formation. Yet what Diamond (1975a) presented
amounted to a fairly comprehensive island biogeo-
graphical theory in its own right, which we term
island assembly theory.
Assembly rules
Diamond adopted the working hypothesis that
through diffuse competition, the component
species of a community are selected, and co-
adjusted in their niches and abundances, so as to fit
with each other and to resist invaders. In his stud-
ies he identified the following patterns or assembly
rules(quoted from Diamond 1975a, p. 423):
1 If one considers all the combinations that can
be formed from a group of related species, only cer-
tain...of these combinations exist in nature.
2 Permissible combinations resist invaders that
would transform them into forbidden combina-
tions.
3 A combination that is stable on a large or species-
rich island may be unstable on a small or species-
poor island.
4 On a small or species-poor island, a combination
may resist invaders that would be incorporated on
a larger or more species-rich island.
5 Some pairs of species never coexist, either by
themselves or as part of a larger combination.
6 Some pairs of species that form an unstable com-
bination by themselves may form part of a stable
larger combination.
7 Conversely, some combinations that are com-
posed entirely of stable subcombinations are them-
selves unstable.
These assembly rules were derived from several
forms of distributional data (e.g. incidence func-
tions, chequerboard distributions) for species and
guilds, alongside data for recolonization of defau-
nated islands. The ‘rules’ were the descriptions and
interpretations of these emergent patterns.
Incidence functions and tramps
Incidence functionsare an essentially simple idea,
reliant on the availability only of good data on
species distributions across a series of isolates. They
show the frequency of a species in relation to values
of a potentially controlling variable, such as area or
isolation (e.g. Watson et al. 2005). However, as orig-
inally employed by Diamond (1975a), they were
richness-ordered incidence functions: plots of
island species number, S, versus the incidence, J, of
a given species on all islands of that value of S.
Unless a huge number of islands is available, it is
necessary to group together islands of a given
range of values of S in order to obtain enough
observations in each class. Figure 5.1 shows the
functions calculated by Diamond for a subset of his
bird species. The legend provides a labelling of the
different distributional patterns, high-Sspecies,
A-tramp, and so on. In order to interpret these
labels some background is needed.
Diamond’s study area lies near the equator, with
the predominant natural vegetation cover being
rain forest. The New Guinea bird species pool con-
sists of about 513 breeding non-marine species.
Offshore, there are thousands of islands of varying
sizes and degrees of isolation, for many of which
bird data were available, in cases including
instances of successful and unsuccessful coloniza-
tion. Some land-bridge islands can be assumed to
have gained much of their avifauna at times of low-
ered sea level, and were thus regarded as supersat-
urated. Others, having undergone disruption by
volcanic action (e.g. Long and Ritter islands), were
regarded as displaced below equilibrium. Others
were regarded as equilibrial. Thus, within the 50
islands studied, although a significant species–area
relationship was recorded, there were statistical
deviations. These deviations appeared intelligible
in relation to historical and prehistorical events,
and were therefore interpreted in terms of long-
term trajectories in avifauna size and composition.
Some species of birds, such as Centropus violaceus,
Micropsitta bruijnii, and Artamus insignis, occur only
on the largest, most species-rich islands. These were
termedhigh-S(orsedentary) species (Diamond
1974), the precise usage being species found on the
four richest of the Bismarck Islands (islands with
81–127 species) plus not more than one from the
next richest category (islands with 43–80 species). At
the other extreme, a small number of species are
absent from the most species-rich islands and are108 COMMUNITY ASSEMBLY AND DYNAMICS