The montane butterflies of the eastern Afrotropics - Biogeography and Ecology of the Rain Forests of Eastern Africa - page 147

elsewhere. There are three options: from other
mountain ranges, from lowland forest, or from
other habitats. These options lead to the following
scenarios, with deduced predicions regarding
distribution pattern.

much smaller during periods with a drier climate.
The age of the belt is very important here, either
before or after the Plio-Pleistocene mountain
building.
(a) Forest belt before the Plio-Pleistocene oro-
genic phase. The mountains present were mainly
the Eastern Arc mountains. In these mountains
the montane forest fauna originated and from
here the fauna dispersed at a much later date over
the other, more recent montane forests. The pat-
tern is: highest concentraion in the Easten Arc
mountains, the numbers decreasing with increas-
ing distance from these mountains; and a twin
relationship between taxa of the Eastern Arc
mountains and of the Central African lowland
forest block (which extends eastward as far as
west Kenya). Other patterns, e.g. a twin relaion-
ship between the mountains of EAT (or at least a
large part including the Easten Arc mountains)
and the lowland forest block that could arise if a
species later extended its range from the Easten
Arc mountains over other mountains as well, are
difficult to distinguish from the next scenario and
therefore not fit to falsify the present scenario.
(b) Forest belt after the Plio-Pleistocene oro-
genic phase. In this case there must have been a
more or less uniform fauna throughout the mon-
tane forests. Undoubtedly part of the fauna
became extinct when the forests retreated to
higher elevaions and distribuion areas of the
species became fragmented. Depending on the
occurrence of speciation as a consequence of the
fragmentaion of the habitat the predicted patten
consists of groups of closely related, allopatric
species without clear hierarchical relaionships
(they all arose 'at the same ime'), or of species
widespread in the montane forests of EAT (and
maybe difereniated here in a number of subspe-
cies) and also occurring in the Central African
forest block. Further, it may be expected that
there are no large differences between the areas
in numbers of taxa and numbers of endemics,
unless there is evidence of large-scale local
exinction (see below).

Origin from other mountain ranges
No coninuous forest connection with areas out-
side Africa is known in Tertiary or Quaternary
times. This leaves only other African mountains
as possible source areas. It has been stated
(Moreau, 1963, 1966; Carcasson, 1964; van
Zinderen Bakker, 1967) that the montane forest
belt was depressed so far that a contact was
established between the Cameroun highlands
(and perhaps mountain ranges in Angola) and the
mountains in EAT. Such a contact was assumed
to have eisted during the Pleistocene at the time
of the northen Ice Ages. If there had been any
substanial contact the predicted pattern should
show a fauna that in the Cameroun highlands and
the mountains in EAT must be similar, with
shared species or at most with sister species.
Relatives of the shared taxa should be found in or
around the source area. Since in this scenario the
montane forests of EAT were also in mutual con-
tact, the distribuion of the species shared with
Cameroun should be more or less even over the
mountains in EAT.

Fragm entation of a once continuous forest belt across
the Afican continent
Several authors (e.g. Carcasson, 1964) have pro-
jected such a belt. There is certainly good
evidence for dramaic climaic changes even or
especially in the near past (Pleistocene) (summary
in Hamilton, 1982). It seems, however, that these
projecions are more based on assumptions about
what would have happened to the forest during
colder or wetter times, than on geological or
palaeontological evidence. If such a belt ever
eisted, and if it eisted long enough, a more or
less uniform forest fauna must have occurred
across Africa, although relief may have caused
local variaions. After fragmentation the forest
fauna was trapped in EAT in refugia along the
coast and in the mountains. The refugial areas
were not extremely small, but may have been

Ecological differentiation
Ecologically the montane forest butterflies can
have originated from three sources: from lowland