Biogeography of East African montane forest millipedes - Biogeography and Ecology of the Rain Forests of Eastern Africa - page 113

The Uluguru Mountains have been sampled
chiefly at the northen end, where accessible from
Morogoro, but most of the available material has
not yet been studied. Endemism appears to be
high: a spirosrepid genus (Hap logonopus), and an
odontopygid (Hofmanies), two oxydesmid genera
(Allocotoproctus and Morogorius), a gomphodesmid
(Uluguia), and a paradoxosomaid (Suohelisoma)
have already been described and yet others are
idenified (R.L.H., unpublished data). Interest-
ingly, these taxa are all monotypic. Although the
oydesmid genera Lyoesmus, Gonepacra and
Rhooesmus occur in the Ulugurus, there is little
indicaion of rampant speciaion such as
Ceratoesmus has produced (although future
eploraions, paricularly in the south, may change
that impression). Rhoodesmus planus has a
subspecies on Pongwe Mountain, and another at
the eastemost end of the Uzungwas, and sister
species of Gonpacra occur in the Ulugurus and
Uzungwas, suggesting substanial recent connec-
ions between the two.
Some elements of the Uluguru fauna occur also
in the Mindu Mountains north and west of Moro-
goro, and at least a local populaion of
Rhooesmus planus represents the auna on
Pongwe Mountain (the only place at which a
rhododesmid coeists with a Ceratoesmus). There
are no known connecions with the millipedes of
the Ngurus or Ukagurus, but of course these
ranges are sill terae incognitae as regards soil
animals.
Proceeding southwest along the mountain arc,
knowledge of the Diplopoda regrettably decreases
exponenially. For the Uzungwa range, only the
viciniy of Mwanihana (above Sanje) has been
sampled, and of extant collecions, little more
than the oxydesmids have been examined. As
already noted, these animals imply close former
connecions with the Ulugurus despite the
present lowland savanna barrier between them.
There is nothing yet on record in the category of
endemic genera, although the oxydesmid genera
Gonepacra and Iingius are nearly so and both
seem to have indulged in local speciaion (three
Uzungwean species in each). There is also some
very meagre evidence that the fauna at the
westen end of the Uzungwas (e.g. around

Mufindi) differs from that at Mwanihana, with
more influences from the 'Katangan' fauna.
So little is known concening the millipedes of
the Poroto Mountains (in which Mount Rungwe
is located), that nothing tangible can be men-
ioned at this point.

Conclusions
It is almost idle to underscore the point that the
Eastern Arc mountains with any remnants of orig-
inal rain forest represent a natural laboratory of
surpassing interest for the study oflocal evoluion
and dispersal: this is already known for many
plant and animal groups. Millipedes compel
attenion, however, because of the more intensive
degree of endemism induced by their biotope
requirements, which for most species are simply
the coninuous eistence of damp broadleaf or-
est. With rather limited vagiliy, millipedes
naturally respect eeological isolaing barriers and
local speciaion tends to be substanial. For
instance, former connecions with the orest
blocks of the Congo and West Africa are evident
only at the family level, there being no millipee
genera in co mmon between the Tanzanian base-
ment mountain faunas and those of the Congo
and Camerounian forests. As noted, differences
between the montane faunas even within
Tanzania tend to be at the generic level, with few
species occurring on more than a single range.
This degree of 'fine-tuning' makes the study of
millipede evoluion paricularly intriguing. An
addiional point is that small soil animals may
survive in remnant patches of rain forest too small
to support large or even moderate sized
vertebrates, so long as enough remains to ensure
adequate substrate dampness.
The raw materials for study are in place
already. Three major problems remain to impede
their uilisaion. First and foremost, adequate
protecion of representaive areas must be in
effect. Secondly, sampling of the fauna must be
undertaken on a comprehensive basis. Thirdly,
somebody must be able to work up the materials
and translate field collecions into published
accounts embodying both basic systemaics and
evoluionay inferences. All three are formidable