The effect of plumage color on the thermoregulatory ...

The effect of plumage color on the thermoregulatory abilities of Lesser Snow Goose goslings BARBARA A. BEASLEY' A N D C. DAVISON ANKNEY Department of Zoology, The University of Western Ontario, London, Ont., Canada N6A 5B7

Can. J. Zool. Downloaded from www.nrcresearchpress.com by 199.201.121.12 on 06/04/13 For personal use only.

Received July 15, 1987 BEASLEY, B. A., and ANKNEY,C. D. 1988. The effect of plumage color on the thermoregulatory abilities of Lesser Snow Goose goslings. Can. J. Zool. 66: 1352 - 1358. A comparison of the reflectance spectra of Lesser Snow Goose (Chen caerulescens caerulescens) goslings revealed that dark-colored goslings (blues) absorbed more visible and near infrared radiation than did light-colored goslings (snows). Thus, we predicted that blues would obtain greater radiative heat loads and expend less energy to thermoregulate in cold, sunny weather but that radiation would be transmitted deeper into light-colored plumages giving snows a thermal advantage under windy conditions. Using doubly labeled water with penned birds at the La PCrouse Bay colony, we found no significant difference in the daily energy expenditures of blues and snows. Blues tended to expend more energy on windy, sunny days and less on calm, cloudy days but the trends were nonsignificant. We conclude that color has a negligible influence on the energy budgets of Lesser Snow Goose goslings relative to the effect of behavioral thermoregulation. BEASLEY,B. A., et ANKNEY,C. D. 1988. The effect of plumage color on the thermoregulatory abilities of Lesser Snow Goose goslings. Can. J. Zool. 66 : 1352 - 1358. Une Ctude de la rkflectance chez la Petite Oie blanche (Chen caerulescens caerulescens) a rCvC1C que les oisons foncCs (forme bleue) absorbent plus de rayons visibles et de rayons voisins de l'infra-rouge que les oisons de couleur claire (forme blanche); cela nous a amenCs A supposer que les oisons de forme bleue absorbent une plus grande quantitC de chaleur radiante et ont besoin de dCpenser moins dlCnergie pour assurer leur thermorCgulation dans des conditions climatiques ensoleillCes et froides, mais que les radiations sont plus facilement absorbkes par un plumage clair, donnant alors l'avantage thermique aux oisons de forme blanche par vent fort. Nous avons utilisk de l'eau doublement marquCe sur des oiseaux emplumCs de la colonie de la baie de La PCrouse et les rksultats de 1'expCrience ont dCmontrC qu'il n'y avait pas de differences significatives dans la dCpense CnergCtique quotidienne chez les oisons de forme bleue et ceux de forme blanche. Les oisons de forme bleue ont tendance A dCpenser plus d'Cnergie les jours ensoleillCs et venteux et moins les jours calmes et nuageux, mais ces tendances ne sont pas significatives. I1 faut donc conclure que la couleur n'a qu'une influence nCgligeable sur le budget CnergCtique des oisons de la Petite Oie blanche, comparativement A l'influence considkrable du comportement thermorkgulateur. [Traduit par la revue]

Introduction The "blue" and "snow" phases of the Lesser Snow Goose (Chen caerulescens caerulescens) are proportionately distributed along a longitudinal gradient in the eastern Canadian Arctic. Blues are more common in the east, snows in the west. This pattern of distribution and recent changes in the color proportions of various colonies (Cooch 1961; Kerbes 1975) have stimulated several ecologists to assess the adaptive significance of plumage coloration in Lesser Snow Geese. Cooch's (1961, 1963) suggestion that blues and snows differ in survival and fitness depending upon environmental conditions was not supported by the data of Cooke et al. (1985) and Rockwell et al. (1985). However, conflicting results from studies of different breeding populations during different time periods are not surprising, considering the large amount of temporal and spatial variability in arctic environmental conditions. Before our study there had been no attempts to investigate the proximate effects of color on Lesser Snow Geese. Heat conservation should be important to arctic breeding birds, especially small, downy precocial young. Therefore, we examined the effect of plumage coloration on heat exchange between Lesser Snow Goose goslings and their environment. Dark-colored plumages reflect less shortwave solar radiation (290 -2600 nm) than do light-colored plumages (Hamilton and Heppner 1967; Lustick 1969, 1971; Heppner 1970; Walsberg et al. 1978; Ellis 1980; Walsberg 1982). Thus, we predicted 'Present address: Department of Biological Sciences, Simon Fraser University, Burnaby, B.C., Canada V5A 1S6. Prinred in Canada I lrnprirne au Canada

that blues would obtain greater radiative heat loads and expend less energy on thermoregulation under cold, sunny conditions than would snows. Most radiation is absorbed by pigments at the feather surface of dark-colored birds so very little is transmitted to the surface of the skin. Light-colored plumages reflect most incoming radiation but the small amount that is not reflected away is also not absorbed by pigments. Instead, it is transmitted deep into the plumage layer so that the heat is generated at a level closer to the skin in light-colored birds than in dark-colored birds (Cena and Monteith 1975; Walsberg et al. 1978; but see Kovarik 1964). Wind dissipates heat more readily from the surface than from deeper layers within the plumage. Thus, under some windy, cold conditions, it is thermally advantageous to be light colored rather than dark (Hutchison and Brown 1969; Walsberg et al. 1978). We predicted, then, that radiation would be transmitted deeper into snow plumages than into blue plumages and that greater heat loads would be retained by snows than by blues under windy conditions. We compared the reflectance properties of blue and snow gosling plumages and then tested our predictions about their relative heat loads. To do this, we compared daily energy expenditures, calculated by using doubly labeled water (Williams and Nagy 1984), of blue and snow goslings that were exposed to weather conditions within the La Perouse Bay colony. Laboratory studies conducted earlier revealed no difference in the thermoregulatory responses of blues and snows when they were exposed to cold temperatures in the dark (B. A. Beasley and C . D. Ankney , manuscript in preparation).

BEASLEY A N D ANKNEY

TABLE1. The effect of color and age on reflectance (two-way ANOVA) in each of the five spectral wavelength ranges for samples taken from the dorsal head region Probability Snows UV (290 -400 nm) Hatch to 8 days


9 -30 days VIS2 (400-500 nm) Hatch to 8 days 9 - 30 days VISl (500-700 nm) Hatchtogdays 9 - 30 days IR2 (700 - 1450 nm) Hatch to 8 days 9-30 days IR1 (1450-2600 nm) Hatchtogdays 9 - 30 days

Blues

Color

Age

Interaction

1.75f 0.85 (n = 4) 2.53 k0.65 (n = 15) 8.75 k3.40 8.53 k0.98 22.25k4.91 15.93f 1.41 40.25 k7.12 35.27f 2.16 51.75f6.50 52.13k3.34

-

NOTE: Values given are mean reflectance (%) f I SE. Sample sizes ( n ) for each color within each age category were the same for each wavelength range.

Methods Reflectance Goslings were from a captive population of Lesser Snow Geese at the Kortright Waterfowl Park (Niska) near Guelph, Ontario, and from the breeding colony at La PCrouse Bay near Churchill, Manitoba. We used 20 snows and 15 blues at ages varying between hatch and 4 weeks. Birds were killed, and skins and attached plumages were removed in one piece, flattened between two sheets of waxed paper, and frozen immediately. Later, frozen plumages were thawed, any adhering blood or fat was scraped off the subcutaneous side of the skin, and the plumage was dried with a hair dryer (no longer than 30 s) to make it stand erect as it does on a live bird. Circular skin samples, 3.5 cm in diameter, with down attached were cut from the dorsal head region and the mid-dorsal body region of each skin. Diffuse spectral reflectance measurements between 290 and 2600 nm of the electromagnetic spectrum were made with a Beckman DK-2A dual-beam ratio-recording spectrophotometer following the procedure of Porter (1967) (except for differences in the preparation of samples). For each sample, the data from the spectrophotometer were divided into five spectral regions: the ultraviolet (UV, 290400 nm), visible-2 (VIS2, 400 -500 nm), visible- 1 (VIS 1, 500700 nm), infrared-2 ( I N , 700- 1450 nm), and infrared-1 (IR1, 1450-2500 nm); the area under the reflectance curve within each region was integrated. Reflectance values were corrected for the amount of sunlight that actually reaches sea level (according to W. B. Porter, personal communication). There is an obvious color change in both morphs after the 1st week of life as new plumage begins to replace down. Therefore, we grouped plumage samples into two age categories: (i) from hatch to 8 days of age and (ii) from 9 to 30 days of age. The general linear model (GLM) procedure of the Statistical Analysis System (SAS Institute Inc. 1982) was used for two-way factorial analyses of variance to determine the effect of color and age on reflectance in each of the five wavelength ranges and over the entire range for samples taken from the head and mid-dorsal body regions. Significance for these and all further statistical tests was accepted at the P < 0.05 level. Daily energy expenditures Pipping eggs were collected from the La Perouse Bay colony. Approximately 12 eggs estimated to hatch on the same day were col-

lected at a time. Eggs or goslings were kept in a portable incubator until 24 h after hatch. Then goslings were sexed, web tagged and leg banded, and kept in a holding pen at the Queen's University Tundra Station. The pen was supplied with a shelter but goslings were brought indoors at night when it rained and (or) when the temperature dropped to low levels. Food (Purina Duck Startena) and water were provided a d libitum. Goslings were imprinted on people to minimize stress caused by handling. Ever 2 days, six goslings, three of each color, were selected for metabolism measurement on the basis of their similar weights and age. Only 2-day-old and 4- to 7-day-old goslings were used. The six goslings were taken to the field site where a flat area of grasses and sedges (mainly Puccinellia phryganodes and Carex subspathacea) was enclosed within a 5 m diameter fence. Low standing willow (Salix) was prominent in the surrounding area. The pen was supplied with a shelter and brooder lamp (powered by a portable generator) to simulate parental body heat. To determine metabolism with doubly labeled water the procedure outlined by Nagy (1975) was followed. A summarized version of this procedure is as follows. Birds were weighed, individually marked with colored plastic leg bands and wing tape, and given an intramuscular (leg) injection of water containing 95 atom percent oxygen-18 and approximately 0.1 to 0.2 mCi hydrogen-3 (1 Ci = 37 GBq). After allowing 1 h for the labeled water to reach equilibrium with the body water (previous tests confirmed that 1 h was sufficient), duplicate 50-pL blood samples from a femoral vein were obtained. These were stored at 6OC in flame-sealed glass heparinized microhematocrit tubes for later analyses. The goslings were then released into the pen and behavioral observations were made during the next 24 h. Using instantaneous scan sampling (Altmann 1974), the activity of each bird was assessed and recorded every minute over hour-long periods (60 observations). Activities included sleeping, comfort movements such as preening and stretching, feeding, walking, and running. We also noted social interactions and whether birds were huddled or alone. We scanned individuals in the same order each time and recorded their behavior as soon as we focused on them. Approximately 900 observations were made on each individual. Half-hour periods between each hour of behavioral observation were used to monitor the microclimates under the shelter and at an exposed area of the pen. Wind speed between the ground surface and 11 cm above ground was determined with a hand-held rotating ane-

CAN. J . ZOOL. VOL. 66, 1988

TABLE2. The effect of color and age on reflectance (two-way ANOVA) in each of the five spectral wavelength ranges for samples taken from the mid-dorsal body region Probability Snows UV (290-400 nm) Hatch to 8 days


9 - 30 days VIS2 (400-500 nm) Hatch to 8 days 9 -30 days VIS 1 (500 -700 nm) Hatch to 8 days 9 - 30 days IR2 (700 - 1450 nm) Hatch to 8 days 9 - 30 days IR1 (1450-2600 nm) Hatch to 8 days 9 - 30 days

Blues

Color

Age

Interaction

1.25 f 0.48 (n = 4) 3.27k 0.77 (n = 15) 4.50k0.50 5.73 k0.91 8.50k0.96 9.00k 1.06

+

26.25 1.1 1 29.60k 1.51 5 1.00+ 1.35 56.07 k 1.74

NOTE: Values given are mean reflectance (%) k I SE. Sample sizes ( n ) for each color within each age category were the same for each wavelength range.

mometer. Direct solar radiation was measured with an Eppley pyroheliometer and a digital LICOR radiometer. Soil surface and air temperatures at 5 and 10 cm above ground were monitored with thermocouples (unshielded) and a Yellow Springs Instruments telethermometer. Maximum and minimum temperatures inside and outside the shelter were also monitored. At the end of the 24 h each gosling was recaptured and reweighed, and a second set of duplicate blood samples were taken. Some blood was taken from an uninjected gosling for measurement of background oxygen- 18. Birds were killed immediately, reweighed on an electric balance to 0.01 g (Mettler PB 300), and then sealed in plastic bags and frozen until they were oven-dried and weighed again to calculate their total body water. Measurements were taken every 2nd day until 48 birds, 24 of each color (and of each age set) had been measured. Blood samples were microdistilled (Wood et al. 1975) to obtain pure water. A portion of each such sample was assayed for tritium activity with a liquid scintillation counter. The remaining distilled samples were sent to K. Nagy at the University of California at Los Angeles for assays of the oxygen-18 by cyclotron-generated proton activation of oxygen-18 to fluorine- 18 with subsequent counting of the gamma-emitting fluorine-18 in a Packard Gamma Rotomatic counting system (Wood et al. 1975). Using the equations of Lifson and McClintock (1966) as modified by Nagy (1975), rates of water flux and CO, production were calculated. Some of the CO, production values that we obtained were out of the range of acceptable values (either negative or unbelievably high). A careful examination of the calculations revealed that errors in the isotope concentration values were the only possible causes of these abemnt values. CO, production values were recalculated using corrected values for isotope concentrations following the procedure suggested by K. Nagy (personal communication (see Beasley 1986). After recalculations, any values of CO, production less than 1 or greater than 7 mL CO, . g-' . h-' were eliminated. Within the remaining sample of 27 birds we assumed that all undetected errors were equally distributed between colors. To convert metabolic rates to joules we assumed that goslings were oxidizing carbohydrates for metabolism and so used the conversion factor of 2 1.14 J/mL CO, (King and Farner 1961). We calculated the proportion of time that individuals spent performing each activity by dividing the number of times that we observed each activity by the total number of observations. Weather conditions measured before

and after each hour-long observation period were averaged for each microhabitat (inside the shelter and outside the shelter). Each of these average values was then multiplied by the total number of times that goslings were seen inside and outside the shelter, respectively, during the corresponding hour of observation. Average conditions experienced by each gosling over the entire 24-h period were calculated from the sum of these products divided by the total number of observations. Statistical analyses were done using the TTEST, GLM, and CORR procedures of SAS (SAS Institute 1982). The amounts of CO, produced by snows and blues each day were converted so that they were not weight specific (they were calculated in the form of mL CO, . g-' . h-' so we multiplied them by the body mass of each gosling to obtain values of mL CO,/h). Most of the variation in energy expenditure was due to the large difference between the body weights of the 2-day-old goslings that were measured in the first four trials and the 4- to 7-day-old goslings that were measured in the last four trials. To eliminate this variation we divided the goslings into two age groups and analyzed the data for each separately. We used regressions to determine whether C 0 , production was influenced by date, body weight, the average temperature experienced, the average wind speed experienced, or the average amount of radiation received from the sun, the brood lamp, or both. Variables that explained a significant amount of the variation were used as covariates in analyses of covariance to compare the CO, production of blues and snows. If predictor variables were correlated, the one that explained the greatest amount of variation was used as the covariate. Student's t-tests were used to compare the average weather conditions experienced by goslings in each age group.

Results Reflectance In the dorsal head region, snows reflected more radiation than did blues in all wavelength ranges except the far infrared, IRl (Table I). Age had no significant effect and there was no significant interaction between age and color but older blues and snows tended to be more similar in reflectance than were younger ones (Table 1). The greatest differences between the head plumage of blues and snows were in the VISl and IR2

1355

BEASLEY AND ANKNEY

TABLE3. Mean proportions of the total observation time spent by snows and blues in each activity (morphs were compared using t-tests and there were no differences (P > 0.2)) Sleeping


Huddled

Comfort activities

Alone Huddled

Stationary Alone feeding

Walking while feeding

Walking and running

Social

Other

Total time (min)

June 17-18 Snow Blue June 19-20 Snow Blue June 21-22 Snow Blueu June 23 -24 Snow Blue June 25 - 26 Snow Blue June 27 -28 Snow Blue June 29-30 Snow Blue July 2-3 Snow Blue Mean Snow Blue Both -

NOTE: Two-day-old birds were tested between June 17 and June 24; 4- to 7-day-old birds were tested after June 25. "Two blues were measured during this trial; in all others, three of each morph were measured.

ranges. There, young snows reflected more than 3.5 times as much radiation as did older blues. In the mid-dorsal body region, the effect of color on reflectance depended on age and was not consistent over the entire wavelength range (Table 2). Young snows reflected significantly more infrared radiation (IR1 and IR2) than did young blues but there was no difference in the other wavelength regions (however, the trend was the same). Older blues and snows did not differ in reflectance in any of the wavelength ranges. Energy expenditures Goslings tested together behaved identically (Table 3). They performed the same activities simultaneously. Thus, blues and snows experienced the same average weather conditions. In general younger birds spent more time huddled than did older birds. The rate of CO, production (Vco,) of 2-day-old goslings was affected significantly by wind speed and temperature (Table 4). These variables were positively correlated (Table 4) so only wind speed, which explained the most variation, was used as a covariate. Blues expended slightly more energy than did snows, but the difference was not significant (Table 5). The effect of wind speed on the rate of CO, production did not vary between morphs (interaction: wind x color: P = 0.53) but wind speed explained more of the variation in the metabolic rates of blues (Vco, = 179 42 wind, P = 0.17, r 2 =

+

0.51) than of snows (Vco, = 91 + 76 wind, P = 0.13, r2 = 0.39). The CO, production of 4- to 7-day-old goslings was influenced by date, temperature, the average amount of radiation received from both the sun and the heat lamp, and the amount of time spent under the heat lamp (Table 6). All four variables were correlated (Table 6) so date was used as the covariate. In this group of birds, blues expended slightly less energy than did snows but the difference was not significant (Table 5). Changes in metabolic rates with date did not differ between morphs (interaction: date x color: P = 0.56) but date explained more of the variation in the metabolic rates of blues (Vco, = 7 11 - 30 date, P = 0.002, r2 = 0.76) than of snows (Vco, = 667 - 21 date, P = 0.25, r 2 = 0.31). Two-day-old goslings that were tested during the first 8 days of the experiment experienced higher wind speeds and a greater amount of solar radiation than did the 4- to 7-day-old birds (Table 7). Goslings in each age group experienced similar average temperatures and they spent similar amounts of time under the brood lamp. Overall, there was no difference in the total amount of radiation that they received.

Discussion The greatest differences in reflectance between blues and snows were in the visible and near infrared portions of the spectrum (14 and 25 % for the body and head regions, respec-

CAN. J . ZOOL. VOL. 66, 1988

TABLE4. Regressions of the rate of COz production (Vcoz; mL/h) of 2-day-old goslings on each of the variablesa shown and a matrix of the correlations between these variables; probabilities and rZvalues are given for each regression equation and probabilities are given in parentheses beneath the correlation coefficients Regression equations


Vco, = Vco, = Vco, = Vco, = Vcoz = Vcoz = Vco, =

Probability

r2

0.4 1 0.20 0.06 0.04 0.19 0.45 0.48

0.07 0.16 0.31 0.35 0.16 0.06 0.05

+ + + + + +

242 9.9 Date 108 1.2 Mass 90 12.8 Temp 144 55.1 Wind 328 - 0.1 Total Rad 158 0.3 Solar 208 0.2 Brood

Date

Mass

Temp

-0.8563 (0. O W )

0.525 1 (0.0796) -0.7664 (0.0036)

Mass

Wind

Total Rad

Solar

Brood

Temp Wind Total Rad Solar "Variables were Date (the trial number), Mass (averaged from measurements taken before and after each trial), Temp (the average ambient temperature, "C), Wind (the average wind speed, mls), Total Rad (the average amount of radiation received from the sun and the brood lamp, W/ml), Solar (the average amount of radiation from the sun only, W/m2), and Brood (the average amount of time spent under the brood lamp, min).

TABLE 5. The daily energy expenditures (mL CO,/h) of snows and blues in two age groups Age 2 days 4to7days

Snows

Blues

Probability

259f 13.4 (n = 7) 522f12.4 (n = 6)

273f 15.9 (n = 5) 506f10.0 (n = 9)

0.51 15 0.3208

NOTE:Mean values ( f 1 SE) for 2-day-old birds were adjusted for the effect of wind speed (mls). Mean values for 4- to 7-day-old birds were adjusted for the effect of date.

tively) and were lower than that reported for dark-colored and white zebra finches (60 % ; Lustick 1969). Therefore, we expected less of a difference between the energy expenditures of blues and snows than the 20% found by Lustick when he measured rates of oxygen consumption. Our comparisons of the energy expenditures of blue and snow goslings indicate that the differences in reflectance were not large enough to convert to a significant difference in their radiative heat loads. Under windy conditions, blues tended to expend more energy than did snows, even though solar radiation was abundant. As wind speed increased, the rates of CO, production of 2-day-old goslings increased. This happened regardless of the correlated increase in temperature with wind speed. The metabolic rates of blues seemed to respond more to wind than did those of snows but sample sizes were too small for the interaction to be significant. Under calm conditions, blues tended to expend less energy than did snows, even though solar radiation was not very

intense. The rate of CO, production of 4- to 7-day-old birds decreased as temperatures and total radiation increased but it also increased with the amount of time that goslings spent under the brood lamp. These relationships indicate that energy expenditure was greatest in cold-stressed birds that were trying to alleviate the cold by spending a lot of time under the brood lamp. Energy expenditures decreased with date, which was negatively correlated with time spent under the brood lamp. The metabolic rates of blues seemed to respond more to date (and therefore to the correlated variables) than did those of snows but sample sizes were too small for the interaction to be significant. Possibly, our experimental conditions were not ideal for finding a difference due to color. It was rarely calm and sunny simultaneously and the average temperatures that goslings experienced were never very cold. Temperatures were rarely more than 5 to 10°C below the lower critical temperature of domestic goslings (about 20°C according to Poczopko 1968). The radiation provided by the brood lamp was necessary for overnight survival of the goslings, particularly when it was damp, but it may have provided more energy than was necessary in the early evenings and late mornings. (Most of the radiation provided by the brood lamp lies in the far infrared wavelength region (1450-3200 nm; Lustick et al. 1980), which would have been absorbed equally by snows and blues.) The time that goslings spent feeding in the pen (and hence the time that they were exposed to weather conditions) was likely low compared with the time that they spend feeding in the wild because their duck chow diets were higher in energy content than are the grasses and sedges to which goslings are normally restricted. Also, their feeding patterns in the wild are probably

BEASLEY AND ANKNEY

TABLE 6. Regressions of the rate of CO, production (Vco,; mL/h) of 4- to 7-day-old goslings on each of the variablesa shown and a matrix of the correlations between these variables; probabilities and r2values are given for each regression equation and probabilities are given in parentheses beneath the correlation coefficients Regression equations


Vco, = 686 Vco, = 700 Vco, = 592 Vco, = 460 Vco2 = 340 Vco, = 546 Vco, = 470

Date

-

+ + +

Probability

r

0.002 0.139 0.045 0.313 0.044 0.120 0.025

0.49 0.16 0.27 0.08 0.28 0.17 0.33

25.4 Date 1.4 Mass 5.8 Temp 30.0 Wind 0.2 Total Rad 0.2 Solar 0.2 Brood Mass

Temp

0.2723 (0.3262)

0.5366 -0.1040 (0.0392) (0.7123) 0.0608 -0.21 17 (0.8295) (0.4489) -0.7252 (0.0022)

Mass Temp

Wind

Total Rad

Solar

Brood

Wind Total Rad Solar 'Variables were Date (the trial number), Mass (averaged from measurements taken before and after each trial), Temp (the average ambient temperature, "C), Wind (the average wind speed. mls), Total Rad (the average amount of radiation received from the sun and the brood lamp. Wlml), Solar (the average amount of radiation from the sun only, Wlml), and Brood (the averdge amount of time spent under the brood lamp, min).

TABLE 7. The average weather conditions experienced by goslings in each age group over the period that they were tested 2 days old 4 to 7 days old (June 17 -24) (June 25 - July 3) n = 12 n = 15 Temperature ("C) Wind speed (m/s) Total radiation (W/m2) Solar radiation (W/m2) Time brooded (min)

13.750.5 2.25 0.1 527 539 342 9 252 5 12

+

13.75 1.0 1.7k0.1 461 +23 198527 238 535

Probability 0.9773 0.0073 0.1399 0.0001 0.7327

NOTE: Values are given as means I SE and the probabilities of differences between groups (obtained from I-tests) are shown. Weather was not monitored on July I because no experiment was run.

closely matched to those of their parents. Thus, we assume that wild goslings need to expend relatively more energy on thermoregulation. They probably also expend more energy on locomotion. If energy demands had been greater and the weather had been calm and sunny simultaneously, the energy expenditures of blues and snows might have been more distinct. Further testing is necessary before that possibility can be dismissed. Goslings seemed to be able to adjust their heat loss behaviorally soon after hatch. In the pen, they sought out warmth wherever it was: from a brood lamp, the sun, or other goslings' bodies. Thus, in spite of their precociousness, they did spend a great deal of time involved in behavioral thermoregulation (on average they were huddled 43% of the time in our study), at least during the first 1.5 week after hatch. To ensure that the weather conditions in the pen were typical of those experi-

enced by wild, free-ranging goslings, microclimate measurements were taken at sites typical of where geese spend much of their time with their families (Boag 1974; Harwood 1975; R. F. Rockwell, personal communication) within the La Pkrouse Bay colony. Weather conditions in the pen were similar to those measured at those sites. More importantly, the measurements revealed that the diversity of available microclimates on the arctic tundra is great enough that goslings should be able to reduce their heat loss via microhabitat selection as do Phainopepla (Walsberg 1982). In our pen study, goslings seemed capable of altering their microclimate to diminish cold stress and thereby compensated for the effects of plumage coloration on thermoregulatory energy expenditure. Observational work is required to assess whether this type of behavioral thermoregulation occurs in the wild and whether it differs between bluesand snows.

CAN. J . ZOOL. VOL. 66, 1988

Acknowledgements This research was

.

and

by the

Wetlands Research Station, the Canadian Wildlife Service, the Affairs, and the Departrnent Of Indian and Sciences and Engineering Research Council of Canada (operating grant to C . D. Ankney). We thank Lynne B r ~ w nfor assistance in the field and Patricia Wood and Kevin Barr for their guidance in the laboratory. Warren Porter, Bob Jeffries, and- Martin Kavaliers loaned essential e a u i ~ m e n tto us. The Queen's University Tundra Biology Station provided living facilities and logistical support. ~~~h~~ coach and an anonymous reviewer made helpful comments about the manuscript.


.

I

I

ALTMANN, J . 1974. Observational study of behavior: sampling methods. Behaviour, 49: 227 - 267. BEASLEY, B. A. 1986. The effect of plumage coloration on the thermoregulatory abilities of Lesser Snow Goose (Chen caerulescens caerulescens) goslings. M. Sc. thesis, University of Western Ontario, London. BOAG,P. T. 1974. A descriptive and functional analysis of posthatching in the Lesser Snow Goose. B.Sc. thesis, Queen's University, Kingston, Ont. CENA,K., and MONTEITH, J. L. 1975. Transfer processes in animal coats. I. Radiative transfer. Proc. R. Soc. London B, 188: 337394. C o o c ~ F. , G. 1961. Ecological aspects of the blue-snow complex. Auk, 78: 72-89. 1963. Recent changes in distribution of color phases of (Chen c. caerulescens) Snow Geese. Proc. Int. Ornithol. Congr. 13: 1182-1194. C. S., ROCKWELL, R. F., and SMITH,J. A. COOKE,F., FINDLAY, 1985. Life history studies of the Lesser Snow Goose (Anser caerulescens caerulescens). 111. The selective value of plumage polymorphism: net fecundity. Evolution (Lawrence, Kans.) , 39: 165 - 177. ELLIS,H. I. 1980. Metabolism and solar radiation in dark and white herons in hot climates. Physiol. Zool. 53: 358 - 372. R. 1967. Radiant solar energy HAMILTON, W. F., 111, and HEPPNER, and the function of black homeotherm pigmentation: an hypothesis. Science (Washington, D.C.), 155: 196- 197. HARWOOD, J. 1975. The feeding strategies of Blue Geese Anser caerulescens. Ph.D. thesis, University of Western Ontario, London. HEPPNER, F. 1970. Metabolic significance of absorption of radiant energy. Condor, 72: 50-59. HUTCHINSON, J. C. D., and BROWN, G. B. 1969. Penetrance of cattle coats by radiation. J. Appl. Physiol. 26: 454 -464.

KERBES, R. H. 1975. Lesser Snow Geese in the eastern Canadian Arctic. Can. Wildl. Serv. Rep. Ser. No. 35. KING,J. R.. and FARNER, D. S. 1961. Energy metabolism, thermoregulation, and body temperaurea In Biology and comparative physiology of birds. Vol. 2. Edited by A. J. Marshall. Academic Press., New York. DD. 215-288. rr KOVARIK, M. 1964. Flow of heat in an irradiated protected cover. Nature (London), 201: 1085- 1087. LIFSON,N., and MCCLINTOCK, R. 1966. Theory of use of the turnover rates of body water for measuring energy and material 1. Theor. Biol- 12: 46-74. LUSTICK, S. 1969. Bird energetics: effects of artificial radiation. Science (Washington, D.C.), 163: 387 -390. 1971. Plumage color and energetics. Condor, 73: 12 1 - 122. LUSTICK, S., ADAM,M., and HINKO,A. 1980. Interaction between posture, color and radiative heat load in birds. Science (Washington, D.C.), 208: 1052 - 1053. NAGY,K. A. 1975. Water and energy budgets of free-living animals: measurement using isotopically labeled water. In Environmental physiology of desert organisms. Edited by N. F. Hadley . Dowden, Hutchinson and Ross, Stroudsburg, PA. pp. 227 -245. POCZOPKO, P. 1968. Studies on the effect of environmental thermal conditions on geese 111. Deep and external body temperature in goslings at thermoneutrality and during short-lasting exposure to cold. Acta Physiol. Pol. 19: 646-660. PORTER, W. B. 1967. Solar radiation through the living body walls of vertebrates with emphasis on desert reptiles. Ecol. Monogr. 37: 273 -295. ROCKWELL, R. F., FINDLAY, C. S., COOKE,F., and SMITH,J. A. 1985. Life history studies of the Lesser Snow Goose (Anser caerulescens caerulescens) IV. The selective value of plumage polymorphism: net viability, the timing of maturation, and breeding propensity. Evolution (Lawrence, Kans.), 39: 178 - 189. SAS INSTITUTE. 1982. SAS user's guide: statistics. SAS Institute Inc., Cary, NC. WALSBERG, G. E. 1982. Coat color, solar heat gain and the conspicuousness in the Phainopepla. Auk, 99: 495-502. WALSBERG, G. E., CAMPBELL, G. S., and KING,J. R. 1978. Animal coat color and radiative heat gain: a re-evaluation. J. Comp. Physiol. B, 126: 21 1-222. WILLIAMS, J. B., and NAGY,K. A. 1984. Validation of the doubly labeled water technique for measuring energy metabolism in Savannah Sparrows. Physiol. Zool. 57: 325 - 328. N. S., WAKAKUWA, WOOD,R. A., NAGY,K. A., MACDONALD, S. T., BECKMAN, R. J., and KAAZ,H. 1975. Determination of oxygen-18 in water contained in biological samples by charged particle activation. Anal. Chem. 47: 646 -650. -

-

-

-

-

-.