Resource Niche and Trade-offs 163much lower than the percentage of light trans-
mitted through early successional tree crowns
(Figure 10.1b). The ability to cast deep shade
requires maintenance of multiple leaf layers, as
well as the tolerance of self-shadin gexperienced
by leaves positioned low within the crown. Adults
of later successional canopy trees have more
steeply inclined leaves, such that even the leaves
at the uppermost layer of the canopy experience
<20% of full sun. At the same time, more steeply
inclined leaves (and also terminal shoots) result
in a shallower regression slope of log (% of full
sun transmitted) plotted against the cumulative
leaf area index (Figure 10.1b). In other words, a
leaf should receive more light and a greater net
carbon gain when leaves above it are displayed at
steeper inclinations. Hence, low light extinction
coefficients allow a deep crown consistin gof mul-
tiple leaf layers that collectively absorb more light,
leaving less light to a small neighbor in its vicinity
(Kitajimaet al.2005).
Greater leaf lifespan of late successional trees
also provides an important physiological basis for
maintenance of multiple leaf layers and com-
petitive ability to cast deep shade. The most
shaded leaves within the crown of late succes-
sional species experience light availability similar
to that in the understory (e.g., much less than 5%
of full sun). In contrast, individual leaves of the
pioneerCecropia longipesexperience much higher
average light levels per unit leaf area, and none
of its leaves persist in light below 10% of full
sun (Figure 10.1b). Thus,Cecropiatrees cannot
invade the space already occupied by deep crowns
of long-lived trees. More studies are needed to
reveal light competition strategies and coexistence
of trees in the uppermost strata of the forest. It
will also be interestin gto examine leaf display and
light extinction by liana crowns that may com-
petitively suppress growth and reproduction of
canopy trees (Avaloset al.2007). In summary, the
light competitiveness of the upper canopy trees
and lianas may be predictable from their func-
tional leaf traits, even though their competitive
dominance cannot be inferred merely by their
positions at a given time.
Smaller trees, including saplings and seedlings,
experience strongly asymmetric light competi-
tion imposed by canopy trees, even though
competition amon gtree seedlin gs is probably
rare in the understory (Svenninget al.2008)
because they occur at low densities (e.g., 1–6
seedlings per m^2 in a neotropical rainforest;
Harmset al.2004). Under the closed canopy
of humid tropical forests, only 0.5–3% of full
sun reaches seedlings and saplings (Chazdon and
Fetcher 1984, Montgomery and Chazdon 2002).
The degree to which they can tolerate shade
and maintain a positive net carbon balance must
be an important determinant of juvenile distri-
bution and abundance. Even slight increases in
shade cast by an understory neighbor may have
large consequences for seedling carbon balance
(Montgomery and Chazdon 2002, Montgomery
2004).
In contrast, juveniles in treefall gaps compete
with each other to pre-empt the higher strata
and to cast shade upon their competitors. Hence,
casual observers may predict that the fastest-
growing individual will be the competitive winner
in a given gap. But is this true? In reality, new
treefall gaps are simultaneously colonized and
occupied by seedlings of early successional and
late successional species, dependin gon dispersal
limitation and other chance events. Thus, species
composition in a gap cannot be predicted simply
from gap size or age (Popmaet al.1988, Hammond
and Brown 1998, Schnitzer and Carson 2001,
Dallinget al.2004). Shade-tolerant tree juveniles
may persist in a newly created gap for many years
after bein gsurpassed by a nearby pioneer tree.
However, the former may eventually grow to dis-
place the latter, possibly after the latter matures
and senesces. Which one should be called the
winner of light competition in this gap? Do gaps
represent a niche position alon gthe li ght gradient
(Denslow 1980), a successional niche (Pacala and
Rees 1998), or merely a phase in transient dynam-
ics (Tilman 1988)?The answers to these questions
differ dependin gon the spatio-temporal scale at
which demographic dynamics are examined in
gaps.
Furthermore, gaps of different sizes do not
create discrete niches. Instead, light availabil-
ity varies continuously across the gap–shade
continuum in relation to heterogeneity of the
overstory canopy and position within each gap
(Brown 1993). Seedlin gli ght requirements also