Seasonal Effects on Metabolism and Thermoregulation in Northern Bobwhite ’

The Condor 99:478-489 0 The Cooper Ornithological Society 1997

SEASONAL EFFECTS ON METABOLISM AND THERMOREGULATION IN NORTHERN BOBWHITE ’

DAVID L. SWANSON AND DUANE l? WEINACHT Department of Biology, Universityof SouthDakota, Vermillion, SD 57069-2390, e-mail: [email protected]

Absfract. Seasonal differences in metabolism and cold hardinessare common among small passerinebirds. However, seasonaladjustments of metabolism and insulationare less well studiedin nonpasserinesand in larger birds. We measuredbasal metabolic rate (BMR), metabolic response to temperature, and maximal capacity for thermogenesis(peak cold- induced oxygen consumption,\iO 2,,,) in late spring/summerand winter in outdoor captive Northern Bobwhite (Colinus virginianus)near the northern boundary of their naturalrange, to determine whether seasonaladjustments in metabolism are a component of acclimatization in this species.Mass, BMR and regressionequations describing diurnal and nocturnal metabolic response to temperaturewere not significantly different between seasons.After metabolic tests below thermoneutrality,cloaca1 temperature (T,) was not dependenton ambient temperature(T,) at either season,and nocturnalT, did not differ significantly between seasons.However, after metabolic testsbelow thermoneutrality,diurnal T, was significantly greater in summer (38.9 k l.l”C) than in winter (37.7 ? l.l”C). Although body mass of winter birds was significantly greater than their body mass in late spring, maximal thermogenic capacity did not differ significantly on a seasonalbasis, and winter bobwhite were only marginally more cold tolerant than late spring birds under severe cold stress. For individual birds tested in both winter and late spring, individual ranking of ‘ir0, Fumwas not consisteptbetween seasons(i.e., birds with a high \;rO, pulllin winter did not necessarilyhave a high VO, sumin late spring). These data suggestlittle seasonaladjustment of metabolism or insulationin the Northern Bobwhite, a relatively large nonpasserinespecies native mainly to regions with relatively mild winter climates. Key words: Northern Bobwhite,Colinus virginianus,metabolism, thermbregulation, seasonal acclimatization,maximal thermogeniccapacity.

INTRODUCTION climatization in small passerines (Dawson and Many small passerine birds are able to inhabit Carey 1976, Dawson et al. 1983b). The precise north temperate regions with relatively severe mechanistic basis for this improved shivering winter climates in spite of their high surface area endurance in winter compared to summer re- to volume ratio, limited capacity for seasonally mains uncertain (Marsh and Dawson 1989, improving insulation and limited ability to store Marsh et al. 1990). Seasonal adjustments in in- fat (Marsh and Dawson 1989). Winter acclima- sulation in these small birds are less marked than tization in these birds is primarily metabolic in metabolic adjustments and play only a minor nature. In small birds, basal metabolic rate role in seasonal acclimatization (Dawson and (BMR) is often, although not always, higher in Carey 1976, Swanson 1991a, Cooper and Swan- winter than in summer (Weathers 1980, Dawson son 1994). These metabolic adjustments allow et al. 1983a, Liknes and Swanson 1996). Max- marked winter improvement in cold resistance imal capacity for thermogenesis (peak, cold-in- relative to summer in passerines inhabiting rel- duced oxygen consumption or summit metabo- atively cold winter climates (Marsh and Dawson lism, VOZsum)increases in winter relative to sum- 1989, Swanson 1990, Cooper and Swanson mer in birds showing marked seasonal changes 1994). in cold tolerance (Hart 1962, Dawson and Smith For large birds (> 100 g), insulation contrib- 1986, Cooper and Swanson 1994). A winter in- utes to winter acclimatization to a greater degree crement of shivering endurance is associated than in small birds (West 1972, Stokkan 1992). with this improved thermogenic capacity and is Seasonal metabolic adjustments in large birds apparently the primary component of winter ac- are less well studied, but appear to be less prom- inent than in small birds, and large birds fail to ’ ’ Received29 July 1996. Accepted20 December show consistent seasonal trends in metabolism 1996. (Beth 1980, Sherfy and Pekins 1994), although

[4781 METABOLISM IN BOBWHITE 479

Weathers and Caccamise (1978) demonstrated where birds were maintained in unheatedoutdoor that seasonal changes in BMR are inversely re- shelters. Quail were transportedfrom hatcheries lated to body mass, with several birds over 200 to Vermillion, Clay County, South Dakota (42” g actually exhibiting winter decreasesin BMR 47’N latitude) and housed in covered, unheated relative to summer. Consequently, the degree to outdoor pens for at least two weeks prior to met- which seasonalmetabolic adjustmentscontribute abolic tests.Food (wild bird seed mix and rolled to acclimatization in larger birds is uncertain. corn) and water were provided ad libitum. Met- We measured seasonal changes in BMR, meta- abolic testswere conductedin late spring/summer bolic response to temperature, and maximal and winter from 1991-1996. For all seasonal thermogenic capacity to determine metabolic metabolic comparisons,the same cohort of birds and insulatory contributions to winter acclima- was tested in late spring/summerand winter, but tization in captive Northern Bobwhite (Colinus separatecohorts of birds were used for measure- virginianus), a relatively large (approx. 200-250 ment of diurnal metabolic responseto tempera- g) nonpasserine. ture and BMR, nocturnal metabolic responseto The Northern Bobwhite is predominantly a temperature, and maximal thermogenic capacity. speciesof the southern United States and north- Birds tested between 26 June and 11 September ern Mexico where winters typically are rather were considered “summer birds,” whereasthose mild (AOU 1983). However, the northern por- tested between 8 January and 20 February were tion of their range encompassesregions with rel- considered “winter birds.” For maximal ther- atively severe winters, including southeastern mogenic capacity, birds were tested during the South Dakota (SDOU 1991). In the northern winter season and from 20-25 May, the latter reaches of the range of the bobwhite, popula- tions fluctuate widely, and this fluctuation is cor- group referred to as “late spring” birds. Bob- related with severity of winter weather (Dins- white were tested without regard to sex because more et al. 1984, SDOU 1991). The role that mean massesof males and females did not differ temperature or other factors play in regulating significantly for any of the cohorts tested, but the northern boundary of bobwhite distribution male : female ratios for various metabolic param- is uncertain. The northern boundary of bobwhite eters were: BMR, 6:3 (summer), 5:3 (winter); distribution in the Great Plains is associatedwith nocturnal metabolic responseto temperature,7:6 the approximate northern limit of the Central (summer), 5:3 (winter); diurnal metabolic re- Great Plains Winter-Wheat and Rangeland Re- sponse to temperature, 6:3 (summer), 5:3 (win- source region (Root 1988a). However, the prox- ter); and maximal thermogenic capacity, 6:6 imate role of vegetation in limiting the range of (summer), 9:9 (winter). All birds were at least 8 the bobwhite is uncertain because dietary com- months old at the time of testing and were con- position varies geographically over the range of sidered adults. the bobwhite (Johnsgard 1973). The northern range boundary also is approximately coincident BMR AND METABOLIC RESPONSETO with the average minimum January temperature TEMPERATURE isotherm of -12°C (Root 1988a), which sug- Oxygen consumption (VO,) was measured by gests that temperature may contribute to regu- open-circuit respirometry with an Ametek S-3A lation of bobwhite distribution. Root (1988b) Oxygen Analyzer according to Swanson (1993). noted that average minimum January tempera- Prior to measurement of VO,, birds were tures at the northern range boundaries for a num- weighed to the nearest gram with a Fisher Model ber of passerine species produced resting meta- 2- 116 balance. Birds were then introduced into bolic rates of approximately 2.5 times BMR. a metabolic chamber fashioned from a 3.8 L Our measurementsof seasonal variation in met- paint can with the inner surface painted black to abolic and insulatory parameters in a climate at provide an emissivity near 1.0. The effective the northern range boundary of the bobwhite volume of the chamber, calculated according to help elucidate the role of temperature in limiting Bartholomew et al. (1981), was 3,774 ml. Tem- bobwhite distribution. perature within the chamber was controlled to 2 METHODS 0.5”C by immersion of the chamber into a bath BIRDS of water/ethylene glycol. Chamber temperature Bobwhite (wild strain) were obtained from hatch- was monitored continuously with a Cole-Parmer eries in South Dakota, Nebraska and Minnesota (Model 8500-40) thermocouple thermometer 480 DAVID L. SWANSON AND DUANE P. WEINACHT and copper-constantanthermocouple previously ygen (helox) was used as the respiratory gas in- calibrated to a thermometer traceable to the U. stead of air. Helox has a higher thermal conduc- S. Bureau of Standards. Flow of dry, CO,-free tivity than air and elicits maximal cold-induced air into the metabolic chamber was initiated pri- 00, at relatively moderate temperatures(Rosen- or to immersion, and flow rates over the test mann and Morrison 1974). Thermogenic capac- period were maintained from 610-755 ml min’ ity tests were conducted between lo:45 and 17: for BMR, and from 9151,425 ml min-l for 00 in both seasons. Birds were placed in the metabolic response to temperature. Flow rates metabolic chamber which was then flushed with were regulated with a Cole-Parmer precision ro- helox for 5 min at 1,590-1,810 ml min-’ before tameter (Model FM082-03ST) previously cali- immersion into the water/ethylene-glycol bath. brated to I 1% accuracy (Swanson 1990). These flow rates were used for all thermogenic Metabolic tests were conducted during both capacity tests and kept oxygen concentration in day (12:40-19:00 CST in summer, 11:20-18:55 excurrent air above 19.7%. Helox cold stress in winter) and night (21:00-03:40 in summer, tests involved exposing individual birds to a de- 19:45-01:15 in winter) to discern circadian dif- creasing series of temperaturesuntil VO, no lon- ferences in metabolism. For measurement of ger increased with a further decreasein temper- BMR, birds were exposed to a chamber temperature. Cold stresstests were initiated at - 12°C ature of 30°C. Individual birds were exposed to in winter and -3 to - 12°C in late spring. During a single temperature from - 10 to 15°C for mea- cold stress tests, chamber temperature was de- surement of metabolic responseto temperature; creased by 3°C at approximately 35 min, and metabolism at only one temperature within this again every 20-30 min thereafter if \;rO, had not range was measured per day for any given in- reached a plateau. For late spring birds, we ini- dividual. Over the course of a season,individual tially began helox cold stresstests at -3°C but birds were tested at no more than three different only two individuals became hypothermic even temperatures within this range, with at least six after prolonged exposure to decreasing helox days between measurements. BMR measure- temperatures even though \;lO, had reached a ments were sometimesobtained on the same day plateau. Therefore, we initiated helox cold stress as another temperature exposure (with at least for the final six individuals tested in late spring 3.4 hr between measurements) or 3-16 days at - 12”C, the same as for winter birds. At both from another temperature exposure. BMR did seasons,metabolic tests were conducted for 60- not differ significantly between these two treat- 90 min or until birds became hypothermic as ments so data were pooled. Nocturnal metabo- indicated by a steady decline in VO, over sev- lism was measured after at least a 5 hr fast to eral minutes. Fractional concentration of oxygen ensure postabsorptive conditions. Birds also in dry, CO,-free excurrent gas was measuredev- were fasted for at least 4 hr prior to daytime ery 60 set over the test period. We calculated measurements. Nocturnal measurements were \;rO, as instantaneous \iO, (Bartholomew et al. conducted for 1 hr, following a 1-hr equilibra- 1981). Mean \;rO, was calculated over succes- tion period after introduction to the chamber.For sive lo-mm intervals over the test period and diurnal metabolism, measurements were con- the highest IO-mm mean was designated\j02_,,,, ducted for 30 min, following a 30-min equili- with the initial 10 min excluded to allow excur- bration period. These equilibration periods were rent O2 concentration to stabilize (Dawson and sufficient to allow ir0, to reach steady-state Smith 1986). All values of iTO, were corrected conditions. The lowest lo-min mean ir0, over to STP the test period was designated ir0, at a partic- Upon removal from the metabolism chamber ular test temperature. Oxygen consumption was at the completion of metabolic tests in both air calculated according to steady-state methods and helox, cloaca1 temperature (20 gauge ther- (Hill 1972, equation 2). mocouple and thermocouple thermometer) and mass were measured. Bobwhite with a T,, < MAXIMAL THERMOGENIC CAPACITY 36.O”C (a value 3°C below mean diurnal T, in Methods for measuring maximal thermogenic winter, and I .7”C below mean diurnal T, in sum- capacity were similar to thosefor BMR and met- mer) were considered hypothermic. Mass loss abolic responseto temperature except that a gas was assumed to be constant throughout the test mixture of approximately 79% helium/21% ox- period. We calculated mass-specific metabolic METABOLISM IN BOBWHITE 481

rates using mass values corrected for mass loss SUMMER 4- over the test period. . Nocturnal 0 Daytime p ~02~=2.30-0.05(Ta) STATISTICS 'c3- 602~=2.60.0.05(Tp) Data are reported as means t SD. Mean values b 0 for BMR, maximal thermogenic capacity, body mass, T,, mass-specific thermal conductance, and time to hypothermia were compared by Stu- dent’s r-test, or by Mann-Whitney U-test if sample variances were unequal (determined by F- test). For maximal thermogenic capacity, we 0 ’ measured the same 12 individuals in both sum- -15 -5 25 35 T&erature (.'c", mer and winter and used paired two-tailed t-tests for comparison of VO, SUmand body mass in FIGURE 1. Relationship of nocturnal and diurnal mass-specificoxygen consumptionto ambient temper- these birds. Comparison of BMR between sexes ature in summer acclimatized bobwhite. The equations and seasonswas by Kruskal-Wallis test, because provided were derived by least squaresregression and variances were not homogenous among groups describeoxygen consumptionas a function of ambient (F-test). We used Fisher’s exact test to compare temperature below thermoneutrality for nocturnal (ire,,) and diurnal (7jOZo)birds. Slopes of nocturnal percentagesof birds becoming hypothermic dur- and diurnal equationswere not significantly different, ing helox cold stresstests in late spring and win- but diurnal intercept was significantly higher (P < ter. Regressions were performed by the method 0.001) than nocturnalintercept. of least squares,and regression lines were compared by ANCOVA. We calculated Pearson’s correlation coefficients for mass, VOXsum, and re- BMR was 3.45 2 0.78 ml O2 min’ (0.95 2 siduals of V02 sumvs. body mass in individual 0.13 ml 0, g-r m-l) in summer (n = 9) and 3.76 quail in late spring and winter to determine % 0.64 ml 0, min’ (1.01 -C 0.11 ml 0, g-’ hrl) whether individual VO, Sumwas consistent be- in winter (IZ = 8). These values were not signif- tween seasons. Statistical significance was ac- icantly different (t,, = 0.87 and 1.15 for total cepted at P 5 0.05. and mass-specific BMR, respectively). The re- RESULTS lationship between mass-specific metabolic rate (ml 0, g-l ln’) and ambient temperature (“C) Bobwhite used for seasonal measurement of below thermoneutrality (Figs. 1 and 2) is best BMR and metabolic response to temperature described by the following equations. consisted of two separate cohorts, the first was tested in summer 1991 and winter 1992 and the summer diurnal: VO, = 2.60 - O.O5(T,), n = second in summer 1992 and winter 1993. Mean 18 (9 birds), R2 = 0.76 mass did not vary significantly between cohorts winter diurnal: VO, = 2.59 - O.O6(T,), IZ = so data were pooled. Mean massesfor summer 18 (8 birds), R* = 0.76 birds (210 ? 38 g, at = 22) and winter birds summer nocturnal: VO, = 2.30 - O.O5(T,), n (228 +- 36 g, IZ = 16) were not significantly dif- = 16 (13 birds), R* = 0.78 ferent (t,, = 1.48). For measurement of maximal winter nocturnal: VO, = 2.12 - O.O5(T,), n thermogenic capacity (winter and late spring of = 18 (8 birds), R2 = 0.72 1996), 12 individuals were measured in both winter and late spring and body mass in winter ANCOVA indicated no significant seasonaldif- individuals (234 2 19 g, it = 12) was signifi- ferences in either slope or intercept for either cantly greater than mass in late spring individ- nocturnal or diurnal VO,. However, at both sea- uals (220 ? 19 g, n = 12; t,, = 3.76, P < 0.01). sons intercepts were significantly higher (P < An additional six individuals were tested for 0.001) during the day than at night and in winter maximal thermogenic capacity in winter, and if diurnal slope was significantly steeper (P < these birds are included in calculations of mean 0.05) than nocturnal slope. Lower Critical Tem- mass, winter mass becomes 231 2 19 g (n = perature (LCT) was calculated by determining 18) which is not significantly different from late the intersection of the line relating nocturnal spring mass (rz8 = 1.61, P = 0.12). VO, to T, below thermoneutrality with a line 482 DAVID L. SWANSON AND DUANE P. WEINACHT

conductance did not differ significantly between seasons(t34 = 0.20) but nocturnal conductance was significantly higher in summer than in winter (tj2 = 2.28, P < 0.05). For those birds where helox cold stresstests were initiated at -12°C (n = 18 in winter, n = 6 in late spring), 27.8% (5/18) were normothermic (Tb > 36°C) on removal from the metabolic chamber (after VO, had attained maximum levels) in winter (removed after 60-70

0 mm), whereas 16.7% (l/6) were normothermic -1 5 -5 25 35 T:mperature (0; in late spring (removed after 90 mm). These per- centages were not significantly different. How- FIGURE 2. Relationship of nocturnal and diurnal ever, mean time to hypothermia in those birds mass-specificoxygen consumptionto ambient temperature in winter acclimatized bobwhite. The equations becoming hypothermic was 62.4 -C 10.3 mm in provided were derived by least squaresregression and winter and 48.6 + 10.5 min in late spring. The describeoxygen consumptionas a function of ambient winter duration is significantly (t,, = 2.53, P = temperature below thermoneutrality for nocturnal 0.02) longer than the late spring duration, which (VO,,) and diurnal (VO,,) birds. Diurnal interceptwas significantly higher (P < 0.001) than nocturnal inter- suggeststhat winter birds are marginally more cept. In addition, diurnal slope was significantly steep- cold tolerant than late spring birds. er (P = 0.05) than nocturnal slope, suggestingim- Maximal thermogenic capacity in late spring proved insulation at night. Neither slopes nor inter- was 20.19 k 3.94 ml 0, mm’ (5.50 + 0.92 ml cepts differed on a seasonalbasis in either nocturnal 0, g-l hr’, n = 12). Winter maximal thermo- or diurnal birds. genie capacity was 22.02 +- 2.14 ml 0, mink (5.74 k 0.62 ml 0, g-l lu’, IZ = 18). These through BMR. Winter LCT was 22.4”C and values did not differ significantly on either summer LCT was 255°C. whole-animal (Mann-Whitney U12,18= 139) or Regressions of T, against T, following meta- mass-specific (tZ8 = 0.87) bases. Furthermore, bolic testsbelow LCT were not significantly dif- maximal thermogenic capacity of the 12 indi- ferent from zero slope in any of the four treat- viduals tested in both winter and late-spring (see ment groups. This indicates that T, is indepen- above) did not differ significantly (tl I = 1.42 and dent of T, below LCT over the range of tem- 0.31 for whole-animal and mass-specific com- peratures studied, so T, values were pooled. parisons, respectively) between seasons(winter Mean T,s of the four treatment groups were: thermogenic capacity = 21.65 k 1.82 ml 0, summer diurnal, 38.9 -t l.l”C (n = 18); winter n-tin-‘, 5.59 t 0.63 ml 0, g-l m-l, rz = 12). diurnal, 37.7 2 l.l”C (n = 18); summer noctur- Metabolic expansibility (VO, JBMR, Dawson nal, 37.0 + l.O”C (n = 16); and winter noctur- and Carey 1976) on a per-bird basis was 5.9 in nal, 37.4 + l.l”C (n = 18). Diurnal T, was sig- both late spring and winter. nificantly higher than nocturnal Tb in summer (tX2= 5.52, P < O.OOl), but not in winter (t,, = 0.93). Summer diurnal T, also was significantly TABLE 1. Diurnal and nocturnal mass-specificther- higher than winter diurnal T, (f34 = 3.23, P -=c mal conductance(C) in captive bobwhite acclimatized to summer and winter. The number of birds used to O.Ol), although nocturnal T, did not differ be- derive the sample size (n) is listed in the “Individual tween seasons (tx2 = 1.39). After BMR mea- birds” column. surement, summer T, was 39.2 ? 0.5”C (n = 9), which was significantly greater than winter T, Indi- vidual (37.6 2 0.8”C, n = 8; t,, = 5.21, P < 0.001). Treatment n birds C (ml 0, s-’ hr-’ “C-‘, Mass-specific thermal conductance (C) was Summer diurnal 18 9 0.070 t 0.009” calculated as C = \jO,/(T, - T,) for metabolic Winter diurnal 18 8 0.069 t 0.008b tests below thermoneutrality (Table 1). Conduc- Summer nocturnal 16 13 0.064 2 0.008 tance was significantly greater during the day Winter nocturnal 18 8 0.058 + 0.007 than at night in both summer (t32 = 2.01, P = 1 Significantly greater than same-season nocturnal C at P = 0.05. h Sigmficantly greater than same-season nocturnal C at P < 0 001. 0.05) and winter (t34= 4.53, P < 0.001). Diurnal ‘Significantly greater than opposite-season nocturnal C at P < 0.05. METABOLISM IN BOBWHITE 483

TABLE 2. Seasonal means (? SD) for BMR and maximal thermogenic capacity for sex classes. Seasonal comparisonswithin sex classeswere by paired t-test, whereas within-season sexual comparisonswere by independent Student’s t-test.

n Mass (~1 BMR (ml 0, min-0 n Mass (g) “O?\“, (ml 0, mm-l) Winter Male 5 231.2 2 41.5 3.89 -c 0.78 6 240.3 5 9.6 21.65 ? 1.86 Female 3 210.1 -c 3.53 3.53 2 0.29 6 227.1 k 24.3 21.65 + 1.95 Late spring/summer Male 6 234.3 i 44.7 3.64 5 0.90 6 228.0 2 17.9 23.08 -+ 2.10 Female 3 189.7 -c 7.4 3.08 5 0.33 6 211.9 2 17.2’ 17.30 r+_3.13b

1 Significantly less (P C 0.05) than mass of wmter females. n Significantly less (P < 0.01) than values for both spring males and winter females.

Although sample sizes for within and between DISCUSSION sex comparisonswere relatively small, for indi- Many large birds, including several Galliformes, viduals measured at both seasons, late spring possessvery effective winter insulation so that males, but not winter males, exhibited signifi- with selection of favorable microclimates (e.g., cantly greater thermogenic capacity than same- snow burrows) thermoregulatory costs are min- seasonfemales on both whole-animal (t,, = 3.76, imal (Marjakangas et al. 1984, Stokkan 1992). P = 0.004) and mass-specific (t,, = 2.98, P = Some of these speciesshow winter improvement 0.014) bases (Table 2). Furthermore, VO, sumin of insulation relative to summer, but in others winter females was significantly greater than that insulation remains seasonally constant (West for late spring females (t,, = 3.57, P < 0.02). 1972, Sherfy and Pekins 1994). For Mute Swans There were no sexual or seasonaldifferences in (Cygnus olor, Beth 1980) and Black Grouse BMR, although again sample sizes were relative- (Tetruo tetrix, Rintimaki et al. 1983), winter ly small (Table 2). conductance exceeds summer conductance at Because maximal thermogenic capacity was measured in 12 individuals in both late spring moderate temperatures.However, winter birds in and winter, we tested whether metabolic rates of both speciesreduce conductanceas T, decreases, individual bobwhite were correlated between whereas conductance in summer birds remains seasons. Whole-animal maximal thermogenic essentially constant with T,. This results in sea- capacity in individual bobwhite was not signif- sonally similar conductance, or lower conduc- icantly correlated between seasons(r = 0.43, P tance in winter, at low temperatures.Basal met- = 0.16), although the trend was in the expected abolic rate (BMR) does not vary consistently on direction (Fig. 3a). However, individual body a seasonal basis in large birds, and winter in- mass was significantly correlated between sea- creases(Beth 1980, Rintimaki et al. 1983, Sher- sons (r = 0.77, P = 0.003, Fig. 3b). We cor- fy and Pekins 1994), seasonal stability rected for possible confounding effects of body (Schwann and Williams 1978) and winter de- mass on thermogenic capacity in late spring creases (Hart 1962, West 1972, Mortensen and birds by plotting log mass against log \;rO, sumr Blix 1986) in BMR relative to summer have all and calculating residuals from the regression been reported. Maximal capacity for heat pro- equation to remove effects of body mass. Winter duction is greater in winter than in summer (13- birds showed no significant relationship between 22%) for pigeons, Columba Zivia (Hart 1962), log mass and log 00, sum,so winter residuals but is unstudied in other large birds. Thus, it were calculated by subtracting mean 00, sum appears that metabolic adjustment occurs to (rather than allometrically predicted \;rO, ,,,) some degree in large birds, but effective cold from measured \;rO, SUm.Residuals of 00, sum resistance in these birds is primarily conferred were not significantly correlated between sea- by effective insulation. This pattern of heavy re- sons (r = 0.15, P = 0.65) indicating that max- liance on insulation for winter acclimatization imal thermogenic capacity in individual bob- contrastsmarkedly with that of small birds that white between seasons(Fig. 3c) was not consis- show only minor seasonalvariation in insulation tent, relative to mass-adjustedmean \jO, SU,,,. and must elevate heat production for thermoreg- 484 DAVID L. SWANSON AND DUANE P. WEINACHT

28 r poorer at both seasons than predicted allome- a . g 26 &ally. Lower critical temperaturesin bobwhite were relatively high and varied little with season 2. . . l . (22.4”C in winter, 25.5”C in summer). These 1 22 . high LCTs are similar to thosereported for many small birds (Weathers and van Riper 1982). Al- lometric equations for LCT for nonpasserine birds, similar to those calculated for passerine birds by Weathers and van Riper (1982), can be generated from allometric equations for BMR (Aschoff and Pohl 1970) and conductance (Aschoff 198 1) in nonpasserines.The allometric equation thus calculated for nonpasserinesis: LCT = T, - 4.24(Mass)0.317 3 u) 235 where LCT and Tb are in “C and Mass is in I P grams. LCT in winter bobwhite is 8.7”C higher . . p220 l . than predicted by this equation, and summer B .‘ bobwhite LCT is 111X higher than predicted. + 205 . Furthermore, although diurnal conductance values for bobwhite are similar to allometric pre- dictions (Aschoff 1981), nocturnal conductance .._ 175 190 205 220 235 250 265 is 45.0% and 52.4% higher than predicted allo- WinterMass (e) metrically in winter and summer, respectively. Summer nocturnal conductance is significantly (10.3%) higher than winter nocturnal conductance, suggesting some improvement in insulation in winter, possibly resulting from the grad- ual season-long molt that summer birds under- went from mid-June through August (pers. ob- serv.), but metabolism must be elevated above basal levels for thermoregulation even at moderate temperatures in both seasons.In addition, r - 0.15 regressionsof metabolism on T, below thermoneutrality do not differ between seasons,which -3.0 d.5 1.5 3.0 WinteriOs,9;9 Residuals suggestslittle winter improvement of insulation. These data indicate that bobwhite are relatively FIGURE 3. Correlations between seasons for (a) poorly insulated at both seasons and must ex- vo* sumr(b) mass, and (c) mass-independentresiduals pend substantial amounts of energy on thermo- of VO, sumin individual bobwhite. Only the mass correlation was significant (P = 0.003), which indicates regulation at cold temperatures. This is a situa- that 7j02 sumin individual bobwhite was not repeatable tion at odds with most other larger Galliformes. between seasons. An alternative explanation for the relatively poor insulation of bobwhite in this study is that captivity resulted in poor plumage condition and ulation even at moderate temperatures. This in- consequentpoor insulation. However, our values dicates that winter acclimatization is primarily a for LCT are similar to those for other relatively metabolic process in small birds (see Marsh and small Galliformes (Brush 1965, Roberts and Dawson 1989 for review). Baudinette 1986), although these studiesalso in- Because the Northern Bobwhite is a relatively volved captive birds. Furthermore, captivity large bird, it might be expected that effective does not preclude seasonal changes in LCT or insulation could obviate the need for marked thermal conductance in other captive Gallifor- seasonal metabolic adjustment. However, this mes (West 1972, Mortensen and Blix 1986, study indicates that insulation in bobwhite is Sherfy and Pekins 1994). Thus, although con- METABOLISM IN BOBWHITE 485 ductance may be artificially high due to poor million, Clay County, South Dakota are -8 and plumage, the minor seasonaldifferences in LCT 24°C respectively (United States National Oce- and thermal conductance in this study suggest anic and Atmospheric Association records). that plumage changes are relatively unimportant Metabolic rates at these temperatures represent to seasonal acclimatization in bobwhite. factorial increments of BMR of 3.0 in winter and Thermoregulatory costs may be reduced by 1.5 in summer for diurnal metabolism, and 2.5 decreasing body temperature, thereby reducing in winter and 1.2 in summer for nocturnal me- the gradient for heat loss to the environment. tabolism. These temperatureselicit thermoregu- Winter, relative to summer, bobwhites exhibited latory costs of 165.9 kJ day-’ in January and significantly lower diurnal, but not nocturnal, T, 15.2 kJ day-r in July. This represents a 10.9- after metabolic tests below thermoneutrality. fold increase in thermoregulatory costsin winter This may provide some energetic benefit for relative to summer and indicates that winter ac- winter birds, but it is of only minor importance climatization in bobwhite must be directed pri- in reducing thermoregulatory costs. For exam- marily toward maintenance of elevated metabol- ple, the 1.2”C difference in mean diurnal T, be- ic rates for prolonged periods. tween winter and summer bobwhite, given sea- Case and Robe1 (1974) provided measure- sonally stable thermal conductance(Table l), re- ments of daily existence energy (EE) for captive sults in an energetic savings at 0°C of only 3% bobwhite held under 1OL:14D photoperiod at O- in winter birds, and this percentage becomes 40°C. In these experiments, bobwhite were con- even smaller at lower ambient temperatures. fined individually in small (48 X 25 X 13 cm) Body mass did not vary significantly between cages over several months. Thus, Case and Ro- seasons within cohorts of bobwhite tested for bel’s (1974) values for EE undoubtedly under- BMR and metabolic response to temperature. estimate actual daily energy expenditure for Body mass also did not differ significantly on a free-living bobwhite. Nevertheless, these values seasonalbasis among cohortstested for maximal should provide a basis for examining the contri- thermogenic capacity. However, in individuals bution of thermoregulatory coststo daily energy tested for maximal thermogenic capacity in both expenditure in winter bobwhite. Case and Robe1 late spring and winter, body mass was signifi- (1974) report a linear relationship between EE cantly greater in winter, but only by 6.2%. Al- and T,, with an EE at 0°C of 205.5 kJ day-‘. though stored fat was not measured directly in Because the relationship between EE and T, was this study, the seasonal stability of body mass linear, EE at the average daily January temper- suggests little difference in stored fat between ature for Vermillion (-8°C) can be calculated seasons. Seasonal stability of fat reserves is a from their equations as 233 kJ day-r. Our esti- condition common to a number of large Galli- mates of winter thermoregulatory costs for bob- formes (Thomas 1982, Mortensen et al. 1985). white comprise 59.4% of EE at 0°C and 70.8% Winter enhancement of storedenergy reserves of EE at -8°C. This indicates that a substantial and insulation apparently are not important com- fraction of winter EE must be devoted to ther- ponents of winter acclimatization in Northern mogenesisin these birds. In fact, if BMR is add- Bobwhite. This suggeststhat metabolic adjust- ed to thermoregulatory costs estimated in our ments may contribute to winter acclimatization studies, metabolic costs exceed EE measure- to a greater extent than insulative adjustments, a ments. Of course, selection of favorable micro- situation similar to small birds. However, basal climates and huddling in coveys (Case 1973) metabolic rate did not vary seasonally in this may reduce thermoregulatory costs, but ther- study. At both seasons,BMR was close to al- mogenesis still comprises a large fraction of lometrically predicted values (Aschoff and Pohl winter energy expenditures. This situation dif- 1970), exceeding predicted BMR by only 1.8 fers from that in a number of larger Galliformes, and 4.4% in summer and winter, respectively. where winter thermoregulatory costs represent From the equations describing VO, as a function only minor fractions of daily energy expendi- of T, below thermoneutrality, crude estimatesof tures (Marjakangas et al. 1984, Stokkan 1992). thermoregulatory costs can be derived (disre- Nevertheless, like BMR and metabolic re- garding radiative and convective effects and heat sponseto temperature,maximal capacity for ther- produced as a by-product of activity). Mean dai- mogenesis did not vary seasonally in bobwhite. ly temperatures for January and July for Ver- Winter VO, sumwas 4.6% below allometric pre- 486 DAVID L. SWANSON AND DUANE P. WEINACHT dictions (Hinds et al. 1993), whereas VO, sumin white’s range, metabolic rates associated with summer was 9.8% below predicted values. The thermoregulation at average minimum January allometric equation of Hinds et al. (1993) was temperaturesfor this region should exceed BMR derived from eight species during spring and by about 2.5 times if Root’s (1988b) findings summer. Thus, there appearsto be little seasonal also extend to nonpasserines. Given that mean metabolic adjustment to cold in Northern Bob- minimum January temperature for Vermillion, white, and thermogenic capacity is not elevated South Dakota is -lo”C, our winter nocturnal above that for other speciesacclimated to warmer equation describing VO, as a function of T, in- periods of the year Furthermore, winter birds dicates a metabolic rate for winter bobwhite of were only marginally more cold tolerant in helox 2.59 times BMR. Thus, our data are consistent cold stresstests that late spring birds. This sug- with the broad-scale relationship between cli- geststhat the lack of seasonaladjustment in me- mate and distribution described by Root tabolism and the relatively poor insulation at both (1988b), which suggeststhat temperature does seasons preclude marked seasonal variation in influence the northern distribution of the North- cold tolerance in bobwhite. em Bobwhite. These data suggestthat winter cli- Metabolic expansibility (VO, ,,,/BMR) in matic conditions in southeasternSouth Dakota birds tested to date ranges from 3.3 (Saarela et may be metabolically challenging for bobwhite, al. 1989) to 8.1 (Dutenhoffer and Swanson and underscore the importance of food avail- 1996). Our values for metabolic expansibility in ability and behavioral thermoregulation to win- bobwhite (5.9 at both seasons)fall well within ter survival of Northern Bobwhite in northern this range. Rough estimates of equivalent air sections of their range. temperatures at helox temperatures eliciting Thermogenic capacity of males and females VO, sumcan be generated by entering values for did not differ in winter, but females showed re- 00, S”rninto equations describing diurnal metab- duced thermogenic capacity in late spring. olism as a function of T, below thermoneutrality Spring thermogenic capacity was measured in and solving for T,. This yields estimated air tem- late May, which correspondsto the egg-laying peratures at VO, sum of -58°C in late period for females. Perhaps the energetic costs spring/summer and -53°C in winter. These tem- of egg production interfere with thermogenic peratures are below actual ambient air temper- performance in female bobwhite. Loss of total atures encountered in the natural environment at body and pectoralis muscle mass associatedwith both seasons, which suggests a metabolic re- egg production in female birds have been re- serve available for acute energy expenditures at ported (Houston et al. 1995), and could influ- both seasons,although convective heat loss cou- ence thermogenic performance (Swanson 1991b, pled with extreme cold temperatures may ap- O’Connor 1995). Indeed, late spring females in proach conditions eliciting VO, sumin winter this study were significantly smaller (7.1%) than bobwhite. winter females (Table 2). For late spring females Over longer periods (e.g., the entire winter), with cold exposure tests initiated at -12°C in however, metabolic rates generally cannot be helox (n = 4), mean time to hypothermia was sustained at levels approaching 00, SUm.Sus- 44.5 + 5.8 min. Two males were subjected to tained metabolic scope in vertebrates (in facto- cold exposure tests initiated at -12°C in helox rial increments of BMR) usually ranges from 1.5 in late spring; one became hypothermic after 65 to 5 (Peterson et al. 1990). The highest sustained min and one remained normothermic for the en- metabolic scope recorded to date is 7.2 for lac- tire 90 min test period. This suggeststhat dif- tating mice (Hammond and Diamond 1992), al- ferences in thermogenic capacity are related to though sustained metabolic scope for cold- variation in cold tolerance in bobwhite. stressedmice was only 4.7 (Konarzewski and Finally, neither VO, sumnor mass-independent Diamond 1994). Furthermore, Root (1988b) residuals of VO, bumwere correlated between found that the northern range boundaries for a winter and late spring, suggesting that maximal number of passerine species were correlated thermogenic capacity in individual quail is not with average minimum January temperatures consistent between seasons.This contrastswith producing metabolic rates of about 2.5 times short-term studies (days to weeks) indicating rel- BMR. Because southeasternSouth Dakota rep- atively high repeatability of maximal oxygen resents the northern extent of the Northern Bob- consumption in individual skinks (Chalcides METABOLISM IN BOBWHITE 487 ocellatus, Pough and Andrews 1984), garter versity of South Dakota Institutional Animal Care and snakes (Thamnophis sirtalis, Garland and Ben- Use Committee. nett 1990), deer mice (Peromyscus maniculatus, LITERATURE CITED Hayes and Chappell 1990), and Belding’s ground squirrels (Spermophilus beldingi, Chap- AMERICANORNITHOLOGISTS ’ UNION. 1983. Checklist of North American birds, 6th ed. American Or- pell et al. 1995). Furthermore, individual maxi- nithologists’ Union, Washington,DC. mal \jO, (both exercise- and cold-induced) in ASCHOFF, J. 1981. Thermal conductancein mammals deer mice was repeatable over 3-month accli- and birds: its dependenceon body size and cir- mation periods, including acclimation to cold cadianphase. Comp. Biochem. Physiol. 69A:61 l- (Hayes and Chappell 1990). However, individual 619. ASCHOFF, J., AND H. POHL. 1970. Der Ruheumsatzvon exercise-induced maximal 00, was only mar- Vogeln als Funktion der Tageszeit und der Kor- ginally consistent, and maximal cold-induced pergrope. J. Ornithol. 111:38-47. VO, was not consistent over l-2 years in free- BARTHOLOMEW,G. A., D. VLECK, AND C. M. VLECK. living ground squirrels (Chappell et al. 1995). 1981. Instantaneous measurements of oxygen consumptionduring pre-flight warm-up and post- Studies of the ecology and evolution of aer- flight cooling in Sphingid and Saturniid moths. J. obic performance assume repeatability of aero- exp. Biol. 90: 17-32. bic performance over time. Our data suggestthat BECH,C. 1980. Body temperature,metabolic rate, and such an assumption may not be justified over insulation in winter and summer acclimatized long periods or changing seasons, at least for Mute Swans (Cygnus odor). J. Comp. Physiol. 136:61-66. VO, sumin bobwhite. The absence of a correla- BRUSH,A. H. 1965. Energetics, temperatureregula- tion between seasonsin individual \;rO, sumin this tion and circulation in resting, active and defeath- study might be explained, in part, by plasticity ered California Quail, Lophorfyx californicus. of winter VO, sumin responseto short-term vari- Comp. Biochem. Physiol. 15:399-421. CASE, R. M. 1973. Bioenergeticsof a covey of bob- ation in winter temperature. Such plasticity has whites. Wilson Bull. 8552-59. been suggested for winter-acclimatized passer- CASE, R. M., AND R. J. ROBEL.1974. Bioenergeticsof ines (Dutenhoffer and Swanson 1996, Swanson, the Bobwhite. J. Wildl. Manage. 38:638-652. unpubl. data). In addition, late spring bobwhite CHAPPELL,M. A., G. C. BACHMAN,AND J. l? ODELL. exhibited a greater range of variance in ir0, sum 1995. Repeatability of maximal aerobic performance in Belding’s ground squirrels,Spermophi- (-30.7 to +33.1%, measured as percent devia- lus beldingi. Func. Ecol. 9:498-504. tion from mass-adjustedmean Vo, ,,,) than win- COOPER, S. J., AND D. L. SWANSON.1994. Seasonal ter birds (- 12.8 to + 10.2%), and this could in- acclimatization of thermoregulationin the Black- fluence seasonalcorrelations of individual \;rO, SUm. capped Chickadee. Condor 36:638-646. DAWSON,W. R.. AND C. CAREY. 1976. Seasonal ac- Thus, metabolic plasticity related to variation in climatization to temperaturein Carduelinefinches. winter temperature or to exigencies of breeding I. Insulative and metabolic adjustments.J. Comp. and nesting may mask season-to-seasoncorre- Physiol. 112:317-333. lations of individual VO, SUm.Alternatively, max- DAWSON,W. R., AND B. K. SMITH. 1986. Metabolic imal thermogenic capacity in individual bob- acclimatization in the American Goldfinch (Car- duelis tristisj. v. 424-434. In H. C. Heller, X. J. white may not be repeatable over long periods. Musacchia and L. C. H. Wang [eds.], Living in This latter possibility merits further investiga- the cold: physiological and biochemical adaptation, and a better test of long-term metabolic re- tions. Elsevier, New York. peatability without the confounding effects of DAWSON,W. R., R. L. MARSH,W. A. BUMMER, AND C. CAREY. 1983a. Seasonaland geographicvari- seasonand breeding status would be to measure ation of cold resistancein House Finches Curpo- thermogenic capacity from one winter to the dacusmexicanus. Physiol. Zool. 56:353-369. next. DAWSON,W. R., R. L. MARSH, AND M. E. YACOE. 1983b. Metabolic adjustmentsof small passerine ACKNOWLEDGMENTS birds for migration and cold. Am. J. Physiol. 245: R755-R767. We thank Michael Weinacht for equipment used to DINSMORE,J. J., T H. KENT, D. KOENIG,l? C. PETER- transportand housebobwhite. We also thank Eric Lik- SEN, AND D. M. ROOSA.1984. Iowa birds. Iowa nes and reviewers for providing valuable commentson State Univ. Press, Ames, IA. earlier versions of this manuscript. This study was DUTENHOFFER,M. S., AND D. L. SWANSON.1996. Re- funded by grants from the American PhilosophicalSo- lationship of basal to summit metabolic rate in ciety and the University of South Dakota Office of passerinebirds and the aerobic capacitymodel for Researchto DLS. Experiments and housingconditions the evolution of endothermy. Physiol. Zool. 69: for bobwhite in this study were approved by the Uni- 1232-1254. 488 DAVID L. SWANSON AND DUANE P. WEINACHT

GARLAND, T., AND A. E BENNETT.1990. Quantitative seasonalacclimatization. J. Comp. Physiol. 165: geneticsof maximal oxygen consumptionin a gar- 298-305. ter snake. Am. J. Physiol. 259:R986-R992. PETERSON,C. C., K. A. NAGY, AND J. M. DIAMOND. HAMMOND,K. A., AND J. M. DIAMOND.1992. An ex- 1990. Sustained metabolic scope. Proc. Natl. perimental test for a ceiling on sustainedmeta- Acad. Sci. 87:2324-2328. bolic rate in lactating mice. Physiol. Zool. 65: POUGH,E H., AND R. M. ANDREWS.1984. Individual 952-977. and sibling-group variation in metabolism of liz- HART, J. S. 1962. Seasonal acclimatization in four ards: the aerobic capacity model for the origin of species of small wild birds. Physiol. Zool. 35: endothermy. Comp. Biochem. Physiol. 79A:415- 224-236. 419. HAYES,J. I?, AND M. A. CHAPPELL.1990. Individual RINTIMAKI,H., S. SAARELA,A. MARJAKANGAS,AND R. consistency of maximal oxygen consumption in HISSA. 1983. Summer and winter temperature deer mice. Func. Ecol. 4:495-503. regulation in the Black Grouse, Lyrurus tetrix. HILL, R. W. 1972. Determination of oxygen consump- Physiol. Zool. 56: 152-159. tion by use of the paramagneticoxygen analyzer. ROBERTS,J. R., AND J. V. BAUDINET~E.1986. Ther- J. Appl. Physiol. 33:261-263. moregulation, oxygen consumption and water HINDS, D. S., R. V. BAUDINETTE,R. E. MACMILLEN, turnover in Stubble Quail, Coturnix pectoralis, AND E. A. HALPERN.1993. Maximum metabolism and King Quail, Coturnix chinensis. Aust. J. Zool. and the aerobic factorial scope of endotherms.J. 34:25-33. exp. Biol. 182:41-56. ROOT,T L. 1988a. Atlas of wintering North American HOUSTON,D. C., D. DONNAN,I? JONES,I. HAMILTON, birds. Univ. Chicago Press, Chicago. AND D. OSBORNE.1995. Changes in muscle con- ROOT,T L. 1988b. Energy constraintson avian dis- dition of female Zebra Finches Poephilu guttutu tributions and abundances.Ecology 69:330-339. during egg laying and the role of protein storage ROSENMANN,M., AND I? MORRISON.1974. Maximum in bird skeletal muscle. Ibis 137:322-328. oxygen consumptionand heat loss facilitation in JOHNSGARD,P A. 1973. Grouse and quails of North small homeotherms by He-O,. Am. J. Physiol. America. Univ. Nebraska Press, Lincoln, NE. 226:490-495. KONARZEWSKI,M., AND J. M. DIAMOND. 1994. Peak SAARELA,S., B. KLAPPER,AND G. HELDMAIER.1989. sustainedmetabolic rate and its individual varia- Thermogenic capacity of greenfinchesand siskins tion in cold-stressed mice. Physiol. Zool. 67: in winter and summer,p. 115-122. In C. Beth and 1186-1212. R. E. Reinertsen [eds.], Physiology of cold adap- LIKNES,E. T., AND D. L. SWANSON. 1996. Seasonal tation in birds. Plenum Press, New York. variation in cold tolerance, basal metabolic rate, SCHWAN,M. W., AND D. D. WILLIAMS.1978. Temper- and maximal capacity for thermogenesisin White- ature regulation in the Common Raven of interior breastedNuthatches Sitta curolinensis and Downy Alaska. Comp. Biochem. Physiol. 60A:3 l-36. Woodpeckers Picoides pubescens, two unrelated SHERFY,M. H., ANDI? J. PEKINS.1994. The influence arboreal temperate residents. J. Avian Biol. 27: of season, temperature and absorptive state on 279-288. Sage Grouse metabolism. Can. J. Zool. 72:898- MARJAKANGAS,A., H. RINTIMAKI,AND R. HISSA. 1984. 903. Thermal responsesin the Capercaillie Tetruo uro- SOUTHDAKOTA ORNITHOLOGISTS UNION.’ 1991. The gallus and the Black Grouse Lyrurus tetrix roost- birds of South Dakota, 2nd ed. Northern State ing in the snow. Physiol. Zool. 57:99-104. Univ. Press, Aberdeen, SD. MARSH,R. L., AND W. R. DAWSON.1989. Avian ad- STOKKAN,K. A. 1992. Energetics and adaptation to justments to cold, p. 205-253. In L. C. H. Wang cold in ptarmigan in winter. Omis Stand. 23:366- [ed.], Advances in comparativeand environmental 370. physiology 4: animal adaptationto cold. Springer- SWANSON,D. L. 1990. Seasonalvariation in cold har- Verlag, Berlin. diness and peak rates of cold-induced thermoge- MARSH,R. L., W. R. DAWSON,J. J. CAMILLIERE,AND nesis in the Dark-eyed Junco (Bunco hyemalis). J. M. OLSON. 1990. Regulation of glycolysis in Auk 107:561-566. the pectoralis muscles of seasonallyacclimatized SWANSON,D. L. 1991a. Seasonaladjustments in me- American Goldfinches exposed to cold. Am. J. tabolism and insulation in the Dark-eyed Junco. Physiol. 258:R71l-R717. Condor 93:538-545. MORTENSEN,A., AND A. S. BLIX. 1986. Seasonal SWANSON, D. L. 1991b. Substratemetabolism under changes in resting metabolic rate and mass-spe- cold stress in seasonally acclimatized Dark-eyed cific conductancein Svalbard Ptarmigan, Norwe- Juncos.Physiol. Zool. 64:1578-1592. gian Rock Ptarmiganand Norwegian Willow Ptar- SWANSON,D. L. 1993. Cold toleranceand thermogen- migan. Omis Stand. 17:8-13. ic capacity in Dark-eyed Juncos in winter: geo- MORTENSEN,A., E. S. NORDOY,AND A. S. BLIX. 1985. graphic variation and comparisonwith American Seasonalchanges in the body composition of the Tree Sparrows.J. Therm. Biol. 18:275-281. Norwegian Rock PtarmiganLagopus mutus. Omis THOMAS,V. G. 1982. Energetic reserves of Hudson Stand. 16:25-28. Bay Willow Ptarmigan during winter and spring. O’CONNOR,T. l? 1995. Metabolic characteristicsand Can. J. Zool. 60:1618-1623. body composition in House Finches: effects of WEATHERS,W. W. 1980. Seasonal and geographic METABOLISM IN BOBWHITE 489

variation in avian standardmetabolic rate. Proc. eycreepers: the Palila (Psitrirostra bailhi) and Int. Ornithol. Congr. 17:283-286. the Laysan Finch (Psittirostra canruns).Auk 99: WEATHERS,W. W., AND D. R. CACCAMISE.1978. Sea- 667-674. sonal acclimatizationto temperaturein Monk Par- WEST, G. C. 1972. Seasonal differences in resting akeets. Oecologia 35:173-183. metabolic rate of Alaskan ptarmigan.Comp. Bio- WEATHERS,W. W., AND C. VAN RIPER. 1982. Temper- them. Physiol. 42A:867-876. ature regulation in two endangeredHawaiian hon-