deriving general principles for reserve systems
from island theory. Increasingly, the resulting lines
of reasoning have been attacked on the grounds:
first, that many valid generalizations are self-obvi-
ous platitudes; secondly, that models are often both
too simplistic and effectively untestable; and,
thirdly, that many theoretical generalizations have
achieved the status of dogma (Doak and Mills
1994). Yet, the general principles derived from
island theory have influenced conservation policy
and they remain pervasive. Because of this, it is
important to understand both the strengths and,
particularly, the weakness of the ‘island’ theories
discussed in this and the previous chapters.
If our focus is on global losses of species diver-
sity, then clearly the loss of—and fragmentation
of—species rich habitats is a key concern. But, the
stochastic consequences of reduced population
sizes are not necessarily simple to understand and
model. Some species can persist well in fragmented
landscapes providing that the areas of remaining
habitat are not too small and isolated. Much may
depend on the previous history of the region
(extent of initial departure from equilibrium
assumptions), the taxa considered, and the scale
sensitivity of the systems, and how the habitat
matrix (the land outside the fragments) is managed.
Simple species–area models, particularly as applied
on regional scales, may not in the end provide reli-
able estimates of how many species really are ‘com-
mitted to extinction’, i.e. still around but heading
for an inevitable nemesis. Some would regard this
assessment as complacent and over-optimistic. For
instance, Pimm and Askins (1995) argue that con-
cerns about the precise causal mechanisms of
extinctions, as expressed here, miss the point and
that, in practice, simple predictions of species losses
based on species–area relationships are actually
fairly reliable. From their analyses of North American
forest birds they suggest that if such estimates err,
they tend to be conservative, i.e. predicting fewer
extinctions than actually occur. We are not con-
vinced that their claim can be generalized, as so
much seems to depend on issues such as system
configuration and the actual drivers of species
threat and extinction (see discussion of the Atlantic
forest analyses, above).
The problem of biodiversity loss (including
population-leveldecline, genetic erosion and species
losses) is a depressingly real one and the pace of
species extinctions is undoubtedly accelerating
(Whitmore and Sayer 1992; Turner 1996). Yet, as
Simberloff et al. (1992) observe, ideas derived from
island theory, and the so-called SLOSS principles
(specifically the benefits of corridors), became
embedded in policies of bodies such as the IUCN
and World Wildlife Fund without having been val-
idated. Thus, policy-makers and legislators have in
cases acted upon such uncertain scientific ‘princi-
ples’. Once released from the bounds of scientific
journals, scientific information is often poorly
understood. It is therefore important not to offer up
general principles and rules where the evidence for
them is equivocal and, in cases, contradictory
(Shrader-Frechette and McCoy 1993).
Nonetheless, we should not simply discard the
theoretical frameworks of island theory, metapop-
ulation theory, etc. They have their relevance to
particular systems and a wider and important role
in helping us to understand the sorts of effects that
might be operative in real-world contexts, and that
therefore should be tested for or monitored. As
Doak and Mills (1994) have put it, we should face
up to the problems rather than use them as an
excuse for ignoring the theoretical debates: ‘it is
incumbent on us to teach such complexity to
managers and nonbiologists, rather than attem-
pting to snow them with undefendable over-
generalizations’.
The notion that species disassembly (as well as
assembly) is fairly structured and predictable, at
least in terms of functional ecological types or
guilds, is one that has reappeared in several guises
in this chapter (e.g. from studies of relaxation,
experimental fragmentation, incidence functions,
nestedness, and edge effects). In the case of the
Manaus habitat fragmentation study, similar struc-
turing was evident both in the initial losses (disas-
sembly) and, in particular circumstances, in the
recovery of bird communities that followed. This
reminds us, first, that habitat change does not have
to be a one-way process, and, secondly, that it
might be possible to target conservation efforts
fairly directly on to particular sets of vulnerable284 ISLAND THEORY AND CONSERVATION