distributions arise (see Box 2.2). In essence, dispersal
hypotheses have it that descendent forms are the
product of chance, long-distance dispersal across a
pre-existing barrier, whereas vicariance biogeogra-
phers envisage species ranges being split up by phy-
sical barriers, often followed by speciation in the now
separate populations (Myers and Giller 1988). Some
vicariance biogeographers seem remarkably dismis-
sive of long-distance dispersal (e.g. Heads 2004),
given the evident importance of the process in island
colonization (Wagner and Funk 1995, McGlone et al.
2001; Winkworth et al. 2002; Cook and Crisp 2005,
McGlone 2005; Rocha et al. 2006) and indeed in post-
separation exchanges between continental land-
masses (e.g. Galley and Linder 2006). Although there
are contexts in which the relative contribution of both
processes can be exceedingly difficult to discern, pro-
viding plenty of room for genuine debate, the polar-
ization of debate (and especially the dismissal of
long-distance dispersal) is unnecessary and has
become a real hindrance to progress. Available data
concerning island origins and environmental histo-
ries illustrate that both types of process operate, and
both contribute to the complexities of biogeographi-
cal patterning (Stace 1989; Keast and Miller 1996;
Winkworth et al. 2002; Galley and Linder 2006).
This point can be exemplified with reference to
the geological history of the Caribbean, introduced
in Chapter 2. Buskirk (1985) set out to employ this
framework in interpreting the phylogenetic rela-
tionships of modern beetles, reptiles, amphibians,
and other terrestrial animals in the Greater Antilles.
The evidence indicates that the Antilles have been
islands throughout most of the Cenozoic and that
therefore most of their faunas were derived from
over-water colonists and not from the fragmenta-
tion of original populations on a large land mass.
However, Buskirk suggests that tectonic events
have affected dispersal and colonization distances.
The shear along the northern Caribbean plate
boundary has moved Jamaica further away from
Central America, and thus modern Jamaican species
that are relicts or radiations from relatively early
colonists have more Central American affinities
than do the endemic species of neighbouring
Antillean islands. Jamaica was largely submerged in
the early to mid-Tertiary, and its modern endemic
fauna lacks the groups that invaded the other
Antilles at about that time. As well as the change in
sea level, the accompanying climatic changes have
been important, and in Jamaica have apparently
selected against members of some animal groups
that required more mesic habitats (Buskirk 1985).
Thus, Buskirk shows that to explain the uniqueness
of Jamaica’s fauna, it is necessary to refer to tectonic
events (horizontal movement and extensive late
Tertiary uplift), combined with Pleistocene cycles of
climate and sea-level change. Both dispersal and
vicariance processes are involved in this history.
Over the past few million years, vertical changes
of land and sea have been of more general signifi-
cance to Caribbean biogeography than horizontal
land movements. An example is provided by
Williams’ (1972) studies of anoline lizards. Anoles
are small green or brown lizards that are more or
less the only diurnal arboreal lizards found in most
of the region. Much of the distributional pattern
within this group is related to the submarine banks
on which today’s islands stand. Each bank has many
islands on it, but as recently as about 7000 BP as a
generalization, many of these islands were joined
together, as they would have been for lengthy peri-
ods during the Pleistocene. The present-day distri-
butions of anoline lizards thus owe much to events
during the low sea-level stands (Williams 1972).
The unravelling of past species movements in
order to determine how and when a species
reached its present insular distribution presents
considerable challenges. Much can depend on
assumptions concerning the powers of long-
distance dispersal of particular species. Many ques-
tions in island biogeography come back to the issue
of dispersal, as it is inherently extremely difficult to
determine the effective long-distance range of
propagules (‘potentially viable units’) (but see
Hugheset al. 1994; Whittaker et al. 1997). We will
return to this theme again; suffice to note for the
moment that although dispersal limitations may
provide parsimonious interpretations of data such
as the anoline lizard distributions, it is important to
remain open to new data and insights into disper-
sal powers of particular species. This point was
driven home by Johnson’s (1980) documentation,
complete with photographs, of the ability ofDISHARMONY, FILTERS, AND REGIONAL BIOGEOGRAPHY 59