166
P.S. Ashton and J.V. LaFrankie
(Ashton 1989) in the aseasonal south, but annually and with different species participating each year in the less rich seasonal regions to the north (Appanah 1985,
Ashton et al. 1988). Mass flowering has recently been shown to be a communitywide phenomenon by two Kyoto graduate students, Kuniyasu Momose and Shoko
Sakai, at the Lambir CfFS site (Sakai et al. 1997).
There appears to be little difference in means of dispersal between forest flora of
aseasonal and seasonal Asia. We conclude that the influence of temporal differentiation on species accumulation is either minor, or is overcome by a more important
influence. For example, mast fruiting satiates predators and reduces seed mortality
among dipterocarps (Janzen 1974, demonstrated by Curran 1994), though comparisons of seed and seedling predation north and south of the Kangar-Pattani line
have yet to be made. Lower seed predation might contribute to higher species richness.
2.3 Density-Dependent Mortality
In most studies, more than half of all species in rain forest communities have proven
not to be habitat specific and, after all factors are taken into account, it appears that
there is a high level of spatial overlap if not ecological complementarity among cooccurring species. A persisting mystery of the tropical forest therefore is why most
tree species maintain high outcrossing rates and genetic variability, which must be
sustained at great cost in the prevailing low density populations and low nutrient
availability (although these costs have yet to be critically documented). If endogenous reasons, such as the need to sustain crossing-over as a means to repair chromosomes are insufficient reason, then the sustainment of high genetic variability
implies that some form of selection is operating and that the conditions necessary
for survival are changing over evolutionary time. These changes are likely to be
biotic, particularly in the uniquely ancient equable climates of the hyper-diverse
forest of the putative refugia. The remaining, and still unanswered possibility is
that the many apparently ecologically complementary tree species populations in
these habitats may each be kept at reduced levels by species-specific interactions
with motile organisms which may limit their fecundity or increase their mortality.
Janzen (1970) and Connell (1971) suggested that seed and seedling predators
may cause density dependent mortality owing to the greater ease with which they
may discover prey which is close to the parent, or to other conspecifics of similar
size to themselves. Evidence for species-specific predation has proven elusive but
there is growing, albeit inconclusive evidence of density-dependent, species-specific pathogenicity (Gilbert et al. 1994). The strongest support has come from studies of individual species' scedling popUlations (Augsberger 1983, 1992; Clark and
Clark 1984; Howe et al. 1985; Webb and Peart, in press).
Attempts to test for density-dependence among all species in community samples
have given mixed results (Hubbell 1980, 1997a; Hubbell and Foster 1990; Wills et
al. 1997; Wills and Condit, in press; Webb and Peart, in press). Hubbell (1997a)
argued that, though present, density dependent mortality is too weak to explain the
P.S. Ashton and J.V. LaFrankie
(Ashton 1989) in the aseasonal south, but annually and with different species participating each year in the less rich seasonal regions to the north (Appanah 1985,
Ashton et al. 1988). Mass flowering has recently been shown to be a communitywide phenomenon by two Kyoto graduate students, Kuniyasu Momose and Shoko
Sakai, at the Lambir CfFS site (Sakai et al. 1997).
There appears to be little difference in means of dispersal between forest flora of
aseasonal and seasonal Asia. We conclude that the influence of temporal differentiation on species accumulation is either minor, or is overcome by a more important
influence. For example, mast fruiting satiates predators and reduces seed mortality
among dipterocarps (Janzen 1974, demonstrated by Curran 1994), though comparisons of seed and seedling predation north and south of the Kangar-Pattani line
have yet to be made. Lower seed predation might contribute to higher species richness.
2.3 Density-Dependent Mortality
In most studies, more than half of all species in rain forest communities have proven
not to be habitat specific and, after all factors are taken into account, it appears that
there is a high level of spatial overlap if not ecological complementarity among cooccurring species. A persisting mystery of the tropical forest therefore is why most
tree species maintain high outcrossing rates and genetic variability, which must be
sustained at great cost in the prevailing low density populations and low nutrient
availability (although these costs have yet to be critically documented). If endogenous reasons, such as the need to sustain crossing-over as a means to repair chromosomes are insufficient reason, then the sustainment of high genetic variability
implies that some form of selection is operating and that the conditions necessary
for survival are changing over evolutionary time. These changes are likely to be
biotic, particularly in the uniquely ancient equable climates of the hyper-diverse
forest of the putative refugia. The remaining, and still unanswered possibility is
that the many apparently ecologically complementary tree species populations in
these habitats may each be kept at reduced levels by species-specific interactions
with motile organisms which may limit their fecundity or increase their mortality.
Janzen (1970) and Connell (1971) suggested that seed and seedling predators
may cause density dependent mortality owing to the greater ease with which they
may discover prey which is close to the parent, or to other conspecifics of similar
size to themselves. Evidence for species-specific predation has proven elusive but
there is growing, albeit inconclusive evidence of density-dependent, species-specific pathogenicity (Gilbert et al. 1994). The strongest support has come from studies of individual species' scedling popUlations (Augsberger 1983, 1992; Clark and
Clark 1984; Howe et al. 1985; Webb and Peart, in press).
Attempts to test for density-dependence among all species in community samples
have given mixed results (Hubbell 1980, 1997a; Hubbell and Foster 1990; Wills et
al. 1997; Wills and Condit, in press; Webb and Peart, in press). Hubbell (1997a)
argued that, though present, density dependent mortality is too weak to explain the