170 Kaoru Kitajima and Lourens Poorter
alone. Most likely, such an equalizin geffect means
greater importance of stochasticity over niches.
Yet, this simple prediction does not explain the
observed match of a species’ position alon gthe
trade-off line with its habitat preference (Poorter
and Arets 2003, Wrightet al. 2003, Poorter
and Bongers 2006). This puzzle may be solved if
the relative fitness value of growth and survival
change non-linearly along the light gradient.
Fitness value of ability to grow fast should be high
only under high light availability, as fast-growing
species can achieve their full growth potential only
when there is sufficient light. In contrast, fitness
value of traits that enhance survival is greater
in lower light, as it takes a long time to recover
from damage incurred by disease, herbivores, and
falling debris in shade. How to integrate growth
and survival in order to evaluate overall perfor-
mance of species is an important challenge left to
future studies.
Differences in individual functional traits
clearly underlie species differences in growth
rates and survival of seedlings. Relative to more
light-demandingspecies,shade-toleranttreestend
to have large seeds, large initial seedling size,
storage cotyledons, dense stem and leaf tissue,
low specific leaf area (SLA, leaf area divided
by leaf mass), low leaf area ratio (LAR, leaf
area divided by whole-plant mass), and high
root:shoot ratio (Kitajima 1994, 1996, Osunkoya
et al.1994, Cornelissenet al.1996, Veneklaas
and Poorter 1998, Poorter 1999, Zanneet al.
2005). Higher stem wood density enhances sur-
vival of seedlings (Augspurger 1984b), probably
because higher tissue density increases biome-
chanical strength for protection against phys-
ical disturbance (Clark and Clark 1985), her-
bivory, and disease (Alvarez-Clare and Kitajima
2007). Large seedling size not only helps seedlings
to emerge from the litter layer (Molofsky and
Augspurger 1992), but also enhances seedling
survival in the understory through a larger pool
of non-structural carbohydrate reserves (which
is the product of carbohydrate concentration
and biomass; Myers and Kitajima 2007, Poorter
and Kitajima 2007). Indeed, carbohydrate pool
size, rather than seed mass, seedlin gmass,
or cotyledon mass, predicts survival and toler-
ance of seedlings following experimental shading
(0.08% of full sun for 2 months) and defoliation
(Myers and Kitajima 2007). These multiple trait
associations together provide the mechanistic
basis for the slower growth as well as higher
survival for seedlings of shade-tolerant species rel-
ative to light-demanding species (Kitajima 1996,
Veneklaas and Poorter 1998).VERTICAL LIGHT GRADIENTS AND
ONTOGENETIC SHIFTS
Ontogenetic shifts, that is, switching of growth
and survival ranks of species between size classes,
may represent a trade-off that contributes to
species coexistence (Baralotoet al.2005). The
null model for this idea is ontogenetic concor-
dance, which is expected when relative differ-
ences amon gspecies in functional traits, such
as allocation patterns and leaf traits, are main-
tained regardless of plant size (Poorter 2007). Do
pioneers always outgrow shade-tolerant species
(null hypothesis), or do ontogenetic shifts in
performance occur (alternative hypothesis) for
physiological reasons? Physically well-defended
leaves (Coley 1988, Poorteret al.2004) and
stems (Guariguata 1998, Van Gelderet al. 2006),
as well as carbohydrate reserves (Kobe 1997,
Poorter and Kitajima 2007), are important for
survival not only at the seedlin gsta ge, but also
at the saplin gsta ge. Across Panamanian tree
species, species that grow fast as a sapling in
the understory also grow fast in gaps (Pearson’s
r=0.61,P<0.001,n=115 species; data from
Welden et al. 1991), but those that grow
fast in the understory are also likely to die
fast (r=0.52,P<0.001,n=108). Hence, the
growth–survival trade-off is as important in
the saplin gsta ge as in the seedlin gsta ge. More
importantly, species’ relative positions alon gthe
growth–survival trade-off lines are generally con-
cordant for saplings and seedlings among tree
species (Gilbertet al.2006).
Yet, there are functional reasons to sus-
pect ontogenetic shifts in growth and sur-
vival rates amon gspecies. Li ght-demandin g
species are thought to have a growth advan-
tage over shade-tolerant species because of their
high LAR. Many light-demanding species are