from a failure to find the appropriate niche space
rather than from a failure to disperse to the island.
He therefore played down the relevance of
turnover. Others have supported this viewpoint
insofar as saying that ‘most genuine turnover of
birds and mammals seems attributable to human
effects’ (Abbott 1983) or that turnover, where it
does occur, mostly involves subsets of fugitive
species and is thus ecologically trivial (Williamson
1981, 1988, 1989a). A similar conclusion was
reached in a study of newly isolated Canadian
woodlots. Once successional effects were removed
from the analysis, turnover within the vascular
flora, if occurring at all, was at a very slow pace
(Weaver and Kellman 1981).
A similar distinction, between the dynamic and
the static, can be made within non-equilibrium
ideas. First, we may recognize disturbed islands:
those in which equilibrium is reached only rarely
because environmental dynamics outpace the
response times of the biota (cf. the disturbance
hypothesis of McGuinness 1984). The biotas may
track changing equilibrium points through time,
but always remain a step or two out of phase. This
idea has been given various labels (e.g. Heaney
1986; Bush and Whittaker 1993; Whittaker 1995),
but within the scheme developed here it is identi-
fied as the dynamic non-equilibrium hypothesis.
We may distinguish this notion from that intro-
duced by Brown (1971) for relictual assemblages
dominated by extinction. In Brown’s system,
although isolates may be losing species on a mil-
lennial timescale, and are thus in a non-equilibrium
condition, on ecological timescales the system of
isolates appears effectively static, i.e. they fit the
static non-equilibriumcondition.
We regard this diagrammatic model as being of
heuristic value although, of itself, it lacks predic-
tive capacity. Moreover, it is often going to be
difficult to assign a study system to a particular
condition such that ready agreement will
be reached amongst ecologists that the data
unequivocally support a particular interpretation
(compare e.g. Bush and Whittaker 1991, 1993, and
Thorntonet al. 1993). See Box 6.1 for an illustra-
tive discussion.
The difference between the dynamic non-
equilibrium hypothesis and the dynamic equilibriumFORMS OF EQUILIBRIA AND NON-EQUILIBRIA 151Table 6.2 Exemplification of classification of studies of island richness and turnover as per Fig. 6.1, showing that different taxa from within a
single island group (Krakatau) support different models, and that in different contexts, data from the same taxa (terrestrial vertebrates, birds,
invertebrates, plants) have been interpreted as supporting different models. (After Whittaker 2004.)
Dynamic, non-equilibrial Static non-equilibrial
- Krakatau plants (Whittaker et al. 1989; Bush and 7. Krakatau reptiles—several species introduced by people, only two species
Whittaker 1991) have been lost, both related to habitat losses (data in Rawlinson et al. 1992) - Krakatau butterflies (Bush and Whittaker 1991) 8. Great Basin mountain tops, North America, small mammals (Brown 1971)
- Bahamas, plants on small islands (Morrison 2002b) 9. Lesser Antilles—small land birds, examined over an evolutionary time-frame
(Ricklefs and Bermingham 2001)
Dynamic, equilibrial Static, equilibrial
- Krakatau birds are a reasonable fit, although showing 10. Krakatau terrestrial mammals—no recorded extinctions to date
clear successional structure in assembly and turnover (Thornton 1996)
(Bush and Whittaker 1991; Thornton et al., 1993) 11. Bahamas, ants on small islands (Morrison 2002a) - Mangrove islets of the Florida Keys, arthropods 12. Oceanic island birds (Lack 1969, 1976 [but see Ricklefs and
(Simberloff and Wilson, 1970) Bermingham 2001]; Walter 1998) - British Isles, birds on small islands 13. Canadian woodlots plant data (if successional effects are
(Manneet al., 1998) removed from the analysis) (Weaver and Kellman 1981)