wardens do you need? Do you have to equip and
manage park headquarters? In many situations,
fewer larger reserves make more sense from these
practical,financial, and inevitably political perspec-
tives. Such considerations often have overriding
importance to decision makers, and many recent
commentators regard the theoretical debates as
having contributed little of direct practical value to
conservation (e.g. Saunderset al. 1991). However,
theimplicationsof different reserve configurations
still demand attention, irrespective of the fate of the
theoretical frameworks or practical realities of land
ownership and opportunity that spawned them.
Dealing with the leftovers
In most cases, natural scientists are unable to exert
great influence on the basic ground-plan of frag-
mentation, and even where they do have a role to
play, it is never a free hand. There is, moreover, a
range of biogeographical, economic and political
considerations which can be crucial to the strategy
adopted. This has been recognized even by the
advocates of the SLOSS ‘principles’. For instance,
the proposed reserve system for Irian Jaya, on
which Jared Diamond acted as an adviser, rightly
paid more attention to pre-existing reserve designa-
tions, distributions of major habitat types, and cen-
tres of endemism, and to economic and
land-ownership issues, than it did to the SLOSS
principles (see Diamond 1986).
Most often, conservationists today are in the
position of dealing with and managing a system of
habitat fragments, and perhaps modifying its con-
figuration slightly. Fragmentation may have been
carried out in a very selective fashion. Thus, on the
whole, farmers take the best land and leave or
abandon the least useful. In much of southern
England, this has meant that most remaining native
woodland is on very heavy, clay-rich soils, with the
lighter soils in agricultural use. The message is that
the species mix in your fragments may not be rep-
resentative of what was in the original landscape.
To give just one example, the short-leaved lime
(Tilia cordata) does not favour the heavier soils, and
it is now known as a species of hedgerows. In the
past it was probably an important forest tree
(Godwin 1975). Equally, the juxtaposition of land-
scape units may be critical. For instance, some
Amazonian forest frogs require both terrestrial and
aquatic environments to be present within a reserve
in order to provide for all stages of the life cycle
(Zimmerman and Bierregaard 1986).
It must also be recognized that humans may
interfere directly in events within reserve systems.
Bodmeret al. (1997) point out that although much
work has focused on extinction rates caused by
deforestation, many of the recorded extinctions in
the past few hundred years have been the result of
over-hunting. They collected data on the relative
abundance of large-bodied mammals in the north-
eastern Peruvian Amazon in areas with, and with-
out, persistent hunting pressure. They found that
species with long-lived individuals, low rates of
increase, and long generation times are most vul-
nerable to extinction. In one reserve in north-
eastern Peru, it was found that nearly half the meat
harvested by hunters in the buffer zones of the
reserve originated from mammals that are catego-
rized as vulnerable to over-hunting. Similarly,
Peres (2001) uses game-harvest data to argue that
there may be synergistic effects of hunting and
habitat fragmentation, such that persistent over-
hunting in fragmented landscapes drives larger
vertebrate populations to local extinction, espe-
cially from smaller habitat patches. Such results can
be used to help formulate appropriate policy
towards such activities.Trophic level, scale, and system extent
The optimal configuration of areas is liable to vary
depending on the type of organism being consid-
ered, and on the spatial scale of the system. If forced
to generalize with regard to trophic level, larger
reserves are more appropriate for large animal
species needing a large area per individual, pair, or
breeding group, and/or requiring ‘undisturbed’
conditions, e.g. the Javan rhino (Rhinoceros sondaicus)
(see Hommel 1990). Species requiring large areas are
often the ones most threatened by us and in need of
protection, e.g. top carnivores, primates, and rhi-
noceros. Whereas top carnivores, such as big cats,
may require large territories, it might be possible to266 ISLAND THEORY AND CONSERVATION