succession theory, and savanna ecology, to name
but a few. The island laboratories once again have
been a testing ground for developing and refining
ideas of broader relevance (Brown 1981, 1986), and
the EMIB was the starting point.
Hence, in the present chapter, we review the
elements that MacArthur and Wilson drew
together in their ecological model, and how well
the EMIB has fared in the light of empirical test-
ing. We go on to review recent developments in
our understanding of the dynamics of species
numbers in isolates, some of which build on, and
others of which differ from the MacArthur–Wilson
dynamic model. As will become evident, this
chapter pays little attention to the compositional
structure of island biotas; rather we are dealing in
the emergent statistical properties, or macroecol-
ogy(Brown 1995) of island assemblages. The reg-
ularities of form in these macroecological
properties point to the existence of underlying
mechanisms or rules governing ecological sys-
tems: the efforts reviewed in the present chapter
have been concerned with identifying what they
might be.
4.1 The development of the equilibrium theory of island biogeography
Several ingredients were combined in the dynamic
equilibrium model (see account by Wilson 1995). The
first is the long-known observation that species
number increases in predictable fashion in relation to
island area (e.g. Arrhenius 1921)—termed here the
island species–area relationship (ISAR)(Box 4.2).
For instance, the zoogeographer Darlington (1957)
offered the approximation that, for the herpetofauna
of the West Indies, ‘division of island area by ten
divides the fauna by two’. A second key ingredient
was that in examining data for Pacific birds,
MacArthur and Wilson (1963) noted that distance
between the islands and the primary source area (i.e.
island isolation) appeared to explain a lot of the vari-
ation. The third ingredient was the notion of the
turnover of species on islands, a force that Wilson
(1959, 1961) had previously invoked on evolutionary
timescales in the taxon cycle (Chapter 9), and which
other biologists had described as occurring on eco-
logical timescales, for instance in analyses of plant
(Docters van Leeuwen 1936) and animalTHE DEVELOPMENT OF THE EQUILIBRIUM THEORY 79Box 4.2 Typology of species-number curves and relationships: an attempt to demystify
It is important to distinguish the four frequently
used macroecological tools discussed in this box.
Within the wider literature, a range of acronyms
have been used for them. We use the acronyms
SAC, SAD, and ISAR for convenience, but the
reader should beware: the ‘A’ stands for three
different terms.
Species abundance distributions: SADs
The species abundance distribution refers to how
the number of individuals is apportioned across the
species present in a community. SADs can be
displayed graphically in several ways, e.g. as a plot
of the frequency of each species against its rank
importance in the community, or as a plot of
the number of species of increasing classes ofabundance (see Magurran 2004). SADs are not a
particularly key tool in island species-numbers
studies, but assumptions concerning the form of
the SAD are important to theoretical mechanistic
arguments linking patterns of richness to species
turnover (see text).
Species accumulation curves: SACs
The species accumulation (or collectors) curve is a
graph of the cumulativenumber of species (yaxis)
with increasing sampling effort (xaxis), where
sampling effort could be variously formulated,
e.g. (1) recording additional individuals within the
chosen sampling area, (2) additional sample plots
of a particular type from within a general locality,
(3) additional contiguous sampling plots.