dispersal events. Examples include Pelargonium,
Wurmbea, and Pentaschistis, each of which has
Australian members that are recently derived from
Cape groups. For a further illustration, Pentaschistis
insularisis one recently derived species, sister to an
African species, which is present on Amsterdam
Island, halfway between Cape Town and Australia.
Galley and Linder (2006) therefore argue that the
formation of biogeographic disjunctions across the
Indian Ocean should be seen as ongoing and not
just a relictual pattern from break-up of the
Gondwanan supercontinent.
In biogeographical terms, New Zealand provides
one of the most contested insular systems, with
some panbiogeographers arguing that the vast bulk
of New Zealand’s biogeography is attributable to
Gondwanan break-up and rafting of the fragments
to their present-day positions. Here again, molecular
phylogenies have made telling contributions to the
story, showing that many New Zealand lineages
have diversified only relatively recently, following
the onset of late Tertiary mountain building (5–2 Ma)
and Pleistocene climatic change (Winkworth et al.
2002). Moreover, as Winkworth et al. (2002) show in
their review, molecular phylogenies are revealing
complex patterns of transoceanic dispersal across the
southern Pacific since the late Tertiary, in many cases
in opposition to present day current systems.
MetrosiderossubgenusMetrosideros, one of the most
widely distributed woody plant groups in the
Pacific, provides a powerful illustration of this.
Analyses based on nuclear ribosomal DNA indicate
that one monophyletic clade underwent a dramatic
range expansion across Polynesia (including
Hawaii) from a Pleistocene dispersal event out of
New Zealand. Wright et al. (2000) interpret this in
terms of changes in wind-flow patterns in the south-
ern hemisphere related to glacial periods.
Some have questioned whether the molecular
clocks can be relied upon, arguing that they might in
effect ‘run fast’ in explosive radiations. Bromham and
Woolfit (2004) test this proposition using 19 inde-
pendent cases of explosive radiations in island
endemic taxa, but find no evidence of a consistent
increase in rates compared to mainland relatives. The
evidence of such recent island-hopping radiations
does not rule out the existence of ancient historical
elements in the biogeography of the southern oceans,
but does conclusively demonstrate that long-distance
dispersal has played a key role in shaping the con-
temporary biogeography of the region, especially for
plants (Sanmartín and Ronquist 2004; Cook and
Crisp 2005; McGlone 2005). Thus, island-hopping on
the grand scale goes on, and has presumably done so
for a long time and by diverse vectors. Not only has
it been the means of colonization of all true oceanic
islands, but it has also been important in shaping the
biogeographical affinities of continental fragment
islands. Not only that, but it is becoming increasingly
apparent that island forms do successfully colonize
mainlands (Santos-Guerra 2001; Carine et al. 2004;
Nicholsonet al. 2005; Rocha et al. 2006), providing
explanations for at least a minority of the disjunctions
in continental biotas on opposite sides of oceans
(Galley and Linder 2006).9.5 Observations on the forcing factors of island evolution
Darwin was right in 1845 to point to oceanic islands
as being of great interest to the ‘philosophical
naturalist’. Evolutionary change on islands has
been shown in this and previous chapters to come
in many forms, from the microevolutionary alter-
ations of morphology in near-shore mice popula-
tions to the macro-scale radiation of lineages. Many
geographical, ecological, and genetic mechanisms
have been invoked in explanation of the evolution-
ary changes recorded. These mechanisms provide
the mode of delivery of change, but what can we
say about the underlying forcing factors of evolu-
tionary change on islands, such as competition,
facilitation, impoverishment, disharmony and
environmental change?
Competition has often been invoked in discus-
sions of island evolutionary change, but as detailed
in the island ecology section of this book such invo-
cations are hard to bring to a definite conclusion.
Even where micro-evolutionary changes appear
attributable to competition, it is difficult to extend
such reasoning to past evolutionary change with
any degree of certainty as the extent of past compe-
tition between two currently sympatric species is
unknowable (Connell’s (1980) ‘ghost of competition238 EMERGENT MODELS OF ISLAND EVOLUTION