172
P.S. Ashton and J.V. LaFrankie
the most wind prone sites, where the diffuse canopy may shelter a richer subcanopy
flora than beneath the denser canopy on lee slopes.
Intermediate disturbance regimes are associated with high species diversity in
reef communities in part because of their heterogeneous light climate at the reef
floor. Rain forest species richness and community composition, and also stand structure and perhaps gap size and frequency, may covary with soil and climate. The
depression of species richness on mesic sites in forests of the calm, aseasonal tropics was already seen in part to be related to increasing canopy density (Ashton
1998b). The most impoverished rain forests here, the Heath forests, occur on podsols.
They have low stature and a simple structure with even, albeit diffuse, canopy and
small gap sizes (Brunig 1974; Gale 1997). The diffuse canopy structure is due to
the small, generally upturned leaves. Forests with the greatest species richness have
tall stature and a similarly diffuse albeit uneven canopy which may be sustained by
patchy, often single tree mortality caused by lightning and occasional catastrophic
drought (see e.g., Leighton and Wirawan 1986; also Gale 1997). The principles of
the Connell hypothesis, therefore, obtain in forests of the aseasonal tropics, but
canopy density, hence light intensity and heterogeneity, may here be determined as
much by the influence of water stress and nutrients on leaf and twig growth, hence
size, shape and density, as by disturbance.
Along the gradient of species richness which occurs from the aseasonal wet to
the strongly seasonal dry tropics, the frequency and intensity of drought mediates
the level of catastrophic mortality, and influence species' survival (Ashton 1993).
Forest stature declines and the mature phase canopy becomes more even with increased rainfall seasonality (Richards 1996), but there is no obvious change in the
frequency of canopy disturbance or the density of canopy trees, unless fire intervenes. The major gradient in mortality patterns, which could influence species richness, is in the understory; this accounts for declines in trees <10 cm dbh from
>7500 to <100/ha (Table 1). Periodic intense drought following seedling establishment, and also fire and browsing, may each be influential. Their relative importance is as yet not known, and is a major objective of our research at Khao Chong
with two, and Huai Kha Khaeng with five dry months, directed by Sarayudh
Bunyavejchewin, Royal Thai Forest Department in collaboration with Toshinori
Okuda, Japanese National Institute of Environmental Studies.
3 Conclusion
Our results already present important implications for conservation planning. Differences in rates of accumulation of species in rainforest tree communities do occur, but may reach a constant level beyond a quite small area. This implies that
attrition of tree populations following forest fragmentation may be very slow and in
large strict conservation areas may not occur at all over realistic time periods.
Variation in species richness correlated with climatically induced variation in
phenology and fecundity occurs at a regional scale, but the relationship is contrary
P.S. Ashton and J.V. LaFrankie
the most wind prone sites, where the diffuse canopy may shelter a richer subcanopy
flora than beneath the denser canopy on lee slopes.
Intermediate disturbance regimes are associated with high species diversity in
reef communities in part because of their heterogeneous light climate at the reef
floor. Rain forest species richness and community composition, and also stand structure and perhaps gap size and frequency, may covary with soil and climate. The
depression of species richness on mesic sites in forests of the calm, aseasonal tropics was already seen in part to be related to increasing canopy density (Ashton
1998b). The most impoverished rain forests here, the Heath forests, occur on podsols.
They have low stature and a simple structure with even, albeit diffuse, canopy and
small gap sizes (Brunig 1974; Gale 1997). The diffuse canopy structure is due to
the small, generally upturned leaves. Forests with the greatest species richness have
tall stature and a similarly diffuse albeit uneven canopy which may be sustained by
patchy, often single tree mortality caused by lightning and occasional catastrophic
drought (see e.g., Leighton and Wirawan 1986; also Gale 1997). The principles of
the Connell hypothesis, therefore, obtain in forests of the aseasonal tropics, but
canopy density, hence light intensity and heterogeneity, may here be determined as
much by the influence of water stress and nutrients on leaf and twig growth, hence
size, shape and density, as by disturbance.
Along the gradient of species richness which occurs from the aseasonal wet to
the strongly seasonal dry tropics, the frequency and intensity of drought mediates
the level of catastrophic mortality, and influence species' survival (Ashton 1993).
Forest stature declines and the mature phase canopy becomes more even with increased rainfall seasonality (Richards 1996), but there is no obvious change in the
frequency of canopy disturbance or the density of canopy trees, unless fire intervenes. The major gradient in mortality patterns, which could influence species richness, is in the understory; this accounts for declines in trees <10 cm dbh from
>7500 to <100/ha (Table 1). Periodic intense drought following seedling establishment, and also fire and browsing, may each be influential. Their relative importance is as yet not known, and is a major objective of our research at Khao Chong
with two, and Huai Kha Khaeng with five dry months, directed by Sarayudh
Bunyavejchewin, Royal Thai Forest Department in collaboration with Toshinori
Okuda, Japanese National Institute of Environmental Studies.
3 Conclusion
Our results already present important implications for conservation planning. Differences in rates of accumulation of species in rainforest tree communities do occur, but may reach a constant level beyond a quite small area. This implies that
attrition of tree populations following forest fragmentation may be very slow and in
large strict conservation areas may not occur at all over realistic time periods.
Variation in species richness correlated with climatically induced variation in
phenology and fecundity occurs at a regional scale, but the relationship is contrary