0198566123.pdf

increase of the species involved. This demonstrated
that equilibrium for slow-growing species (i.e.
those with a low intrinsic rate of population
increase) should be tested over longer periods, a
point equally clear from Schoener ’s (1983) review
of empirical data. Another interesting feature was
that in the presence of disturbance, turnover was
mostly accounted for by ephemeral populations
(arguably trivial ecologically). This held even in
equilibrial simulations, a finding again in concor-
dance with empirical evidence, in the form of bird
data from Skokholm (Williamson 1989a,b).
Figure 6.1 is thus offered as a restatement of
island ecological theory, in which the ‘narrow’ view
of EMIB is represented in the broader subject space

against other ‘narrow’ view alternatives. The four

extreme positions on turnover are thus dynamic

equilibrium, static equilibrium, dynamic non-equi-

librium, and static non-equilibrium. In practice,

they may be difficult to distinguish with the data

available. Moreover, the response times, dispersal

abilities, and trophic status of different taxa may

mean that comparative studies of different taxa

from the same set of islands adopt differing posi-

tions in this framework (Table 6.2; Bush and

Whittaker 1993). Once again, following Haila’s

argument, the characteristics of the target

taxon/group in relation to the spatial scale of the

system will largely dictate the applicability of these

labels to case study systems. Similarly, Crowell

FORMS OF EQUILIBRIA AND NON-EQUILIBRIA 155

HFLM Disturbance

'Hurricane

Joe'

LFHM

Sterilizing

eruptions

Interactive

equilibrium

Non-interactive

equilibrium

Time

Empirical data for Krakatau, heavily generalized. Includes successional features ('waves')

Projected trends to non-interactive equilibrium

Projected trend to interactive equilibrium, due to declining I rather tha rising E. N.B. In

practice this and the non-interactive form may not be separable due to pseudoturnover

Role of major environmental disturbances (e.g. a hypothetical 'Hurricane Joe') setting back

succession and altering I and E projections

High frequency, low magnitude events

Low frequency, high magnitude events

E

HFLM

LFHM

Rate

I

Figure 6.2A model of island immigration and extinction incorporating disturbance. The figure provides a highly generalized representation of
data for the recolonization of Krakatau and three hypothetical projections. (From Bush and Whittaker 1993.)