Journal of J Comp Physiol B (1988) 157:783-790 Comparative Systemic,Biochemical, and Environ- Physiology B .%~"~o',o.. Springer-Verlag 1988 Thermoregulation, gas exchange, and ventilation in A delie penguins (Pygoscelis adeliae) Mark A. Chappell and Sherrie L. Souza Department of Biology, University of California, Riverside, California 92521, USA Accepted October 14, 1987 Summary. Adelie penguins (Pygoscelis adeliae) ex- Introduction perience a wide range of ambient temperatures (T,) in their natural habitat. We examined body tem- Recent research has revealed considerable flexibili- perature (Tb), oxygen consumption (l)o~), carbon ty and variation in the responses of avian ventila- dioxide production (Vco2), evaporative water loss tory systems to changing ambient temperature (rhH2o), and ventilation at T~ from --20 to 30 ~ (Bucher 1981, 1985; Bech and Johansen 1980; Body temperature did not change significantly be- Brent et al. 1984; Kaiser and Bucher 1985; Chap- tween -20 and 20 ~ (mean Tb=39.3 ~ Tb in- pell and Bucher 1987). These responses presumably creased slightly to 40.1 ~ at Ta= 30 ~ Both l)o~ reflect the dual function of the respiratory system and l/co2 were constant and minimal at Ta between as a pathway for both gas exchange and, secondar- -10 and 20 ~ with only minor increases at -20 ily, for heat loss. At high ambient temperatures and 30 ~ The minimal 1/o2 of adult penguins (T,), many species greatly enhance evaporative (mean mass 4.007kg) was 0.0112ml/[g-min], heat loss by panting. At low Ta, increased thermo- equivalent to a metabolic heat production (MHP) genesis is supported by elevated rates of oxygen of 14.9 Watt. The respiratory exchange ratio was consumption (lZo2) and carbon dioxide production approximately 0.7 at all Ta. Values of rhn2o were (lZco2). These increased gas exchange requirements low at low T~, but increased to 0.21 g/rain at 30 ~ may be accommodated by increasing lung ventila- equivalent to 0.3% of body mass/h. Dry conduc- tion (minute volume, lz0, oxygen extraction (Eo~), tance increased 3.5-fold between -20 and 30 ~ or both. Minute volume is itself a function of respi- Evaporative heat loss (EHL) comprised about 5% ration frequency (f) and tidal volume (Va'). In some of MHP at low Ta, rising to 47% of MHP at Ta = birds (e.g., pigeons, prairie falcons, parrots, and 30 ~ The means of ventilation parameters (tidal chukar partridges), Eo~ remains high but relatively volume [VT], respiration frequency [f], minute vol- constant at Ta below thermoneutrality, and in- ume [l/I], and oxygen extraction leon]) were fairly creasing oxygen demand is accommodated by in- stable between -20 and 10 ~ (VT did not change creases in V~ (Bech et al. 1985; Bucher 1981, 1985; significantly over the entire Ta range). However, Kaiser and Bucher 1985; Chappell and Bucher there was considerable inter- and intra-individual 1987). Other species (e.g., mallard ducks and Euro- variation in ventilation patterns. At Ta = 20-30 ~ pean coots) show a contrasting pattern: Eo~ in- f increased 7-fold over the minimal value of creases with increasing Vo2 at Ta below thermoneu- 7.6 breaths/rain, and 1~I showed a similar change. trality (Bech et al. 1985; Brent et al. 1984). Because Eo~ fell from 28-35% at low Ta to 6% at Ta= temperature effects on 1)I and Eo2 have been stud- 30 ~ ied in a limited number of species, the physiologi- cal, phylogenetic, and/or ecological foundations Abbreviations: C thermal conductance; EHL evaporative heat and adaptive significance, if any, for these differ- loss; Eo2 oxygen extraction ;frespiratory frequency; MHP met- ences are unclear. abolic heat production; rh~i~o evaporative water loss; LCT High Eo2 at low T~ is often interpreted as an lower critical temperature; RE respiratory exchange ratio; T. ambient temperature; Tb body temperature; 1/o2 rate of oxygen adaptation for reducing the metabolic cost of ther- consumption; l)co2 rate of carbon dioxide production; l?~ inspi- moregulation by decreasing respiratory heat loss ratory minute volume; VT tidal volume (Johansen and Bech 1983). However, in the studies 784 M.A. Chappell and S.L. Souza: Thermoregulation and ventilation in penguins performed to date the estimated energy savings due lus within 45 s of removing the bird from the chamber. Penguins to high Eo: are minor - ca. 3-6% of total thermo- were fasted at least 12 h (usually >20 h) before experiments. All measurements were made during local daylight (070~1900 regulatory costs (Bucher 1981; Chappell and Palmer time). Bucher 1987). Moreover, it is difficult to separate Oxygen consumption ([2o2), carbon dioxide production the influence on Eo: of low Ta from that of high (~2co2), and evaporative water loss (rhH2o) were measured using 12o:, since all of the species studied thus far show open-circuit respirometry. Flow rates of dry air (12 1/min at substantial thermogenic 1/o~ at low Ta- low Ta to 25 1/min at high Ta) were regulated _+ i % with Tylan mass flow controllers. A fraction (50-100 ml/min) of the excur- To address these problems, additional compar- rent air was diverted through a Viasala humidity sensor, dried ative data are needed, particularly from birds with and passed through an Anarad AR-50 CO2 sensor, scrubbed thermoneutral zones which extend to very low T,. of CO2 with Ascarite and redried, and passed through an Ap- In these species, 12o: is 'decoupled' from Ta, allow- plied Electrochemistry S-3A 02 sensor. Before and after mea- surements, all sensors were referenced against dry air diverted ing study of temperature effects on ventilation from upstream of the respirometry chamber. The humidity without the confounding influence of changing re- probe resolved 0.1% R.H. and was calibrated over salt solu- quirements for oxygen uptake. Accordingly, we in- tions to within 1% of actual R.H. The CO2 sensor was cali- vestigated thermoregulatory physiology, ventila- brated weekly with a precision reference gas to an accuracy tion, and gas exchange in the Adelie penguin (Py- of 0.005% COa. Both CO2 and Oz analyzers resolved concen- trations of 0.001%. During experiments, 02 concentration was goscelis adeliae). This species routinely experiences always > 20.3% and CO2 concentration never exceeded 0.6%. subzero air and seawater temperatures in its natu- Chamber humidity was kept below 50%, except at Ta < 10 ~ ral habitat, and is well insulated. Consequently it Ventilation and gas exchange data were taken onl.y after birds shows little elevation of /2o: at T~ as low as had been held at one T, for at least lh and Vo2 was low and stable. Gas exchange data were recorded by a microcom- --20 ~ During the nesting season, Adelies may puter equipped with an analog-to-digital converter. The sample also experience considerable heat stress. On warm, interval was 5 s, which allowed 53 min of uninterrupted mea- sunny days, both chicks and adults may pant for surements between pauses for data storage. 12o2 was calculated hours at a time (Murrish 1973, 1982). as: 12o2 = 12. (No2 - Feo2) /[ M (1 - Feo2)] (l) where [2 is flow rate (STPD), Fio 2 is the fractional concentration Materials and methods of 02 in incurrent air, Feo2 is fractional 02 concentration in Animals. We made preliminary measurements in July 1986 on excurrent air, and M is body mass. Vco2 was calculated as : adult birds from a colony maintained at Sea World, in San Pco2 =/2" (Feco 2-- JUlcoyM (2) Diego, California. These penguins (N= 8) were obtained from wild populations, but had been held in captivity for 10 to where Feco2 is the fractional concentration of CO2 in excurrent 15 years. They were kept in a large group aviary maintained air and Fie% is the fractional concentration of CO2 in incurrent at -5 to 5 ~ and were fed a diet of fish and squid supple- air. Rates of evaporative water loss (rhH2O) at Ta > 0 ~ were mented with vitamins. We also obtained data from wild pen- computed by substituting water vapor concentrations for CO2 guins collected during December 1986 and January 1987 in concentrations in Equ. (2). Data on vapor pressures and vapor the breeding colony on Torgersen Island, near Palmer Station densities were obtained from Tracy et al. (1980). No condensa- off the west coast of the Antarctic Peninsula (64~ tion within the flow system was observed at Ta > 0 ~ Plexiglas 64~ These birds (N= 6) were captured in hand nets, held adsorbs small quantities of water vapor, which can decrease in the laboratory for 2 to 3 days, and released after measure- the accuracy of water loss measurements in conditions of low ments were completed. Most of the wild penguins were obtained humidity, flow rate, and rhH~o. Potential errors from this source from small 'resting' groups; none were captured while attend- were minimized by the relatively high humidities and flow rates ing nests. Ventilation data from wild birds was much more we used, and by the 1 2 h equilibration period between mea- uniform than those from captives, and I2o2 of wild birds was surements. At T, _< 0 ~ frost formed in the excurrent air line, substantially lower than the IZo~ of captives (see Results). Ac- preventing direct measurement of rhn o. Accordingly we used cordingly, we restricted most of our analyses to wild penguins. Equ. (56) in Calder and King (1974) 2 to compute water loss Unless otherwise specified, all reported measurements refer to from ~2o: at these T,. This approximation, which predicts rhn~o wild penguins. from Ta and metabolic rate, may not be completely accurate for Adelie penguins. However, any errors are probably unim- Measurements. Rates of gas exchange were measured in a plexi- portant for calculations of heat balance because, at T~ < 0 ~ glas respirometry chamber (internal volume 92 1) placed in a evaporative heat loss is typically < 6% of metabolic heat pro- larger environmental chamber which controlled ambient tem- duction.