random genetic changes, drift, and selection (e.g. for
different feeding niches). A degree of differentiation
would then have been evident between the separate
populations. At some point, individuals of one of
the derived populations flew to an island already
occupied by a slightly differentiated population.
The result of this might sometimes have been an
infusion of the newcomers into the established
population, but where the two populations
exhibited a greater degree of behavioural separa-
tion, hybridization would have been limited.
Selection would have favoured members of the two
groups that fed in different ways from each other
and so did not compete too severely for the same
resources.
Recent research has shown that female finches
appear able to distinguish between acceptable and
unacceptable mates on the basis of beak size and
shape and also on the different patterns of songs
(Grant 1984). This may have been the means by
which females were able to select the ‘right’ mates,
thus breeding true, and producing progeny that
corresponded with the ‘peaks’ rather than the ‘val-
leys’ of the resource curves. Such characteristics
would have a selective advantage in the popula-
tions over time, thus allowing two or more species
to exist in sympatry.
To put these points together: the radiation of the
lineage has taken place in the context of a remote
archipelago, presenting extensive ‘empty niche’
space, in which the considerable (but not excessive)
distances between the islands has led to phases of
interisland exchange only occasionally. Differing
environments have apparently selected for differ-
ent feeding niches both between and within islands.
Thereafter, behavioural differences between forms
maintain sufficient genetic distance between sym-
patric populations to enable their persistence as (to
varying degrees) distinctive lineages.
As Grant (1984) cautioned, this is not the only
way in which speciation can occur. Theoretically a
lineage may have split into two non-interbreeding
populations on a single island, as suggested by the
model of competitive (sympatric) speciation
(Chapter 8), provided that the environment pre-
sented two or more sufficiently distinct resource
peaks (i.e. was sufficiently heterogeneous). Further
progress has been made in understanding the
interactions and exchanges that may occur between
sympatric populations of Galápagos finches, high-
lighting the importance of sympatric episodes in
lineage development (Grant and Grant 1996b), but
before discussing this work we will outline the case
of the Hawaiian equivalents to Darwin’s finches.
The Hawaiian islands have formed as a narrow
chain from a hotspot, which appears to have been
operational for over 70 million years, although the
oldest high island of the present group, Kauai, is
only about 5.1 million years old (Table 2.3). It is
believed that the hotspot has never been closer to
North America than it is today and its position
relative to Asia has probably also been stable. Mole-
cular clock data suggest that few lineages exceed
10 Ma, i.e. the pre-Kauai signal in the present biota
of the main islands is actually rather limited
(Wagner and Funk 1995; Keast and Miller 1996).
At least 20 natural avian colonizations have been
suggested (Tarr and Fleischer 1995), and endemics
comprise approximately 81% of the native avi-
fauna. There have been other avifaunal radiations
on Hawaii. They include the elepaio (Chasiempis
sandwichensis), a small active flycatcher endemic to
Hawaii. Distinctive subspecies occur on Kauai and
Oahu, and a further three occur within the
youngest and largest island, Hawaii itself (Pratt
et al. 1987). The native thrushes of Hawaii are
placed in the same genus (Myadestes) as the soli-
taires of North and South America. Five species are
recognized by Pratt et al. (1987), each occurring on
its own island or island group. One was last seen in
1820, and three of the remaining species are
severely endangered, with populations—if they
survive—numbering fewer than 50 individuals
(Ralph and Fancy 1994).
The Hawaiian honeycreepers (or honeycreeper-
finches) have shown an even greater radiation
than Darwin’s finches. They are a monophyletic
endemic group perhaps best considered a subfam-
ily, the Drepanidinae (see Pratt 2005). For a long
time it was believed that this radiation had resulted
in 23 species in 11 genera. From a single type of
ancestral seed-eating finch, the group had radiated
to fill niches of seed-, insect-, and nectar-feeding
species, with a great variety of specialized beaks222 EMERGENT MODELS OF ISLAND EVOLUTION