Condor 8491-96 0 The Cooper Ornithological Society 1982

ENERGETICS AND SPREAD-WINGED BEHAVIOR OF IN FLORIDA

WILLARD W. HENNEMANN, III

ABSTRACT. -Metabolic rates and body temperatures of four captive Anhingas (Anhingu ) were measured over a range of ambient temperatures. In ad- dition, Anhingaswere observed in the field in an attempt to correlate the frequency of behaviors, such as wing-spreading,with environmental conditions, suchas the intensity of solar radiation and ambient temperature. Anhingas have low basal metabolic rates and high thermal conductancesfor their mass. The frequency of spread-wingedbehavior increasesat higher solar intensities and is inversely cor- related with ambient temperature. Anhingas assuming spread-winged postures orient with their backs to the sun, maximizing the surface area exposed to inso- lation and maintaining an angle of incidence of approximately 90”. In contrast, engagedin gular flutter face into the sun, thereby minimizing the surface area exposedto insolation and increasingthe angle of incidence. Anhingas spread their wings to dry the plumage and to absorb solar energy during cool weather. They thereby supplement their low metabolic rates, compensate for their high thermal conductances,and replace heat lost via evaporation and convection due to wetting of the plumage. _

Although spread-winged behavior is well surements were made in a constant-tempera- known in , its function is not estab- ture chamber with a volume of 326 1, through lished (Clark 1969, Kennedy 1969). Suggested which air flowed at rates of 4.4 to 6.8 l/min. functions include the synthesisof vitamin D (Flow rates were measured with a Brooks (Kennedy 1968, 1969); the removal of ecto- R215-B rotameter after the removal of CO, parasites (Kennedy 1969); skin conditioning and water; all values were corrected to STPD.) during molt (Potter and Hauser 1974); feather I measured metabolic rates when the subjects maintenance (Kahl 197 1); wing-drying (Rijke were in a postabsorptive state between the 1968, Siegfried et al. 1975, Berry 1976); and hours of 09:OO and 18:00 during fall, 1979. thermoregulation (Heath 1962, Clark 1969, Metabolic runs were terminated only after Kennedy 1969). The wing-drying hypothesis steady state conditions had been reached is especially attractive because Casler (1973) (Christensen 1947, Heusner 1955). I moni- and Rijke (1968) found that thesebirds possess tored the birds’ deep body temperaturesduring wettable plumage, which allows water to pen- metabolic measurements using Mini-mitters etrate the air spacesnext to the skin. This re- (Mini-mitter Co., Indianapolis, IN): minia- duces their buoyancy and facilitates under- ture, temperature-sensitive, AM radio trans- water stalking of prey. mitters, which were coated with beeswax and I examined the metabolism of Anhingas fed to the subjects. I also measured cloaca1 (Anhingu anhingu) in the laboratory, and their temperatures after each metabolic trial using behavior in the field, to determine if the birds a Schultheis thermometer. Both methods of use spread-winged behavior in drying the measuring body temperature yielded statisti- wings or in thermoregulation. cally similar results and data obtained by the two methods are used interchangeably METHODS throughout the text. I hand-raisedfour Anhingas(two adult females Mean basal rates of metabolism were esti- with an averageweight of 1.12 kg; two juvenile mated from independent measurementsof the females averaging0.93 kg), housedthem in an minimal rate of oxygen consumption within outdoor aviary in which they were exposed the region of thermoneutrality. At tempera- and became acclimated to weather conditions tures below thermoneutrality, I estimated of Gainesville, Florida, and maintained them minimal thermal conductances(C,) for each on a diet of whole smelt and minnows. I mea- from the relationship C, = M/(T, - T,) sured temperature-specific rates of metabo- using simultaneous measurementsof its met- lism for each individual over a range of am- abolic rate (M), body temperature (T,,), and bient temperatures (0 to 40°C) in an open flow ambient temperature (T,). I then compared system employing a Beckman G2 or Applied measuredvalues for mean basalmetabolic rate Electrochemistry S-3A oxygen analyzer. Mea- (Mb) with the value expected from body mass

1911 92 WILLARD W. HENNEMANN, III

ometer. The behavior recorded at the end of each 5-min period was considered one obser- vation of that behavior. If the individual under observation behaved in more than one way during the 5-min period, I recorded the frac- tion of time devoted to each action. I then calculated the frequency of each behavior in percent, i.e., (no. of observations of the be- havior)/(total no. of observations) X 100, un- der each specific set of environmental condi- tions. For example, if I made 20 observations of Anhingas while the intensity of solar radia- tion was below 222 W/m2 and T, was between 0 and 5°C and if birds were preening during two of them, then preening was considered to represent 10% of the behavior observed under those environmental conditions. I then ex- amined thesedata for correlations between the FIGURE 1. The relationshipof metabolic rate and body frequency of certain behaviors and environ- temperature of four Anhingas to environmental temper- mental conditions. ature. Measurementswere made during daylight hours in fall, 1979. Each symbol representsone measurement for RESULTS one individual. The dashedline describesthe relationship for juveniles, the solid for adults. The number above each The relationship between the metabolic rates line is its slope. and body temperatures of adult and juvenile Anhingas and ambient temperature is pre- sentedin Figure 1. Basal metabolic rates, ther- (W) for a nonpasserinein the active phase of mal conductances,and data concerning body its diurnal cycle, usingthe equation of Aschoff temperature are listed in Table 1. Daytime T, and Pohl(l970) M, = 5.14 W-o.27in which M, ranged from 38.5 to 42.O”Cin both age groups. is in cm3 0,/g. h and W is in g. I also compared I did not see gular flutter during any of the the mean measuredvalues of C, with the value metabolic trials. The slopes of the regression predicted for a bird of this body mass using lines fitting the data presentedin Figure 1 can- the equation of Lasiewski et al. (1967) not be used as estimates of thermal conduc- C/W = 0.85 Wmo.5in’ which C’ is “wet ther- tance because they do not extrapolate to T,. mal conductance” in cm3 0,/g. h.“C (McNab This suggeststhat Anhingas combine physical 1974); W is in g. (i.e., postural and plumage adjustments) with I watched the activities of a resident popu- chemical (i.e., metabolic) thermoregulation lation of 10-l 5 Anhingas at Lake Alice in (McNab 1980). Gainesville. Observations were made at all At low T,, both in the field and during met- times of the day and in all seasonsduring 1978 abolic trials, Anhingas commonly folded their and 1979. At 5-min intervals I recorded (1) necks in a tight S-shape and pressed them the amount of time that each bird under ob- tightly against the upper body. Occasionally, servation spent in such activities as wing- birds laid the head and neck along the back spreading,preening, foraging, and gular flutter, and covered them with the wings. Both pos- (2) the orientation of each individual with re- tures should reduce thermal conductance by spect to sun and wind, (3) ambient (shade) reducing surfacearea and I, therefore, labelled temperature, and (4) the intensity of solar ra- them heat conservingpostures (Table 2). diation measured with an Epply PSP radi- My field observations are summarized in

TABLE 1. Thermoregulation in adult and juvenile Anhingas.*

M,asa c, as a Mb percentof C,. percentof (cm’O>/g-h) predictedM,** (2) (cm’O>/g.h .“C) predictedC,***

Adults 0.59 77 39.1 28.0 0.0446 190 Juveniles 0.60 74 39.9 32.0 0.040 1 177 Combined 0.59 76 39.8 29.5 0.0454 184

* M, = basalmetabolic rate, T, = body temperature,T,, = lower critical temperature,C, = minimal thermal conductance. ** M, predictedby the equatmnof Aschoffand Pohl (1970). *** C, predictedby the equationof Lasiewskiet al. (1967). ANHINGA WING-SPREADING 93

Table 2. Behavior in general, and wing-spread- ing in particular, was highly dependent on weather. Anhingas assumed spread-winged postures much more frequently during cool than during warm weather. The frequency of this behavior was negatively correlated with T, (Fig. 2, Y = -0.85, P < 0.01, 9 df), but pos- itively influenced by the intensity of solar ra- diation (Table 2). When the intensity was less than 222 W/m* (cloudy days with mostly dif- fuse radiation), wing-spreading constituted only 3.8% of the observed behavior, but it ac- counted for 47.1% of the latter when intensity exceeded 222 W/m* (T, < 30°C in both cases, P < 0.001, test for the equality of two per- centages,Sokal and Rohlf 1969). When solar FIGURE 2. Frequency with which Anhingas orient away from the sun and exhibit spread-winged behavior radiation was low (e.g., during early evening), as a function of environmental temperature. The behav- birds just emerging from the water often ioral frequency(ordinate) is the percent of the bird’s total fanned their wings and tail continuously (fan behavior devoted to orienting away from the sun or drying, Table 2) rather than hold them steady spreadingits wings within a specificrange of temperatures (as was usually the case). (abscissa).Each symbol representsthe mean of at least 30 observations. The upper line describesthe frequency of Thirty-four percent of the observed wing- orientation away from the sun; the lower line the fre- spreading involved birds that appeared to be quency of spread-wingedbehavior. The equation above wet, 49% involved individuals that appeared each line describesthe relationship. The half-filled sym- to be dry, and 17% involved individuals whose bol indicates that the frequenciesof both behaviors were plumage condition could not be determined. identical. Wet birds usually held their wings completely outstretched, whereas dry individuals often Anhingas oriented differently while heat- held them closer to the body or occasionally stressed(gular fluttering), facing into the sun drooped at their sides. During such behavior 53% of the time, obliquely (side to the sun) Anhingas commonly held their wings and 36% of the time, and away from the sun only body perpendicular to the incident sunlight, 11% of the time (55 observations). This ori- flattening out when the sun was high, standing entation is statisticallynonrandom (x2 = 43.12, erect when the sunwas low. Most birds spread- P < 0.001, 2 df). Gular flutter occurred when ing their wings also oriented with their backs T, was between 25 and 30°C but only when to the sun: in only 16 or 3.4% of 468 obser- the intensity of solar radiation exceeded 700 vations did they orient differently. In these 16 W/m2. All other observations of gular flutter cases,all of the birds had wet plumage, most occurred when T, was above 30°C. It was sig- (14/l 6) were facing into the wind, and most nificantly more common in bright sunlight, ( 14/ 16) were noted when the intensity of solar being 5.6% of the observed behavior when ra- radiation was less than 222 W/m*. The fre- diation was less than 222 W/m* as compared quency with which Anhingas oriented with to 23.8% of the behavior when it exceeded222 backsto the sun was also negatively correlated W/m* (T, > 30°C; test for the equality of two with T, (Fig. 2; r = -0.83, P < 0.01, 9 df). percentages, P < 0.01). Gular flutter was

TABLE 2. Frequenciesfor the behavior of Anhingas under various environmental conditions.

Environmental conditions Frequency of behaviors (%) Solar Total Sunning Using Using heat radiation number of and gular conserving Fan (W/m’) (2) observations Sunnina” preeninC Preeninp. Perchine flutter Swimmine “Ostllres dwine

>222 <30 860 40.8* 6.3 9.1 22.1 2.6 16.9 1.6 0.0 <222 <30 387 1.6 2.2 13.0 43.6 0.0 5.4* 31.8* 2.5* >222 >30 111 8.8 0.0 17.1 32.0 23.8* 18.2 0.0 0.0 <222 >30 25 10.0 0.0 28.0 39.5 5.6 16.9 0.0 0.0

1,383 26.5 4.6 11.8 29.1 3.5 14.0 9.4 1.1

s Behavioral frequency refers to the incidence of a specific behavior expressed as a percentage of the total number of behavioral observations (see text for further explanation). DSunning refers to spread-winged behavior in the absence of preening. c Sunnrng and preening refers to spread-winged behavior accompanied by preening. *Significantly different (P < 0.05) from the other percentages in the column (test for the equality of two percentages, Sokal and Rohlf 1969). 94 WILLARD W. HENNEMANN, III never accompanied by wing-spreading. Ori- the plumage rapidly when sunlight is unavail- entation during gular flutter (predominately able. Anhingas often expose the skin during facing the sun) was similar to that during ex- wing-spreading by erecting wing and back posure to strong winds in which individuals feathers, especially while wet. Rapid drying usually faced into the wind, thereby exposing should be facilitated by a combination of a minimum of surface. feather erection, the black feather pigmenta- Anhingas spent only 14.0% of the time for- tion, and exposure of the maximum surface aging in the water during the study. Most of area to the sun via wing extension and ori- their time (more than 80% of the total obser- entation. Although the remiges may not ad- vations) was spent perching (preening, sun- sorb water to the extent of the breast feathers ning, or just sitting). They spent significantly studied by Casler (1973) and Rijke (1968), the lesstime in the water when T, was below 25°C wing coverts may be wettable and it may be (9.8% of 783 total observations) than when it these which a bird is attempting to dry by fan- exceeded 25°C (18.8% of 600 total obser- ning when the sky is overcast. Although the vations, test for the equality of two percent- wingsappear water laden when birds leave the ages,P < 0.001). water, the contention that Anhingas must dry After leaving the water, Anhingas shook ex- their plumage before they can fly (McAtee and cesswater from their plumage and almost in- Stoddard 1945) is not supported by my ob- variably spread their wings. However, they servations of birds flying, albeit weakly, from also spread their wings before entering the the water’s surface. This suggeststhat wing- water (with dry plumage). Seventy-one percent drying serves primarily as an adaptation to of all trips to the water, and 90% of all such conserve metabolic heat. trips when solar radiation exceeded 222 Other data suggestthat spread-winged be- W/m2, was preceded by wing-spreading.This havior also has an overlapping thermoregu- value increasedto 94% when consideringonly latory function: those occasionswhen T, was below 30°C and (1) Dry Anhingas spread their wings more solar radiation exceeded 222 W/m2. often than do wet birds. (2) The frequency of this behavior was in- DISCUSSION versely correlated with T, and might comprise My observations indicate that Anhingas are as much as 88% of the behaviors observed highly effective predators, often capturing when T, wasless than 10°C (Fig. 2). In contrast, three to five fish in a period of 5 to 10 min. the frequency was only 15% at temperatures This may explain why most of their time is above 30°C. devoted to activities other than foraging. An- (3) Anhingas are jet black on the back and hingas, therefore, can be considered as “time wings. Dark plumage may absorb more solar minimizers” with respect to feeding habits radiation than pale (Hamilton and Heppner (Schoener 197 1). Their reduced buoyancy, due 1967, Lustick 1969) depending on the wind to water-absorbing plumage, may improve speed (Walsberg et al. 1978) and the angle of their efficiency as underwater predators by al- incidence (Lustick 1980). That Anhingas nor- lowing them to remain relatively motionless mally gain heat by means of insolation is ob- underwater. However, the temporary increase vious because gular flutter appeared in free- in thermal conductancethat must result from living birds when T, was between 25 and having a layer of water adjacent the skin must 27.5”C and the intensity of solar radiation was subject them to thermoregulatory stress, es- moderate, but did not appear in my captive pecially on cold days (Mahoney, in press;Hen- birds even when chamber temperatures ex- nemann, pers. observ.), a predicament similar ceeded 3 7°C. to that suffered by land birds exposedto heavy (4) Most birds oriented away from the sun rains (Kennedy 1970). This too may explain (and perpendicular to the incident radiation) why Anhingas spendso little time in the water, while spreadingtheir wings, but faced the sun especially during cool weather. Considering during heat stress(gular flutter). Orienting per- the high rates of heat loss that must occur in pendicular to incident radiation would enable individuals with water-saturated plumage, it Anhingas to take full advantage of the heat would be energetically advantageousfor An- absorbing qualities of the black plumage by hingas to dry as quickly as possible after for- increasingthe surface exposed to the sun and aging,thereby re-establishinga layer of air next by optimizing the angle of incidence (Lus- to the skin. Wing-spreading may be a behav- tick 1980). Orienting toward the sun would ioral adaptation to this end. Such a function reduce the amount of surface exposed to it, as is suggestedby the appearanceof this behavior demonstrated for Herring Gulls (Larus argen- in wet birds on overcast days, and by their fan- tutus; Lustick et al. 1978), and maximize the drying behavior, which almost certainly dries angleof incidence, making the birds effectively ANHINGA WING-SPREADING 95 white with respectto the absorption of radiant be required for this purpose, and the appear- energy (Lustick 1980). ance of wing-spreading is not in any way as- (5) Anhingashave a high C, in air (I 77-l 90% sociated with soaring. of that predicted for a nonpasserine of their Although wing-spreading apparently has mass; Table 1) and this increaseswhen they both thermoregulatory and wing-drying func- are wet. This high rate of heat loss could be tions in Anhingas, it may have different func- offset by the absorption of radiant energy dur- tions in other birds. (The use of such behavior ing insolation. for thermoregulation in Anhingas may be an (6) Anhingas also have a low M, (74-77% evolutionary result of the development of of that predicted for a nonpasserine of their water adsorbing plumage as a foraging aid.) mass;Table l), which may reduce heat stress also have plumage that holds in summer, but may be maladaptive in cool water, but only the distal portions of their or wet weather. feathers are wettable, and they consequently (7) Anhingas spread their wings before en- maintain a layer of air next to the skin while tering the water. This may enable them to store foraging (Casler 1973). This difference in heat passively to offset the high thermal con- feather structure affords cormorants greater ductance associatedwith wetting. insulation while immersed and may explain (8) Anhingas have a high lower critical tem- why they appear to spend more time in the perature (Table 1, Fig. 1) and probably en- water than do Anhingas. This, plus data on the counter temperatures below thermoneutrality Cape (Phalucrocorax capensis; for most of the winter in northern Florida. Siegfried et al. 1975) on four speciesof South Consequently, they may have difficulty mak- African cormorants (P. capensis, P. neglectus, ing up heat lossesvia metabolism and balanc- P. lucidus, and P. africanus; Berry 1976), and ing their energy budgetsin many parts of their my personal observations of the Double- winter range. crested Cormorant (P. auritus), all of which That birds can take advantage of solar ra- indicate that cormorants orient into the wind diation to reduce the metabolic costs of en- when they assume spread-winged postures, dothermy is well documented (Morton 1967; suggestthat such behavior functions only in Heppner 1969, 1970; Lustick 1969, 1971; drying the wings. Ohmart and Lasiewski 1971; De Jong 1976). The high C, and wettable plumage of Anhin- ACKNOWLEDGMENTS gas may cause thermoregulatory problems I thank J. Anderson, J. Kaufmann, C. Lanciani, S. Thomp- during cool or wet periods, but may be bene- son, and B. McNab for their advice and criticisms of both ficial in preventing heat stressduring tropical the work and the manuscript. This paper representspart of a larger study funded by grants from the Frank M. summers. The spread-wingedposture offers a Chapman Memorial Fund, Sigma Xi, and the National metabolically inexpensive, behavioral means ScienceFoundation (DEB-7906056). of reducing heat loss (thermal conductance). In Gainesville, Anhingas are near the northern LITERATURE CITED limit of their winter range (Palmer 1962). The data on Mb and C, presented here (Fig. 1) ASCHOFF, J., AND H. POHL. 1970. Der Ruheumsatz von Vbgeln als Funktion der Tageszeit und der KBrper- suggestthat Anhinga is poorly adapted to cool g&se. J. Omithol. 111:3847. climates and may explain why it only occurs BERRY,H. H. 1976. Physiologicaland behavioral ecology in tropical and subtropical areas. Spread- of the Cape Cormorant Phalacrocoraxcapensis. Ma- winged behavior may be an essentialadapta- doqua 9:5-55. CASLER,C. L. 1973. The air-sac systemsand buoyancy tion in balancing the energy budgets of these of the Anhinga and Double-crestedCormorant. Auk birds, supplementing metabolic energy with 90:324-340. insolation, compensating for heat lost via CHRISTENSEN,B. G. 1947. A simple apparatus for si- evaporation and convection due to the wetting multaneousmeasurement of the standardmetabolism of the plumage, and reducing daily energy re- and the respiratory quotient in small laboratory an- imals, supplementedwith investigations on normal quirements so that lesstime need be spent for- and hypophysectomisedrats. Acta Physiol. Stand. aging (i.e., wet). 12:l l-16. Wing-spreading of Anhingas is not a mech- CLARK,G. A., JR. 1969. Spread-wing postures in Pel- anism for “dumping heat” since it never ac- ecaniformes,Ciconiiformes, and Falconiformes. Auk 86:136-139. companies gular flutter. In addition, spread- DE JONG, A. A. 1976. The influence of simulated solar winged sunning in Anhingas is not simply a radiation on the metabolic rate of White-crowned means for realigning the feathers after soaring Sparrows.Condor 78: 174-179. as proposed for vultures and other soaring HAMILTON,W. J., III, AND F. HEPPNER.1967. Radiant birds by Houston (1980). Although Anhingas solar energy and the function of black homeotherm pigmentation: an hypothesis.Science 155:196-l 97. occasionallysoar, they maintain spread-winged HEATH,J. E. 1962. Temperature fluctuation iti the Tur- posturesfor much longer periods than would key Vulture. Condor 64~234-235. 96 WILLARD W. HENNEMANN, III

HEPPNER,F. H. 1969. Bird feathers and radiation. Sci- MCATEE,W. C., AND H. L. STODDARD.1945. Wettable ence 164:202. water birds. Auk 62:303-304. HEPPNER,F. H. 1970. The metabolic significanceof dif- MCNAB,B. K. 1974. The energeticsof endotherms.Ohio ferential absorption of radiant energy by black and J. Sci. 74:370-380. white birds. Condor 7250-59. MCNAB,B. K. 1980. On estimatingthermal conductance HEUSNER,A. 1955. Technique de m&surede courte dur&e in endotherms. Physiol. Zool. 53:145-156. de la consommation d’oxyg*ne et de l’activitk du rat. MORTON,M. L. 1967. The effects of insolation on the Etude du circuit de Benedict. C. R. Sot. Biol. Paris diurnal feeding patterns of the White-crowned Spar- 149~2254-2257. rows (Zonotncka’ leucophrys gambeliz) Ecology HOUSTON.D. C. 1980. A uossible function of sunning 48:690-694. behavior by Griffon V&ures, Gyps spp., and other OHMART,R. D.. AND R. C. LASIEWSKI.1971. Roadrun- large soaringbirds. Ibis 122:366-369. ners: energy conservation by hypothermia and ab- KAHL, M. P. 197 1. Spread-wing posturesand their pos- sorption of sunlight. Science 172~67-69. sible functions in the Ciconiidae. Auk 88:715-722. PALMER,R. S. 1962. Handbook ofNorth American birds. KENNEDY, R. J. 1968. The role of sunbathingin birds. Vol. 1. Yale Univ. Press, New Haven, CT. Br. Birds 61:320-322. POTTER.E. F.. AND D. C. HAUSER. 1974. Relationship KENNEDY, R. J. 1969. Sunbathingbehavior of birds. Br. of gnting’and sunbathing to molting in wild birds. Birds 62:249-258. Auk 91:537-563. KENNEDY,R. J. 1970. Direct effects of rain on birds; a RIJKE, A. M. 1968. The water repellency and feather review. Br. Birds 63:40 l-4 14. structure of cormorants, Phalacrocoracidae.J. Exp. LASIEWSKI,R. C., W. W. WEATHERS,AND M. H. BERN- Biol. 48:185-189. STEIN. 1967. Physiological responsesof the Giant SCHOENER,T. W. 1971. Theory of feeding strategies. Hummingbird, Patagona gigas. Comp. Biochem. Annu. Rev. Ecol. Syst. 2:369404. Physiol. 23:797-8 13. SIEGFRIED,W. R., A. J. WILLIAMS,P. G. H. FROST,AND LUSTICK,S. 1969. Bird energetics:effects of artificial ra- J. B. KINAHAN. 1975. Plumage and ecology of cor- diation. Science 163:387-390. morants. Zool. Afr. 10:183-192. LUSTICK,S. 1971. Plumagecolor and energetics.Condor SOKAL,R. R., AND F. J. ROHLF. 1969. Biometry. Free- 73:121-122. man, San Francisco. LUSTICK,S. 1980. Interaction between posture,color and WALSBERG,G. E., G. S. CAMPBELL,AND J. R. KING. 1978. the radiative heat load in birds. Science 208: coat color and radiative heat gain: a re-eval- 1052-1053. uation. J. Comp. Physiol. 126:21l-222. LUSTICK,S., B. BATTERSBY,AND M. KELTY. 1978. Be- havioral thermoregulation: orientation toward the Department of Zoology, University of Florida, Gainesville, sun in Herring Gulls. Science200:81-82. Florida 3261 I. Accepted for publication 15 December MAHONEY,S. Thermal energeticsof Anhingas and Dou- 1980. ble-crestedCormorants. In Proceedingsof South Af- rican Seabird Symposium. Capetown Univ., Cape- town. In press.