Thermoregulatory Responses of Bridled and Juniper Titmice to High Temperature ’

The Condor 100:365-372 0 The Cooper Ornithological Society 1998

THERMOREGULATORY RESPONSES OF BRIDLED AND JUNIPER TITMICE TO HIGH TEMPERATURE ’

WESLEYW. WEATHERS AND ERICK GREENER Department of Avian Sciences, University of California, Davis, CA 956164532, e-mail: [email protected]

Abstract. Bridled Titmice (Baeolophus wollweberi) and Juniper Titmice (B. ridgwayi) occur sympatrically in southeasternArizona, with Bridled Titmice preferring habitats that are more heavily vegetated,moister, and cooler than those occupied by JuniperTitmice. To assesswhether these differences in habitat preference have physiological correlates, we measuredthe oxygen consumption,evaporative water loss, and body temperatureof postbreeding titmice at ambient temperaturesbetween 24-44°C. Bridled Titmice were less tolerant of heat than Juniper Titmice and had significantly higher rates of metabolic heat production and evaporative water loss, but not body temperature,at ambient temperatures above 40°C. These differences were entirely attributableto the Bridled Titmouse’s smaller body size (10 versus 15 g), and the differences vanishedwhen rates were expressedper unit metabolic mass (mass raised to either the 2/3 or 314power). Within the thermoneutralzone, the rate of evaporative water loss (EWL) was significantly lower in Juniper Titmice than Bridled Titmice, even after accounting for the difference in body size. Reduced EWL is characteristicof speciesfrom hotter, drier habitatsand suggeststhat physiology plays a role in these species’ habitat preferences. Key words: Baeolophusridgwayi, Baeolophuswollweberi, energy metabolism, evaporative water loss, habitat selection, heat stress.

INTRODUCTION forming meta-analyses on multiple two-species Organisms are inextricably linked to their phys- comparisons (Gurevitch et al. 1992). This latter ical environment through a continuous and re- approach is especially useful given that conclu- ciprocal exchange of matter and energy. Al- sions from multispecies comparisonsbased upon though this exchange is widely regarded as fun- modem comparative methods may change when damental to the evolution of life histories (Fisher the underlying phylogeny is revised (Garland et 1930, Bartholomew 1958, Dunham et al. 1989), al. 1991). its importance in influencing species distribu- In this study, we compare physiological re- tion, abundance, and habitat choice remains sponsesto high temperature in two congeners, poorly known (Root 1988a, 1988b). Studies of the Bridled Titmouse (Baeolophus wollweberi) congeneric species provide mounting evidence and the Jumper Titmouse (B. ridgwuyi). Bridled for small but significant physiological differ- Titmice weigh about one-third less than Juniper ences having adaptive value in specific environ- Titmice (Dunning 1984) and although their evo- ments (e.g., Hinds and Calder 1973, Hayworth lutionary relationship is unclear, both species and Weathers 1984, Hinsley et al. 1993). Al- presumably descended from a recent common though these studies suggest that physiology ancestor(Gill et al. 1989). The Bridled Titmouse may have a pervasive role in determining spe- is typically found in mid-elevation, mesic juni- cies distribution, they are based upon two-spe- per-pine woods throughout its more southerly cies comparisons,which have been criticized on distribution (Fig. 1) (Miller et al. 1957, R. Hutto, statistical grounds (Garland and Adolph 1994). pers. comm.) and thus it may encounter gener- However, two-species studies remain valuable ally cooler temperatures than those that prevail tools in physiological ecology becausethey pro- at the lower elevations preferred by the Juniper vide unique insights into organismal perfor- Titmouse (Cicero 1996). In the Chiricahua mance. They can even be used to study adap- Mountains of southeasternArizona where both tation (sensu evolutionary processes) by per- species are common residents, Bridled Titmice generally breed at higher elevations than Jumper ’ ’ Received21 July 1997. Accepted 7 January 1998. Titmice (1,500-2,300 versus 1,350-1,700 m) * Current address:Division of Biological Sciences, (E. Greene, pers. observ.) and thus experience The University of Montana, Missoula, MT 59812. markedly cooler temperatures.Even in the zone

13651 366 WESLEY W. WEATHERS AND ERICK GREENE

measured were presumably heat acclimated. They had startedto molt, but we did not quantify its intensity.

PHYSIOLOGICAL MEASUREMENTS We determined metabolic heat production and evaporative heat loss of captive adult Bridled (n = 11) and Juniper (n = 9) Titmice by measuring their oxygen consumption (VO,) and evaporative water loss (EWL) at stable air temperatures between 24 and 44°C. We captured birds with mist nets during the morning and transported them to the laboratory at SWRS where they were housed individually in 20-L glass terraria with live meal worms (Tenebrio) and wax worms (Gulleria) provided ad libitum. Both species proved to be tame, readily ate the available food and, for the most part, maintained weight. Bridled Titmice held in captivity more than one FIGURE 1. Ranges of Bridled Titmice and Juniper day exhibited a mean mass change of -0.03% Titmice in North America (based upon Cicero 1996, (range -4.6 to 8.0%). The comparable value for and Ficken and Necedal, in press). Juniper Titmice was -4.5% (range -10.5 to 0.0%). of elevational overlap, Bridled Titmice prefer We measured VO, and EWL at night (21:00- sites that are more heavily vegetated, moister, 02:00) on fasting birds (minimum of 4 hr post- and average 3.3”C cooler than those occupied by prandial) that were resting in the dark. Measure- Juniper Titmice (Greene and Weathers, unpubl. ments typically began the day of capture with data). These differences in habitat use were ini- individual birds being subjected to 2-3 different tially attributed to interspecific competition and randomly selected temperatures per night over interspecific territoriality (Dixon 1950, Marshall l-3 nights. Individual birds were measured on 1957). More recent studies indicate that direct average (mean -C SD) 4.0 -t 1.8 times (range l- contest competition does not appear to be im- 6) for the Bridled Titmouse and 3.7 + 1.2 times portant as these species can overlap during the (range 2-5) for the Juniper Titmouse. The me- breeding season without aggressive interactions tabolism chamber consisted of a 1.5-L glass jar (Gaddis 1987) and they do not respond to equipped with a wooden perch, mineral oil trap heterospecific song during territory establish- for urine and feces, a thermocouple for record- ment or during the breeding season (E. Greene, ing air temperature (T,), and air inlet and outlet unpubl. data). In this study, we examine the pos- ports. Air temperature within the chamber was sibility that differences in the species’ physio- controlled to within t- 0.2”C of the desired level logical tolerance to heat and aridity many un- by placing the chamber in a dark, temperature- derlie their breeding habitat segregation in controlled cabinet. The cabinet was equipped southeasternArizona. with a viewing port and light that allowed us to assessthe bird’s behavior at the end of the me- METHODS tabolism runs. We studied the thermoregulatory responses of To measure metabolism, birds were weighed postbreeding Bridled and Juniper Titmice to to the nearest 0.05 g with a calibrated electronic temperature during July 1989 at the Southwest- balance and placed in the respirometry chamber em Research Station (hereafter SWRS) located at a preset temperature. In most cases we al- in the Chiricahua Mountains of southeasternAr- lowed 1 hr to elapse after temperature within the izona near Portal, Arizona (32”O’N, 109”lO’W). metabolism chamber had stabilized before mon- During our study, shade air temperature at itoring 00, and EWL for at least 5 min. How- SWRS between 09:30-17:30 averaged 32.0 2 ever, for measurementson Bridled Titmice at T, 5.1”C (range 26.4-39.8”C). Thus the birds we > 41.5”C, the equilibration time was reduced to TITMOUSE THERMOREGULATION 367

> 4O”C, which is sufficiently low that evaporation should not have been impeded. (Above 4O”C, relative humidity averaged 11.8 2 2.9% and was unrelated to T,: r = 0.14, P = 0.45.) 36 l" li ' The fractional concentration of 0, in inlet and outlet chamber air (both dry and CO,-free) was determined with an Applied Electrochemistry S3-A oxygen analyzer, and VO, was calculated with Eq. 2 of Hill (1974). We corrected 90, to STPD and calculated rates of metabolic heat production assuming 20.1 kJ of heat were pro- duced per liter of 0, consumed by fasted birds. We monitored temperature of the metabolism -24 26 32 36 40 44 chamber with a 36-gauge thermocouple sus- Ambient temperature (“C) pended about 5 cm above the bird and connected FIGURE 2. Relation of colonic body temperature to a Bailey/Sensortek model Bat-12 thermocou- (above) and metabolic heat production (below) to am- ple thermometer. Thermocouple calibration bient temperaturein Bridled Titmice (solid circles) and (both T, and T,,) was against a mercury ther- Juniper Titmice (open squares). Line of equality de- mometer traceable to the National Bureau of picted in top panel. See text Equations l-3 for regres- Standards. sion statistics. DATA ANALYSIS an average of 47 min (range 33-58 min) because Because we had only a limited number of birds these birds became severely stressedby longer available for study, we had to make multiple exposures.At the end of each run, the light with- measurements on individuals in order to de- in the cabinet was turned on and the bird’s be- scribe each species’ physiological response to havior was simultaneously noted. The bird was temperature. This pseudoreplication (Hurlbert then quickly removed from the chamber and its 1984) is a virtually inevitable consequence of body temperature (T,,) measured within 30 set working with uncommon organisms. To assure by inserting a sheathed40-gauge thermocouple complete sample independence for data plotted into the colon such that no temperature drop oc- in Figures 2-4, we would have had to capture a curred upon slight withdrawal. The bird was total of 62 individual titmice, which was not reweighed and the mean of the initial and final possible at our study site. We accommodatedthe mass was used in calculations. pseudoreplication by using repeated-measures Dry, CO,-free air passedthrough a rotameter (calibrated with a bubble meter, Levy 1964) and 50 , I into the metabolism chamber at between 0.4 to r I . k 0.75 L min-l, which resulted in chamber air at- ;‘ .- cn 40 . l taining 99% equilibrium in 9.2-17.3 min (La- . siewski et al. 1966). The dew point of outlet E : chamber air was determined to the nearest O.l”C with a General Eastern model 1lOODP dew point hygrometer.Evaporative heat loss was calculated from dew point (Bernstein et al. 1977), assuming that 2.43 kJ of heat were dissipated for each g of water evaporated.Because buildup of chamber humidity due to an animal’s evaporative water loss can interfere with evaporative cooling at high temperatures (Lasiewski et al. 1966), we employed higher air-flow rates (mean Ambient temperature (OC) = 0.63 L min-I) at air temperatures > 40°C. FIGURE 3. Relation of evaporativewater loss to am- This maintained the chamber water vapor pres- bient temperaturein Bridled Titmice (solid circles) and sure at 0.97 + 0.25 kPa (mean -C SD) when T, Juniper Titmice (open squares). 368 WESLEY W. WEATHERS AND ERICK GREENE

1.0 of both specieswere awake, but with their beaks closed and no visible signs of panting. At slight- ly higher temperatures the two species’ re- sponsesdiverged. At T,s > 40.3”C, all Bridled Titmice were awake and panting vigorously with their wings held away from their bodies. Juniper Titmice, however, did not begin to pant until T, exceeded 40.6”C, and even then they never drooped their wings as did Bridled Titmice. These behavioral observations are consistent c 0 with our physiological data and indicate that Bridled Titmice are more severely stressedby 0.0 __ 24 20 32 36 40 44 temperatures above 40°C than Juniper Titmice.

Ambient temperature (“C) HEAT PRODUCTION FIGURE 4. Ratio of heatlost by evaporationto met- Metabolic heat production (I$,,) of most bird abolic heat productionin BridledTitmice (solid cir- species is independent of T, within the thermo- cles) and JuniperTitmice (open squares). neutral zone (TNZ) and increases linearly with T, above the upper critical temperature (T,,) ANCOVA (SAS 1992) to compare least squares (Weathers 1981). Both titmice species fit this regressionsdescribing the species’ physiological pattern and we used linear relations to described responses to high temperature. These analyses their l$,, versus T,. We attempted to identify revealed individual birds to be statistically each species’ T,, objectively using a dual re- equivalent (probabilities for individual effects gression procedure similar to that of Yeager and were 0.14-0.69). Thus, no significant bias was Ultsch (1989). This proved unsuccessful, how- introduced by repeatedly sampling the same in- ever, because regressions for both species at dividuals. Means reported for various parame- higher T,s had negative slopes. Based upon the ters were calculated using average values for in- birds’ behavior (see below) and visual inspec- dividual birds. tion of the data (Fig. 2), the T,, of both species Comparing rates of metabolic heat production appears to lie between 3%40°C. Accordingly, and evaporative water loss of speciesthat differ we fit regressions to data above 38°C using in body size is somewhat problematic. Tradi- weighted regression to compensate for the Bri- tionally, mass independent values have been cal- dled Titmouse’s heteroscedasticdata (Neter et al. culated by dividing physiological rates by body 1990). The analyses yielded the following rela- mass raised to the 0.75 power (Kleiber 1947), tions: although an argument can be made for using an Bridled Titmouse: exponent of 0.67 (Heusner 1982). We avoided this controversy by adjusting rates using both k,(mW g-r) = -53.2 + 2.04T, (1) exponents. The results were statistically equiv- (9 = 0.420, syx = 4.23, s,,= 0.370, n = 11, N= 26) alent. Throughout the text, n denotes the number of Jumper Titmouse: individuals and N the number of measurements. For means, N and n are identical. Means are re- l$,,(mW g-l) = -28.8 + 1.25T, (2) ported 2 SD. (r2 = 0.327, sy,= 2.21, s,,= 0.582, n = 7, N= 12) RESULTS These equations differ neither in intercept (AN- BEHAVIORAL RESPONSESTO HIGH COVA F,,,g = 1.20, P > 0.2) nor slope (F,,,, = TEMPERATURE 1.78, P > 0.2). However, lack of significant dif- Most individuals of both species were asleep at ference is attributable to the convergence of the the end of metabolic runs conducted below data sets at lower temperatures. Above 40°C 38°C. Birds slept either in an upright posture or mean a,,, of the two species is significantly dif- with their heads tucked into their scapular feath- ferent (Table 1). There is no significant differ- ers. Between 3%40.3”C, nearly all individuals ence in mean A, above 40°C however, when TITMOUSE THERMOREGULATION 369

TABLE 1. Mean (SD) of measurementsmade at ambient temperaturesexceeding 40°C for 11 Bridled and 6- 7 Juniper Titmice.

Bridled Juniper Variable Titmouse Titmouse ta P Ambient temperature,“C 41.4 (0.4) 41.9 (0.5) 2.05 0.06 Body temperature,“C 43.1 (0.4) 42.7 (0.4) 2.05 0.06 Heat production, mW g-l 29.8 (4.4) 24.4 (1.9) 2.88 0.01 Heat production, mW gm314 53.0 (7.9) 47.8 (3.2) 1.56 0.14 Heat production, mW ge2j3 63.7 (9.9) 57.9 (3.8) 0.99 0.34 Evaporative water loss, mg g-l hr-’ 28.3 (7.1) 21.8 (2.3) 2.22 0.04 Evaporative water loss, mg g-3’4 hr-’ 50.4 (12.7) 42.7 (4.4) 1.19 0.16 Evaporative water loss, mg geu3hr ’ 61.0 (15.3) 53.4 (5.5) 1.19 0.25 Heat loss/heatproduction 0.60 (0.09) 0.59 (0.10) 0.23 0.80

a Two-tailed test. df = 15 or 16

fi,,, is expressed per unit metabolic mass, i.e., Mean T, calculated for measurements above body mass raised to either the 213 or the 314 40°C also did not differ between species (Table power (Table 1). 1). We calculated basal metabolic rate (BMR, mW_‘g) at thermoneutral temperatures (24- EVAPORATIVE WATER LOSS 38°C) using mean values for individual birds. Mean evaporative water loss (EWL) measured BMR was significantly higher in Bridled than in at T,s between 24-35°C was significantly higher Jumper Titmice (Table l), but the difference in in Bridled Titmice and the difference persisted BMR was insignificant when the difference in when EWL was expressedon a mass indepen- the species’ body mass was taken into account dent basis (Table 2). Above 32°C EWL of both (Table 2). species increased exponentially (Fig. 3), as is typical of homeotherms. At T,s above 40°C the BODY TEMPERATURE mean EWL of Bridled Titmice was significantly The body temperature (Tb) of both species in- higher than that of Juniper Titmice, but this dif- creased linearly with ambient temperature (Fig. ference vanished when EWL was expressedon 2). Equations fit to data above 32°C for each a mass-independentbasis (Table 1). speciesdiffered neither in slope (F,,,, = 0.86, P Regressionsof log,, mg g-r hr-’ EWL versus > 0.3) nor intercept (F,,,, = 0.51, P > 0.4). The T, for data at T,s > 32°C did not differ signifi- common equation for pooled data above 32°C cantly for the two species either in slope (F,,,, is: = 0.32, P > 0.5) or intercept (F,,38 = 0.90, P > 0.3). The common regressionfor the pooled data Tb = 29.6 + 0.32T, (3) is: (r2 = 0.767, sYX= 0.72, sb = 0.023, n = 20, log ,0 EWL = -2.4076 + O.O91T, (4) N = 62) (12 = 0.881, log,, syx= 0.1117, Sb = 0.004,

TABLE 2. Mean (SD) body mass, basal metabolic rate, and evaporative water loss measuredwithin the thermoneutral zone for 9 Bridled and 7 Juniper Titmice.

Bridled Variable Titmouse P P Body mass, g 10.0 (0.8) 14.8 (1.0) 10.52 0.001 Basal metabolic rate, mW g-l 22.4 (1.9) 19.3 (0.7) 3.95 0.002 Basal metabolic rate, mW gm3” 39.9 (3.9) 37.8 (1.2) 1.27 0.22 Basal metabolic rate, mW gm2j3 47.9 (4.9) 47.3 (1.5) 0.30 0.76 Evaporative water loss, mg g-’ hrr 7.4 (0.7) 5.2 (1.2) 4.26 0.001 Evaporative water loss, mg ge314hrr ’ 13.1 (1.3) 10.2 (2.3) 2.95 0.01 Evaporative water loss, mg gmu3hr ’ 15.8 (1.6) 12.8 (2.9) 2.50 0.02

a Two-tailed test, df = 14. 370 WESLEY W. WEATHERS AND ERICK GREENE

II = 19, IV = 60) 1970), and EWL averaged 133 and 102%, respectively, of that predicted allometrically (Wil- EVAPORATIVE HEAT LOSS liams 1996). When the species’ difference in The ratio of heat lost by evaporation to meta- body size is accountedfor, however, BMRs were bolic heat production (&/II,,,) provides an over- statistically equivalent, whereas EWL of Juniper all index of the response to heat. As is typical Titmice averaged 20% lower than that of Bridled of birds generally, the &&I, ratio of titmice in- Titmice (Table 1). Juniper Titmice in Arizona’s creased exponentially at higher temperatures Chiricahua Mountains occupy habitats that are (Fig. 4). At T,s > 32”C, there was no significant more xeric than those preferred by Bridled Tit- difference between the two species in either in- mice and this is reflected in their significantly tercept (F,,,, = 0.38, P > 0.5) or slope (F,,,, = lower EWL. Whether this is an evolved response 0.53, P > 0.5). The combined data can be de- is unclear, but it implies that for these species scribed as: moisture is more important than temperature in log,, H&I, = -3.2078 + O.O72T, (5) establishing habitat preferences. At high ambient temperatures, resting meta- (r2 = 0.894, log,, sYx= 0.0824, s,, = 0.003, bolic rate (RMR) increased more slowly in both 12= 19,N= 60) speciesthan expected for birds of their size. The slope of the line relating RMR to ambient tem- At T,s above 4O”C, the mean I$&,, ratio of both species was identical (Table 1). perature (i.e., the coefficient of heat strain, h,) in Bridled Titmice averaged 69% of that predicted DISCUSSION allometrically (Weathers 1981) and in Juniper We hypothesized that because Bridled Titmice Titmice only 50% of that predicted. This prob- prefer cooler habitats than those occupied by Ju- ably reflects the fact that the titmice were heat- niper Titmice, they would be less heat tolerant. acclimated prior to study, whereas the speciesin Our data support this hypothesis. Bridled Tit- Weathers’ (1981) allometric analysis were not. mice exhibit greater increases in resting meta- Heat acclimation reduces h, (DeShazer et al. bolic rate and evaporative water loss at ambient 1970, Marder et al. 1989), and measurementsof temperatures above 40°C than do Juniper Tit- titmice unacclimated to heat are needed to as- mice, and behaviorally they appear more certain if these species’ responses are unusual. stressedby high temperature. Although signifi- Regardless of whether the titmouses’ low h, is cant, these differences are entirely attributable to an evolved response,it has (general sense) adap- the Bridled Titmouse’s smaller body size. We tive value. The increase in RMR which occurs found no evidence which would suggestthat Ju- at temperatures above the upper critical temper- niper Titmice are better adapted to high temperature has two principal sources:the muscular ef- ature than Bridled Titmice, except by virtue of fort involved in panting and elevated body tem- their larger size. Apparently, the high tempera- perature. Because metabolic heat production ex- tures encountered by these speciesin southeast- acerbatesheat stressby increasing the total heat em Arizona can be accommodatedfully by ge- load, reducing its magnitude may have adaptive neric parid physiology. In contrast, our data sug- value (Weathers 1981). The value of a relatively gest that aridity may help shape these species’ low rate of increase in RMR is apparent as Bri- microhabitat preferences. dled Titmice must increase EWL at high tem- Aridity is a significant selective force in the peraturesmuch more than do Juniper Titmice in evolution of avian physiology as both basal metabolic rate (BMR) and evaporative water loss order to dissipate their extra heat production. (EWL) within the thermoneutral zone are rela- Thus compared with Bridled Titmice, Juniper tively low in desert species (Dawson and Ben- Titmice have the dual advantage of expending nett 1973, Dawson 1984, Williams 1996). Nei- both less energy and water at high temperatures. ther of the titmice we studied is a desert species Although these differences are entirely attribut- and both had average or above average BMR able to the Juniper Titmouse’s larger size, they and EWL. BMR of Bridled and Jumper Titmice improve this species physiological performance averaged 116 and 119%, respectively, of that in its dryer and somewhat hotter preferred hab- predicted allometrically (Aschoff and Pohl itat. TITMOUSE THERMOREGULATION 371

ACKNOWLEDGMENTS to do two-species comparativestudies: limitations on inferring adaptation. Physiol. Zool. 67:797- We thank Wade Sherbrooke and all the staff of the 828. SouthwesternResearch Station of the American Mu- GARLAND,T., JR., R. B. HUEY, AND A. E BENNETT. seum of Natural History for their unflagging support. 1991. Phylogeny and thermal physiology in liz- JosephWilliams provided helpful commentson an ear- ards: a reanalysis.Evolution 45:1969-1975. lier draft of the manuscriptand Thomas Famula pro- GILL, E B., D. H. FUNK, AND B. SILVERIN.1989. Pro- vided invaluable help with SAS. This research was tein relationships among titmice (Baeolophus). supportedby a Life SciencesResearch Foundation Fel- Wilson Bull. 10:182-197. lowship to EG, and National Science Foundation GUREVITCH,J. L., L. L. MORROW,A. WALLACE,AND J. grants DEB 80-22765, BSR 85-05490, and BSR 87- S. WALSH. 1992. A meta-analysisof competition 18195 to WWW. in field experiments.Am. Nat. 140:539-572. HAYWORTH,A. M., AND W. W. WEATHERS.1984. Tem- LITERATURE CITED perature regulation and climatic adaptation in Black-billed and Yellow-billed Magpies. Condor ASCHOFF,J., AND H. POHL. 1970. Der Ruheumsatzvon VGgelen als Funktion der Tageszeit und der KBr- 86: 19-26. pergrbsse.J. Ornithol. 111:38-47. HEU~NER,A. A. 1982. Energy metabolism and body BARTHOLOMEW,G. A. 1958. The role of physiology size. I. Is the 0.75 mass exponent of Kleiber’s in the distributionof terrestrialvertebrates, p. 81- equation a statistical artifact? Resp. Physiol. 48: 95. In C. L. Hubbs [ed.], Zoogeography.Am. As- 1-12. soc. Adv. Sci., Washington,DC. HILL, R. W. 1974. Determination of oxygen consump- BERNSTEIN,M. H., D. M. HUDSON,J. M. STEARNS,AND tion by use of the paramagneticoxygen analyzer. R. W. HOYT. 1977. Measurement of evaporative J. Appl. Physiol. 33:261-263. water loss in small animals by dew-point hygrom- HINDS,D. S., ANDW. A. CALDER.1973. Temperature etry. J. Appl. Physiol.: Resp., Environ., and Ex- regulationof the pyrrhuloxia and the Arizona car- ercise Physiol. 43:382-385. dinal. Physiol. Zool. 46:55-71. CICERO,C. 1996. Sibling species of titmice in the HINSLEY,S. A., I? N. FERNS,D. H. THOMAS,AND B. Baeolophus inornatus complex (Aves: Paridae). PINSHOW.1993. Black-bellied Sandgrouse(Pter- Univ. Calif. Publ. Zool. 128:1-127. ocles orient&s) and Pin-Tailed Sandgrouse(Pfer- DAWSON,W. R. 1984. Physiological studiesof desert ocles alchutu): closely related specieswith differ- birds: present and future considerations.J. Arid ing bioenergetic adaptationsto arid zones. Physi- Environ. 7:133-155. 01. Zool. 66:20-42. DAWSON,W. R., AND A. E BENNE?T.1973. Roles of HURLBERT,S. H. 1984. Pseudoreplicationand the de- metabolic level and temperatureregulation in the sign of ecological field studies.Ecol. Monogr. 54: adjustment of western plumed pigeons (Lopho- 187-211. phups ferruginea) to desert conditions. Comp. KLEIBER,M. 1947. Body size and metabolic rate. Biochem. Physiol. 44A:249-266. Physiol. Rev. 27:51l-541. DESHAZER,J. A., K. A. JORDON,AND C. W. SUGGS. LASIEWSKI,R. C., A. L. ACOSTA,AND M. H. BERNSTEIN. 1970. Effect of acclimationon partitioningof heat 1966. Evaporative water loss in birds I. Charac- loss by the laying hen. Trans. Am. Sot. Agr. Eng. teristicsof the open flow method of determination, 13:82-84. and their relation to estimatesof thermoregulatory DIXON, K. L. 1950. Notes on the ecological distribu- abilitv. Comn. Biochem. Phvsiol. 19:445-457. tion of the Juniper and Bridled Titmice in Arizo- LEVY, A.’ 1964.’ The accuracy‘of the bubble meter na. Condor 52:140-141. method for gas flow measurements.J. Scientific DUNHAM,A. E., B. W. GRANT,AND K. L. OVERALL. Instrum. 41:449-453. 1989. Interfaces between biophysical and physi- MARDER,J., Y. ARIELI,AND J. BEN-ASHER.1989. De- ological ecology and the population ecology of fense strategiesagainst environmental heat stress terrestrial vertebrate ectotherms. Physiol. Zool. in birds. Israel J. Zool. 36:61-75. 62:335-355. MARSHALL,J. T 1957. Birds of pine-juniper woodland DUNNING,J. B. 1984. Body weights of 686 speciesof in southernArizona and adjacent Mexico. Pacific North American birds. Western Bird Banding As- soc. Monogr. 1. Eldon Publishing, Cave Creek, Coast Avifauna 32. AZ. MILLER,A. H., H. FRIEDMAN,L. GRISCOM,AND R. T. MOORE. 1957. Distributional check-list of the FICKEN,M. S., AND J. NECEDAL.In press. Bridled Tit- mouse (Parus wollweberi). In A. Poole and E Gill birds of Mexico, Part II. Pacific Coast Avifauna [eds.], The birds of North America. The Academy 33. of Natural Sciences,Philadelphia, and The Amer- NETER,J., W. WASSERMAN,AND M. H. KUTNER.1990. ican Ornithologists’ Union, Washington,DC. Applied linear statistical models, 3rd ed. Irwin, FISHER,R. A. 1930. The genetical theory of natural Homewood, IL. selection. Clarendon Press, Oxford. ROOT, T 1988a. Energy constraintson avian distri- GADDIS,l? K. 1987. Social interactions and habitat butions and abundance.Ecology 69:330-339. overlan between Junioer and Bridled Titmice. ROOT, T 1988b. Environmental factors associated SouthhesternNat. 32: i97-202. with avian distributional boundaries.J. Biogeog- GARLAND,T., JR., AND S. C. ADOLPH.1994. Why not raphy 15:489-505. 372 WESLEY W. WEATHERS AND ERICK GREENE

SAS INSTITUTE.1992. SAS user’s guide: statistics, evaporative water loss in birds. Auk 113:457- Version 6.08. SAS Institute, Inc., Cary, NC. 472. WEATHERS,W. W. 1981. Physiological thermoregula- YEAGER,D. P, AND G. R. ULTSCH.1989. Phvsiological tion in heat-stressedbirds: consequencesof body regulation and conformation: a BASIC program size. Phvsiol. Zool. 54:345-361. for the determination of critical Iuoints. Phvsiol., WILLIAMS,Jr B. 1996. A phylogenetic perspective of Zool. 62: 888-907.