dijkstra, JR Lasky, RA Mauck, GH Visser, JB Williams Funct Ecol 20, 491–499 (2006).
42
AF Bennett J Exp Biol 160, 1–23 (1991).
43
JA Gessaman, KA Nagy Condor 90, 662–668 (1988).
44
CM Bishop Proc R Soc London B 266, 2275 (1999).
45
JR Arens, SJ Cooper Physiol Biochem Zool 78, 579–589 (2005).
46
AF Bennett, JA Ruben Science 206, 649–654 (1979).
47
M Kersten, T Piersma Ardea 75, 175–187 (1987).
48
S Daan, D Masman, A Groenewold Am J Physiol 259, R333–R340 (1990).
49
T Piersma Integr Comp Biol 42, 51–67 (2002).
50
JP Hayes J Comp Physiol B 159, 453–459 (1989).
51
F Bozinovic Physiol Zool 65, 921–932 (1992).
52
DS Hinds, CN Rice-Warner Physiol Zool 65, 188–214 (1992).
53
P Koteja Comp Biochem Physiol A 87, 205–208 (1987).
54
A Sparti Physiol Zool 65, 77–96 (1992).
55
DL Swanson, KL Olmstead Physiol Biochem Zool 72, 566–575 (1999).
56
CR Taylor, ER Weibel Respir Physiol 44, 1–10 (1981).
57
ER Weibel, LD Bacigalupe, B Schmitt, H Hoppeler Respir Physiol Neurobiol 140, 115–132 (2004).
58
AF Skutch Pac Coast Avifauna, pp. 31 (1954).
59
GA Bartholomew, D Vleck, CM Vleck J Exp Biol 90, 17–32 (1981).
60
RW Hill J Appl Physiol 33, 261–263 (1972).
61
K Schmidt-Nielsen Animal Physiology: Adaptation and Environment (Cambridge Univ Press, Cambridge, UK, 1997).
62
S Saarela, B Klapper, G Heldmaier Physiology of Cold Adaptation in Birds, eds C Bech, RE Reinertsen (Plenum, New York), pp. 115–122 (1989).
63
MA Chappell, M Zuk, TS Johnsen Funct Ecol 10, 578–585 (1996).
64
J Felsenstein Am Nat 125, 1–15 (1985).
65
T Garland, PH Harvey, AR Ives Syst Biol 41, 18–32 (1992).
66
M Westoby J Ecol 83, 892–893 (1995).
67
T Price Philos Trans R Soc London Ser B 352, 519–529 (1997).
68
A Muñoz-García, JB Williams Physiol Biochem Zool 78, 1039–1056 (2005).
69
CG Sibley, JE Ahlquist Phylogeny and Classification of Birds (Yale Univ Press, New Haven, CT, 1990).
70
T Garland, AW Dickerman, CM Janis, JA Jones Syst Biol 42, 265–292 (1993).
Information & Authors
Information
Published in
Classifications
Copyright
© 2007 by The National Academy of Sciences of the USA.
Submission history
Received: August 14, 2007
Published online: December 26, 2007
Published in issue: December 26, 2007
Keywords
Acknowledgments
We thank Thomas Dijkstra, Ana María Jiménez, and Jennifer Ro for assistance in the field and the laboratory and Lisa Miller, Brianne Addison, Rubi Zambrano, Betzi Pérez, and Ben Lascelles for support in Gamboa. Ed Hice (University of California, Riverside) built the flight wheel, and Jill Soha of the Borror Laboratory (Ohio State University) provided a microphone. Staff from the Smithsonian Tropical Research Institute, especially Raineldo Urriola and Orelis Arosemena, greatly facilitated our research. We also thank the anonymous reviewers for their helpful comments. This study was funded by National Science Foundation Grant IBN 0212587 and by the University of California, Riverside.
Notes
This article is a PNAS Direct Submission.
This article contains supporting information online at www.pnas.org/cgi/content/full/0707683104/DC1.
Authors
Competing Interests
The authors declare no conflict of interest.
Metrics & Citations
Metrics
Citation statements
Altmetrics
Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFGet Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to get full access to it.