N45
2014 Northeastern Naturalist Notes Vol. 21, No. 4
B. Heinrich
Redpoll Snow Bathing: Observations and Hypothesis
Bernd Heinrich*
Abstract - A flock of 100–150 Carduelis flammea (Common Redpolls) made at least 252 cavities and
short tunnels in fluffy snow over 5 consecutive days in western Maine at temperatures from -26 °C to
-14 °C. They produced neither structure during 4 days when temperatures were above freezing and the
snow was wet, nor during a month of other variable weather. In the same area during two winters, Carduelis
tristis (American Goldfinches), Carpodacus purpureus (Purple Finches), Pinicola enuncleator
(Pine Grosbeaks), and Coccothraustes vespertinus (Evening Grosbeaks) exhibited no snow tunneling
behavior at any time even though, excluding the Pine Grosbeaks and Goldfinches, I observed them
throughout the entirety of 2 winters. Common Redpolls did not use the observed snow cavities as
shelters, nor did the structures appear to be related to foraging. Social stimulation and play are likely
proximate stimuli for the behavioral release of the Common Redp olls’ snow-manipulating behavior.
Introduction. Carduelis flammea L. (Common Redpoll) lives year-round on tundra and
taiga north of Hudson Bay where they are Betula spp. (birch), Alnus spp. (alder), and Picea
spp. (spruce) seed specialists (Knox and Lowther 2000). They move south in winters
when their main food fails. In the winter of 2001–2002, flocks of these birds arrived in
Adirondack Park in mid-January and started tunneling in the snow (Collins and Peterson
2003). The authors concluded that the birds stimulated each other in their activity, and were
apparently doing it for enjoyment. Palmer (1949) and Furness (1987) reported similarly
described bathing activity of Common Redpolls in the Northeast. In Europe, small northern
birds, including Common Redpolls and Parus cinctus Boddaert (Siberian Tits), have been
reported to take up residence in snow burrows in order to escape wind and cold (Korhonen
1981, Novikov 1972, Sulkava 1968,). The function, timing, and reason for the Common
Redpolls’ snow-manipulation activity remain unclear. In the winter of 2012–2013, Common
Redpolls, 4 other northern finch species, and 4 non-finch species were present for
nearly constant observation for weeks at a time, and I report here on the incidence of their
snow-bathing and snow-burrowing behavior and make comparative observations between
Common Redpolls and other concurrent bird species during snow availability in 2 winters.
Methods. During winter 2012–2013, I observed irruption of Common Redpolls near Mt.
Blue in western Maine at 2 bird feeders that were continually supplied with black sunflower
seed during the course of observations from 5 November 2012 to late February 2013. I
observed the feeders and their surrounding 0.4-ha clearing, bounded by forest, from cabin
windows that permitted panoramic views in all directions (Fig. 1). Fluffy snow suitable for
burrowing was available almost continuously from the first snowfall, on 5 November 2012,
to 19 February 2013. I made counts and measurements of the snow depressions after the
birds had left the clearing and erased each one to avoid double-counting them. I measured
snow density by scooping up specific volumes (2 to 4 liters) of snow and comparing each
volume to the amount of water produced when the sample melted.
No Common Redpolls were present in winter 2013–2014; however, I observed 4 other
species of finches— Carduelis tristis L. (American Goldfinch), Carpodacus purpureus
Gmelin (Purple Finch), Pinicola enuncleator L. (Pine Grosbeak), and Coccothraustes vespertinus
(Cooper) (Evening Grosbeak)—for possible snow burrowing activity during the
two winters.
*Department of Biology, University of Vermont; Burlington, VT 05405; bheinrich153@gmail.com.
Manuscript Editor: Jean-Pierre Savard
Notes of the Northeastern Naturalist, Issue 21/4, 2014
2014 Northeastern Naturalist Notes Vol. 21, No. 4
N46
B. Heinrich
Results. The first snow fell on 5 November in 2012, and the first Common Redpolls, a
flock of about 30, appeared at the feeder on 16 November. Common Redpolls continued to
arrive at the feeder in similar-sized groups until mid-January, and their numbers increased
to about 50 during that time and then to over 100 by 20 January 2013. The snow was fluffy
and temperatures ranged from -8 °C to -24 °C, during the 5 days (starting on 23 January
2013) prior to the main observations reported here when 50–150 Common Redpolls came
regularly. A flock of up to 22 Evening Grosbeaks and 20 Pine Grosbeaks were also daily
visitors during that time; during the 2013–2014, however, winter, the Common Redpolls
and Pine Grosbeaks were absent, but I observed Evening Grosbeaks, American Goldfinches,
and Purple Finches daily, with 3 to >12 of each species present at the feeders. All of
the finches visited the same feeders and loitered in the same trees and bushes in the 0.4-ha
clearing as the Common Redpolls. Details of conditions and behaviors varied greatly from
day to day.
23 January 2013: Four minutes after sunrise (07:06 hours), ~150 Common Redpolls
and flocks of Evening and Pine Grosbeaks arrived at the feeders. After their early morning
feeding, the Common Redpolls started hopping on the surface of 17 cm of powdery snow.
They concentrated their activity in and around nearby (less than 100 m away), 1–2 m-tall Prunus
virginiana L. (Choke Cherry) bushes. The wind was blowing ~16 km/h. and the temperature
was -26 °C. Although the feeders were still loaded with food, and seed was spilled onto the
snow under them, no food was available where the birds were seemingly randomly hopping
about on the snow. However, soon one bird after another ducked head first into the snow
and then, while fluttering with its wings, it proceeded a few centimeters forward within the
snow. Although this behavior looked like the activity typical of bathing birds, it differed in
being accompanied by an additional forward locomotion and without repeated head-lifting,
Figure 1. Photograph taken from an observation window of a gathering of Common Redpolls at a
Chenopodium sp. (goosefoot) inflorescence in part of the clearing where the events in this report
took place.
N47
2014 Northeastern Naturalist Notes Vol. 21, No. 4
B. Heinrich
subsequent shaking, and feather preening. After each burrowing episode, the birds popped
up and hopped again on top of the snow, having left a groove or a tunnel in the snow (Fig. 2).
After 1 h, all the birds left the snow, to either return to the feeders or to perch almost motionless
in small groups in the trees along the edge of the clearing. By 15:00 hours, all the birds
had left the site.
Figure 2. Photographs of burrows or snow tunnels made by Common Redpolls 23–27 January 2013.
2014 Northeastern Naturalist Notes Vol. 21, No. 4
N48
B. Heinrich
There were 51 burrowing depressions where numerous individuals had gathered in small
groups on the snow (Table 1). Most depressions were grooves 4–5 cm wide, 6–20 cm long,
and usually 4–6 cm deep. Five of them were tunnels with clear entrance and exit holes and
with a bridge of snow left intact in between. On the other hand, those hopping along on the
top of the snow under the feeder left only tracks.
24 January 2013: The temperature dropped to -28 °C during the night, and about a hundred
Common Redpolls again arrived after first light. As before, they repeated their snow
bathing shortly after their initial feeding frenzy, but strong wind obliterated the record of
their ground activity.
25 January 2013: In the morning, at -23 °C, again shortly after the initial dawn-time
feeding, the Common Redpolls gathered on the top of the snow in the same patch of Choke
Cherry bushes where they produced a total of 59 snow furrows. A group of them then
gathered in another patch of Choke Cherry bushes on another side of the cabin, where they
proceeded to make 54 more furrows. At a third location, birds gathered in another small
patch of Choke Cherries and created 16 furrows there. Birds created 17 more furrows at the
periphery of the clearing, but only 3 of them were in the open areas of the field, for a total
of 149 fresh furrows and tunnels that morning (Table 1). As during the 2 previous days, the
birds engaged in burrowing/bathing activity only after the morning feeding. Snow bathing
ended by 10:00 hours that morning, but small groups of the birds continued to come and
feed (but not burrow) throughout the day until about 14:00 hour s.
26 January 2013: Temperatures were similar to the previous day and there was, as
before, sunshine and no wind. Fresh tracks of the Common Redpoll’s activity were again
inscribed on the snow an hour after their morning feeding when they had dispersed onto the
snow, but this time there was no snow burrowing.
Table 1. Summary of daily observations of times and conditions for the Common Redpoll snow
tunneling/burrowing behavior in 2013. Approximately 100–200 Common Redpolls were present on
every morning during this time. Soft snow was present throughout the study, temperatures refer to
dawn and afternoon, and burrowing occurred before noon.
Date Temperature (°C) Conditions # of burrows
23 Jan -26 to -23 Clear 51
24 Jan -28 to -21 Clouds and strong wind ?
25 Jan -23 to -15 Clouds and sun 149
26 Jan -21 to -12 Clear 0
27 Jan -21 to -12 Clear 42
28 Jan -18 to -6.5 Clear 0
8 Feb -20 to -15 Light snow/windy 0
9 Feb -16 to -14 Snowing 0
10 Feb -15 to -9 Clear 0
11 Feb -9 to -7 Light snow/clear 0
12 Feb -3 to +2 Overcast 0
13 Feb -4 to +2 Clear to overcast 0
14 Feb -7 to 0 Partial overcast 0
15 Feb -7 to +4.5 Overcast 0
16 Feb 5.5 to -1 Snowing 0
17 Feb -12 Snowing 0
18 Feb -20 to -17 Clear and windy 0
19 Feb -12 to +2.0 Thin overcast 0
N49
2014 Northeastern Naturalist Notes Vol. 21, No. 4
B. Heinrich
27 January 2013: Temperatures and weather were similar to the last 2 days, but the birds
made only 42 new depressions or tunnels. This was the last day that snow bathing occurred.
Throughout the 4 days in late January when at least 100 Common Redpolls had made
the 252 snow tunnels, all burrowing was within a radius of about 150 m of the two feeders,
and although 5 other bird species also used the feeders, none of them engaged in similar
behavior. The maximum numbers of each bird species present at once was: 30 Evening
Grosbeaks, 20 Pine Grosbeaks, 9 Poecile atricapilla L. (Black-capped Chickadee), 5 Sitta
canadensis L. (Red-breasted Nuthatch), 1 Sitta carolinensis Latham (White-breasted Nuthatch),
and 2 Cyanocitta cristata L. (Blue Jay). Like the Common Redpolls, these birds
routinely landed on the snow under the feeders when seeds were spilled there, but unlike
the Common Redpolls, none of them descended onto the ground beyond the feeders within
sight of the cabin and none engaged in snow bathing.
During the time that the Common Redpolls made tunnels and depressions in the snow,
the other species often roosted high in the trees in the surrounding forest, leaving no tracks,
burrows, or furrows in or on the snow. In contrast, I routinely observed the tracks and overnighting
snow caves of the much less common Bonasa umbellis L. (Ruffed Grouse).
I continued to record weather and snow conditions and bird presence until 19 February
(Table 1). During this time, Common Redpolls were present in approximately the same
numbers as before, but I did not observe snow bathing after the end of January. I classified
snow density (water/snow ratio) as: light and fluffy = 1/22, when the Common Redpolls
left shallow furrows while hopping along on the snow; 1/12, when they walked on the snow
without sinking in; and 1/5.5 and 1/2.5 when it was wet. I recorded temperatures of -18 °C
–4.5 °C, and skies were clear and overcast during the study per iod.
I made observations of the same feeders from November 2013 to late January 2014. At
least 5 cm of snow fell on the night of 10 December, and fluffy snow was available throughout
most the rest of the study period; temperatures dropped to as low as -27 °C but usually
stayed between -12 °C and -20 °C. During these 5 months, up to 20 Evening Grosbeaks,
15 American Goldfinches, and 8 Purple Finches came to the feeder daily and loitered in
the neighboring trees. As during the previous winter, I observed no snow bathing by any of
these finches or by the same coterie of non-finch species observed the previous year: Blackcapped
Chickadees, Red-breasted and White-breasted Nuthatches, and Blue Jays.
The data do not point to any direct stimulus in the environment that induced the Common
Redpolls’ bathing movements involved in the tunneling behavior. With hundreds of
the birds in the area around the feeders for many days in all sorts of weather and over a
wide range of temperatures, I identified no environmental factor as an obvious behavioral
trigger. However, it may be noteworthy that individuals within the Common Redpoll flocks
appeared to be strongly cuing on each other. They arrived in tight flocks, remained so while
near the feeders, and departed the study area en masse. For example, when one bird flew to
feed on a weed stalk that I had not seen a bird visit in weeks, another and another followed
it until there was a crowd of dozens on and around the plant. The flock then flew successive
circuits, leaving and returning in unison. Similarly, bathers apparently attracted each other
and also acted in unison.
Discussion. The above observations confirm and extend previously reported snow bathing/
snow burrowing behavior of Common Redpolls in Maine (Palmer 1949), but the conditions
under which they occurred constrain various hypotheses of the purpose (e.g., “bathing”;
Furness 1987). As far as I know, there are no references of any of the 4 finch species routinely
observed at the Maine study site for 3 winters to either bathe or to overnight in snow. The lack
of snow bathing by the other finch species I observed simultaneously with the Common Red2014
Northeastern Naturalist Notes Vol. 21, No. 4
N50
B. Heinrich
polls indicates that the Common Redpolls’ behavior is not common finch behavior. Furthermore,
the bathing behavior did not appear to be triggered by any discernible stimulus except
the presence of others doing the same thing. What triggers the behavior and what is its function?
Foraging. Cade (1953) described a single observation of Common Redpolls feeding on
snow-covered weeds in snow tunnels, although there was no evidence that they had made
them. During my observations, I saw no vegetation where the birds made tunnels or depressions.
There was no evidence that the Common Redpolls were seeking food because the
behavior occurred only immediately after they had had an ad libidum feeding opportunity
at the nearby feeders. In one instance, approximately 20 Common Redpolls gathered to
pick at the tops of a Chenopodium sp. (goosefoot) inflorescence that was exposed above the
snow (Fig. 1). They returned to this site at least 20 times in succession in repeated short,
in-unison flights, but none of the birds made an attempt to burrow into the snow to obtain
seeds from the mostly still-buried inflorescences. The observed tunneling behavior cannot
be explained as a foraging strategy because it did not occur in the presence of food, but
rather in its absence.
Bathing. In one report (Palmer 1949), Mrs. E.A. Anthony of Mount Desert Island, ME,
reported Common Redpolls on the roof of her garage plunging head first into wet snow and
fluttering their wings like birds taking a bath. The behavior suggested to the author that the
birds were bathing. In the present observations, however, there was never any melt-water
in the snow when the birds were fluttering in it and making tunnels. To the contrary, in my
observations, when the snow was wet they did not bathe. Furthermore, there was never any
bill-preening of the feathers, a behavior typically associated with the stereotyped bathing
mvements of birds (B. Heinrich, pers. observ.). Regarding my observations, it is also unlikely
that this one finch species would or could have soiled their feathers when all the others
were also feeding on dry sunflower seeds at times when the ground was always “clean”
because of snow cover.
Shelters. Small birds are vulnerable to cold because their large surface/volume ratio
accelerates heat loss (Dawson et al. 1993). They may compensate by huddling in groups
or using shelters. Auriparus flaviceps Sundevall (Verdin) build winter nests and huddle in
them communally, thereby gaining considerable energy economy (Buttemer et al. 1987).
Sitta pygmaea Vigors (Pygmy Nuthatch) huddle in existing tree holes overnight (Guntert et
al 1988), and Picoides pubescens L. (Downy Woodpecker) and P. villosus L. (Hairy Woodpecker)
excavate tree holes in the winter and use them specifically for repeated overnight
sheltering (Kilham 1984). Regulus satrapa Lichtenstein (Golden-crowned Kinglets) travel
in groups and form a huddle on tree branches at dusk (Heinrich 2004).
Rather than being a by-product of some function, Common Redpolls may make furrows
and burrows in the snow as shelters for energy economy and warmth. The con-specific
Ruffed Grouse submerge themselves into loose snow at night (Robbins 1901), and also rest
in these shelters on stormy days (B. Heinrich, pers. observ.). However, in the Maine woods,
I have never observed a snow furrow or snow cave on/in the snow that might have been
made by a small bird, which suggests that Common Redpolls do not construct such shelters
here.
Common Redpolls presumably overnight during northern winters without the use of tree
cavities. How they survive cold winter nights on their northern home in the Arctic tundra is
not known, although it has been suggested they overnight in snow tunnels (Clement 1968,
Knox and Lowther 2000). In northern Finland, experimentally caged Common Redpolls
examined in winter “experienced discomfort when sheltered in snow” and died if forced
N51
2014 Northeastern Naturalist Notes Vol. 21, No. 4
B. Heinrich
to remain in snow burrows for more than an hour; instead they appeared to prefer to perch
side-by-side (Korhonen 1981). At sub-zero tempertures in East Greenland, Plectophenax
nivalis L. (Snow Bunting) overnighted in narrow cracks of cliffs and in scrapes on the lee
side of snow banks in drifting wind, where they become buried (Meltofte 1983). In the present
observations, the Common Redpolls invariably flew off over the forest in the evening,
and came flying back to the feeders from the forest at dawn; none ever stayed overnight in
the clearing where they had made their snow holes. The birds never remained in their burrows
for more than several seconds, even in wind and at temperatures as low as -26 °C, and
none returned to them in the evening; thus, the functional proximal significance for making
snow cavities in order to overnight or escape cold is doubtful.
Social Stimulation and Play. A commonly employed explanation for apparently senseless
behavior is to define the activity as play. Play provides practice for biologically relevant
behavior before it is needed, and includes any behavior motivated by enjoyment rather than
anticipated reward. Play by birds, such as that in Corvus corvax L. (Common Raven; Heinrich
and Smolker 1998) is an activity that, though not proximally utilitarian, has a likely
utilitarian function expressed in a non-utilitarian context. A similar narrative may apply to
the Common Redpolls’ behavior.
Common Redpolls are one of the smallest boreal finches (~13 g), and because of their
small size are thought to have evolved relevant physiology and behaviors that maintain their
energy balance and temperature regulation during winter in their high Arctic snow-swept
tundra habitat. My observations demonstrate that Common Redpolls not only have the
ability, but also the propensity for burrowing in snow. Although these behaviors are likely
derived from bathing movements, they endow the birds with a behavioral repertoire that
is potentially useful for surviving deep cold in the subnivian zone in the Arctic. However,
small birds in the forest might be vulnerable to the common subnivian ground predators,
such as Blarina brevicauda (Say) (Short-tailed Shrew). Furthermore, given that during each
of the 2 winters during which I conducted my study, there were 1 or 2 days of snow-melt
followed by deep frost, I suspect that any small bird spending the night under snow in such
a temperature regime of alternating rain or thawing with freezing would entail the lethal
risk of becoming trapped under an ice crust. That is, the birds’ overall costs of overnighting
under the snow may be greater in this climate rather than in the high Arctic, and they desist
from going through their entire overnighting behavioral repertoire. Undoubtedly, winter
thaws and variable weather make the survival strategy of subnivian overnighting very risky
for small birds.
Literature Cited
Buttemer, W.A., L.B. Astheiner, W.W. Weathers and A.H. Hayworth. 1987. Energy savings attending
winter-nest use by Verdin (Auriparus flaviceps). Auk 104:531–535.
Cade, T.J. 1953. Sub-nival feeding of the redpoll in interior Alaska: A possible adaptation to the
northern winter. Condor 55:43–44.
Clement, R.C. 1968. Common Redpoll. Pp. 407–421 , In A.C. Bent and O.L. Austin (Eds.). Life Histories
of North American Cardinals, Grosbeaks, Buntings, Finches, Sparrows, and Allies. Part 1.
US National Museum Bulletin 237, Vol. 1. Smithonian Institution Press, Washington, DC.
Collins, J.E., and J.M.C. Peterson. 2003. Snow burrowing by Common Redpolls (Carduelis flammea).
The Kingbird 53(i):13–22.
Dawson, W.R., R.L. Marsh, and M.E. Yacoe.1993. Metabolic adjustments of small passerine birds for
migration and cold. American Journal of Physiology 245:R755–R767.
Furness, G. 1987. Common Redpolls excavating snow burrows and snow bathing. The Kingbird
(Spring):74–75.
2014 Northeastern Naturalist Notes Vol. 21, No. 4
N52
B. Heinrich
Guntert, M., D. Hay, and R.P. Balda. 1988. Communal roosting in the Pygmy Nuthatch: A winter
survival strategy. Proceedings of the International Ornithological Congress 19:1 964–1972.
Heinrich, B. 2004. Overnighting of Golden-crowned Kinglets in winter. Wilson Bulletin115:123–124.
Heinrich, B., and R. Smolker. 1998. Play in Common Ravens (Corvus corax). Pp. 27–44, In M. Beckoff
and J.A. Byers (Eds.). Animal Play: Evolutionary, Comparative, and Ecological Perspectives
Cambridge University Press, Cambridge, UK. [ # PP.?].
Kilham, L. 1984. Life History Studies of Woodpeckers of Eastern North America. Nuttall Ornithological
Club, Arlington, VA. 240 pp.
Knox, A.G., and P.E. Lowther. 2000. Common Redpoll (Carduelis flammea). No. 543, In A. Poole and
G. Gill (Eds.). The Birds of North America. BNA, Inc., Philadelphia, PA.
Korhonen, K. 1981. Temperature in the nocturnal shelters of the Redpoll (Acanthis flammea L.) and
the Siberian Tit (Parus cinctus Budd.) in winter. Annales Zoologici Fennici 18:165–168.
Meltofte, K. 1983. Arrival and pre-nesting period of the Snow Bunting, Plectophenax nivalis, in East
Greenland. Polar Research 1:185–198.
Novikov, G.A. 1972. The use of under-snow refuges among small birds of the sparrow family. Aquilo
Seriologica Zoologica 13:95–97.
Palmer, R.S. 1949. Maine Birds. Bulletin Museum of Comparative Zoolog y, vol. 102.
Robbins, R.C. 1901. Ruffed Grouse, Bonasia umbellus, in snow. Auk 18:190–191.
Sulkava, S. 1968. On small birds spending the night in the snow. Aquilo Seriologica Zoologica
7:33–37.