The montane butterflies of the eastern Afrotropics - Biogeography and Ecology of the Rain Forests of Eastern Africa - page 148

without a geographic pattern of phylogenetic rela-
tionships. Whatever its origin, once a montane
forest species had come into being it could follow
its own course. According to opportuniies
offered it could disperse through and differeniate
in the montane habitats in EAT, possibly obscur-
ing its origin. Thus, we must consider the follow-
ing scenario.

forest butterflies (in a broad sense, not ascending
over 1500 m); from butterflies living in open
habitats (woodland to treeless habitats); and from
other montane forest butterflies. Scenarios for an
origin from other mountain areas and for frag-
mentaion of a coninuous forest belt with adap-
taion from lowland to highland condiions have
just been dealt with. They both have a strong
geographical component. Here we are concened
with the possibility that the montane forest butter-
flies arose within EAT from butterflies living in
lowland forest or in non-forest habitats. It has, of
course, also a geographical component since dif-
ferent habitats cannot occupy exactly the same
place, but they form a mosaic, a patchwork within
EAT. Consequently, twin relaionships can be
epected within EAT and between different
habitats, and not between EAT and some area
outside. In the case of origin from lowland forest
butterflies the diference with the fragmentaion
scenarios is that it is irrelevant whether or not
there ever was a coninuous orest belt. What
counts is the presence of an ancestor at lower
elevaions in EAT, and how this ancestor came
there is unimportant. To come into consideraion
for this scenario a montane forest species must
have a sister species in forests at lower elevaions
but in the same area. Of course, if such situaions
eist they can also be eplained by a secondary
dispersal of the lowland forest species into EAT
after fragmentaion of a coninuous belt; but apart
from this being an extra assumpion, why should a
lowland forest species do so if it could not survive
there before without adaping itself to montane
forest?
It seems unlikely that the old mountains of the
Easten Arc, if they carried montane forest from
an early date, had no butterfly fauna at all. There
was plenty of ime for this fauna to develop rom
the fauna around, apart from the possibility of a
contact with orests to the west. If so, the Easten
Arc mountains could be expected to have the
highest number of montane forest butterflies.
Later this fauna could have spread to other
mountains where in the meanime also montane
forest species could have evolved from species
with other habitat preference. As a result a mosaic
of partly overlapping ranges can be expected

Intenal development
If the montane forests of EAT ever were con-
inuous, irrespecive of a possible contact with
lowland forest or montane forest in Cameroun,
they must have had a more or less uniform fauna.
In that case we may expect to find widespread
species as well as monophyleic groups of
allopatric species together covering most or all of
the mountain ranges. A complicaing factor could
be large-scale local exincion. Several authors
(e.g. Livingstone, 1975; van Zinderen Bakker,
1982) maintain that during the height of the last
Glacial (about 18 000 BP) the climate was not only
cooler but also much drier and most of EAT was
covered with open woodland or arid vegetaion.
Forests were found only in the wettest areas such
as the Easten Arc mountains and the Westen
Border Range. If this really happened there
should be a high concentraion of endemics in
these refugial areas, and few or no endemics in
the other mountain ranges. If the montane forest
fauna was uniform before it became exinct except
in the refugia, there should be either many shared
species or many sister relaionships between the
Eastern Arc mountains and the Westen Border
Range.
The scenarios are not mutually exclusive: all
could have occurred separated in ime, or some
even simultaneously but in different parts of the
area.

Notes on the genera

All species and subspecies of butterflies that in
EAT are restricted to montane forest are listed in
Appendix 8.2. Data sources are the same as for
the montane grassland butterflies. Addiional
references are menioned below.