Many animals migrate great distances.  The mechanisms that
they use to find their way over changing terrain and feature-less
expanses, is a fundamental question in the study of animal behavior.
In this laboratory, students repeated a classic experiment that set the
stage for understanding the cues that migratory birds in particular
can use to find their way across unfamiliar areas in order to reach
their navigational goal.

Our Study Species. Racing pigeons, better known as "homing" pigeons, have a long
and honorable history in assisting the communication of humans.  These birds are
a domesticated form of the rock dove (Columba livia), that have been selected
to enhance the wild bird's natural ability to fly quickly to its roost and nesting site.
This predictable behavior and the ease with which this species is kept in captivity makes
it a perfect model organism for the study of the physiological mechanisms that
enable homing behavior.  Our study subjects were loaned to us by pigeon racing
enthusiast Ms. Genie McGee of Sunshine Lofts in Hernando, MS.

Navigation Cues.  In order to find their way home from Oxford, the pigeons must
have the ability to not only know compass direction, but also to sense the way
in which they have been displaced from their home (that is, they must have some
sort of internalized map of their own position on the planet relative to their home loft).
On bright sunny days, pigeons use a solarcompass.  Because the sun
follows an path across the sky from east to west at a regular rate, the pigeons use the
time of day (from their internal clock) and the sun's position to figure out compass
directions.  When it is overcast and the sun is not visible, pigeons appear to have
an internal magnetic compass that responds to the magnetic fields on the face of the
Earth.  Pigeons can also learn local landmarks if they have visited the area previously.
Our study pigeons had never been to Oxford and were tested on a sunny day.

Release Methods.  Each of 32 pigeons was released separately so that their
choice of compass direction was an independent event. After recording the
pigeon's leg band identification number, the student chose a release direction
by using a three digit compass direction from a random number table.  The bird
was released by gently tossing it a few feet into the air.  The movements of the
pigeon were tracked by several observers until it had disappeared from sight.
The "degrees from North" of the point on the horizon where the bird was last seen
was considered their choice of "home" direction.

Behavioral Observations. Before choosing a final direction,
most of the birds sampled various directions, sometimes repeatedly. Pigeon racers
call the zig-zagging or back and forth flying of released birds "routing". Three
pigeons landed immediately, either on the transport boxes holding the other pigeons,
or on the ground nearby.  Several birds flew back over us after having flown
in another direction.  Almost all the birds choose the western half of our release
site as the first direction they wished to fly.

Statistical Results.  Most but not all of our pigeons oriented in the general direction
of their home loft in Hernando.  None of them crossed the horizon at the exact
compass orientation towards their home (320 degrees from North).  To conclude
that pigeons have a significant ability to orient in the home direction, circular
statistics must be applied to the data.  Circular statistics tell us the probability that
our pigeons just randomly happened to choose the directions that they did.  We
already know that pigeons can navigate home.  The purpose of the statistical test
is to confirm that our class' study design was sufficient to demonstrate this process
conclusively.

This circular figure has 360 degrees in total and is oriented so that 0 degrees is North
(note that 0 and 360 are the same point).  The pigeons' home loft was in the direction
of 320 degrees.  The block dots stacked up outside the circle each indicate a
disappearance orientation for one pigeon.  The average direction that our pigeons
flew was about 256 degrees...or 64 degrees more west than they should have if they
were flying directly home.  Some birds flew in the direction opposite to home
when we last saw them.  Such pigeons eventually figure out their error and arrive
home a bit later than the others.
    The statistical analysis by the Animal Behavior students found that pigeons
did not orient randomly (Rayleigh Test, z = 12.45, P < 0.001), and were significantly
oriented in the direction of their home loft (V test, V = 2.35, P < 0.01).  The
P levels indicate the likelihood that pigeons happened to orient this way by chance
alone.  Pigeons that were kept in a covered crate previous to release did not appear
to orient less well than pigeons that could see out of their crate.  Pigeons hatched
this year were more likely (39%) to orient within 30 degrees of the home direction
than older pigeons (13%).  Pigeons that had active nests back in the home loft,
were on average 14 degrees more direct in their orientation towards home.

Conclusions.  The racing pigeons clearly were capable of orienting in the general
direction of their home loft, even though they had never been to Oxford, Mississippi
before.  They did not on average fly directly in the compass direction of their home,
although two birds were very close.  Because we could not see the birds as they passed
the tree tops on the horizon, we may have missed the birds as they routed closer in the
home direction.  Ms. McGee later reported that she observed some of her pigeons
flocking together as they flew north along Interstate 55, a north-south roadway
that passes Hernando, their home location.  It appears that some of the pigeons may
have waited for the other released birds to join them.  Indeed for pigeon races, the
birds are usually released in large groups.  We were surprised to see that the young
pigeons, which had not participated in races, had a more accurate orientation
direction average than more experienced, older racers.  Our sample size is a bit small to be
sure of this effect, but it is worth investigating further in future releases.  Finally,
pigeons with a reason to hurry home, appeared to orient more directly to the loft,
although sample size is small in this comparison as well.
    In addition to being enjoyable, this laboratory exercise demonstrated the techniques
that are used to study navigation ability in animals.  Students had to learn how to
describe map orientation, and calculate a special class of statistics for dealing with
'circular', rather than linear, data.  The students learned proper handling methods for
birds, confronted how to use a random number table to avoid biasing their release results,
and developed a respect for the important history of human reliance on homing pigeons.

Acknowledgments.  We are grateful to Ms. Genie McGee of Sunshine Lofts for
sharing her pigeons, time and considerable pigeon knowledge with us.  The
American Racing Pigeon Union can put you in touch with local pigeon racers.
Their website is www.pigeon.org.  This study was conducted with the approval of the
Institutional Animal Care and Use Committee.