The conservation of the forest resources of eastern Africa: past influences, present practices and future needs - Biogeography and Ecology of the Rain Forests of Eastern Africa - page 304

minimise water loss?' The short reply must be that
maimum water retenion will occur with the least
disturbed primary forest; but, as yet, 'In
Tanzania, watershed management is given very
little attenion' (Temu, 1984).

which cover habitat as well as species diversiies
would be opimal. This.is academic in the case of
eastern Africa as forest reserves are already highly
fragmented and of small size. Recent study in the
Usambaras has revealed the impoverished nature
of the avifauna of small isolated forest patches
(Newmark, 1991a).
Reserves should be big enough to contain
populaions of viable size. 'How big is a viable
populaion?' and 'how big an area does such a
populaion require?' are unanswered quesions
for African forest taxa; but see Collar & Stuart
(1985) for pracical ideas or orest birds. Popula-
tion geneicists believe 250 individuals may be
necessary to avoid problems of loss of geneic
heterogeneity (e.g. Wilcox, 1984; but see the
review of Roche & Dourojeanni, 1984, and works
in Frankel & Soule, 1981; Soule, 1986 and
Tsingalia, 1990). Lovejoy (1984) and Lovejoy &
Oren (1981) suggested that a reserve should be
big enough 'to maintain the characterisic diver-
sity of its component communiies', and gave a
rule of thumb that the required minimum area is
twice the space needed by a viable populaion of
the top carnivore. For leopard 500 m2 may be
needed, which is obviously not possible in eastern
African forests! Luckily leopard (and other top
canivores - some larger raptors) can range out-
side forest cover, but several other lesser car-
nivores and most herbivores cannot.
The short answers to quesions of how big an
area is needed to conserve genetic diversity must
be 'as many big areas as possible'.
Small conservaion areas may be more efecive
at retaining species if they are close to larger areas
which can act as a source of immigraion or
recolonisaion. Linkages in the form of corridors
or buffer zones, a good example of which is a
dendriic riverine forest strip (e.g. Harris, 1984),
are important to maintain geneic transfer
between forest patches and allow recolonisaion.
The corridor concept figured prominently in the
planning of a protected area network in the forests
of India and elsewhere in the world (Rodgers &
Panwar, 1988; Hobbs, Saunders & Hussey,
1990). Newmark (1991a) discusses the biology
and importance of corridors for bird movement in
the Usambara Mountains. Helliwell (1976) and

he esign of orest proteaed areas

The concept of an opimum design for protected
areas arose from island biogeographical theory
and from the geneics of small populaions. How
big an area, and how many individuals of a spe-
cies, are needed to ensure long-term viabiliy of
target populaions, are questions addressed by
these ideas. Design, though, is meaningless
unless there is capability to implement design on
the ground. The present patten of fragmentaion
is largely related to a past and present lack of
effecive land use planning. Short-term 'local'
economic interests frequently dominate over
longer term 'naional' conservaion interests.
The locaion of the protected area may be
important. Many authors have stressed the advan-
tages of conserving areas considered to be species
refuges or refugia (e.g. Diamond & Hamilton,
1981; Grubb, 1982; Prance, 1982). Refugia are
'Centres of Richness and Endemism' (Diamond,
1985) and have higher than normal species diver-
siies and numbers of endemics. This topic is
largely concened with the East TanzaniaKenya
refuge. Refugia are areas of persistent climate,
little subject to long-term patterns of change. You
get more species conserved for longer ime
periods per square kilometre (or per dollar
invested) in refugia than elsewhere (Myers,
1982).
The quesion of how big an area to conserve is
central to conservaion planning. Iniial theory
suggested the larger the better, and that one large
area was better than several small ones (e.g. May,
1975; Miller, 1978). Other authors (e.g.
Simberloff & Abele, 1976) stated that a number
of smaller reserves would conserve more species
than one large area. Certainly there is much
empirical evidence to show that when reserve size
is reduced species will become locally exinct, and
that such species loss is proporional to area
reducion. A large number of good sized reserves