The Condor 102:635-644 0 The Cooper Omithologxal Society 2WO

SEASONAL ENERGETICS OF MOUNTAIN CHICKADEES AND JUNIPER TITMICE ’

SHELDON J. COOPERS Department of Biology, Utah State University,Logan, UT 84322-5305

Abstract. I used behavioral,meteorological, and laboratorymetabolism data to calculate daily energy expenditure (DEE) in seasonallyacclimatized Mountain Chickadees (Poecile gambeli) and Juniper Titmice (Baeolophusgriseus). Analyses of laboratory metabolic data revealed that foraging energy requirementswere not significantly higher than alert perching energy requirements. Respective DEE of chickadees and titmice were 48.8 kJ day-’ and 48.3 kJ day-’ in summer and 66.3 kJ day-’ and 98.7 kJ day-’ in winter. DEE as a multiple of basal metabolic rate (BMR) was 2.31 in summer chickadeesand 1.91 in summer titmice. DEE was 2.70 times BMR in winter chickadeesand 3.43 times BMR in winter titmice. The marked increase in calculated DEE in winter birds compared to summer is in contrastto a pattern of increasedDEE in the breeding seasonfor severalavian species.These data suggest that winter may be a period of even greater stringency for small birds than previously believed. Key words: Baeolophusgriseus, daily energy expenditure,energy metabolism,Juniper Titmouse,Mountain Chickadee,Poecile gambeli.

INTRODUCTION stantial increase in adult energy demand and Small passerine birds that overwinter in cold subsequently, DEE is highest during breeding. temperate regions require prolonged energy ex- DEE during the breeding seasontypically equals penditure for regulatory thermogenesis. In ad- or exceeds that during winter in birds that have dition, the onset of winter decreases foraging been studied. Although energetic demands may time due to shorter days and may reduce the not be higher in winter than during other periods availability of foraging substratesdue to heavy of the year, the conditions in which they must snow or ice cover. Concurrently with these sea- be met are much harsher. sonal changes in photoperiod and climate, cold Root (1988b) calculated the resting metabolic temperate-wintering passerinesundergo season- rate of 14 species whose metabolism as a func- al acclimatization that facilitates thermoregula- tion of ambient temperature was available from tory homeostasis. Previous studies of seasonal the literature at the minimum January tempera- acclimatization in passerine birds have focused ture at each species’ northern range boundary. primarily on seasonal variation in basal metab- That northern boundary metabolic rate (NBMR), olism, cold tolerance, maximal thermogenic ca- which includes basal metabolism (BMR) and pacity, and substrate metabolism (Dawson and thermoregulatory metabolism, is equal to 2.45 Marsh 1989, Marsh and Dawson 1989a, 1989b). times the BMR for each of the 14 species. The These studieshave generally collected metabolic total DEE of those birds must be somewhat data for individuals over a very short time pe- greater that 2.45 times basal, because the birds riod (up to a few hours). must also expend energy for foraging, digestion, Seasonal patterns observed thus far in avian and other activities. Thus, birds may be limited DEE support two alternative hypotheses (Mas- to overwintering in regions where they do not man et al. 1986, Weathers and Sullivan 1993). have to raise their DEE beyond slightly greater The “reallocation hypothesis” predicts little sea- than 2.45 times basal levels. However, this pro- sonal variation in DEE. The “increased demand posed limit of 2.45 has been criticized by Castro hypothesis” holds that breeding results in a sub- (1989) and Repasky (1991). In order to deter- mine the role of DEE on biogeographic patterns in birds, closely related species with different 1Received 8 September 1999. Accepted 7 April northern range distributions need to be exam- 2000. 2 Current address: Department of Biology, Univer- ined. sity of Wisconsin-Stevens Point, Stevens Point, WI The Mountain Chickadee (Poecile gambeli) 54481-3897, e-mail: [email protected] and the Juniper Titmouse (Baeolophus grisems)

16351 636 SHELDON J. COOPER

are small, largely nonmigratory passerine birds ed singing and grooming. I cannot be certain that occupy regions of western North America. that each of my time-budget samples for Moun- The northern limits of the distributions of Moun- tain Chickadees within one season was of a dif- tain Chickadees and Jumper Titmice lie in north- ferent individual because not all birds observed em British Columbia (6O”N latitude) and south- were banded. However, I made a conscious ef- em Idaho (44”N latitude), respectively (Godfrey fort to avoid sampling the same individual twice 1986, Cicero 1996). These limits coincide with within a season and to sample as many individ- -23°C and - 12°C January average minimum uals within a 3.2 km2 area per study site. In ad- isotherms, respectively (Root 1988a). Thus, the dition, Juniper Titmice adults remain in pairs DEE of these two species may be important in year-round and also maintain year-round terri- determining their northern range distribution. In tories (Dixon 1949). Thus, I was able to observe this study I compare the DEE of seasonally ac- both banded and unbanded pairs within their climatized Mountain Chickadees and Juniper own territories for relatively long periods of Titmice. I postulated that DEE in seasonally ac- time. climatized birds would not be stable due to in- creased thermostatic costs in winter and that METEOROLOGY DEE for titmice would be near or below 2.45 X Concurrent with my time-budget measurements, BMR. I monitored the birds’ diurnal thermal environ- ment with a meteorological station. Microcli- METHODS mate sensorswere mounted on metal poles and STUDY SITE AND SPECIES were placed 2 m above ground level (snow level The field portions of this study took place be- in winter) within 25 cm of a tree trunk. The 2- tween 5-8 February, 1996 for winter measure- m height was chosen as typical foraging/perch- ments, and between 31 July and 3 August, 1996 ing sites from field observations in 1993 to for summer measurements.Field data for Moun- 1996. This is especially important for wind tain Chickadees were recorded in the Bear River speed measurement because wind profiles vary Mountains, Cache County, Utah (41”52’N, with height measured from the ground (Camp- 11 l”34’W) at an elevation of 2,200 m. The study bell and Norman 1998). For Mountain Chicka- site consisted of mixed conifers and quaking as- dees, I placed the meteorological station near pen (Populus tremuloides). Field data for Juni- subalpine fir (Abies Zusiocarpa),and for Juniper per Titmice were recorded in the Raft River Titmice, I placed the meteorological station near Mountains, Box Elder County, Utah (41”50’N, Utah juniper. These trees were the most fre- 113”25’W) near Rosette, Utah at an elevation of quently used for foraging by the respective bird 1,850 m. The study site primarily consisted of species (pers. observ.). Meteorological variables Utah juniper (Juniperus osteosperma) with measured were: (1) air temperature (T,) (with a sparsely scattered singleleaf pinyon pine (Pinus shaded 36-gauge copper-constantanthermocou- monophylla). The titmouse study site is at the ple), (2) operative temperature (T,) (with a 3.5- northern edge of their range (Cicero 1996). cm diameter copper sphere thermometer painted flat gray; Bakken et al. 1985, Walsberg and TIME-ACTIVITY BUDGETS Weathers 1986), and (3) wind speed (u) (with a I collected 16 time-budget samples totaling 67 Thornwaite model 901 cup anemometer). Sensor min of observation for summer chickadees and outputs were monitored at 60-set intervals, av- 16 time-budget samples totaling 87 min of ob- eraged every 60 min, and recorded with a servation for winter chickadees.I collected nine Campbell Scientific CR10 electronic datalogger. time-budget samples totaling 60 min of obser- Thermocouples were calibrated with a thermom- vation for summer titmice and eight time-budget eter traceable to the U.S. Bureau of Standards. samples totaling 80 min of observation for win- The cup anemometer was factory calibrated. ter titmice. Samples were distributed throughout Details of nocturnal microclimate measure- the day in order to achieve uniform coverage of ment can be found in Cooper (1999). In brief, the birds’ active day. I observed focal individ- for nocturnal microclimate measurement, I used uals for 2-30 min (mean + SE = 5.3 + 0.4) and T,, measured inside nest boxes occupied by a recorded the time spent in three activities single bird for both chickadees and titmice. (perching, foraging, or flying). Perching includ- Wind speed was measured on different nest box- DAILY ENERGY EXPENDITURE IN CHICKADEES AND TITMICE 637 es which did not contain a bird and was always serted into the inlet port of the metabolic cham- zero. ber. Individuals were weighed and then placed inside the metabolic chamber where they LABORATORY METABOLISM perched on l-cm wire mesh placed 3 cm above MEASUREMENTS a l-cm layer of paraffin oil used for the collec- I measured the metabolic heat production of tion of fecal material. Oxygen consumption chickadees and titmice by measuring their oxy- (VO,) was then measuredusing open-circuit res- gen consumption (VO,) at stable air tempera- pirometry with an Ametek Model S-3A oxygen tures between - lo” and 30°C. The birds used in analyzer (AEI Technologies, Pittsburgh, Penn- these measurements were captured during sum- sylvania). Air was drawn through the metabolic mer and winter of 1995 and 1996. Birds were chamber with a diaphragm pump and was dried transportedfrom the field to Logan, Utah, where with indicating Drierite (anhydrous CaSO,) and they were housed in individual cages (0.3 X 0.3 scrubbedof CO, with Ascarite. Outlet flow rates X 0.3 m) and held in a temperature-controlled of dry, CO,-free air were maintained at 442-450 environmental chamber (3 X 3 X 2.5 m) for a mL min-’ by a Matheson precision rotameter maximum of one week. The chamber tempera- (Model 604), which was calibrated to 2 1.0% ture and photoperiod were programmed to fol- (Brooks vol-u-meter), and located downstream low a cycle that approximated the season and from the metabolic chamber. These flow rates study site to which the birds had been accus- yielded changes in oxygen content between in- tomed. While in captivity, birds were provided flux and efflux gas of 0.3 and 0.7%, and main- with food (Tenebrio larvae and wild bird seed) tained oxygen content of efflux gas above and water as needed. Birds tested from 1 June 20.2%. Fractional concentration of oxygen in ef- to 25 August were designated “summer birds,” flux gas was determined from a 100 mL mini and those tested from 20 November to 10 Feb- subsample passed through the oxygen analyzer. ruary were designated “winter birds.” Oxygen Measurements of the efflux gas were recorded consumption was not measured for any summer every 15 set on a computer using Datacan 5.0 birds that were noticeably molting. data collection and analysis software (Sable Sys- I measured 30, during the active phase of the tems International, Henderson, Nevada). Evap- daily cycle on fed birds at rest in darkened me- orative water loss (EWL) was determined over tabolism chambers to estimate energetic costs of a 60-min timed interval by measuring the in- daytime maintenance plus the cost of alert crease in mass of a downstream absorbent train perching, and on fed birds in metabolism cham- containing Drierite. Low permeability Bev-a- bers (equipped with a dish of wild bird seed) Line tubing was used to connect the metabolism exposed to normal fluorescent room lighting to chamber to the downstream absorbent train. All estimate energetic costs of daytime maintenance weighings were made on an analytical balance plus the cost of foraging. Nighttime mainte- (Mettler H51AR). At the end of each metabo- nance-energy requirements were estimated from lism trial, birds were removed from the chamber previous \;rO, measurements during the rest and body temperature (Tb) (2 O.l”C) was re- phase on fasted birds resting in the dark (mini- corded by inserting a 30-gauge copper-constan- mum of 4 hr since last meal) (Cooper 1998). tan thermocouple into the cloaca to a depth (ap- Measurements were made on individual birds proximately IO-12 mm) where further insertion using a 3.8 L metabolic chamber fashioned from did not alter temperature reading. a paint can. The inside of the metabolic chamber VO, and EWL were measured on individual was painted flat black to provide an emissivity birds exposed to a single randomized tempera- near 1.0. Metabolic chamber temperature was ture in the dark and also in normal room light- regulated within + 0.5”C by placing it in a tem- ing. Individuals were given 24-hr rest in be- perature-controlled environmental chamber. tween VO, measurements. All individuals were Metabolic chamber temperature was monitored tested within one week of capture. Individual continuously throughout each test with an Ome- birds were placed in the metabolic chamber for ga thermocouple thermometer (Model Omni IIB, a total of 2 hr. The first hour was an equilibration previously calibrated to a thermometer traceable time and ir0, was measured over the last 60- to the U.S. Bureau of Standards) attached to a min of the trial. Oxygen consumption was cal- 30-gauge copper-constantan thermocouple in- culated as steady state VO, using Eq. 4a of 638 SHELDON J. COOPER

Withers (1977). All values were adjusted to STP I assumed that active phase basal metabolism Rates of metabolic heat production were calcu- was 1.2 times BMR for chickadees and titmice lated assuming that 20.1 k.I of heat were pro- in order to correct each activity for the daily duced per liter of oxygen consumed for both fed cycle. and fasted birds (Gessaman and Nagy 1988). ESTIMATING ENERGY COSTS UNDER FIELD TIME-ACTIVITY LABORATORY ESTIMATE OF CONDITIONS DEE Equation (1) usually provides mean DEE values I calculated the DEE of seasonally-acclimatized within 5% of the mean DEE determined by dou- chickadees and titmice using time-budget, me- bly labeled water (DLW) of free-ranging birds, teorological, and laboratory metabolism data provided certain criteria are met. First, mainte- from the following equation: nance and activity costs must be determined for the study population(s) at the same season as DEE = (t,&n)+ (tapEi,,)+ &,,ri,> + (hEi,,) (1) time budgets are recorded (Weathers and Sulli- where t representsduration (in hours) of the ac- van 1993). Secondly, maintenance and activity tivity phases and of the type of activity, and fi costs must be evaluated under field conditions is the energy requirements for a given activity using heat transfer theory that uses standard op- (in kJ hrl). The subscriptsrepresent the time of erative temperature to calculate thermoregula- day (p = nighttime) or the type of activity (m = tory costs (Buttemer et al. 1986, Weathers and maintenance metabolism, ap = active perch, fo Sullivan 1993). I calculated the complex thermal = foraging, and fl = flight). In equation (l), the environment encountered by birds in this study first bracketed term, nocturnal energy expendi- by calculating standard operative temperature ture, consists of basal and thermoregulatory en- (T,,) on the basis of the measured field T, and ergy requirements of a sleeping bird. The second wind speed (u) using Bakken’s (1990) general- bracketed term represents maintenance-energy ized passerine T,, scale: requirements plus active perching-energy re- T,, = T, - (1 + 0.26~“.~) (T,, - T,) (2) quirements of a daytime bird. The third brack- eted term represents maintenance-energy re- I did not use DLW to determine DEE in chick- quirements plus foraging-energy requirements of adees and titmice due to the difficulty in recap- a daytime bird. The second and third bracketed turing marked individuals within 24 to 48 hr. terms subsume thermoneutral and thermoregu- latory energy requirements during the bird’s ac- STATISTICAL ANALYSES tive phase and includes the heat increment of Time-activity budgets and time-activity labora- feeding (HI). The fourth bracketed term repre- tory estimatesof DEE were compared using Stu- sents flight-energy requirements of a daytime dent t-tests because variances were equal (F- bird. I used Carlson and Moreno’s (1992) doubly tests for equality of variance). Regression lines labeled water estimate of short flight costs in were fit by the method of least squares. Slopes Willow Tits (Poe& montunus) to determine the and intercepts of regressionlines were compared energy cost of flight (Ej,) in eq. (1). Thus, flight by ANCOVA. Significance was determined at costs were calculated as 11.7 times nighttime the P < 0.05 level for all analyses. All analyses basal metabolic rate (BMR). BMR data were were performed with SPSS 6.1 (Norusis 1989). taken from Cooper (1998). I related laboratory All data are presented as mean 2 SE. measurementsof fi(,,, fiap, and I& directly to the RESULTS 60-min recordings of microclimate measure- ments associated with each bird’s diurnal and WEATHER nocturnal phases,respectively. In order to deter- During the summer and winter study period, no mine total daily energy costs of each activity precipitation fell. In summer, mean T, was 15.1 (perching, flight, foraging, nocturnal mainte- -C l.O”C for chickadees and 19.5 2 0.9”C for nance) for chickadees and titmice, I subtracted titmice. In winter, mean T, was -3.6 t 0.5”C basal metabolism from each activity. Because for chickadees and -5.7 -C 1.2”C for titmice basal metabolic rate averages 20-25% higher (Fig. 1). These temperatures are within normal during the active phase of the daily cycle than T, ranges for each study site (Utah State Climate during the rest phase (Aschoff and Pohl 1970), Center, Logan, Utah). In summer, mean wind DAILY ENERGY EXPENDITURE IN CHICKADEES AND TITMICE 639

time intervals were used to calculate laboratory estimates of DEE. Chickadees and titmice spend over 50% of their active day foraging in both summer and winter (Table 1). The time budgets of the two species were similar. There were no significant differences either between species or seasonally (Table 1).

LABORATORY METABOLIC RATES Under the conditions of my laboratory metabo- lism measurements (isothermal metabolism chamber with no significant shortwave radiation or forced convection), T, is the same as standard operative temperature (T,,). Although normal fluorescent room lighting illuminated the meta- ._‘ bolic chamber used to determine foraging costs, this would amount to a negligible amount of ir- radiance received by the bird due to construction 06 00 14 00 22 00 04.00 06 00 14 00 22 00 04 00 of the chamber. For example, Verdins (Auripa-

TlME TIME YUSjlaviceps) exposed to normal fluorescent

Summer Winter room lighting in glass metabolic chambers were subject to an irradiance of < 3 W rn-? (Wolf and FIGURE 1. Temperature and wind speed (u) for summer Mountain Chickadees (A), summer Juniper Walsberg 1996). The relationship between oxy- Titmice (B), winter Mountain chickadees(C), and win- gen consumption (VO,) and standard operative ter Juniper Titmice (D). Air temperature(T,) is repre- temperature are described for fed summer birds sented by solid lines and operative temperature(T,) is (Fig. 2) and winter birds (Fig. 3). representedby dashedlines. The comparison of slopes and intercepts by analysis of covariance (ANCOVA) of these re- speed was 0.6 2 0.2 m set-’ for chickadees and gression equations allow comparison of perch- 0.9 k 0.1 m secl for titmice. In winter, mean ing and foraging energy costs. For summer wind speed was 0.2 5 0.1 m set- I for chicka- chickadees,neither slopes (F,,,, = 1.6, P > 0.2) dees and 0.3 -C 0.1 m sect’ for titmice (Fig. 1). nor intercepts (F,,,, = 1.7, P > 0.2) were sig- nificantly different between perching and for- TIME-ACTIVITY BUDGETS aging-energy requirements. For summer titmice, In summer, chickadees and titmice began for- neither slopes (F,,,, = 0.0, P > 0.9) nor inter- aging around 05:OO and went to roost around cepts (F,,,, = 0.0, P > 0.9) differed between 19:00, making their active day about 14 hr long. perching and foraging-energy requirements. For In winter, chickadeesand titmice began foraging winter chickadees, both slopes (F,,,, = 23.2, P around 07:30 and went to roost around 16:30, < 0.01) and intercepts were significantly differ- making their active day about 9 hr long. These ent between perching and foraging costs (F,,,, =

TABLE 1. Percentageof the active day that seasonallyacclimatized Mountain Chickadeesand JuniperTitmice spent in various activities. Sample sizes are the number of 2-30 min observations for the indicated focal individuals. t-tests were performed on arscine transformedpercentages.

Summer Winter Mountain Juniper Mountain Juniper Chickadees Titmice Chickadees Titmice Percentageof active day spent (n = 16) (n = 9) (n = 16) (n = 8) Perching 27.5 2 5.2 34.5 ? 7.3 39.4 F 6.7 40.5 t 4.1 Foraging 68.7 2 4.7 61.5 ? 6.8 53.0 i- 5.4 54.9 -t 5.2 Flying 3.8 -c 0.4 4.0 -c 0.3 7.6 2 1.2 4.6 % 0.2 640 SHELDON J. COOPER

l -Fora ing A 0 --- Pert 1.Ing

. . .._‘

. --- . -10 0 10 20 30 40 ! EN20 : 16 r . : 12 0 I2 8 a f--:;:“1‘ . 4 0 : 0 -.._-.._ 01 . \__‘ -10 6 Ib 20 30’ 40 50 -10 0 10 20 30 40 ! Standard operative temperature (“C) Standard operative temperature (“C)

FIGURE 2. Relationship between oxygen consump- FIGURE 3. Relationship between oxygen consump- tion and standardoperative temperature for summer-ac- tion and standard operative temperature for winter- climatized Mountain Chickadees (A) and Juniper Tit- acclimatized Mountain Chickadees (A) and Juniper mice (B) during the active phase of their daily cycle. Titmice (B) during the active phase of their daily cy- Dots representactive birds under lit conditions(forag- cle. Dots represent active birds under lit conditions ing) and open circles representbirds under dark condi- (foraging) and open circles represent birds under dark tions (perching). Linear regressionequations were, for- conditions (perching). Linear regression equations aging chickadees:VO, = 11.97 - O.l8T,, (n =, 15, i were, foraging chickadee: VO, = 11.91 - 0.36T,, (n = 0.55, P < O.OOl), perching chickadees: VO, = = 13, r2 = 0.79, P < O.OOl), perching chickadees: 12.27 - 0.26T,,.(n = 16, r2 = 0.72, P < O.OOl), ir0, = 15.46 - O.l7T,, ,(n = 14, r* = 0.64, P < foraging titmice: VO, = 9.73 - 0.22T,, {n = 11, rZ = O.OOl), foraging titmice: VO? = 11.46 - 0.30T,, (n 0.79, P < O.OOl),and perching titmice: VO, = 9.63 - = 10, r* = 0.87, P < O.OOl), and perching titmice: 0.21T,, (n = 15, rZ = 0.83, P < 0.001). \jO? = 11.39 - 0.30T,, (n = 10, r2 = 0.78, P < 0.001).

36.9, P < 0.001). For winter titmice, neither ks = W - WV, - T,,) slopes (F,,,, = 0.0, P > 0.9) nor intercepts (F ,,,, = 0.0, P > 0.9) differed between perching and where M is metabolic rate and E is evaporative foraging-energy requirements. heat loss (assuming 2.429 J of heat for each mg In order to determine the effect of activity on of water evaporated). Thermal conductance val- plumage disruption, I calculated overall thermal ues (Table 2) were compared using ANOVA

conductance (&) for individuals using the equa- (F, 90 = 7.6, P < 0.001). Pairwise mean com- tion of Bakken (1976): parkons were made using Fisher’s LSD; winter

TABLE 2. Thermal conductance(mW g-l “C-l) for fed birds in lighted conditions (foraging) and for fed birds in dark conditions (perching). Sample sizes are in parentheses.

Summer Winter Mountain Juniper Mountain Juniper Condition Chickadees Titmice Chickadees Titmice Perching 1.44 2 0.11 (15) 1.04 r 0.12 (11) 2.20 -c 0.15 (12) 1.52 2 0.12 (9) Foraging 1.56 + 0.11 (14) 1.17 ? 0.12 (9) 1.65 ? 0.13 (13) 1.49 t 0.15 (8) DAILY ENERGY EXPENDITURE IN CHICKADEES AND TITMICE 641

TABLE 3. Daily energy budget of seasonallyacclimatized Mountain Chickadees and Juniper Titmice as cal- culated by the TAL method.

Summer Winter Mountain Juniper Mountain Juniper Variable (kJ day-t) Chickadees Titmice Chickadees Titmice DEE 48.8 + 0.6 48.3 t 1.0 66.3 -+ 1.5 98.7 % 0.1 Basal metabolism” 23.3 + 1.0 27.9 + 1.5 25.4 + 0.9 27.8 2 0.4 Nocturnal thermoregulation 6.5 i 0.1 4.5 f 0.1 14.5 + 0.0 32.9 ? 0.0 Alert oerchin@ 4.2 2 1.1 3.8 + 1.3 10.3 i 1.0 15.6 ? 0.7 Foragmgb v 9.2 2 1.0 6.4 ? 0.9 11.3 f 1.1 17.1 + 0.6 Flying 5.6 2 0.7 5.7 2 1.2 4.8 2 0.5 5.3 + 0.1

*Data are calculated for field conditmns mcorporating the circadnn rhythm m basal metabolism h Data include thermoregulation and heat mcrement of feeding. chickadees that were perching had significantly spent foraging in winter. This is in spite of a higher thermal conductance than all other birds. 36% and 204% increase in DEE in chickadees and titmice, respectively. This lack of seasonal TIME-ACTIVITY LABORATORY (TAL) variation in foraging time suggeststhat foraging ESTIMATE OF DEE efficiency increases in winter in these species. Daily energy expenditure estimated by the TAL This is probably due to the availability of food method averaged 48.8 and 48.3 kJ day-’ for caches stored during the fall by these species summer Mountain Chickadees and Juniper Tit- (Sherry 1989). However, this lack of seasonal mice, respectively (Table 3). Daily energy ex- variation in foraging time may also be due to penditure averaged 66.3 and 98.7 kJ day-’ for the high amount of time spent caching food in winter chickadees and titmice, respectively (Ta- summer birds in this study. For example, pre- ble 3). For both chickadees and titmice, DEE dicted foraging time based on Bryant and Wes- was significantly higher in winter compared to terterp (1980) would be 32% and 24% of the summer (chickadees, tie = 11.0, titmice, t,, = active day in summer chickadees and titmice, 34.5; both P < 0.001). Juniper Titmice weighed respectively. Actual time spent foraging in sum- significantly more than Mountain Chickadees in mer birds was 37% higher than predicted for both summer and winter (Cooper 1998), and the both chickadees and titmice. mass difference confounds direct comparison of DEE. However, the difference in body mass can ACTIVITY HEAT AND THERMOREGULATION be removed by converting DEE to units of kJ By comparing the regression equations relating g-0.63 day-‘, where Mm0.63is the interspecific metabolism to T,, for fed, daytime birds resting scaling of DEE (Weathers and Sullivan 1989). in the dark with those for fed, daytime birds ex- Using 11.7 g as the mean daily mass of chick- posed to light, the energetic cost of physical ac- adees and 17.4 g for titmice (unpubl. data), I tivity can be calculated. In summer chickadees, computed the mass-adjusted DEE for summer and both summer and winter titmice, regression and winter-acclimatized individuals. Summer equations did not differ significantly. For winter chickadees mass-adjustedDEE (10.4 ? 0.13 kJ chickadees, the slopes and intercepts were sig- gmO_“iday- I, n = 16) was significantly higher nificantly different. Winter chickadees in illu- than mass-adjustedDEE of summer titmice (8.0 minated chambers had lower metabolism than ? O.l6 kJ gmo6”day- ‘, n = 9; t,, = 11.1, P < those resting in the dark. How actively-foraging 0.001). In winter, mass-adjusted DEE was sig- birds can possibly have lower metabolism than nificantly lower for chickadees (14.1 t 0.3, kJ inactive perching birds is unclear. One possible g-0.63day- ‘, n = 16) than titmice (16.3 5 0.1 kJ explanation is that perching chickadees became gmo.63day-l, II = 8; tz2 = -3.5, P < 0.01). stressedwhen they were deprived of food during their normal period of foraging activity. These DISCUSSION data indicate that heat produced as a by-product TIME-ACTIVITY BUDGETS of activity substitutes for thermoregulatory re- Mountain Chickadees and Juniper Titmice did quirements and indicates that chickadee and tit- not increase the percentage of the active day mouse behavior has no net energy cost at cold 642 SHELDON J. COOPER temperatures. A similar circumstance applies to sible confounding variable with my study is that Yellow-eyed Juncos (Bunco phaeonotus) 1 did not collect time-budgets during the peak of (Weathers and Sullivan 1993), to the foraging the breeding season and therefore do not know behavior of winter Verdins (Webster and Weath- whether my TAL DEE calculations would ers 1990), and terrestrial locomotion in cold-ex- change. However, during the summer period posed White-crowned Sparrows (Zonotrichia when I collected time budgets, individuals were leucophrys) (Paladin0 and King 1984). storing food items, resulting in increased forag- Complete substitution of thermoregulatory ing times relative to predicted times and proba- demands due to exercise heat in birds may not bly resembling foraging times of adults feeding occur in some birds due to plumage disruption nestlings. For example, the amount of time spent that causes increased thermal conductance foraging by summer birds in this study is very (Nomoto et al. 1983). For chickadees and tit- close to that recorded for Yellow-eyed Juncos mice, thermal conductance does not vary be- feeding nestlings and fledglings (Weathers and tween perching and foraging birds (except in- Sullivan 1989). In order to estimate whether my creased conductance in perching winter chicka- DEE calculations would change during the dees). This demonstratesthat very little plumage breeding season, I related the equations relating disruption occurs while foraging in these birds. VO, to T,, using wind speed from this study Lack of plumage-layer disruption has recently (Fig. 1) and mean air temperatures(T,) from the been demonstrated in eastern House Finches study sites during May (Utah State Climate Cen- (Carpoducus mexicanus) hopping on a treadmill ter) to calculate T,,. May is when breeding be- at 0.5 m see’ (Zerba et al. 1999). Heat produced gins in these species(pers. observ.). For activity as a by-product of activity may substitute for periods, I used mean maximum T,, and for rest- thermoregulatory requirements in birds more ing periods I used mean minimum T, at the re- regularly than previously believed. spective study sites. DEE calculated with May climate data is 57.3 kJ day-’ in Mountain Chick- SEASONAL VARIATION IN DEE adees and 53.2 kJ day-’ in Juniper Titmice. Summer DEE values for chickadees and titmice Thus, DEE may be lo-17% higher in breeding are 86.8% and 63.8%, respectively of predicted birds. However, these values are still markedly DEE based on body mass (Nagy 1987). Winter lower than winter values. DEE values for chickadees and titmice are The markedly increased DEE in winter rela- 118.0% and 130.4%, respectively of allometri- tive to summer contrasts with data from most tally predicted DEE (Nagy 1987). Although passerines tested to date. For birds where DEE these winter values exceed allometric predic- was measured with DLW, only male dippers tions, they are within 1% of mass-specificvalues (Cinclus cinclus) and White-crowned Sparrows of DEE for winter-acclimatized Black-capped have increased DEE in winter compared to Chickadees determined using doubly labeled breeding (Bryant and Tatner 1988, Weathers et water (Karasov et al. 1992). DEE values from al. 1999). All other passerinesin which DEE has this study are within 8% to 15% of 24-hr me- been measured seasonally have relatively stable tabolism measurements using total collection DEE or markedly increased DEE during the feeding trials of Mountain Chickadees and Ju- breeding season (Weathers and Sullivan 1993). niper Titmice (unpubl. data). In addition, be- Two possible factors may explain the seasonal cause heat produced as a by-product of activity changesin DEE found in chickadeesand titmice substitutesfor thermoregulatory requirements in in this study. First, the birds in this study were these species, DEE values should not be signif- exposed to much colder environmental temper- icantly affected by the relatively small sampling atures, and therefore increasing thermoregula- time of perching or foraging activity. tory costs compared to other birds so far tested The “increased demand hypothesis” holds (with the exception of dippers) (Bryant and Tat- that breeding results in a substantial increase in ner 1988). Secondly, winter-acclimatized Moun- adult energy demand and consequently, DEE is tain Chickadees have 13% higher thermal con- highest during breeding. Data from the present ductance (nighttime values) and winter-accli- study indicate that winter, due to its increased matized Juniper Titmice have 26% higher ther- thermoregulatory costs, represents a substantial ma1 conductance than allometrically predicted energy increase compared to summer. One pos- (Cooper 1998). Thus, the relatively poor insu- DAILY ENERGY EXPENDITURE IN CHICKADEES AND TITMICE 643 lation of these birds, especially of titmice, in- getics of foraging and free existence in birds. Int. &nithol. Con&.-17:292-299. creases their thermoregulatory costs. BUTTEMER.W. A.. A. M. HAYWORTH.W. W. WEATHERS. AND k A. ~AGY. 1986. Time-budget estimatesof ROLE OF DEE ON NORTHERN RANGE LIMITS avian energy expenditure: physiological and me- DEE as a multiple of basal metabolic rate teorological considerations. Physiol. Zool. 59: (BMR) was 2.3 1 in summer chickadees and 1.91 131-149. CAMPBELL,G. S., AND J. M. NORMAN.1998. An intro- in summer titmice. DEE was 2.70 times BMR duction to environmental biophysics. 2nd ed. in winter chickadees and 3.43 times BMR in Springer-Verlag, New York. winter titmice. The winter values exceed the CARLSON,A., AND J. MORENO.1992. Cost of short suggested northern boundary metabolic rate flights in the Willow Tit measured with doubly (NBMR) of 2.45 times BMR. The total DEE of labeled-water. Auk 109:389-393. CARLSON,A., J. MORENO,AND R. V. ALATALO.1993. these birds would be expected to be somewhat Winter metabolism of coniferous forest tits (Pari- greater than 2.45 times basal, because the birds dae) under arctic conditions: a study with doubly must also expend energy for digestion and flight. labeled water. Ornis Stand. 24:161-164. For winter-acclimatized Siberian Tits (Poecile CASTRO,G. 1989. Energy costsand avian distributions: cinctus) and Willow Tits (Poecile montunus) limitations or chance?-a comment. Ecology 70: 1181-1182. tested from their northern January isotherm, CICERO,C. 1996. Sibling species of titmice in the Pa- DEE was 2.55 and 2.50 times BMR, respective- rus inornatuscomplex (Aves: Paridae). Univ. Ca- ly (Carlson et al. 1993). Thus, data for chicka- lif. Pub. Zool. 128:1-217. dees and tits appear to closely conform to a sug- COOPER, S. J. 1998. The role of cold acclimatization gested NBMR of Root (1989b). However, the on the biogeography of the Mountain Chickadee (Pm-usgambeli) and the Juniper Titmouse (Parus data for titmice exceeds the proposed NBMR of ridgwayi). Ph.D. diss., Utah State Univ., Logan, 2.45 times BMR and suggests that although a UT. NBMR may exist, the multiple may be higher COOPER,S. J. 1999. The thermal and energetic signif- than 2.45 times BMR. icance of cavity roosting in Mountain Chickadees and Juniper Titmice. Condor 101:863-866. ACKNOWLEDGMENTS DAWSON,W. R., AND R. L. MARSH. 1989. Seasonal acclimatization to cold and seasonsin birds, p. James Gessaman, Kim Sullivan, Wayne Wurtsbaugh, 83-94. In C. Beth and R. E. Reinertsen IEDS.~. Walter Koenig, and one anonymousreviewer provided Physiology of cold adaptation in birds. Pien& many useful suggestionson earlier drafts of this man- Press, New York. uscript. Kim Sullivan generously loaned me the me- DIXON, K. L. 1949. Behavior of the Plain Titmouse. teorological equipment used in this study. Thanks to Condor 51:110-136. Keith Dixon for providing information on study sites GESSAMAN,J. A., AND K. A. NAGY. 1988. Energy me- and to Jaclyn Cooper and Glenn Wilson for field as- tabolism errors in gas-exchange conversion fac- sistance.This study was supportedin part by a North tors. Physiol. Zool. 61:507-513. American Bluebird Society studentresearch grant and GODFREY,W. E. 1986. The birds of Canada.Natl. Mus. a Sigma Xi Grant-in-Aid of Research.Birds were cap- Canada, Ottawa. tured under federal (#PRT-779300) and state KARASOV,W. H., M. C. BRITTINGHAM,AND S. A. TEM- (2COLL1401) scientific collecting permits. PLE. 1992. Daily energy expenditure by Black- capped Chickadees (Parus atricapillus) in winter. LITERATURE CITED Auk 109:393-395. ASCHOFF,J., AND H. POHL.1970. Rhythmic variations MARSH,R. L., AND W. R. DAWSON.1989a. Avian ad- in energy metabolism. Fed. Pro& 29:1541-1552. iustmentsto cold. v. 205-253. In L. C. H. Wang BAKKEN.G. S. 1976. A heat transfer analvsis of ani- LED.], Advances in comparativeand environment2 mals: unifying conceptsand the applicationof me- physiology. Springer, Berlin. tabolism data to field ecology. J. theor. Biol. 60: MARSH,R. L., AND W. R. DAWSON.1989b. Energy sub- 337-384. strates and metabolic acclimatization in small BAKKEN,G. S. 1990. Estimating the effect of wind on birds, p. 105-l 14. In C. Beth and R. E. Reinertsen avian metabolic rate with standardoperative tem- [EDS.], Physiology of cold adaptationin birds. Ple- perature. Auk 107587-594. num Press, New-York. BAKKEN,G. S., W. R. SANTEE,AND D. J. ERSKINE. MASMAN. D.. M. GORDIJN.S. DAAN. AND C. DIJKSTRA. 1985. Operative and standardoperative tempera- 1986. Ecological energetics of the kestrel: field ture: tools for thermal energetics studies. Am. estimates of energy intake throughout the year. Zool. 25:933-943. Ardea 74:24-39. -_ BRYANT,D. M., AND P TATNER.1988. Energetics of NAGY. K. A. 1987. Field metabolic rate and food re- the annual cycle of dippers Cinclus cinch. Ibis quirement scaling in mammals and birds. Ecol. 130:17-38. Monogr. 57: 11l-128. BRYANT,D. M., AND K. R. WESTERTERP.1980. Ener- NOMOTO, S., W. RAUTENBERC,AND M. IRIKI. 1983. 644 SHELDON J. COOPER

Temperatureregulation during exercise in the Jap- breeding-season energetics of nonmigratory anese Quail (Colurnix coturniw juponica). J. White-crowned Sparrows. Auk 116:842-847. Comp. Physiol. B 149519-525. WEATHERS,W. W., AND K. A. SULLIVAN.1989. Juvenile NORUSIS, M. J. 1989. SPSS for windows. SPSS Inc., foraging proficiency, parental effort, and repro- Chicago. ductive success.Ecol. Monogr. 59:223-246. WEATHERS,W. W., AND K. A. SULLIVAN.1993. Sea- PALADINO,E, AND J. R. KING. 1984. Thermoregulation sonal patterns of time and energy allocation by and oxygen consumption during terrestrial loco- birds. Phvsiol. Zool. 66:51l-536. motion by white-crowned sparrows Zonotrichia WEBSTER,M.-D., AND W. W. WEATHERS.1990. Heat leucophrysgambeli. Physiol. Zool. 57:226-236. producedas a by-product of foraging activity con- REPASKY,R. R. 1991. Temperature and the northern tributes to avian thermoregulation.Physiol. Zool. distributions of wintering birds. Ecology 72: 63:777-794. 2274-2285. WITHERS,F? C. 1977. Measurementof irO,, iX02, and ROOT, T. 1988a. Atlas of wintering North American evaporative water loss with a flow-through mask. birds. Univ. Chicago Press, Chicago. J. Appl. Physiol. 42:120-123. ROOT,T. 1988b. Energy constraintson avian distribu- WOLF, B. O., AND G. E. WALSBERG.1996. Thermal tions and abundance.Ecology 69:330-339. effects of radiation and wind on a small bird and SHERRY,D. E 1989. Food storing in the Paridae. Wil- implications for microsite selection. Ecology 77: 2228-2236. son Bull. 101:289-304. ZERBA,E., A. N. DANA, AND M. A. LUCIA. 1999. The WALSBERG,G. E., AND W. W. WEATHERS.1986. A sim- influence of wind and locomotor activity on sur- ple technique for estimating operative environ- face temperature and energy expenditure of the mental temperature.J. Therm. Biol. 11:67-72. eastern House Finch (Curpoducus mexicanus) WEATHERS,W. W., C. R. OLSON,R. B. SIEGEL,C. L. during cold stress. Physiol. Biochem. Zool. 72: DAVIDSON,AND T R. FAMULA.1999. Winter and 265-276.