11. Patterns of Tree Species Diversity
165
few instances where immigration and extinction rates have achieved their equilibrium.
Our data are beginning to show a surprising level of congruence at global level
(Table 2). Forests in similar habitats in the aseasonal tropics of the Ecuadorian
Amazon and Northwest Borneo, both areas which appear to have escaped major
climatic change during the Pleistocene, appear to have comparable richness, while
forests in Panama, Zaire, and Thailand show a gradient of species richness in relation to rainfall seasonality, irrespective of their apparent relative insularity. These
early results imply that forests reach an equilibrium richness according to their
habitat, irrespective of their island biogeography. The Sinharaja forest in Sri Lanka
is a major exception and also forests in eastern Borneo which, by comparison with
those in northwest Borneo in similar habitats, have markedly fewer species (Ashton
1984). These two areas of aseasonal rain forest, so different in extent, nevertheless
share a history of climatic change during the last glaciation. Sri Lankan and East
Bornean forests may therefore still have been in a phase of species accumulation
prior to human influence.
Very small islands nevertheless, as Hubbell (1995) has shown, have substantially reduced richness and also steeper dominance-diversity curves than in adjacent mainland forests, even where they have only become isolated during the Holocene. Moreover, and consistent with the theory of island biogeography, the dominant species are unpredictable.
We infer therefore that island biogeography has little influence on tree species
diversity at community scale in areas greater than a certain minimum which is at
present unknown. This may in part be due to the very long period needed for tree
species richness to reach equilibrium in forests of substantial area, but also because
richness there eventually reaches an asymptote in relation to specific physical conditions.
2.2 Temporal Variability in Fecundity
Established juveniles have a competitive advantage over germinants in all plant
communities (see, e.g., Brown and Whitmore 1992 for tropical forest). Coexistence
of ecologically equivalent species will therefore be prornoted if species respond to
different flowering cues, including endogenous cues, such that opportunities for
successful establishment occur in different years for different species (Chesson and
Warner 1981). This effect will be enhanced if seed dispersal is restricted, as in the
case of most rain forest species. In this case, our data run contrary to theory. In
south-east Asia, high tree species richness is closely associated with aseasonality of
rainfall and drought. Climate changes from aseasonal to seasonal with 2-3 dry
months, but without change in mean annual rainfall, across a narrow boundary, the
Kangar-Pattani Line, which crosses the Thai-Malaysian frontier (Whitmore 1984).
Though there is little change in generic composition, species richness is approximately halved in the seasonal climate (Table 2). Counter to expectations, canopy
dipterocarps mass flower and synchronously mast fruit at intervals of 4-7 years
165
few instances where immigration and extinction rates have achieved their equilibrium.
Our data are beginning to show a surprising level of congruence at global level
(Table 2). Forests in similar habitats in the aseasonal tropics of the Ecuadorian
Amazon and Northwest Borneo, both areas which appear to have escaped major
climatic change during the Pleistocene, appear to have comparable richness, while
forests in Panama, Zaire, and Thailand show a gradient of species richness in relation to rainfall seasonality, irrespective of their apparent relative insularity. These
early results imply that forests reach an equilibrium richness according to their
habitat, irrespective of their island biogeography. The Sinharaja forest in Sri Lanka
is a major exception and also forests in eastern Borneo which, by comparison with
those in northwest Borneo in similar habitats, have markedly fewer species (Ashton
1984). These two areas of aseasonal rain forest, so different in extent, nevertheless
share a history of climatic change during the last glaciation. Sri Lankan and East
Bornean forests may therefore still have been in a phase of species accumulation
prior to human influence.
Very small islands nevertheless, as Hubbell (1995) has shown, have substantially reduced richness and also steeper dominance-diversity curves than in adjacent mainland forests, even where they have only become isolated during the Holocene. Moreover, and consistent with the theory of island biogeography, the dominant species are unpredictable.
We infer therefore that island biogeography has little influence on tree species
diversity at community scale in areas greater than a certain minimum which is at
present unknown. This may in part be due to the very long period needed for tree
species richness to reach equilibrium in forests of substantial area, but also because
richness there eventually reaches an asymptote in relation to specific physical conditions.
2.2 Temporal Variability in Fecundity
Established juveniles have a competitive advantage over germinants in all plant
communities (see, e.g., Brown and Whitmore 1992 for tropical forest). Coexistence
of ecologically equivalent species will therefore be prornoted if species respond to
different flowering cues, including endogenous cues, such that opportunities for
successful establishment occur in different years for different species (Chesson and
Warner 1981). This effect will be enhanced if seed dispersal is restricted, as in the
case of most rain forest species. In this case, our data run contrary to theory. In
south-east Asia, high tree species richness is closely associated with aseasonality of
rainfall and drought. Climate changes from aseasonal to seasonal with 2-3 dry
months, but without change in mean annual rainfall, across a narrow boundary, the
Kangar-Pattani Line, which crosses the Thai-Malaysian frontier (Whitmore 1984).
Though there is little change in generic composition, species richness is approximately halved in the seasonal climate (Table 2). Counter to expectations, canopy
dipterocarps mass flower and synchronously mast fruit at intervals of 4-7 years