Treefall Gaps and Plant Species Diversity in Forests 197foci for regeneration and succession in both tem-
perate and tropical forests (e.g., Richards 1952,
Beard 1955, Bray 1956, Webb 1959). Following
the development of non-equilibrium explanations
for the maintenance of diversity (e.g., Connell
1978, Huston 1979), a number of authors more
fully developed the concept of gap-phase regen-
eration into an important theory to explain the
maintenance of species diversity in tropical forests
(e.g., Ricklefs 1977, Whitmore 1978, Denslow
1980, 1987, Hartshorn 1980, Brokaw 1982,
1985a, 1985b, Orians 1982, Hubbell and Foster
1986). The variety of ideas on the role of gaps in
the maintenance of diversity can be synthesized
into a single “ga phy pothesis.” The formation of
canopy gaps by the death of one to a few canopy
trees creates sufficient resource heterogeneity to
allow for resource partitioning and niche differen-
tiation (sensuGrinnell 1917), or releases sufficient
resources (e.g., light and nutrients) to permit the
establishment or reproduction of plant species
that would otherwise be excluded from the forest
in the absence of gaps.
While much of the contemporary theory for the
ga phy pothesis was formulated in the late 1970s
and early 1980s, community-level empirical tests
of this hypothesis remained uncommon until the
late 1990s (e.g., Hubbellet al. 1999, Kobe 1999,
Schnitzer and Carson 2001). Given the historical
importance of the gap hypothesis, it is surprising
that few studies explicitly tested this hypothesis
in tropical forests. Previous tests have focused on
the capacity of plants to partition resources or to
respond rapidly to high resource availability. Col-
lectively, these tests suggested that the diversity
of some plant growth forms were maintained by
gaps, whereas that of other growth forms was not.
Consequently, evidence for the ga phy pothesis as
a general mechanism to maintain plant species
diversity is likely to depend upon the functional
grou por growth form under study.
The goal of this chapter is to summarize new
and previously published data to provide a coher-
ent picture of the role of treefall gaps in trop-
ical forest regeneration and dynamics, and to
determine the degree to which there is empiri-
cal support for the gap hypothesis. Specifically,
we will describe and examine the following key
topics: (1) the main processes and pathways of
gap-phase regeneration; (2) the models and mech-
anisms by which treefall gaps are proposed to
maintain diversity; (3) the empirical evidence
for the gap hypothesis as applied to major plant
growth forms or functional groups; (4) variation
in the impact of gaps across broad environmental
gradients; and (5) the relationshi pbetween the
ga phy pothesis and the intermediate disturbance
hypothesis. We restrict our definition of the gap
hypothesis to canopy gaps that are formed by
the standing death or toppling of one to a few
trees or a significant limb-fall. Larger-scale stand
replacement disturbances are outside the scope of
the ga phy pothesis. We acknowledge that there
is a continuum of disturbance events that create
small gaps to events that create very large clear-
ings (Liebermanet al. 1989, Vandermeeret al.
1996); nonetheless, single- to several-tree canopy
gaps are the most frequent type of disturbance
in many tropical forests (e.g., Hubbell and Foster
1986).PROCESSES AND PATHWAYS OF
GAP-PHASE REGENERATION
The process of gap colonization can be divided into
four discrete categories, only some of which are
mutually exclusive.
1 From seed.Plants recruit from seeds that were
present in the soil seed bank prior to gap formation
(Dallinget al. 1998) or are dispersed into the gap
soon thereafter by wind or animals (Schemske and
Brokaw 1981, Levey 1988, Wunderleet al. 2006).
Seeds of shade-intolerant pioneer species typically
require the high light and temperature conditions
of gaps for germination (Pearsonet al. 2003a).
Once established, these pioneers can partially fill
gaps via extremely rapid growth rates (Brokaw
1985a). Shade-tolerant species may also recruit
from seed immediately following ga pformation
(Kitajima and Poorter Chapter 10, this volume).
2 From advance regeneration. Shade-tolerant
seedlings and saplings present in the understory
prior to gap formation grow rapidly to fill the gap
(Uhlet al. 1988).
3 From vegetative reproduction.Trees or shrubs
within the ga por lianas pulled into the ga p
by fallen trees produce numerous clonal shoots