sanguinea(the apapane) and Vestiaria coccinea(the
i’iwi), two of the more widespread taxa, show an
intriguing lack of differentiation, and it has been sug-
gested that this may reflect relatively recent range
expansion. This may have occurred in response to
the arrival and spread of the tree Metrosideros poly-
morpha, or possibly because of the loss of competitors
due to the extinction of other avian taxa. If these
interpretations are correct, these two species may
represent the early stage of a new cycle of dispersal
and differentiation (Tarr and Fleischer 1995).
Having set out the basics of the Hawaiian pic-
ture, we can pick up the thread of the discussion of
the interactions and exchanges between sympatric
populations by reference to comparative studies
between the Hawaiian and Galápagos lineages.
Studies of morphological variation within popula-
tions of Darwin’s finches have revealed that there is
considerable variation in body size and beak traits
within populations of ground finches (Geospiza),
and that the amount of variation itself differs
among populations (Grant 1994). In studies of six of
these species, Grant and Grant (1994) found
evidence for the limited occurrence of introgressive
hybridization, allowing gene flow between popula-
tions of different species and contributing to the
intraspecific variation observed.
Grant (1994) set out to establish from the study of
524 museum specimens of the 7 species of Hawaiian
honeycreepers with finch-like bills (which might be
termed honeycreeper-finches) whether the same
hybridization process was at work in a lineage that
has diversified to a much greater extent. The meas-
urements demonstrated that variation within pop-
ulations of Hawaiian honeycreeper-finches was less
than among the Geospiza. The single Hawaiian
species with both sympatric and allopatric popula-
tions did not show greater variation in the
sympatric population, as would have been expe-
cted if hybridization was effective. It was con-
cluded that there was no evidence of hybridization
occurring within the past 100 years. Grant reported
that G. C. Munro had suggested in 1944 that
Rhodacanthis flavicepsmight be a hybrid produced
from R. palmeriandR. psittacea, and indeed the only
morphological hint of hybridization among the
Hawaiian species came from the close similarity
found between R. flaviceps and R. palmeri. If
R. flavicepsis indeed of hybrid origin, then hybridiza-
tion must have been going on for a long time, as the
species is recognizable in fossil material.
Grant (1994) put forward two hypotheses which
might explain why the hybridization evident in the
Galápagos finches was not apparent in the Hawaiian
equivalents.●Geological and electrophoretic studies indicate
that the honeycreeper-finches have been present in
the Hawaiian archipelago for a much longer
period, perhaps three times longer than have the
finches of the Galápagos, and over this time have
diversified further and have evolved prezygotic
(behavioural) and/or postzygotic isolating mecha-
nisms (Table 8.1).
●The Hawaiian honeycreeper-finches have evolved
greater dietary specializations in the generally less
seasonal and floristically richer Hawaiian islands. In
an environment with more distinctive resource
peaks, and following the line of reasoning outlined
in the section on competitive speciation, stabilizing
selection for specialist feeding may have provided
strong selective pressure against an ecological niche
intermediate between the two parental species.Grant put forward a model to capture the impor-
tant distinctions in a hypothetical phylogeny for the
radiation of sympatric taxa (Fig. 9.7). An early phase
of divergence is characterized by occasional genetic
contact through interbreeding, followed by a genet-
ically independent phase. The duration of the phase
of introgression will depend on the ecological isola-
tion attained, which will be related to environmen-
tal heterogeneity. From genetic distance measures of
finch species known to hybridize from the
Galápagos and from North America, Grant (1994)
estimates the apparent duration of the phase of
introgression, i.e. of hybridization, as up to 5 million
years. He speculates that it may have been occur-
ring over the whole of the estimated 2.8 million
years of the radiation of Darwin’s finches and quite
possibly for much of the 7.5 million years estimated
by some (but see Tarr and Fleischer 1995) for the
diversification of the Hawaiian honeycreepers.
All such figures involve error margins and in the
case of the Hawaiian species the evidence for such
a model is very largely indirect. Hybridization has224 EMERGENT MODELS OF ISLAND EVOLUTION