(bi-coordinate magnetic navigation) varying across the Earth’s surface or by celestial
information. Both the geomagnetic and celestial parameters (elevation angle to
certain star configurations and sky rotation) can be manipulated in the laboratory
by using large magnetic coils (see above) and a planetarium sky. Studies have been
performed with passerines where course shifts have been recorded as a response to
simulated geomagnetic and geographical displacements.
Large-scale displacement experiments with ringed birds, performed mainly
during 1930–70, have been used to study navigation abilities. The main aim is
to find whether displaced birds maintain the same heading and end up in the
“wrong” place (expected in clock-and-compass orientation), or whether they
change their heading in accordance with the displacement and end up in the cor-
rect place (i.e. true navigation to a specific goal). Some of the most spectacular
experiments involved over 15,000 starlings transported from the Netherlands in
autumn to release sites in Switzerland and Spain. In more recent years the orienta-
tion of caged passerine migrants has been studied during lateral displacements by
ship relative to their intended migration route (e.g. Åkesson et al. 2001a), asking
the same type of questions.
7.6.4Selection experiments
BothZugunruhe(restless behavior shown at migration times) and direction are
encoded in the birds’ genetic migration program (Berthold 1996), and are proba-
bly exposed to strong selection. The length and intensity of nocturnal migratory
activity can be studied in cages recording the bird’s jumping activity. The inheri-
tance of migratory activity has been studied both by selection- and cross-breeding
experiments demonstrating that the phenotypic character can be changed in only
a few generations. Furthermore, cross-breeding experiments with migratory
Blackcaps Sylvia atricapillafrom European populations with different migratory
directions (SW- and SE- in autumn) show that the migratory direction is inherited
in an intermediate fashion in the offspring (Berthold 1996).
7.6.5Circular statistics
A special type of statistics called “circular statistics” is required to analyze circular
data (e.g. Batschelet 1981), such as departure directions of migrating birds
recorded by radio-telemetry or migratory activity recorded in Emlen funnels.
Since we deal with directions that can be represented on a circle we cannot use
linear statistics to treat these data. This can be illustrated by the angular differ-
ence between 10and 350being only 20, with a mean direction of 0(north),
while the arithmetic mean of 180would erroneously indicate a mean direction
toward south. Vector addition is used to calculate the mean vector for a group of
174 |Migration