Herpetofauna of the eastern African forests - Biogeography and Ecology of the Rain Forests of Eastern Africa - page 192

to the mainland coastal forests of easten Africa.
Some species are able to eist in both the lower
forests on the coast and those of the mountains.
These include the anurans Metnsophyne
micranotis, Lptopelis lavomaculatus and
Sp elaeophyne methneri, the lacerid Gastropholis
prasina, a chameleon Rhampholeon brvicauatus
and the snakes Aparallactus wenei and
Philothamnus macrops. The origins of these are
difficult to determine. At least two of the anurans,
M. micranotis and S. methnei, escape the hotter,
drier condiions of coastal orest by burrowing
into crevices in soil and inside rotten logs.
Were these forest species which were able to
adapt to drier condiions at the edge of former
forest? This would seem likely in the case of S.
methnei and R. brvicauatus because they have a
generally ypical Easten Arc distribuion patten,
and the coastal locality at which they were col-
lected, Kilwa, once was much more forested than
at present.
A few of the lowland species from drier areas
may slowly be peneraing the orests. A new spe-
cies of gecko in the Lygotylus somalius species
group, L. howelli, has been discovered in ground-
water forest on Zanzibar (Pasteur & Broadley,
1988). L. schflei uluguuensis is the name given
to an isolated populaion in the Ulugurus, but the
species is generally one of hot, dry areas, with
other subspecies from woodland in Kenya. It may
now be isolated in the Ulugurus, but have moved
in from the drier porions of that range in which
woodland is found below moist forest. A parallel
situaion may eist in the lizards Agama montana
and A. agama usambarae. Moody (1980) notes hat
the family Agamidae is absent rom the ropical
forest belt. While these two agamas are found on
rocky slopes at the forest edge, their origins would
therefore appear to be outside the forest.
The colonising ability of various forest species
can be examined by studying the fauna of the
orests on the volcanic mountains of easten
Africa. To the northwest of the Easten Arc on
Mount ilimanjaro, Mount Meru and the
Ngorongoro Crater Highlands, very few forest
forms are found. The snake fauna contains no
orest species from the Easten Arc. What was
formerly regarded as a Kilimanjaro endemic (but

a non-forest species) and given specific status as
Crotaphopeltis kagelei is now treated as a subspe-
cies of a more widely occurring species, Dip-
sadoboa shrevei (Rasmussen, 1986). Only the
chameleon Chamaeleo udis (Ngorongoro and
Meru), and the skink Lptosiaphos kilimnsis, with
a wide mountain orest distribuion, have been
able to invade these forests and no species of
amphibian or repile seems to have evolved in situ
on these mountains. With the excepion of
Leptosiaphos kilimensis and the snake Dasypeltis
atra, there was little colonisaion of Mount Kili-
manjaro and Meru forests. Species from the for-
ests of the Easten Arc mountains were unable to
cross the relaively dry, open habitat which separ-
ates them from the volcanic mountains.
There is now much evidence to suggest that
climaic condiions were once moist enough to
have permitted at least coniguous, if not com-
pletely coninuous, forest belts from much of west
to easten porions of the African coninent. Dur-
ing the Pleistocene, this easten forest was
evidently much further fragmented, and as · condi-
ions changed, suitable condiions or moist trop-
ical forest largely remained only on the crystalline
block and associated mountains. These forests
have been termed 'refugia', and they first received
intensive study when Moreau (1966) coined the
term 'Tanganyika-Nyasa' system when consider-
ing their avifauna. Much earlier, Loveridge
(1933, 1952) had emphasised the importance of
the isolaion of the UsambaraUluguru forests,
though evidence available to him at that ime sug-
gested that there was less connecion between
those mountains and the Uzungwas than we now
know to be the case. From evidence based on the
distribuion of amphibians and repiles as well as
birds (Stuart, 1983) it now seems clear that the
Uzungwa Mountains as well as others in the south
must be included in any postulated refugia.
Certain plateau forests, such as Rondo and
Kiono, may also have retained sufficient moist
forests to permit the survival of some forest-
dependent species or to have been colonised by
these later.
Laurent (1973) notes that recurrent phases of
conracion and epansion of forest provided
opportuniies for geographic isolaion which may