Tropical Forest Community Ecology

40 Paul V.A. Fineetal.

y = 1.745x −1.9919

R^2 = 0.4754

0

1

2

3

4

5

2.0 2.4 2.8 3.2 3.6

Log minimum biome area (km^2 )

Log current tree diversity

P<0.0189

Figure 3.5 Log current tree diversity

plott ed with th eminimum siz eof each

biom eduring th ePl eistoc en e. Th esmall est

areas for each of the 11 biomes in the two

Pleistocene maps (21,000 and

6000 years ago) were log-transformed and

plotted with log current tree species richness.

and extinction rates, this conclusion is unsatisfy-
ing because it ignores the possibility that current
biome diversities are not at equilibrium – that
is, at levels attained if current areas were stably
maintained for tens of millions of years.
By incorporating th ehistory of biom esiz e
into th eanalysis, w efound significant corr ela-
tions between the integral of area over time and
current tree diversity, whether evaluated from
the Miocene, Oligocene, or Eocene. Fine and Ree
(2006) reported similar results when they cor-
related the same tree species diversity estimates
with area–time composites based on five sepa-
rate interpretations of past climates based on
five sources. Together, these results suggest that
th ecombin ed siz eand long evity of a biom ear e
important factors in explaining its current species
richness. Examined alone, however, neither fac-
tor produced a significant trend. The idea that
tim eis important is hardly n ew: Wallac e(1876),
Willis (1922), and others have claimed that the
extraordinarily high diversity of tropical rain-
forests is due to the stability or greater age of
these forests. Many have also noted that most plant
lineages appear to have originated in the tropics
(Cran eand Lidgard 1990, Latham and Rickl efs
1993, Juddet al. 1994, Ricklefs 1999, 2004,
Wiens and Donoghue 2004). Other large-scale
studies have noted that area and species diversity
are positively related. Tiffney and Niklas (1990)
found that fossil plant species richness correlated
with the overall land area of the northern hemi-
sphere at 12 slices in time between 410 Ma and
10 Ma, and a recent study of palynological data by
Jaramilloet al. (2006) found that neotropical tree

diversity peaked in the Eocene when we propose

that tropical forests covered the largest amount

of area.

Extinction

It is difficult to ass ess th er elativ eimportanc eof

speciation versus extinction as underlying causes

of the result that area size integrated over time cor-

relates with extant tree diversity (Figure 3.4). The

significant correlation between minimum biome

area during the Pleistocene and extant tree diver-

sity is suggestive that extinction caused by glacial

cooling and drying is on eimportant factor in

explaining extant tree diversity patterns. How-

ever, another important consideration might be

th eamount of latitudinal shift in each biom edur-

ing glacial periods. For example, the tropics also

decreased in area during the Pleistocene, but trop-

ical refugia remained within earlier tropical biome

borders (Bush 1994). In contrast, temperate and

boreal refugia were located closer to the equa-

tor during glacial advances, and largely outside

of their pre-glacial period borders (Figure 3.1).

If trees disperse more slowly than their moving

refugia (cf. McLachlanet al. 2005), temperate

tree species extinction may have been higher than

that attributable solely to the overall reduction in

biom ear ea.

The tropical conservatism hypothesis

Th eGAAH that w et est h er eis in som eways

similar to the tropical conservatism hypothesis