and Wilson were not the first to recognize or theorize
about the species–area effect—that plots of species
number versus area have a characteristic form, dif-
fering between mainland and island biotas—but it
was they who developed a fully fledged theory to
explain it: the equilibrium theory of island bio-
geography, based on a dynamic relationship
between forces accounting for immigration to
islands (afforced on remote islands by speciation)
and losses of species from them.
Whether on islands or continents, the form of the
relationship between area and species number is of
fundamental biogeographical importance. Over the
past half century, there have been numerous island
biogeographical studies that have set out to
describe and understand the form of the
species–area curve, and the influence of numerous
other variables on species richness. What have
these studies shown? Are we any closer to under-
standing the factors that control species richness in
patches and islands, and how they vary in signifi-
cance geographically? We will show that
MacArthur and Wilson’s equilibrium theory has
actually proved remarkably difficult to test and that
while its heuristic influence remains strong, its pre-
dictive value appears limited. In large measure this
may be because of insufficient attention to the scale
of the study system.
The study of Community assembly and dynamics on
islands(Chapter 5) has interested natural scien-
tists since at least the mid-nineteenth century,
which explains why the botanist Melchior Treub
took the opportunity to begin monitoring the recol-
onization of the Krakatau islands just 3 years after
they were sterilized in the eruptions of 1883
(Whittakeret al. 1989). It emerges that island biotas
typically are not merely a ‘random’ sample of main-
land pools. There is generally some degree of struc-
ture in the data. Important insights into the
processes producing such structure have come
from Jared Diamond’s work on birds on islands off
the coast of New Guinea. Diamond (1975a) formu-
lated a set of ‘assembly rules’ based largely on dis-
tributional data. In this work he invoked a strong
role for interspecific competition in structuring
ecological assemblages on islands, but he also
recognized a role for long-term ecological (and
evolutionary–ecological) processes. As we will see
later, this approach became embroiled in heated
controversy, focused both on our ability to measure
pattern and on the causal interpretation of the
patterns detected. Other approaches have since
been developed that aim to detect and interpret
structure in island assemblages. One important
form of structure is nestedness, the term given to
the situation where upon ranking a set of island
species lists by species richness, each species set is
found to represent a proper subset of the next larger
species set. Again, the measurement of nestedness
and its causal interpretation turn out to be less than
straightforward, but it appears clear that a signifi-
cant tendency to nestedness is common within
island (and habitat island) archipelagos.
Much of the research on island assembly has
been based around ‘snapshot’ data, capturing dis-
tributions at a particular point in time. However,
the dynamics of island biotas have also been the
focus of study and debate. For instance, Diamond
developed the idea from his New Guinea studies
that some species are ‘supertramps’, effective at col-
onizing but not at competing in communities at
equilibrium, whereas others are poor colonists but
effective competitors. This is a form of successional
structuring of colonization and extinction.
Successional structure can be recognized through
other approaches, for instance by analysis of the
dispersal characteristics of the flora of the Krakatau
islands through time, or by analysis of the habitat
requirements of birds and butterflies appearing on
and disappearing from the same islands. Krakatau
is unusual in having species–time data for a variety
of taxa, and to some extent they tell different sto-
ries, yet stories linked together by hierarchical links
between taxa and by the emergent successional
properties of the system.
Within this book, a recurring theme is the impor-
tance of scale, and we open the final chapter of this
section, on Scale and island ecological theory, with
a consideration of the scale of relevance of theoreti-
cal constructs such as the MacArthur–Wilson
model. There has been an increasing realization
that ecological phenomena have characteristic spa-
tial and temporal signatures, which tend to be
linked (Delcourt and Delcourt 1991; Willis and6 THE NATURAL LABORATORY PARADIGM