age and with gap area (Lawton and Putz 1988). In
10-m^2 gaps, leaf area recovers to 50% of the mature
phase leaf area index of 5.1 in 3 years. For 120-m^2
gaps, 75% recovery of leaf area occurs in 3 years and
90% recovery within 6 years. The recovery of leaf
area has implications for the recovery of productive
capacity, nutrient immobilization, and moderation of
the erosive power of rainfall.
Landscape pattern of natural disturbance and regener-
ation. Natural disturbance imposes a pattern upon
vegetation; Hubbell (1979) likened forests to a palimp-
sest, a parchment written on, erased, and written over.
Just as reuse of a particular parchment was not a ran-
dom matter, the natural disturbances of forests and
their recovery are not equally likely everywhere. Gaps
in elfin forest of Monteverde are clumped; too many
gaps have centers within 20 m of the centers of gaps
of similar age for gap formation to be considered a
random process (Lawton and Putz 1988). This pattern
may be due to the scale of the impact of gusts; nearby
trees may fall simultaneously. One mechanism for the
latter is that asymmetric crown growth into gaps
increases the likelihood of treefall in that direction
(Young and Hubbell 1991). The loss of neighbors may
also increase exposure to wind in settings such as the
Monteverde elfin forests.
Whatever the mechanisms, aggregation of distur-
bance imposes spatial structure on stands at a scale
larger than the individual tree, which should influ-
ence the design of ecological studies. In the elfin
forest, if there are more gap centers within 20 m of
others than expected by chance, then there are dis-
turbed patches of forest 40—50 m (4—5 emergent tree
heights) across. This patchiness (0.25 ha) is large rela-
tive to the size of elfin forest stands, which are typi-
cally 2—10 ha. If the height of the canopy provides a
natural metric, and taller forests show similar aggre-
gation of disturbance, then a spatial pattern with a
scale of around 1 ha should exist for a 30-m-tall for-
est. Since intensive sampling for vegetation attributes
and ecosystem processes commonly occurs at smaller
scales, attention to the disturbance components in
ecosystem structure is needed for accurate extrapo-
lation to watershed and landscape scales.
Research in the future. The problem of species diver-
sity remains a critical issue in both theoretical and
applied ecological arenas. A montane area such as the
Cordillera de Tilaran offers a different perspective on
the evolution and ecological maintenance of bio-
logical diversity from lowland regions. Environmen-
tal gradients are more dramatic in mountains, and
offer clearer opportunities to examine partitioning
of resources.Dramatic variation exists in vegetation structure
and composition within the Cordillera de Tilaran, but
questions of pattern and process have been little ex-
plored. The coupling of geomorphic and vegetation
processes deserves particular attention. Landslides
are dynamic sculptors of steep slopes, and the veg-
etation patch dynamics they initiate is conpicuous
throughout the range. We know little of the temporal
scales involved in the patch dynamics, and nothing
of the feedbacks between local vegetation structure
and slope stability. The coupling of wind exposure,
vegetation structure, and vegetation dynamics also
deserves quantitative examination at the landscape
scale. Exposure to the trade winds establishes con-
spicuous gradients of water availability and mechani-
cal stress in the Cordillera, but its influence on the
tempo and spatial scale of forest dynamics is not docu-
mented. Other questions of ecological stability in-
volve the persistence of distinctive vegetation types
such as swamps, elfin ridgecrest forests, and xeric
cliffsides. Although ambitious programs of long-term
monitoring are needed to address these questions,
others maybe accessible to small, well-designed stud-
ies of critical components.9.1.2. Leeward Cloud Forest
General forest description. A long-term study was ini-
tiated in 1987 to characterize the structure, composi-
tion, and dynamics of a study area (1480-1550 m) in
the leeward cloud forest of the lower montane wet
forest (Lawton and Dryer 1980). The 4-ha study area
is located in the designated Research Area of the
MCFP, about 3 km west of the visitor center (Fig. 9.8).
The forest is composed of trees 15-30 m tall, with a
well-developed subcanopy. Epiphytes are diverse
and abundant (Nadkarni 1986a, Ingram and Nadkarni
1993). In the study area, tree density and size were
measured (diameter at breast height, DBH) to 0.1 cm
and height estimated to 1 m for all trees >10 cm DBH).
Branch surfaces in the crown interior of nearly all
trees >30 cm DBH support large hemiepiphytes and
bryophytes, herbs, and woody shrubs in interwoven
root-humus mats up to 30 cm thick. The climate of
the study area follows the description of leeward slope
forest described in Chapter 2, Physical Environment.Forest soils. Four soil pits (1 x 1 x 2 m deep) on the
forest floor in the study area were excavated. The con-
tinually moist soils of the forest floor are derived from
volcanic rhyolites and classified as typic dystrandept
(Vance and Nadkarni 1990; Table 9.1). These volcani-
cally derived soils are considered to be fairly fertile and
recently deposited and to share characteristics with
other tropical cloud forests at similar elevations.313 Ecosystem Ecology and Forest Dynamics