KEEPING WARM 
production required for the ptarmigan 
to maintain a high body temperature is 
the same at 32° in summer as it is at 
-13° in winter. 
Many small birds wintering in cold 
climates also appear to reduce their 
heat-production requirement by seeking 
protected roosts in dense vegetation, 
crevices in snow, or cavities in rocks or 
trees. These roosts protect against direct 
exposure to precipitation and the chill- 
ing effects of wind, and they reduce 
radiant heat losses to the night sky. S. 
Charles Kendeigh of the University of 
Illinois noted that the body heat lost by 
house sparrows (Passer domesticus) in 
wooden roost boxes warmed the air in- 
side the box and contributed to a sub- 
stantial energy saving during the night. 
Observations by Brina Kessel of the 
University of Alaska of the activity of 
chickadees (Parus atricapillus) in 
Alaska suggested that these birds re- 
mained in their roosts for longer periods 
during the daylight hours and foraged 
for shorter periods during cold spells (at 
temperatures of about —58°) than on 
milder days. This behavior would re- 
duce the high heat losses incurred by the 
movement of plumage during flight, 
thereby conserving energy reserves. 
In addition, some small birds, most 
notably chickadees, may not expend the 
necessary energy to maintain high body 
temperatures at night. Susan Chaplin of 
Cornell University found that body tem- 
peratures of chickadees exposed to 32° 
in the laboratory dropped eighteen to 
twenty-two degrees during the night. 
This nocturnal reduction in body tem- 
perature — known as hypothermia — is a 
normal phenomenon from which the 
birds recover spontaneously the next 
morning. If it occurs under natural con- 
ditions in the field, hypothermia can 
save the birds up to 23 percent of the 
estimated energy required to maintain 
normal temperatures. 
Although small birds may use in- 
creased insulation, protected roosts, 
and/or nocturnal hypothermia to re- 
duce the requirements for heat produc- 
tion and husband their energy resources 
in winter, they must also be able to 
sustain high levels of heat production for 
long periods. We believe that they prob- 
ably shiver continuously all winter ex- 
cept during periods of flight, when mus- 
cle contractions associated with flying 
also produce heat substantially above 
basal levels. The general question of 
how small birds produce high levels of 
heat for long periods has not received 
much attention, and it was this area that 
excited our curiosity on our cross-coun- 
try ski trip. 
Our interests coincided with the re- 
search of William R. Dawson, who was 
then our professor at the University of 
Michigan and who has spent nearly 
twenty years investigating avian meta- 
bolic adaptations to cold climates. Join- 
ing forces with him in an attempt to 
discover more about the heat-generating 
capacities of birds, we selected the 
subfamily Carduelinae, which includes 
goldfinches, house finches, pine siskins, 
crossbills, redpolls, and pine grosbeaks, 
for our studies. The cardueline finches 
are ideal for our purposes since the 
majority weigh less than 30 grams, the 
geographical distributions of many spe- 
cies are primarily northern, or boreal, 
and populations of many species are 
year-round residents in cold climates. 
American goldfinches ( Carduelis 
tristis), the first subjects of our study, 
weigh 12 to 15 grams and are widely 
distributed throughout North America. 
The winter distribution of the eastern 
subspecies extends northward into 
southern Canada, but some individuals 
may migrate into the southern states in 
winter. Banding records accumulated 
by the U.S. Fish and Wildlife Service 
indicate that members of the local popu- 
lation in southern Michigan, where we 
worked extensively, may remain in the 
area all year. Interestingly, however, in- 
dividuals may migrate in some years 
and not in others: a bird banded in 
Alabama one winter was recovered in 
Michigan the following January. 
Our first goal was to characterize the 
thermogenic, or heat-generating, capac- 
ities of goldfinches. In our experiment, 
both temperature and time were factors, 
since the ability to produce large 
amounts of heat and to sustain this high 
level of thermogenesis is important for 
survival in the cold. We devised a cham- 
ber in which the temperature of the air 
surrounding the bird was lowered from 
approximately 32° to -94° over a pe- 
riod of an hour and a half. Although 
these birds probably never encounter 
such severe cold stress in nature, this 
technique elicited the desired large 
thermogenic response. We felt that the 
technique would also prove useful in 
comparisons of the thermogenic capaci- 
ties of species or of individuals within 
the same species captured in various 
seasons; J.S. Hart had previously estab- 
lished that differences in the ability to 
withstand severe cold are demonstrably 
correlated with differences in survival 
time at more moderate temperatures. 
As the air temperatures in the cham- 
ber cooled, the goldfinches elevated 
their metabolic heat production, mea- 
sured as oxygen consumption, to offset 
the progressively higher rate of heat 
loss. By the time the air temperatures 
fell to —94°, the birds were producing 
heat at about 5.5 times the basal level to 
maintain a gradient of approximately 
198 degrees between body and air tem- 
perature. Individual birds were retained 
in the chamber as long as oxygen con- 
sumption rose or remained stable, which 
indicated heat production was adequate 
for maintenance of body temperature, 
and were removed from the chamber 
when oxygen consumption fell, indicat- 
ing the birds were no longer able to 
sustain heat production. 
When goldfinches captured in sum- 
mer in southern Michigan were sub- 
jected to this test, they were able to 
withstand only a few minutes at tem- 
peratures below —76° before oxygen 
consumption dropped and they became 
severely hypothermic. In contrast, we 
were impressed to learn that the gold- 
finches trapped in winter could sustain 
necessary levels of heat production for 
six to eight hours below —76°. Sampling 
birds from our area of Michigan every 
month of the year we found that the 
seasonal variation in thermogenic ca- 
60 
