Energetics of Toucans, a Barbet, and a Hornbill: Implications for Avian Frugivory

Home , Emerald toucanet, Hornbill, Pici (taxon), Ramphastos, Saffron toucanet, Toco toucan

The Auk 118(4):916-933, 2001

ENERGETICS OF TOUCANS, A BARBET, AND A HORNBILL: IMPLICATIONS FOR AVIAN FRUGIVORY

BRIAN K. MCNAB 1 Departmentof Zoology,University of Florida,Gainesville, Florida 32611, USA

ABSTRACT.--Rateof oxygenconsumption was measuredin eight speciesof toucans,one barbet, and one hornbill to examinefactors that influencetheir energyexpenditures. Basal rate of metabolismin thosespecies, supplemented with measurementsfrom three wood- peckersand three mousebirds,is correlatedwith body massand either food habitsor climate:temperate acorn- and insect-eatingspecies have basal rates that are -70% greaterthan tropicalfruit-eating species. The temperate,insect- or acorn-eatingspecies are woodpeckers, so level of basal rate is also correlated with familial affiliation. The toucans, barbet, and hornbill havebasal rates similar to thoseof frugivorouspigeons and bats,which collectivelyare low by avianstandards. The effectsof climate,food habits,and family affiliationon basal rates in endothermsare difficult to separate,given the restricteddata set available.Earlier conclusionsthat the basalrates of birds greatlyexceed those of mammalsare confounded by characterinteractions that influencedependence of basalrate on body mass.The largest toucanshowed a remarkableability to reduceenergy expenditureat low ambienttemper- atureswithout reducingcore body temperature,possibly as a result of peripheralvasocon- striction.Received 7 June1999, accepted16 April 2001.

BIRDS HAVE BASAL rates of metabolism that at low latitudesto high levelsof solarradiation; vary with many factors,including body mass. McNab (1988a)suggested that the basalrate of They are describedas having higher basal rates birds is correlated with food habits; McNab than mammals,with passerineshaving higher and Bonaccorso (1995) showed that basal rate basal rates than other birds (Lasiewski and in insectivorous birds is correlated with for- Dawson 1967,Aschoff and Poh11970),and they aging style; and McNab (2000) demonstrated have a circadian rhythm in which the rate is that the basalrate of frugivorouspigeons in the highest,even when inactive,during the normal South Pacific is correlated with island size. period of activity (Aschoffand Pohl 1970). A Here I present data on temperatureregula- recent analysisby Reynoldsand Lee (1996) ar- tion and rate of metabolismof eight speciesof gued that passerinesdo not havehigher basal toucansthat belongto five of the six genera,on rates than other birds, a conclusionfirst sug- one speciesof African ground-barbet(Trachy- gestedby Prinzingerand Hanssler(1980), al- phonusdarnaudiO, and on Blyth's Hornbill (Ac- though the data used by Reynolds and Lee erosplicatus), a member of the palaeotropical were not stipulated and potentially included family Bucerotidae.The toucan family Ram- measurements that do not meet strict standards phastidaeis limited in distributionto the Neo- for basal rate (McNab 1997). Nevertheless, tropics, which along with woodpeckers(Pici- much interspecificvariation is found in basal dae), honeyguides (Indicatoridae), jacamars rate, even when body size, taxonomy,and time (Galbulidae), puffbirds (Bucconidae),and bar- of day are taken into consideration.Although bets (Capitonidae) constitutethe order Picifor- the extent to which that residual variation in- mes.Recent analyses, however, have suggested cludesmeasurement error is unclear, the larg- that New World barbetsconstitute a subfamily est proportion is correlatedwith the natural (Capitoninae) of the Ramphastidae (Prum history of the speciesstudied. Thus, Weathers 1988), whereas African barbets are placed in (1979) showed that latitude (and climate) has the family Lybiidae and Asian barbets in the an effect on basal metabolic rate in birds; Ellis family Megalaimidae (Sibley and Ahlquist (1980, 1984) demonstrated that basal rate is cor- 1990). related with plumage color in speciesexposed Toucans are of special interest becauseof their commitmentto frugivory (Remsenet al. 1993), although large speciesalso opportunis- E-mail: [email protected] tically prey on birds and mammals (Skutch

916 October2001] EnergeticsofAvian Frugivores 917

1983, Boinski 1987, Mitchell et al. 1991). Pro- Zoo) occursin southeasternBrazil to Paraguay and pensity for carnivory in sometoucans is most northern Argentina. (8) Red-Billed Toucan (Ram- marked in the highlands,which may compen- phastostucanus; one male and one female studied at satefor a reducedfruit availability(Galetti et the Parque Ambiental y Zoo16gicoGustavo Rivera) al. 2000). African ground-barbetsfeed on fruit occursprincipally in the tropical lowlands of South America north of the Amazon River. (9) Toco Toucan and insects (Short and Horne 1980). Hornbills, (Ramphastostoco; one male and one female borrowed traditionally placedin the order Coraciiformes, from the Lowry Park Zoo) lives in South America representan ecotypeconvergent with toucans, from the Guianasand Brazil southto Paraguayand large arborealspecies, such as A. plicatus,being northern Argentina. (10) Blyth'sHornbill (Acerospli- highly frugivorous(Kemp 1995). catus;one male and one female,semicaptives studied Thesedata permit an examinationof effectof at Christensen ResearchInstitute, Madang, Papua body size,food habits,behavior, and climateon New Guinea) inhabits Burma east to Sulawesi, New energeticsof tropical nonpasserinefrugivores Guinea, the BismarckArchipelago, and the Solomon through a comparison of toucan energetics Islands, although some of its distribution east of with those of the barbet, hornbill, and three New Guinea may be due to human transport (D. W. Steadman pers. comm.). mousebirds (Bartholomew and Trost 1970, Somecaptive individuals used in this study were Prinzinger et al. 1981, Prinzinger 1988). Tou- caughtin the wild, whereasothers were the product cans are also compared with woodpeckers of captivebreeding. They ate a variety of fruits with (Kendeighet al. 1977, Weatherset al. 1990), the protein and vitamin supplements and lived under only other piciformsthat have had their ener- the local natural photoperiod and temperaturere- geticsmeasured. In addition,I comparethe en- gimes. Measurements in Venezuela occurred in Au- ergy expenditures of those frugivores with gust and December, in Papua New Guinea in July, those of frugivorous pigeons (McNab 2000) and in Florida in fall or spring. and fruit bats belongingto the family Ptero- Experimentaltechniques.--Measurements of rate of metabolism,as oxygenconsumption, occurred in an podidae (Bonaccorsoand McNab 1997,McNab open system.Room air was suckedthrough a cham- and Armstrong 2001, McNab and Bonaccorso ber, the size of which varied with the size of a bird: 2001). a 329 L hollow-walled chamber was connected to a thermoregulatedwater bath; a 2 and a 44 L chamber, METHODS the temperaturesof which were regulatedby placing the chamber into the hollow-walled chamber; or a 15 Birds.--! studied 10 speciesof birds. They were (1) and a 27 L chamberwas placed a temperature-con- D'Arnaud's Ground-Barbet (Trachyphonusdarnaudii; trolled refrigerator.The barbet was measuredin a 2 one male and one female borrowedfrom the Lowry L chamber, most toucans in the 15 or 44 L chamber, Park Zoo, Tampa, Florida), a speciesthat occursin and the Toco Toucan both in the 44 and 329 L cham- EastAfrica from the Sudanto Tanzaniain dry brush bers. Hornbills were confined in the 27 L chamber country. (2) Saffron Toucanet(Baillonius bailloni; one during measurements.After being removedfrom the male and one female borrowed from the Central chamber, air was sequentially scrubbedof CO2 (by Florida Zoological Park, Lake Monroe, Florida) lives indicating soda lime) and water (by indicating silica in the Atlantic rainforest of Brazil. (3) Groove-Billed gel). The air stream then went to a flowmeter and to Toucanet(Aulacorhynchus sulcatus; one male and one an Applied ElectrochemistryS3A-II OxygenAnalyz- female studied at the ParqueAmbiental y Zoo16gico er, the outputsof which went to a stripchartrecorder. Gustavo Rivera, Falc6n, Venezuela) occurs in north- The flow rates varied with chamber size: flow rates ern Colombia and Venezuela. (4) Emerald Toucanet were 0.6 to 0.9 L min • in the 2 L chamber, 1.0 to 1.5 (Aulacorhynchusprasinus; one male and one female L mint in the 15 L chamber, 4.1 to 6.6 L mint in the borrowedfrom the Lowry Park Zoo) lives in south- 27 L chamber, 2.1 to 2.7 L min -• in the 44 L chamber, ern Mexico and Central America south to Venezuela, and 4.5 to 9.0 L min -t in the 329 L chamber. Those Colombia, Ecuador, and Peru. (5) Spot-Billed Tou- rates were high enoughto ensurean adequatemix- canet (Selenideramaculirostris; two males and one fe- ture of air in the chamber,as was demonstratedby male borrowed from the Lowry Park Zoo) is found independence of the calculated rate of metabolism in Brazil south of the Amazon to northeastArgen- from flow rate. tina. (6) Black-NeckedAra•ari (Pteroglossusaracari; To minimize the influence of activity, minimal one male and one female studied at the ParqueAm- rates of metabolism that lasted ---2 min were used; biental y Zoo16gicoGustavo Rivera) lives from Ven- individual runs usually lasted 2-3 h. The experi- ezuela and the Guianas souththrough southernBra- mentsreported here representeda total of 1,572h of zil. (7) Red-BreastedToucan (Ramphastosdicolorus; measurements;usually two experimentswere run at one male and one femaleborrowed from Lowry Park the same time. The cloacal temperature and body 918 BRIANK. MCNAB [Auk, Vol. 118 mass of the birds were measured at the terminus of rate, was significant.If not, all measurementswere eachexperiment, preferentially when rate of metab- pooled and used in calculationof the mean and SE olism was minimal to determine whether low rates (n = number of measurements). An examination of reflectlow body temperatures.The range in ambient data shows(see below) that no significantdifference temperatureused permitted the zoneof thermoneu- existedin basal rate among individuals in 8 of the 10 trality, basal rate of metabolism,and minimal ther- species. The range in individual basal rates, ex- mal conductance to be estimated. Measurements in pressedrelative to valuesexpected from body mass, mostspecies were madeboth at night and during the within a specieswas -•8%, and pooled estimatesof day to determine whether rate of metabolism basal rates are within 1% of those obtainedby aver- showeda circadian rhythm. aging means of the individual performances.The Thermal conductance was estimated from C = two exceptionsreflect either a noticeablesexual di- 17o2/AT,where C is thermalconductance (cm 3 0 2g-• morphism in body mass (A. plicatus)or possiblya h-• øC-%Vo2 is rateof metabolism(cm 3 02 g-• h-•), sexuallybased differencein behavior (S. maculiros- and AT is the temperaturedifferential between body tris). A further complicationoccurs when no differ- and ambient (McNab 1980). Thermal conductancesat ence among individuals is found in mass-specific temperaturesbelow thermoneutralityoften were not basal rate, but a differenceis found in body mass, constant: conductances tend to decrease with a fall which means that a difference occurs in total basal in ambienttemperature. Under thosecircumstances, rate (R. dicolorus).In thosecases, a speciesmean was similar conductanceswere grouped togetherover a derived from averagingindividual means. range in ambient temperatures.Mean conductances I used an ANCOVA package(Super ANOVA; Aba- were placedin the figuresas the slopeof a curvethat cus Concepts, Berkeley,California) to determine extrapolated to the mean body temperature in the whether basal rate of metabolism or minimal ther- samplewhen rate of metabolismequaled zero, as re- mal conductancewas correlatedwith body mass,cli- quired by the Scholander-Irvingmodel (Scholander mate, food habits, and family affiliation in the spe- et al. 1950).Thus, in Trachyphonus,three mean ther- cies studied here supplementedwith data from the mal condutances were calculated, one that corre- literature on mousebirdsand woodpeckers.To facil- sponded to the lower limit of thermoneutrality,one itate thoseanalyses, basal rate and minimal conduc- that representedminimal thermal conductance,and tance(dependent variables) were enteredin separate one that was intermediate (Fig. 1A). All measure- analyseswith total units as logarithms,as was body ments of metabolismwere correctedto standard dry mass, to convert these power functions into linear temperatureand pressure. equations. Various authors (Nickerson et al. 1989) have ad- When classifyingspecies into ecologicalor behav- vocated a mathematical technique to determine an ioral categories,difficulties are encounteredbecause intersectionbetween two curves,which potentially thoseaspects of the life of speciesare morecomplex could be used to determine the lower limit of ther- than simple categorieswill reflect. However,if the moneutrality and the slope of the curve below ther- categoriesreflect precisely the diversity in habits, moneutrality.The difficultieswith its applicationto most statisticalcells will contain only one species. metabolism data are that (1) the lower limit of ther- Furthermore,seasonal and geographicalvariations moneutrality does not correspondto the thermal in food habits contributesto complexityof classifi- conductanceat lower temperatures(B. bailloni);(2) cation. The best that can be done is to recognize thermal conductance decreases with a fall in ambient broad trophic or climatic categoriesas an approxi- temperature(T. darnaudii,A. prasinus,R. dicolorus,R. mation, which canbe attained by lumping categories tucanus,R. toco),which means that the slope of the that do not show statisticallysignificant differences pooled fitted curve below thermoneutralityis not and rerunning the statisticalanalyses. In theseanal- thermal conductance(McNab 1980);or (3) the rate of yses,the resultsare presentedafter all insignificant metabolism below thermoneutrality decreases(R. factors are discarded. Statistical significancewas toco). The technique used by Yeager and Ultsch _•0.05. (1989) to estimateintersection of two curvesdepends on their linearity,which is oftennot the casefor me- RESULTS tabolism at temperaturesbelow thermoneutrality. Any error involved in estimatingthe lower limit of Trachyphonusdarnaudii.--The two D'Arnaud's thermoneutralityby other than a mathematicaltech- Ground-Barbetsused in this study weighed nique is not greaterby more than 1-2øC,which has 36.6 q- 0.12 g. All measurements(n = 50) were no appreciableeffect on the conclusionsstated here. Statistics.--Toalleviate the potential problem of made at night. The range of thermoneutrality having only a few individualsof a speciesavailable, was narrow, only from 35 to 38øC (Fig. 1A), an F-test was used to determine whether the inter- which reflectsits small size.Within that range, individual variation in the rate of metabolism in the basal rate of metabolismequaled 1.07 ___0.023 zone of thermoneutrality,an estimate of the basal cm3 0 2g-• h -• (n = 16), which is 71%of thebas- October2001] EnergeticsofAvian Frugivores 919 al rate expectedfrom the curve for nonpasser- equaled40.4 + 0.15øC(n = 8). At ambienttem- ines (Aschoff and Pohl 1970). Basal rate did not peratures below 17øC, thermal conductance differ (F = 2.00, df = i and 14, P = 0.18) be- equaled0.069 + 0.0044cm 3 02 g • h • øC-• (n = tween thoseindividuals, their meansbeing 69 5), which is 94% of the value expectedfrom the (n = 8) and 73% (n = 8) of the valuesexpected averagemass. It did not differ (F = 0.26, df = from body mass, the mean of which is 71%. i and 3, P = 0.64) between the two toucanets. Body temperatureat ambienttemperatures be- Aulacorhynchussulcatus.--The two Groove- low 35øCequaled 38.3 -+ 0.15øC(n = 38). The Billed Toucanets had a mean mass that was increasein rate of metabolismat temperatures 131.7 + 2.18 g (n = 29). Thermoneutrality ex- below thermoneutralitydid not conformto one tended from 23 to 31øC(Fig. 2A), within which simple relationship (Fig. 1A and above); the basal rate of metabolism equaled 0.923 + lower limit of thermoneutrality was set by a 0.0624cm 3 02 g-• h -• (n = 7), which is 86% of conductanceequal to 0.278 _+0.0134 cm 3 02 g • the value expectedfrom mass. Basalrate did h • øC • (n = 10), whereas minimal thermal not differ (F; 0.45, df; i and 5, P = 0.84) be- conductanceat ambient temperaturesbetween tweenthe two individuals,which equaled87 (n 5 and 25øCwas 0.145 _+0.0039 cm3 02 g-• h -• -- 5) and 84% (n; 2) of the expectedvalues øC-• (n = 20), which is 103% of the value ex- (mean= 86%).Three daytime measurements in pected from the curve describedby Lasiewski thermoneutralitywere higher (1.32 _+0.023 cm 3 et al. (1967) and 57% of the conductance that 02 g-• h-•). Body temperatureat ambienttem- defines the lower limit of thermoneutrality. peratures between 15 and 25øC was 38.8 -+ Minimal conductancesdid not differ (F = 1.27, 0.38øC(n; 7) at night and 40.7 _+0.28øC (n = df = 1 and 18, P = 0.28) between the two 8) during the day. Mean thermal conductance individuals. at night was 0.059 + 0.0032cm 3 02 g-• h Baillonius bailloni.--Two Saffron Toucanets (n = 7), which is 80% of the value expected were measured,a male that weighed 136.2 --- from mass, whereas conductanceduring the 0.79 g (n = 25) and his daughterthat weighed day was 0.066 + 0.0032cm 3 02 g-• h 129.8 + 0.81 g (n = 22). The zone of thermo- = 5), some of the differencereflecting a differ- neutrality extendedfrom 19 to 34øC(Fig. lB). ence in body temperature,38.4 versus40.4øC, The basal rate in the male was 0.906 + 0.0297 respectively.The nocturnal conductancedid cm3 02 g-• h -• (n = 15), which is 85% of the val- not differ (? = 0.18, df = i and 5, P = 0.69) be- ue expected from mass, whereas the female tween the two individuals. had a basal rate equal to 0.989 ---0.0311 cm 3 02 Aulacorhynchusprasinus.--The two Emerald g • h • (n = 13), which is 92% of the expected Toucanetshad a meanbody massequal to 174.7 value (mean of the individuals = 89%). Within + 1.47 g (n = 56). The zone of thermoneutrality this zone, no difference (F = 3.33, df = I and extended from 23 to 34øC (Fig. 2B). Basal rate 26, P < 0.08) existedbetween thoseindividuals of metabolismwas 0.854 -+ 0.0186 cm3 02 g in the mass-specificbasal rates. The meanbasal h • (n = 20), or 86% of the value expectedfrom rate of this toucanetwas 0.948 cm3 02 g • h •, mass. No difference (F = 2.51, df = 1 and 18, P which is 89%of the valueexpected from a mean -- 0.13) existed in basal rate between the two mass of 133.0 g. In spite of the differencein individuals studied,which equaled83 (n = 10) body mass,no differencewas found (F = 0.42, and 89% (n = 10) of the valuesexpected from df = i and 26, P = 0.52) between the total basal mass (mean = 86%). The daytime rate of me- rates of the male (123.9 + 4.29 cm3 02 h •) and tabolismin thermoneutralitywas 1.17 + 0.073 the female (128.0 __+4.61 cm3 02 h-•), the mean cm3 02 g • h -• (n = 5). Body temperatureat being 126.0 cm3 02 h-L During daytime, the night was independentof ambienttemperature rate of metabolismin thermoneutralitywas al- below 31øCand equaled38.1 --+0.07øC (n = 41); most twice the night rate: 1.71 _+0.072 cm3 O2 during the day it was higher: 40.5 + 0.30øC(n g • h • øC • (n = 8). Body temperatureduring = 7). Two thermal conductancesare described nighttime equaled36.7 + 0.15øC(n = 31) at am- for the Emerald Toucanet, one that defined the bient temperaturesbelow 31øC,the femalehav- lower limit of thermoneutrality,0.057 _+0.0024 ing a slightly higher body temperature(37.2 ___ cm3 0 2 g • h -• øC • (n = 10), and the other that 0.21øC In = 12]) than the male (36.5 -+ 0.17øC was minimal, 0.046 ---0.0010 cm 3 02 g • h • øC-• [n = 19]). During the day, body temperature (n = 13), which is 79% of the value expected 920 BRIAN K. MCNAB [Auk, Vol. 118

ß ß ß ß •J40•13_ ee be ß ß ß .• /438 IJ.J

5 J36•o

4- Trachyphonusdamaudii

A I I I I I I 10 20 30 4O AMBIENT TEMPERATURE (øC)

0 42 G 0 ,j 0 • -- 0 0 -- 40 n

38 UJ ß ß ß .e e ee 3• • o 34 rn .-.. 4.0 i I •ee" ßee - '

Baillonius bailloni ø I

• 2.0 o o o o

o o o _ . ß ß

ß ß ß

B 0.0 I I I I I • 10 20 30 40 AMBIENT TEMPERATURE (øC) October2001] EnergeticsofAvian Frugivores 921 from mass. Minimal conductance did not differ speciesdo not give an internally consistentpic- (F = 0.10, df = 1 and 11, P = 0.76) between ture of its energetics. those two toucanets. Pteroglossusaracari.--The two Black-Necked Selenideramaculirostris.--The female Spot- Aragarishad a mean massof 200.7 + 2.04 g (n Billed Toucanethad a meanbody massequal to = 33). Thermoneutrality extended from 29 130.5 + 1.51 g (n = 13), whereasthe two males through 35øC (Fig. 3B). Within that tempera- had massesequal to 142.7 + 0.63 g (n = 13) and ture range,basal rate equaled0.794 ___0.054 cm 3 144.5 + 2.26 g (n = 13). The zone of thermo- 02 g-1 h • (n = 11), which is 83% of the value neutrality extendedfrom 24 to 34øC(Fig. 3A). expectedfrom mass;no difference(F -- 0.13,df Within that range, the mass-specificbasal rate = 1 and 9, P -- 0.73) was found in between the was correlated (F -- 15.92, df = 1 and 17, P < two individuals, which had basalrates equal to 0.001) with the sex of the individual: it was 1.35 82 (n = 6) and 85% (n = 5) of the values ex- + 0.020 cm3 02 g-1 h-1 (n = 14) in the two males, pectedfrom mass(mean = 84%). During the which did not differ from each other (F = 1.37, day, rate of metabolismat thosetemperatures df = 1 and 12, P = 0.26), and 1.51 +_0.046 cm3 was 1.20 + 0.103 cm3 02 g-1 h • (n = 3), which O2 g • h 1 (n = 5) in the female.The two males was different (F = 12.00, df -- 1 and 12, P = had rates that were 127 (n = 7) and 130% (n = 0.005) from nocturnal measurements.Body 7) of valuesexpected from massand that of the temperaturewas independentof ambienttem- female was 141% of the expectedrate. If those peraturesbetween 11 and 35øC:at night it was are basalrates, they are by far the highest,cor- 38.0 + 0.17øC(n = 17), whereasduring the day rected for body mass, measuredin toucans. it was appreciablyhigher, 40.9 + 0.28øC(n = Even after measurementswith overt activity 13). Thermal conductancewas highly variable, were excluded, rate of metabolism in thermo- with no differencebetween day and night (F = neutrality during the day was higher than at 2.89, df = 1 and 15, P = 0.11) or between the night: 1.69 _+0.007 cm3 02 g-• h -1 (n = 4). two individuals (F = 2.16, df = 1 and 5, P = Nocturnalbody temperaturesin that species 0.20); the mean equaled0.089 + 0.0055cm 3 02 were highly variable, especially within the g-1 h-1 øC 1 (n = 17), which is 148% of value zone of thermoneutrality(Fig. 3A): at ambient expectedfrom body mass. temperaturesbetween 17 and 35øC,body tem- Ramphastos dicolorus.--The Red-Breasted perature was bimodally distributed, either Toucanwas representedby two individuals, a equaling 38.6 + 0.16øC(n = 12), which was al- male, whose masswas 356.7 + 2.08 g (n = 22), ways associatedwith a lower rate of metabo- and a female,whose masswas 301.1 ___0.52 g lism, or 41.1 ___0.10 øC (n = 23), which was as- (n -- 22). All measurementswere madeat night. sociatedeither with a higher or a lower rate of The zone of thermoneutralityextended from 20 metabolism.At ambient temperatures<15øC, to 34øC(Fig. 4A). Within that range no differ- nocturnalbody temperaturewas 41.2 + 0.26øC ence (F = 0.17, df = 1 and 25, P = 0.69) in mass- (n = 6) (Fig. 3A). The higher nocturnal body specificbasal rate was seenbetween the two in- temperatures were similar to diurnal body dividuals in spite of differencein their masses; temperatures,which equaled 41.1 + 0.15øC(n mean basal rate was 0.689 _+0.0229 cm 3 02 g-• = 4). Thermal conductance in the female was h • (n = 27), which is 82% of the valueexpected 0.095 _+ 0.0019 cm3 02 g-1 h-1 oC-1 (n = 4), from an averagemass of 328.9 g, or 83% of the which is 128%of the valueexpected from mass, value expectedfrom the male'smass and 81% whereas in the males it was 0.084 _+ 0.0040 cm 3 of the value expectedfrom the female'smass. 02 g-1 h-1 oC-• (n = 8), which is 118% of the Thosedata mean that the male had a higher (F expected value (Fig. 3A). The data from this = 5.29, df = 1 and 25, P = 0.03) total rate (242.9

FIG. 1. Body temperatureand rate of metabolismas a functionof ambienttemperature in (A) two D'arnaud's Ground-Barbets(Trachyphonus darnaudii) and (B) two Saffron Toucanets(Baillonius bailloni). At ambienttemperatures below thermoneutrality,thermal conductanceequals the slopeof the fitted curves,the valueof whichis indicatedby a decimalfraction. All measurementsin theground-barbet were made at night. 922 BRIANK. MCNAB [Auk,Vol. 118

•; 2.0

•; 1.0

A n- 0.0

2.0 o o ,-•?;,, • • • • • • • • •," 136 0

Aulacorhynchusprasinus 0 , , o ' Ioday I '•'•'• o o • '•-• 0

0.0 I I I I I I 0 10 20 30 40 AMBIENT TEMPERATURE (øC) FK;,2, Bodytemperature and rate o:[ metabolism as a :[unctiono:[ ambient temperature 'in (A) twoGroove- BilledToucanets (Aulacorhynchus sulcatus) and (B) two EmeraldToucanets (Aulacorhynchus prasinus). October2001] EnergeticsofAvian Frugivores 923

x •'-• n 42 n ß ee C•'e o• •' • .•ß 8o• I I I I , I-' Seleniderao male-daymaculirostris I x ß+ female-daymale-night - e• xfemale-night+ + _

10 20 30 40 AMBIENT TEMPERATURE (øC)

o ¸ o

3.0 i i i i

2.0- o••'•o PteroglossusI aracariI o ß o o

o o 1.0-

o.o 10 2o 3o 4o AMBIENT TEMPERATURE (øC) FIG.3. Bodytemperature and rate of metabolismas a functionof ambienttemperature in (A) threeSpot- Billed Toucanets(Selenidera maculirostris) and (B) two Black-NeckedAragaris (Pteroglossusaracari). 924 BRIAN K. McN^B [Auk, Vol. 118

2.0 I t I

Ramphastos dicolorus I.1o '.•

1.0 o o oß o o ß ß o

A 0,0_ , 20 30 4O AMBIENT TEMPERATURE (øC)

O0 O0¸ 0 0 0 0 ß 42n o 44p o o oe 38 • ß ß I I I I j 36 rn

Ramphastos tucanus o day .o• o ø ß night

o o o o•oo

B I I I lO 20 30 4O AMBIENT TEMPERATURE (øC) October2001] EnergeticsofAvian Frugivores 925 f

+ 10.34 cm3 0 2 h -1) than the female (209.9 _ association with a low body temperature 9.97 cm3 02 h-i), which raises the question of (34.6øC). which ratesare ecologicallyand evolutionarily Ramphastostoco.--Of two Toco Toucans,the "relevant," total or mass-specific(see McNab male had a mean massequal to 634.8 ___1.00 g 1999). The mean total basal rate in this species (n = 31) and the female had a mass equal to was 226.6 +_ 10.2 cm3 0 2 h -1. 527.6 + 1.00 g (n = 30). Thermoneutralityex- Body temperature at ambient temperatures tended from 17 to 34øC (Fig. 5A). Within that from 9 to 30øCequaled 39.2 + 0.09øC(n; 32). region, rate of metabolismwas quite variable, Thermal conductancein the male was higher (F but because all measurements were made at = 34.81, df = 1 and 12, P -< 0.001) than found in night, variability was not due to a circadian the female (Fig. 4A) 0.046 + 0.002 cm3 02 g-• h -1 rhythm. One factor that contributedto the var- øC 1(n = 6), which is 102%of the value expected iation was that some measurements were made from the male's mass, and 0.034 _+0.002 cm3 0 2 in a 44 L chamber and others in a 329 L cham- g-1 h-• oC-1(n = 6), or 69%of the valueexpected ber. The pooled rate measured in the large from the female's mass. If a unified conductance chamberwas 0.750 + 0.022 cm3 0 2 g-• h • (n = is preferred, it would be 0.041 cm3 0 2 g-1 h-t 19), which is 105% of the rate expectedfrom a øC 1,which is 87% of the value expectedfrom a mean massof 581.2 g. The mass-specificrate in 328.9 g mass. the male was 0.707 + 0.026 cm3 0 2 g-1 h-1 (rt = Ramphastostucanus.--The two Red-Billed 10), which is 101% of the expectedvalue, and Toucanshad a mean massequal to 420.3 ___8.27 did not differ (F = 3.69, df = I and 17, P = 0.07) g (n ; 24). The zone of thermoneutralityex- from that in the female, which was 0.798 ___ tended from 25 to 35øC (Fig. 4B). Within that 0.040 cm3 0 2 g • h -• (n: 9), or 108% of the ex- range, the basal rate, measured at night, pected value. Total rate of metabolismalso did equaled 0.729 ___0.0154 cm3 0 2 g-• h -• (n = 9), not differ (F = 1.09, df; I and 17, P = 0.31) which is 93% of the value expectedfrom mass. between the individuals, the mean of which is No difference (F; 1.24, df; 1 and 7, P = 0.30) 434.9 + 19.1 cm 3 0 2 h -1. The rate measured in was found in basal rate between the two indi- the smaller chamberwas higher (F = 15.00, df viduals,which had basalrates equal to 92 (n = = I and 36, P = 0.0004) than that measured in 6) and 96% (n = 3) of the expectedvalues (mean the larger chamber:0.873 _+0.022 cm 3 0 2g • h • = 94%). During the day, rate of metabolismin (n ; 19), which is 122% of the expected rate. thermoneutralitywas somewhathigher: 0.780 When those toucans were enclosed in the + 0.0379 cm3 0 2 g • h -• (n = 7). Body temper- smaller chamber,they appearedto show much ature at ambient temperaturesbetween 10 and anxiety;the bestestimate of basalrate therefore 33øCwas higher during the day (39.6 + 0.32øC was measuredin the large chamber.Body tem- [n = 15]) than at night (36.8 + 0.41øC[n = 9]; perature at ambient temperaturesbetween 0 Fig. 4B). Conductancethat correspondedto the and 32øCalso reflected chamber size, although, lower limit of thermoneutrality was 0.061 + surprisingly, body temperature was slightly 0.0034 cm3 0 2 g • h i oc-t (n = 6). At ambient lower (F = 23.44, df; I and 56, P -< 0.0001) in temperatures<17øC, thermal conductancede- the smaller chamber: 38.6 --- 0.08øC (n ; 29) creasedto 0.045 + 0.0009 cm3 0 2g • h • øC-1(n versus 39.1 + 0.08øC (n = 28). = 3), which is 101% of the expectedvalue and Chamber size had a qualitativeeffect on rate 74% of the conductance that defines the lower of metabolismat temperaturesbelow thermo- limit of thermoneutrality.A similar value,0.042 neutrality (Fig. 5A). When Toco Toucanswere cm3 0 2 g-• h -1 øC-•, was measuredat 11øCin confined in the smaller chamber at ambient

FIG. 4. Body temperatureand rate of metabolismas a function of ambienttemperature in (A) two Red- BreastedToucans (Ramphastos dicolorus) and (B) two Red-Billed Toucans(Ramphastos tucanus). Circled points in the case of R. tucanus indicate rates that were associated with an intermediate thermal conductance. All measurementsin R. discoloruswere made at night; no differencewas found in rate of metabolismbetween the male and female in thermoneutrality,but a marked differencewas found at lower ambienttemperatures (see text). 926 BRIANK. MCNAB [Auk, Vol. 118

13_ •o ø¸ o • _._._•4240L...• •__• o - I--

1.5 i I I • • • / •36 O

• . o largechamber ee ß smallchamber 1.0 • IRamphastøstøcø .0,,•..•_• .,O 0 ' ' 'ß ' ' ..po o • 0 0

0.,50 • ...aJ•.-"

o.oA ,. , , , i i 0 10 20 30 4O AMBIENT TEMPERATURE (C)

O 40 • o o ø© 1.5 I I I I I /'"'•//q38 • o Aceros plicatus

1.0

o o• o 0.5 o • /o 0% o _ oß O,• •0 O O v-

B 0.C i i I I I I 20 30 40 AMBIENT TEMPERATURE (øC)

FIG.5. Bodytemperature and rate of metabolismas a functionof ambienttemperature in (A) two Toco Toucans(Ramphastos toco) and (B) two Blyth'sHornbills (Acerosplicalus). All measurementsin thosespecies were made at night. October2001] EnergeticsofAvian Frugivores 927 temperaturesbetween 5 and 17øC,thermal con- 4), which is 104% of the value expectedfrom ductanceequaled 0.041 _+0.0014 cm 3 0 2g-• h -1 the mean mass. øC-• (n = 12), which is 116% of the value expected from the mean mass;the two individ- DISCUSSION uals showed no difference in that thermal conductance (F = 0.05, df = I and 10, P = 0.83). In An examination of the variation in the stan- the larger chamberat temperaturesbetween 12 dard energeticsof toucansand other nonpas- and 17øC, rate of metabolism conformed to a serinefrugivores throws somelight on factors conductanceequal to 0.034 _+0.0026 cm 3 02 g-1 that influence basal rate of metabolism and h-1 oC I (/• = 7), which is 96% of the expected thermal conductancein those species.Those value; no difference (F = 0.64, df = i and 5, P factorspotentially include body size,food hab- = 0.46) was found between the two individu- its, climate, and taxonomic affiliation. als. At lower ambient temperatures,however, The responseof somespecies to experimen- rate of metabolismfell without any decreasein tal procedurescomplicates analysis of their en- corebody temperaturewhen toucanswere con- ergetics.That wasmost apparent in two species fined in the large chamber(Fig. 5A). The de- containedin this study. $elenideramaculirostris creasein rate of metabolismat low tempera- showed a bimodal distribution of nocturnal tures reflected a radical reduction in thermal body temperatures,the modesappearing to re- conductance: at 0.8øC, thermal conductance flect two distinct physiologicalstates. At am- was approximatelyequal to 0.010cm 3 02 g • h • bient temperatures<17øC, body temperature øC-• (n = 2), which is only 28% of the value ex- conformed to the higher level (Fig. 3A), al- pected from mass.The responseto low ambi- thoughmeasured at night. Thosedata raise the ent temperatures was reversed by shifting possibilitythat the unusuallyhigh rate of me- those toucans from one chamber to the other, tabolism in thermoneutralityin this toucan is and then restored to the original value by re- not representativeof standardconditions, and turning the toucanto the original chamber,so therefore is not basal (see McNab 1997). Clear- the responseof that toucanto low ambienttem- ly, more work on the energeticsof that species peraturesreflected chamber size. is required to clarify its characteristics.Second- Acerosplicatus.--Of the two Blyth'sHornbills ly, the high rates of metabolismof Ramphastos studied here, the female weighed 1,597.4 ___ tocomeasured in the smallerchamber may have 13.69g (n = 16) and the male 1,965.8+ 35.91 g reflectedanxiety produced by the enclosureof (n = 6). All measurementswere made at night. those large toucansinto a small space.Those Thermoneutrality extended from 35 to 16øC concerns have led to the exclusion of the mea- (Fig. 5B), or possiblyas low as 12øCin the fe- surements on $. maculirostris and the small male. Within this zone the basal rate of the fe- chamber values for R. toco in the following male was 0.495 + 0.0161cm 3 O2 g-• h -• (n = 13), analyses. which is 91% of the expectedvalue, and is Basalrate of metabolism.--Dataon basal rate, greater(F = 5.72, df = 1 and 16, P = 0.03) than climate, food habits, and family affiliation in that of the male, which was 0.430 + 0.0090 cm 3 sevenspecies of toucans,Trachyphonus, and oth- O2g • h t (n = 5), or 83% of the expectedvalue. er members of the order Piciformes, namely Therefore,in spite of a larger mass,the total woodpeckers,as well as tropical nonpasserine basal rate in the male (838.7 + 39.8 cm3 02 h -1) frugivores, including mousebirds and the was not different (F = 1.08, df = I and 16, P = hornbill, are summarized in Table 1. The basal 0.31) from that of the female (789.9 _+24.7 cm3 metabolic rates of toucans were similar to, or 02 h-•), the mean of which is 814.3 + 28.9 cm3 somewhatbelow, those predicted by the As- 02 h -1. For interspecificcomparisons, the mean choff-Pohlcurve for birds otherthan passerines basalrate would be 0.463 cm3 02 g I h 1,which and well below the Reynolds-Leecurve (Fig. 6). is 87% of the rate expectedfrom the meanmass The greatestreduction in basal rate compared of 1,781.6 g. Body temperature was 40.1 _+ with standardcurves was found in Trachyphon- 0.11øC(n ; 21) at ambient temperaturesbe- us, which is only 71% of the value expected tween 6 and 33øC. The thermal conductance from mass by the Aschoff-Pohlcurve. Blyth's that definedthe lower limit of thermoneutrality Hornbill, which is frugivorous, and mouse- equaled0.021 + 0.0009cm 3 0 2g-• h -• øC-• (n = birds (Urocoliusmacrourus, Colius striatus, and 928 BRIAN K. MCNAB iAuk, Vol. 118

o o o October2001] EnergeticsofAvian Frugivores 929

3 _ t Ramphastidae .,.,,_.2 ß Lybiidae Reynolds-Lee

/XO BucerotidaePicidae ., •/ ...XX.. ,.,. --.,• Aschoff-Pohl _ [3Ccliidae ...... • _

2 3

LogloBody Mass (g)

FIG. 6. Log•0total basal rate of metabolismas a functionof log•0body massin frugivorousnonpasserine birds. The standardcurve for birds otherthan passerines(Aschoff and Poh11970)and for all birds (Reynolds and Lee 1996) are included. Data from Table i and referencestherein.

C. castanotus),which feed on fruit and leaves, with omnivorous,frugivorous, or fruits mixed had basal rates (Bartholomew and Trost 1970, with other foods have indistinguishablebasal Prinzinger et al. 1981, Prinzinger 1988), com- rates, as do those with insectivorous and acorn- pared to the Aschoff-Pohlcurve, similar to, or eating habits. Thus, when those categories below,those found in toucans(Fig. 6). The few were lumped into two categories,log•0 basal data availableon woodpeckers,namely two Eu- rate was correlatedwith log•0body mass (F; ropean species(Jynx torquillo[Wryneck] and 1346.87, df = i and 12, P -< 0.0001) and food Dendrocoposmajor [Great Spotted Woodpeck- habits (F = 82.93, df = 2 and 12, P -< 0.0001; R2 er]) and the North AmericanAcorn Woodpeck- = 0.991).Then, I2o2 (cm 3 O2h-•) •- 4.34f g0.786, er (Melanerpesformicivorous), indicated higher wheref is a dimensionlesscoefficient that rep- basalrates than expectedfrom mass(Kendeigh resentsfood habits. In this analysis,fruit-eat- et al. 1977,Weathers et al. 1990)and muchhigh- ing nonpasserines(toucans, barbet, and horn- er residuals than found in toucans and mou- bill) have basal rates that are 59% of those of sebirds(Fig. 6). acorn- and insect-eating species (woodpeck- In an analysisof covariance,log•0 basal rate ers), that is,f = 1.00 for insectand acorn-eaters was not correlated with climate (P = 0.57), food andf = 0.59 for frugivores. habits (P = 0.99), and family affiliation (P = The analysis of avian frugivory can be ex- 0.30) when entered together with log•0body panded further by comparingenergetics of tou- mass. However, when food habits were entered cans,barbet, and hornbill with frugivorouspi- alonewith mass,log•0 basal rate was correlated geonsand doves(Columbidae) and flying foxes both with log•0body mass (F = 552.83, df = 1 (Pteropodidae).Those data, expressedin mass- and 8, P -< 0.0001) and food habits (F = 21.59, specificrates, are shown as a function of body df = 5 and 8, P = 0.0002;R 2 = 0.995). Species massin Figure 7. Notice that the data generally 930 BRIANK. MCNAB [Auk, Vol. 118

1.0 ß pigeons -; ß toucans : 0 pteropodidsI • hornbill '

0.0

all-mammal 4

Log10Body Mass (g)

FIG.7. Log10mass-specific basal rate of metabolismas a functionof logs0body massin a barbet,toucans, a hornbill,fruit pigeons(McNab 2000),and flying foxes(McNab and Armstrong2001, McNab and Bonac- corso2001). Also indicatedare the standardcurves for nonpasserines(Aschoff and Pohl 1970),all birds (Reynoldsand Lee 1996), and all mammals (McNab 1988a). fall on top of eachother. The highestresiduals tively; R2 = 0.993).No differencewas foundbe- in basalrate are foundin two flying foxes(Dob- tween tropical, subtropical,subtropical-tropi- sonia moluccensisand Pteropus vampyrus), cal, or arid distributed species. Therefore, whereas the lowest residual is found in a fruit climateswere lumped into two categories,tem- pigeon (Duculapacifica)--that is, the highest perate and tropical. When climate categories basal rates are found in two mammals and the were just two, log•0basal rate is correlatedwith lowest in a bird. In other words, no clear dif- log•0body mass(F = 1346.87,df = i and 12, P ferencesexist in basalrates among frugivorous -< 0.0001) and climate (F = 82.93, df = 1 and 1; toucans,pigeons, and flying foxes(except for R2 = 0.991).Then, I2o2 (cm302 h-•) = 2.55c g0.786, somesmall [<40 g] nectarivorouspteropodids where temperate specieshave c = 1.70 and that were not includedbecause they are impre- tropical speciesc = 1.00. The difficulty with cisethermoregulators [Bonaccorso and McNab that conclusion is that food habits and climate 1997, McNab and Bonaccorso2001]). Nonpas- in this sampleare not independent:temperate serinefrugivores thus have low basal rates,al- speciesare insect-or acorn-eaters,whereas the thoughthe largesttoucans tend to haveslightly tropicaland subtropicalspecies are frugivores. higher basal rates,possibly associated with an So,the high basalrate of temperatewoodpeck- increasedlevel of carnivory. ers may simply reflect their insectivorousor When climateis enteredonly with log•0body acorn-eatinghabits. (In this context,measure- mass,log•0 basal rate correlatedwith both fac- ments of energeticsof fruit-eating, neotropical tors (F = 21.20, df = 4 and 9, P -< 0.0001, and woodpeckers,such as some speciesof Celeus F = 1050.75,df = i and 9, P -< 0.0001,respec- and Melanerpes,would be valuable.) October2001] EnergeticsofAvian Frugivores 931

1.5

1.0

ß Ramphastidae ß Lybiidae zN Bucerotidae [] Coliidae 0.5

I I I I 1 2 3 Logl0Body Mass (g) FIG.8. Log•0total thermal conductance asa functionof logt0 body mass in frugivorousnonpasserine birds. The standard curve of Lasiewski et al. (1967) is indicated.

The parametersof avianenergetics often are in the so-called "comparative method," any also correlatedwith phylogeny(Bennett and more than one can allocatethem exclusivelyto Harvey 1987),but whetherancestry limits per- "behavior" or "ecology." How to partition in- formanceis opento question;besides, no qual- teractions to various causitive agents is pres- ity phylogenyexists for toucans(D. W. Stead- ently unclearand may be impossible.Besides, manpers. comm.). In thisstudy, log•0 basal rate is it possibleto conceiveof a toucanhaving a is correlatedboth with log•0body mass(F = low basal rate becauseof ancestry if the low 354.45; df = i and 9, P -• 0.0001) and family rate is ecologicallyunacceptable? If not, pres- affiliation (F = 28.09; df = i and 4, P -• 0.0001, enceof a low basalrate in toucansis m6relikely R2 = 0.995),but only if climateand food habits relatedto frugivory (and tropicaldistribution) are not included.As pointedout by Westobyet of toucans.The allocationof responsibilityfor al. (1995), many factorsinteract to determine a low basalrate to "phylogeny"needs the dem- complexcharacter states, such as basal rate of onstration of a functional association of that metabolism.For example,most toucanshave character state with the evolution of some dis- massesbetween 130 and 400 g, are frugivorous, tinctivetoucan property (e.g. the evolutionof aretropical in distribution,and have low basal high basal ratesin mammalswas associated rates, as well as sharing a common ancestry, with evolution of a eutherian form of repro- whereas studied woodpeckersare insect- or duction;McNab 1986,Lillegraven et al. 1987). acorn-eating,are temperate in distribution,and Minimal thermalconductance.--The only fac- havehigh basal rates. One cannot allocate those tor influencing1og•0 minimal thermal conduc- interactingcharacter states exclusively to phy- tance amongseven toucans, a barbet, a horn- logenyas the causitiveagent, as is usuallydone bill, and three tropical mousebirdswas log•0 932 BRIANK. MCNAB [Auk, Vol. 118 body mass (F = 116.04, df = i and 10, P -< BARTHOLOMEW, G. A., AND C. H. TROST. 1970. Tem- 0.0001; R2 = 0.921; Fig. 8). perature regulation in the SpeckledMousebird, A remarkable observation on conductance Colius striatus. Condor 72:141-146. BENNETT, P.M., AND P. H. HARVEY. 1987. Active and was that the largesttoucan, R. toco,had a very low thermal conductance at low environmental resting metabolismin birds: Allometry,phylogeny and ecology.Journal of Zoology (London) temperatures,thereby reducingheat loss and 213:327-363. rate of metabolism without decreasingcore BOINSKI,S. 1987. Birth synchronyin squirrel mon- body temperature.A similar reductionin rate keys (Sairnirioerstedi). Behavioral Ecology and of metabolismat low environmentaltempera- Sociobiology21:393-400. tures has been described in some tropical BONACCORSO,E J., AND B. K. MCNAB.1997. Plasticity mammals of intermediatesize, namely a tree- of energeticsin blossombats (Pteropodidae): kangaroo(Dendrolagus matschiei), the red pan- Impact on distribution. Journalof Mammalogy da (Ailurusfulgens),and severalarboreal viver- 78:1073-1088. ELLIS, H. [. 1980. Metabolism and solar radiation in rids (McNab 1988b,1995); it was explaineddue dark and white herons in hot climates.Physio- to an enhanced peripheral vasoconstriction. logicalZoology 53:358-372. The effect of a small chamber, noticed here in ELLIS,H. [. 1984.Energetics of free-rangingseabirds. the TocoToucan, may be due to claustrophobic Pages203-234 in SeabirdEnergetics (G. C. Whit- anxiety and the consequentoccurrence of a tow and H. Rahn, Eds.). Plenum Press, New high heart rate, high blood pressure,and ab- York. senceof peripheral vasoconstriction.Failure to GALETTI,M., R. LAPS,AND M. A. PIzO. 2000. Frugi- reducerate of metabolismat low temperatures vory by toucans(Ramphastidae) at two altitudes as a result of confinement in small chambers in the Atlantic forest of Brazil. Biotropica 32: also has been reported in a lemur (McNab 843-851. KEMP, A. 1995. The Hornbills, Bucerotiformes. Ox- 1988b) and a viverrid (McNab 1995). ford University Press,New York. KENDEIGH, S.C., V. R. DOL'NIK, AND V. M. GAVRILOV. ACKNOWLEDGMENTS 1977.Avian energetics.Pages 127-204 in Graniv- orousBirds in Ecosystems,their Evolution,Pop- I thank the many peoplewho made this work pos- ulations, Energetics,Adaptations, Impact, and sible. They include Sra. M. Santana and Sr. A. San- Control. (J. Pinowskiand S.C. Kendeigh,Eds.). tana, ParqueAmbiental y Zoo16gicoGustavo Rivera, International Biological Programme 12, Cam- Falc6n, Venezuela; Alexis Arends, Universidad Na- bridge University Press, Cambridge, United tional Experimental Francisco de Miranda, Coro, Kingdom. Venezuela;Lex Salisbury,Lowry Park Zoo, Tampa, LASWIESKI,R. C., AND W. R. DAWSON. 1967. A re-ex- Florida; Fred Antonio, Central Florida Zoological amination of the relation between standard met- Park, Lake Monroe, Florida; and Frank J. Bonaccorso, abolicrate and body weight in birds.Condor 69: National Museum and Art Gallery,Boroko, National 13-23. Capital District, PapuaNew Guinea. Measurements LASWIESKI, R. C., W. W. WEATHERS, AND M. H. BER- were made on toucansat the Parque Ambiental y STEIN.1967. Physiological responses of the Giant Zoo16gicoGustavo Rivera, Falc6n, Venezuela; on Hummingbird, Patagonagigas. Comparative Bio- toucansand the barbet in the Department of Zoolo- chemistryand Physiology23:797-813. gy, University of Florida; and on the hornbill at the LILLEGRAVEN,J. A., S. D. THOMPSON, B. K. MCNAB, ChristensenResearch Institute, Madang, PapuaNew ANDJ. L. PATTON.1987. The origin of eutherian Guinea.I greatlyappreciate the criticalevaluation of mammals.Biological Journal of the LinneanSo- this manuscriptby E J.Bonaccorso, D. W. Steadman, ciety 32:281-336. C. Bosque,and two anonymousreviewers. A pre- MCNAB,B. K. 1980.On estimatingthermal conduc- print of a paper on the food habits of toucanswas tance in endotherms.Physiological Zoology 53: kindly given to me by M. Galetti, UniversidadeEs- 145-156. tadual Paulista,Rio Claro, Brazil. The figures in this MCNAB,B. K. 1986.Food habits, energetics,and the article were producedby Laurie Walz and Jennifer reproductionof marsupials.Journal of Zoology Piascik. (London) 208:595-614. MeN^B, B. K. 1988a. Food habits and the basal rate LITERATURE CITED of metabolismin birds. Oecologia77:343-349. MeN^B, B. K. 1988b.Energy conservationin a tree- ASCHOFF,J., AND H. POHL.1970. Rhythmic variations kangaroo (Dendrolagusrnatschiei) and the red in energy metabolism. Federation Proceedings panda (Ailurusfulgens). Physiological Zoology 29:1541-1552. 61:280-292. October2001] EnergeticsofAvian Frugivores 933

MCNAB,B. K. 1988c.Complications inherent in scal- tanotus)during fastingand torpor.Comparative ing the basal rate of metabolism in mammals. Biochemistryand PhysiologyA 69:689-692. Quarterly Review of Biology63:25-54. PRINZINGER,R., AND I. HANSSLER.1980. Metabolism- MCNAB,B. K. 1995.Energy expenditure and conser- weight relationshipin somesmall non-passerine vation in frugivorousand mixed-diet carnivor- birds. Experentia36:1299-1300. ans. Journalof Mammalogy 76:206-222. PRUM,R. O. 1988.Phylogenetic interrelationships of MCNAB,B. K. 1997. On the utility of uniformity in the barbets (Aves: Capitonidae) and toucans the definition of basal rate of metabolism.Phys- (Aves: Ramphastidae) based on morphology iological Zoology 70:718-720. with comparisonsto DNA-DNA hybridization. MCNAB, B. K. 1999. On the comparativeecological Zoological Journal of the Linnean Society 92: and evolutionarysignificance of total and mass- 313-343. specificrates of metabolism.Physiological and REMSEN,J. V., JR., M. A. HYDE, AND A. CHAPMAN. BiochemicalZoology 72:642-644. 1993.The dietsof Neotropicaltrogons, motmots, MCNAB,B. K. 2000. The influenceof body mass,cli- barbets, and toucans. Condor 95:178-192. mate,and distributionon the energeticsof South REYNOLDS,P.S., AND R. M. LEE III. 1996. Phyloge- Pacificpigeons. Comparative Biochemistry and netic analysisof avian energetics:Passerines and Physiology127A :309-329. nonpasserinesdo not differ. American Natural- MCNAB, B. K., AND M. I. ARMSTRONG.2001. Sexual ist 147:735-759. dimorphismand the scalingof energeticsin fly- SCHOLANDER, P. E, R. HOCK, V. WALTERS, F. JOHN- ing foxesof the genusPteropus. Journal of Mam- SON, AND L. IRVING. 1950. Heat regulation in malogy 82:709-720. some Arctic and tropical mammals and birds. MCNAB, B. K., AND F. J. BONACCORSO.1995. The en- BiologicalBulletin 99:237-258. ergeticsof Australasianswifts, frogmouths,and SHORT,L. L., AND J. E M. HORNE. 1980. Ground bar- nightjars.Physiological Zoology 68:245-261. bets of EastAfrica. Living Bird 18:179-186. MCNAB, B. K., AND F. J. BONACCORSO.2001. The me- SKUTCH,A. E 1983. Birds of Tropical America. Uni- tabolismof New Guinean pteropodid bats.Jour- versity of TexasPress, Austin. nal of ComparativePhysiology B 171:201-213. SIBLEY,C. G., AND J. E. AHLQUIST.1990. Phylogeny MITCHELL, C. L., S. BOINSKI, AND C. P. VAN SCHAIK. and Classificationof Birds.A Studyin Molecular 1991. Competitiveregimes and female bonding Evolution. Yale University Press, New Haven, in two speciesof squirrel monkeys(Saimiri oer- Connecticut. stediand S.sciureus). Behavioral Ecology and So- WEATHERS,W. W. 1979. Climatic adaptationin avian ciobiology28:55-60. standard metabolicrate. Oecologia42:81-89. NICKERSON, D. M., D. E. FACEY, AND G. D. GROSS- WEATHERS, W. W., W. D. KOENIG, AND M. T. STAN- MAN. 1989. Estimatingphysiological thresholds BACK.1990. Breeding energeticsand thermal with two-phaseregression. Physiological Zool- ecology of the Acorn Woodpeckerin central ogy 62:866-877. coastal California. Condor 92:341-359. PRINZINGER,R. 1988.Energy metabolism, body-tem- WESTOBY,M., M. R. LEISHMAN, AND J. M. LORD. 1995. peratureand breathingparameters in nontorpid On misinterpreting the "phylogenetic correc- Blue-Naped Mousebirds Urocoliusmacrourus. tion". Journalof Ecology83:531-534. Journal of Comparative Physiology B 157:801- YEAGER,D. P.,AND G. R. ULTSCH.1989. Physiological 806. regulationand conformation:A BASICprogram PRINZINGER, R., R. G(DPPELL,A. LORENZ, AND E. for the determinationof critical points. Physio- KULZER.1981. Body temperatureand metabo- logicalZoology 62:888-907. lism in the Red-Backed Mousebird (Colius cas- AssociateEditor: C. Bosque