...from cowpats to South America, it is difficult to see
what is not or at some time has not been an island.
(Mabberley 1979, p. 261)The islands move horizontally and vertically and thereby
grossly modify the environment on and around them. Life
forms too must evolve, migrate, or become extinct as the
land changes under them.
(Menard 1986, p. 195)2.1 Types of islands
Islands come in many shapes and sizes, and their
arrangement in space, their geology, environments,
and biotic characteristics are each extremely vari-
able, simply because there are vast numbers of
them in the world. If this makes them marvellous
experimental laboratories for field ecologists and
biogeographers, it also means that attempting to
make generalizations about islands holds the dan-
ger—almost the certainty—that one will be wrong!
Defining the term ‘island’ for the purposes of
this book is not as straightforward as it seems.
Many biogeographical studies treat isolated
patches of habitat as de factoislands, but if we take
the simplest dictionary definition then an island
is a ‘piece of land surrounded by water’. This
seems simple enough, yet some authors regard
land areas that are too small to sustain a supply of
fresh water as merely beaches or sand bars rather
than proper islands: the critical size for a supply
to be maintained being about 10 ha (Huggett
1995). At the other end of the scale, the distinction
between a continent and an island is also fuzzy, in
that larger islands assume many of the character-
istics of continents. Indeed, Australia as an entity
is considered an island continent and is rarely
treated in biogeographical analysis as being an
island (for an exception, see Wright 1983).
Where ought the line to be drawn? If New
Guinea is taken as the largest island—somewhat
arbitrarily, given that many would consider
Greenland an island (in fact it is a set of three
islands unified by an icecap)—then islands in the
sea constitute some 3% of the Earth’s land area
(Mielke 1989). Much of island biogeography—and
of this book—is concerned with islands signifi-
cantly smaller than New Guinea, and commonly
treats such substantial islands as New Guinea and
Australia merely as the ‘mainland’ source pools. To
some extent this reflects a distinction between
island evolutionary biogeography, mostly con-
cerned with larger (and typically oceanic) islands,
and island ecological biogeography, mostly con-
cerned with other types of island.
For present purposes we may divide islands
into two broad types: true islands, being land
wholly surrounded by water; and habitat islands,
being other forms of insular habitat, i.e. discrete
patches of habitat surrounded by strongly con-
trasting habitats (Table 2.1). True islands can in
turn be subdivided into island continents
(Australia), oceanic islands, continental frag-
ments, continental shelf islands, and islands in
lakes or rivers.●Oceanic islandsare those that have formed over
oceanic plates and have never been connected to
continental landmasses.
●Continental fragmentsare those islands that by
their location would pass for oceanic islands but in
terms of their origin are actually ancient fragments10CHAPTER 2