leaves in the aftermath of the storm and also
feeding on the fleshy bracts of a storm-resistant
native liane. Pteropus tonganus, on the other hand,
seems to have adopted the unfortunate behavioural
response of entering villages to feed on fallen fruit,
where they were killed by domestic animals or
people.
Piersonet al. (1996) make the point that island
species of the tropical cyclone belt have evolved
with recurrent cyclones. It appears in this case that
the species of narrowest geographical range actu-
ally has evolved more effective responses to
cyclone events. Given that flying foxes can live for
20 years or so, individuals may experience several
severe storms in a lifetime. Such events in them-
selves are unlikely to lead to extinction of the
species, provided that other threats, such as habitat
loss and hunting, are constrained. Where the two
are combined, however, extinction becomes more
likely (Christian et al. 1996). Although wind
damage is typically patchy, it seems reasonable to
suggest that areas of high topographic complexity,
e.g. volcanic cones and deep valleys, are the most
likely areas to retain patches with some foliage, and
so should be given priority in reserve design.
6.5 Future directions
... in practice it is next to impossible to find a set of
islands without environmental change or evidence of the
severe impact of man’s activities
(Case and Cody 1987, p. 408)We have learnt a lot from studying island ecology
under the ruling dynamic, equilibrial paradigm.
We can continue to do so, whether focusing on
improving macroecological models for island
species number by incorporating energetics or by
new approaches to nestedness and assembly struc-
ture. In this chapter we have reviewed evidence for
departures from the dynamic equilibrium condi-
tion. Of course, some degree of variance can be
accepted within the equilibrium paradigm, as sys-
tems of rapid response rates and small amplitude
environmental fluctuations may swing rapidly
back in to balance, such that most of the time an
island could indeed be close to equilibrium.
Application of dynamic, equilibrial models does
not demand a precise fixed value (Simberloff 1976).
However, depending on the relative timescales of
changes in the island environment and response by
the biota, we contend that islands may variously
be: always in equilibrium; in equilibrium most of
the time but occasionally disturbed from equilib-
rium; or never in equilibrium, but always lagging
behind changes in environment.
The emphasis on habitat determinism by Lack
and others has been opposed by many supporters
of dynamic island biogeography. They see it as
returning the subject to a narrative form. The pres-
ence of many endemics on very isolated islands
clearly demonstrates that even among a taxon of
generally good dispersers, e.g. birds, there comes a
point where isolation is sufficient to render immi-
gration and colonization a rare and essentially
probabilistic process. However, there is no need to
draw such a stark distinction between dynamic and
static views of islands. Studies reviewed in this
chapter suggest that is possible to build bridges
between these differing positions and that very
often the determinants of species number and com-
position lie somewhere between the essentially
stochastic, dynamic equilibrium and more ‘static’
behaviour, and require a more complex explanation
of community organization on islands than
provided for by the EMIB.
Conceivably, some island systems (large and
remote) might respond to a severe El Niño event,
for example, by exhibiting considerable flux in pop-
ulation sizes, biomass, and productivity, without
much impact on species turnover rates. In other
cases (small and near-shore islands), a severe cold
snap in the winter might have little effect on bio-
mass and productivity but might greatly impact on
populations of small birds, leading to measurable
alteration of species turnover rates. On the whole,
there appears to have been insufficient attention
paid to systems departing from the assumptions of
the dynamic equilibrium paradigm. Similarly, as
exemplified by Krakatau, where disturbances are
severe, or environmental change is very consider-
able, there may be long-term ecological changes
which unfold over the course of decades or
centuries. In such cases, a successional model of
island ecology can provide a more appropriateFUTURE DIRECTIONS 161