In a first attempt to explain the evolution of the avian navigational syste
m, Bellrose suggested that compass mechanisms and the ability for true navi
gation had developed in connection with migration across increasing distanc
es. Yet birds use compasses, the mosaic and the navigational maps even clos
e to home and for homing. This means that those mechanisms must have develo
ped for orientation within the home range, with the necessity to optimize t
he everyday flights acting as selective pressure. In view of this, any atte
mpt to reconstruct the evolution of the avian navigational system must star
t out with the non-flying ancestors of birds.
Considering the requirements of orientation by landmarks and by using a com
pass, compass orientation with the help of the magnetic field appears to be
the simplest mechanism; consequently, it must be assumed to belongs to the
most ancient orientation strategies. The magnetic compass is wide-spread a
mong animals, but it appears to function according to different principles
among the various groups of vertebrates so that it is unclear whether birds
inherited their magnetic compass from their reptilian ancestors or develop
ed a mechanism of their own. The same is true for the sun compass. The cruc
ial role of the magnetic compass in the ontogenetic development of the sun
compass might indicate a similar relationship for the phylogenetic developm
ent.
Over short distances within the home range, orientation based solely on com
pass orientation appears possible, using the strategy of route reversal, wi
th non-straight routes being integrated. Since this strategy accumulates er
rors, it becomes inaccurate over longer distances, thus causing selective p
ressure to use local site-specific information. This leads to the formation
of the mosaic map, a mechanism that includes landmarks as well as compass
orientation. Today, the mosaic map of landmarks is a mechanism by itself, e
stablished according to innate learning principles that associate informati
on on path integration with site-specific information, thus forming a direc
tionally oriented mental representation of the distribution of landmarks. T
he navigational map is formed by applying the same principles to factors of
the nature of gradients; it thus appears to have developed from the mosaic
map. Whether or not it is a special development of birds associated with t
heir flying ability is unclear. Because the birds probably inherited the ba
sic mechanisms of orientation from their ancestors, one would expect these
mechanisms to be similar in all birds. For the mechanisms involving learned
components, this means that they are established following common rules. B
irds improved those mechanisms and adapted them to their specific needs.
Migration is assumed to have begun with non-directed search movements for r
egions offering better conditions. At this stage, the already existing mech
anisms of homing were sufficient for navigation between the various areas.
When these first movements turned into regular migration between two region
s, the migratory program began to evolve, starting out with spontaneous ten
dencies in a preferred direction. The magnetic compass may have served as f
irst reference system for the migratory direction; later celestial rotation
, indicated by the changing pattern of polarized light during the day, obta
ined its important role in indicating the reference direction geographic So
uth. In the course of Lime, sophisticated migration programs with changes i
n direction, controlling time programs, responses to trip ger mechanisms et
c. developed. The migratory direction and distance, i.e, the amount of migr
atory activity, continue to be subject to selective pressure so that birds
can respond to the environmental conditions in an optimal way. The transiti
on from daytime migration to night migration did not require new mechanisms
, as the magnetic compass can be used at any time of the day. Later however
the star compass evolved, which is to be considered a special development
of night-migrating birds, with its way of functioning well adapted to the s
pecific needs of migrants. Birds also developed the ability to derive infor
mation on celestial rotation from the rotating stars at night and to transf
er this information directly to the star compass. Since migratory habits ev
olved many times independently among birds, the same has to be assumed for
the specific mechanisms of migratory orientation. This means that they need
not necessarily be identical in all bird migrants. We are to expect conver
gent developments, however leading to mechanisms of the most suitable type.