The montane butterflies of the eastern Afrotropics - Biogeography and Ecology of the Rain Forests of Eastern Africa - page 144

African species form the sister group of the
Holarcic or Palaearcic Arnnina as a group.
Thus a Palaearcic ancestry of Issoria seems most
likely.
There are nine species with a northern (or
northeasten) ancestry, i.e. more than one fifth
(21.4%) of all species in the area that are resric-
ted to montane grassland. In three species the
African populaions are considered conspecific
with populaions outside Africa. In the other spe-
cies the ancestry is more remote. Not surprisingly,
the northern element is srongest in Ethiopia
where two species reach their southern limit in
Africa (Golias erate, Aryreus hyperbius), and where
three more species with northen ancestry occur,
in total more than half of the montane grassland
species in Ethiopia.

inally South African and epanded their range to
the north.
In some instances the extenal connecions in
different directions concen the same species. For
the majority of the species (30; 71.5 % ) there is no
evidence of a non-African ancesry nor of an
ancesry outside EAT.

Ecological relationships

This secion examines if species whose geo-
graphic ancesry is obscure may have originated
on the spot from species with other habitat prefer-
ences. Therefore we shall restrict ourselves to the
genera whose species have an African ancestry.

Metisella. Most Metisella species live in forested
areas, but an apparently monophyleic group of six
species is found in open habitats. Only one of
these, M. carsoni, is restricted to montane grass-
lands, so its habitat preference seems derived. Its
sister species is, however, sill unknown. It could
be the sympatric omosa or an allopatric species.

Wes ten conneaions
There are three instances of a connecion
between EAT and Cameroun: Golias eleao, Issoria
baumanni and Gapys disjunaus (through connexivus
and bamendanus). In these cases the complete
absence of relaives in Cameroun points to an
East African origin of the Cameroun populaions.
It is interesing to note that the first two species
are known from a few localiies in Zaire west of
the area under study: they may have a wider
habitat tolerance than the other species. The case
of the Gapys species is less clear since the exact
relaionship between disjunaus, connexivus and
bamenanus is uncertain.

Keestes. Since there are no further species
restricted to montane grassland among the 18
species ofthe genus, the preference for montane
grassland in barbeae seems derived. Since its
range is so wide and potential sister species are
found in South as well as in East Africa, it is of
little help in elucidaing barberae's origin.

Gapys. The species are found in montane grass-
lands as well as in woodland. A switch in habitat
preference is not needed to understand the
present situaion, i.e. the ancestors of the two spe-
cies in Appendix 8.1 could also have lived in mon-
tane grasslands.

Southen conneaions
Here we consider only connecions with South
Africa, so we leave aside the case of Issoria
smaragdera which occurs just south of the area
under study in easten Zimbabwe. There are
remarkably few such connecions (four only; 10%
of the montane grassland species of EAT):
Keestes barberae, Golias eleao, Gapys disjunaus and
Aaizera stellata. In all cases it is the same species
occurring in EAT as well as in South Africa, no
sister group situaions being known. Since the
species concened have relaives n EAT (except
Golias eleao, which has a northen origin), there is
no reason to suppose that the species were orig-

Hapendyreus. Only one of the three species in
South Africa occurs in open habitats in general,
all 14 other species (including all species in EAT)
live in montane grassland. So also the ancestors of
the species in EAT probably preferred this kind
of habitat.

Aaizera. The only other coninental African spe-
cies overlaps the distribuion range of stellata over