model. Multiple colonization, as exemplified by the
yellow warbler, occurred in a few other cases, but
overall the results provide remarkably good sup-
port for the original model as a fair representation
of the temporal sequence followed by many
species. Some 20 lineages colonized the Lesser
Antilles between 7.5 and 10 Ma, with individual
island populations typically persisting for about
4 million years, and with single island endemics
typically amongst the oldest species, as required by
the model. Some species have undergone phases of
secondary expansion. Phases of expansion and con-
traction do not appear to be correlated across
species and contrary to Pregill and Olson (1981)
many do not correlate with glacial climate transi-
tions. Indeed, many of the cycles predate the most
intense glaciation episodes of the last 2 million
years. However, some 17 lineages appear to have
colonized the Lesser Antilles between about 0.75
and 0.55 Ma, which suggests that although taxon
cycles occur independently of climatic fluctuations,
the onset of the most extreme Pleistocene climatic
fluctuations may indeed have influenced the com-
position of the Lesser Antillean avifauna.
In conclusion, Ricklefs and Bermingham (2002)
find evidence for expansion phases occurring at
intervals on the order of 10^6 years, and that these
episodes are not synchronous across independent
lineages. They interpret these data as supporting
the idea that taxon cycles are related to changes in
specific aspects of population–environment rela-
tionships, such as host-parasite interactions: as
demanded by the ‘counter-adaptation’ argument of
Ricklefs and Cox (1972). Future analyses of the phy-
logenies of the parasites may allow a more rigorous
test of this mechanism.
Caribbean anoles
The anoline lizards of the Caribbean comprise about
300 species, or between 5 and 10% of the world’s
lizards (Roughgarden and Pacala 1989). On
Caribbean islands they are extremely abundant,
replacing the ground-feeding insectivorous birds of
continental habitats. The distribution of species/sub-
species relates best not to the present-day islands but
to the banks on which the islands stand, separated
from other banks by deep water. The anoles have
long provided one of the classic illustrations of
character displacement, whereby island banks have
either one intermediate-sized species or two species
of smaller and larger size respectively. This has been
taken to indicate that where two populations have
established, interspecific competition between them
has resulted in divergence of size of each away from
the intermediate size range so that they occupy
distinct niche space (Brown and Wilson 1956). This
progresses to the point at which the benefit of further
reduction in competition is balanced by the
disadvantage of shifting further from the centre of
the resource distribution (Schoener and Gorman
1968; Williams 1972). However, Roughgarden and
Pacala (1989) contend that within the islands of the
Antigua, St Kitts, and Anguilla banks in the northern
Lesser Antilles, a series of independently derived
facts contradict this model, and instead support a
form of taxon cycle (see also Rummel and
Roughgarden 1985).
The anolid taxon cycle of the northern Lesser
Antilles begins with an island occupied by a
medium-sized species, which is joined by a larger
invader from the Guadeloupe archipelago. Rather
than the two species moving away from the middle
ground, as in the character displacement scenario,
both species evolve a smaller size. The medium-
sized species becomes smaller as it is displaced
from the middle ground by its larger competitor.
This opens up niche space in the centre of the
resource axis, selecting for smaller size in the
invader, which thus approaches the medium
(presumably optimal) size originally occupied by
the first species. The range of the original resident
species therefore contracts, culminating in its
extinction. At this point, the invader has taken
up occupancy of the niche space of the first species
at the outset of the cycle. The islands remain in the
solitary-lizard state until (hypothetically) another
large invader arrives.
The key points in the Roughgarden and Pacala
(1989) analysis are as follows. Intensive scrutiny of
the systematics of anoline lizards over a period of
four decades resulted in a phylogenetic tree com-
prising five main groups. The northern Lesser
Antilles are populated by the bimaculatusgroup,THE TAXON CYCLE 215