0198566123.pdf

Following the island analogy, the first factors to
be included are isolation and area. Increased iso-
lation of reserves reduces migration into them from
other reserves. If a reserve is created by clearance of
surrounding habitat, then it follows that on initial
isolation the immigration curve should be
depressed. The contiguous area of habitat is also
reduced and thus extinction rate should increase.
At the point of creation, therefore, the habitat island
contains too many species (it may even gain fugi-
tive displaced populations), and the result is that it
becomessupersaturated. It follows that it should in
time undergo ‘relaxation’to a lower species num-
ber, a new equilibrium point (Fig. 10.5). Given
knowledge of isolation and area of patches, it
should be possible to estimate the number main-
tained at equilibrium in a variety of configurations
of habitat patches. On these grounds, Diamond and

May (1981) favoured larger rather than smaller

reserves, short rather than long interreserve dis-

tances, circular rather than elongated reserves

(minimizing edge effects), and the use of corridors

connecting larger reserves where possible (Fig. 10.6).

These widely cited suggestions spawned a largely

theoretical debate in which much appeared to hang

on the validity and interpretation of the EMIB.

Having considered the EMIB, MVP, and metapopu-

lation ideas ahead of it, some of the limitations of

the SLOSS approach will be immediately apparent

and so can be dealt with quickly.

The simplest criticism of this approach follows

the line that the EMIB is a flawed model. Hence it

provides no firm foundation for the development

of conservation policy. If policy-makers adopt such

theories as though providing formal rules, as some

did in this case, this criticism is justified (Shafer

264 ISLAND THEORY AND CONSERVATION

Near

patchI

“Mainland”

samplel Small

patch E

Far

patchI

Large

patchE

Biotic collapse

0

0

Immigration rate (

I)

d c b a

Extinction rate (

E)

Species number

Figure 10.5According to the assumptions of the EMIB, reducing area causes supersaturation as immigration rate declines and extinction rate
rises. This causes loss of species, i.e. ‘relaxation’ to a lower equilibrium species number from a →b→c→d. In the extreme scenario of ‘biotic
collapse’, the immigration rate is so low that the equilibrium species number is near zero. (Redrawn from Dawson 1994, Fig. 4.) In practice this
abstract theory may lack predictive power as to either numbers lost or rate of loss, being founded on assumptions of the system passing from
equilibrium (pre-fragmentation), to non-equilibrium (at fragmentation), to a new equilibrium condition (when IandEbalance again):
circumstances which may not apply in complex real-world situations. This is not to deny species losses, they do occur, but very often they are not
a random selection, but instead are drawn from predictable subsets of the fauna or flora, structured in relation to: particular ecological
characteristics of the species in question; habitat or successional changes in the isolate; and the nature and dynamics of the matrix around it.
Equally, isolates may acquire increased numbers of some species, or gain new species, sometimes of an ‘undesirable’ nature judged in terms of
their impact on the original species.