although closely related and hybridizing to a
limited extent (through the activities of males of the
sexual form), are considered separate species. Their
coexistence appears to be stable in the short term at
least, as the proportions of their two populations
were similar in 1993 to a previous survey in 1986.
Hanleyet al. note that the asexuals are actually
extremely heterozygous relative to the sexual pop-
ulation. This may seem odd at first glance, but
apparently all parthenogenetic vertebrates with a
known origin have arisen from a hybridization
event, in large part accounting for their higher lev-
els of heterozygosity. In this case, far from the asex-
ual form suffering from competition with the
sexual, it is the parthenogenetic form that has the
wider distribution within the atoll (perhaps indica-
tive of hybrid vigour), with different clones having
overlapping but different habitat preferences, and
with the sexual form more or less restricted to the
lagoon beach. The coexistence of the two forms
appears to be facilitated by the specialization of the
lesssuccessful sexual taxon to beach habitats.
Hybridization
According to figures cited by Grant and Grant
(1994), speciation through hybridization may
account for as many as 40% of plant species (all
allopolyploids arise in this way), but is considered
to be much rarer in animals. However, cross-breed-
ing in animals is not so rare, and there have been
numerous studies of hybrid zones, wherein
hybridization takes place between two recognized
species. When successful, it provides a form of gene
flow between them (cf. Kaneshiro 1995, above).
Linking the themes of pollination and hybridiza-
tion, Parrish et al. (2003) report evidence of
hybridization of dioecious fig species on islands in
the Sunda Strait region of Indonesia (including the
Krakatau islands), which indicates a breakdown in
pollinator specificity on the islands (as hypothe-
sized by Janzen 1979), possibly linked to temporary
shortages of the matching fig wasp pollinator
species. Hybridization is also considered to be com-
mon within the Hawaiian flora. Indeed, even in the
species-rich genus Cyrtandra, which was once
thought not to hybridize much (if at all), recent
studies suggest it to be quite common (Sohmer and
Gustafson 1993).
Grant and Grant (1994) studied the morphologi-
cal consequences of hybridization in a group of
three interbreeding species of Darwin’s finches on
Daphne Major (Galápagos) between 1976 and 1992.
Geospiza fortisbred with G. scandensandG. fuligi-
nosa. The hybrids in turn backcrossed with one of
the recognized species. Interbreeding was always
rare, occurring at an incidence of less than 5% of the
populations, but nonetheless was sufficiently fre-
quent to provide new additive genetic variance into
the finch populations two to three orders of magni-
tude greater than that introduced by mutation.
During the course of their studies Grant and Grant
(1996b) noted a higher survival rate among hybrids
following the exceptionally severe El Niño event of
1982–83. This event led to an enduring change in
the habitat and plant composition of the island,
which appears to have opened up the food niches
exploited by the hybrids, and provided for most of
the back-crossing observed during the study. Their
work thus incidentally demonstrates that signifi-
cant fluctuations in climate can have important eco-
logical and evolutionary consequences.7.5 Peculiarities of pollination and dispersal networks on islands
We have established that islands tend to be species
poor and disharmonic, that island colonists go
through founding events and subsequent expan-
sion in numbers, and that evolutionary ecological
mechanisms subsequently include density compen-
sation, niche expansion, and niche shifts. To pro-
vide another example of density compensation or
over-compensation, densities of anoline lizards as
great as two individuals per square metre have
been measured on some Caribbean islands (Bennett
and Gorman 1979). Such high densities must, it is
reasoned, lead to a greater relative level of intraspe-
cific competition for resources than is usual. In such
circumstances, we can expect to see selection for the
exploitation of new food resources, such as nectar,
pollen, fruits, or seeds. Here we review evidence for
the distinctive ways in which island pollination and
dispersal networks have developed.PECULIARITIES OF POLLINATION AND DISPERSAL NETWORKS ON ISLANDS 179