CENOZOIC BIODIVERSITY 163eccentricity of orbit, have periods of 41 ka and
100 ka, respectively; they are complemented by
the annual timing of the minimum Earth-Sun
distance which varies with a 21 ka period. These
orbital oscillations lead to variations in insola-
tion and thus ambient temperatures. Such tem-
perature variations would have been greater
towards the poles, but led to eustatic sea-level
changes that were global in effect (Bennett 1997;
Cronin & Raymo 1997).
In a series of previous papers, Rosen (1981,
1984,1988) has already suggested how an inten-
sification of glacioeustatic cyclicity may have
promoted diversification within the IWP focus.
In essence, a type of species diversity pump was
envisaged whereby taxa created in the outer
islands of the western Pacific and eastern Indian
oceans during sea level lowstands were sub-
sequently 'pumped' into the central Indonesian
focus during highstands. In time, the latter
region became a form of refugium for numerous
sympatric species. A similar mechanism of
species production by repeated cycles of allo-
patry and secondary sympatry over at least a
10 Ma period has recently been suggested for the
temperate zone floras of eastern Asia (Qian &
Ricklefs 2000). Such floras have twice as many
species as their counterparts in eastern North
America, and it is thought that this is the product
of repeated climatic and sea-level changes over
a geographically much more heterogeneous
terrain (Qian & Ricklefs 2000).
We are also beginning to appreciate more
fully the effect of Milankovitch oscillations on
species ranges; such orbitally forced range
dynamics (ORD) may well be the basis of a
range of macroecological phenomena (Gaston
& Blackburn 2000). In a wide-ranging review,
Dynesius & Jansson (2000) demonstrated that
ORD tended to be larger and more pronounced
in high- than low-latitude regions. Although this
might be taken as a strong indication that the
formation of ecological isolates, a necessary first
step in the process of allopatric speciation,
would be more likely in the high latitudes, these
authors have argued that this is not in fact the
case. Instead, they contend that the short, stable
periods between high-latitude oscillations were
not long enough for the process of gradual
speciation to be completed; isolates were either
brought back together again or they became
extinct. Such a process is essentially reversed in
the tropics, where much smaller ORD promotes
the formation of isolates and drives the process
of allopatric speciation (Dynesius & Jansson
2000).
In an alternative model, Chown & Gaston
(2000) have pointed out that many tropical taxa
are stenotopic and can, in theory, range large
distances within a more or less constant temper-
ature belt defined by 25°N and 25°S. High-
latitude taxa can achieve large ranges too, but
these tend to be eurytopes buffered to withstand
a considerable amount of climate change. Thus
it is the tropical taxa that may well be the most
vulnerable to temperature change; repeated
temperature oscillations through the late
Neogene could have caused far more range
disjunctions in the low- than high-latitude
regions (Chown & Gaston 2000).
Although the precise mechanisms have yet to
be substantiated, it is likely that ORD played a
key role in the Neogene diversification event
(Bennett 1997). It should also be emphasized
that their effects were not necessarily confined
to just tropical regions. Some high-latitude and
polar clades obviously radiated through the
Neogene too (e.g. Crame & Clarke 1997), where
both sea-level and temperature oscillations
would have been magnified in certain regions. It
is perhaps the superimposition of ORD on other
factors, such as greater habitat heterogeneity
and productivity, that makes their presence most
keenly felt in the tropics.
In formulating his diversity pump model,
Rosen (1984) supposed that much the same sort
of process had operated in the ACEP focus as
the IWP one, though on a somewhat smaller
scale. From a comprehensive stratigraphical
study of Neogene molluscs in the Caribbean
region, we know that diversity did indeed rise
steadily from approximately 14 to 5 Ma BP, but
then levelled off substantially (Jackson &
Johnson 2000). As there is good evidence to
show that over the last 5 Ma extinction rates
were higher in this region than anywhere else in
the tropics, there must also have been an
equivalent burst of originations (Allmon et al
1996; Jackson et al 1999). For Caribbean corals,
we know that diversification increased from
approximately 16 to 4 Ma BP, but then from 4 to
1 Ma BP a peak of originations preceded a peak
of extinctions (Budd et al 1996; Allmon 2001).
Between 9 and 1 Ma BP there were >120 coral
species in the Caribbean, as compared with 62 at
the present day. Patterns of faunal change in the
Western Atlantic Neogene are obviously
complex and so too must be the environmental
changes that underpin them. Nevertheless it is
likely that origination and extinction patterns in
this region were strongly affected by the pro-
gressive closure of the Central American
Seaway and rise of the CAI (see above).
Although deep-water circulation through the
seaway was blocked at 3.6 Ma BP, and shallow-
water circulation at 3.0 Ma BP, the slowly rising