0198566123.pdf

species would be lost and might not be replaced by
immigration for some years (e.g. Paine 1985).
Average this across the whole island assemblage and
the island will, at any one time, fail to ‘sample’ a sig-
nificant number of (principally) the rarer birds found
in the mainland habitat patch (i.e. they will be
absent). Preston showed that a log–normal series of
abundance should give rise to zvalues of approxi-
mately 0.263, towards the low end of the range of
values then available from islands and above those
of continental patches. These differences therefore
pointed to a role for isolation via population migra-
tion, and, in short, Preston’s theorizing, combined
with his analysis of island data, led him inevitably to
a focus on turnover in his 1962 article, foreshadow-
ing by a few months several of the elements of
MacArthur and Wilson’s (1963) paper.
For present purposes these points of detail are not
critical, yet it remains of compelling interest to theo-
reticians to determine the best species abundance
distribution (SAD) model, because SADs are foun-
dational to much ecological theory (e.g. Hubbell
2001). One crude way of analysing this indirectly is
by examination of the form of the ISAR, as on theo-
retical grounds it is expected that a plot of Sversus
logAshould produce a straight line where the loga-

rithmic series applies, whereas a log Sversus logA

plot should produce a straight line where the

log–normal series applies. By reference to the argu-

ment in the previous paragraph, it might be antici-

pated that the former situation would be found to

apply for large samples (or areas) and the latter for

small samples (or areas). While noting exceptions,

both Williamson (1981) and Rosenzweig (1995) con-

clude that empirical studies from islands show a

fair degree of support for Preston’s model: the

species–area plot is more usually linear on a log–log

plot than in any other simple transformation.

Although this has a wider significance, for the

moment it is sufficient to note that the position taken

on species–abundance patterns by Preston, and by

MacArthur and Wilson, was entirely reasonable.

The distance effect

MacArthur and Wilson (1963, 1967) predicted that

ISARs would become steeper with increasing geo-

graphic isolation. Unfortunately, the increased

impoverishment of island biotas with increasing

isolation is confounded with variations in other

properties of islands, particularly their area. The

distance effect has turned out to be difficult to test;

THE DEVELOPMENT OF THE EQUILIBRIUM THEORY 83

Increasing isolation

diminishes passive

sampling by islands

Number of species, arithmetic scale

Abundance of individual species, logarithmic scale

b a

Veil line

Increased sampling effort

by biologist

Figure 4.3Preston’s log–normal species–abundance relationship. The portion of this hypothetical abundance distribution that is sampled
may be a function of either active sampling effort (by the biologist) or passive sampling effort (by an island). Only the portion to the right of the
veil line will be sampled.