the major vehicle for facilitating research, which gen-
erally fostered small-scale, low-budget science. There-
fore, only those projects that brought their own infra-
structure with them and that restricted themselves to
largely observational data collection have maintained
research projects that produced ecosystem-level re-
search in Monteverde.
Existing information on ecosystem ecology in
Monteverde is extremely patchy. Of the areas of eco-
system research, only five are covered in this book:
(1) hydrology and soils (Chap. 2, Physical Environ-
ment), (2) vegetation classification (Chap. 3, Plants),
(3) bird community structure (Chap. 6, Birds), (4) for-
est structure and dynamics (this chapter), and (5)
nutrient cycling (this chapter). Here we report on for-
est structure, composition, and dynamics of two pri-
mary forest types, the windward elfin woodland and
the leeward cloud forest; summarize nutrient cycling
research that has focused on ecological roles of canopy
communities in whole-forest nutrient cycles; compare
these patterns with other tropical montane forests; and
outline areas for future research.
9.1. Forest Structure, Composition,
and DynamicsKnowledge of structure and floristics is necessary to
understand tropical montane forest ecosystem ecol-
ogy (Grubb et al. 1963, Tanner 1977, Heaney and Proc-
tor 1990, Lugo and Lowe 1995). Forest structure has
been investigated in Monteverde to understand for-
est dynamics, the light environment, nutrient cycling,
seed dispersal, and the epiphyte community. Long-
term data exist on forest structure and composition
from two primary forest types: the windward elfin
woodland and the leeward cloud forest (Lawton and
Dryer 1980), both located within the Monteverde
Cloud Forest Preserve (MCFP).9.1.1. Windward Cloud Forest Dynamics
The very wet windward cloud forests along the crest
of the Cordillera de Tilaran illustrate extreme ex-
amples of dynamic forests. The trees that perch on
hogback ridges, cling to precipitous slopes, and stand
amid the pools on soggy swampland fall with great
frequency during the storms of the windy season (see
Chap. 2, Physical Environment). The terrain and cli-
mate are in some ways favorable to plant growth;
water does not appear to be a limiting factor, and
mineral nutrients seem to be in good supply. In con-
trast, conditions may be extraordinarily stressful in a
mechanical sense, as severe storms regularly batter the
Cordillera. Several times a year, the dwarfed forestson the peaks, saddles, and ridgecrests most exposed
to the trade wind flow are enveloped in howling,
hurricane-force winds, and cloud-wrack. The batter-
ing of forests by storms has long been of interest to
plant ecologists and foresters (Cowles 1899, White
1979, Pickett and White 1985). Hurricanes have pro-
vided lessons on the impact of catastrophic storms on
tropical forests of the Caribbean region (Beard 1946,
1949, Wadsworth and Englerth 1959, Lugo et al. 1983,
Brokaw and Walker 1991, Zimmerman et al. 1994).
The patterns of damage and recovery may provide
insight into the diversity of forest structures within
the Cordillera, and reflect on general ideas about tropi-
cal ecosystem stability and the relationship between
natural patterns of disturbance and resource partition-
ing in complex communities.The ontogeny of dwarf forest stature. The situation
in the Cordillera de Tilaran differs from hurricane-
dominated forests of the Caribbean. We focus on the
dwarfed, or elfin, forests characteristic of wind-
exposed peaks and ridges along the crest of the Cor-
dillera. If one hikes from the Information Center of
the MCFP to the Continental Divide at the pass into
the Valle de Perms Blancas, and looks south along the
broad saddle of Brillante to the slopes and peak of
Cerro Ojo de Agua beyond, one sees an archipelago
of smooth-canopied elfin forest on ridgecrests em-
bedded in a sea of taller and rougher canopied cloud
forest in the intervening sheltered coves and hollows.
Similar dwarfed cloud forests are widespread through-
out the tropics, particularly in the trade wind belts
(Shreve 1914, Brown 1919, Richards 1952, Beard
1955, Howard 1968, Leigh 1975).
Why are these elfin forests dwarfed? Early botani-
cal explorers were impressed with the inclement cli-
mate of such areas. Expedition accounts contain a
litany of complaints about the mud, rain, and absence
of sun for days, so it is not surprising that they em-
phasized the impact on trees of the very environ-
mental features that made their own travels difficult
(Weberbauer 1911, Seifriz 1923, Lane-Poole 1925,
Gleason and Cook 1927, van Steenis 1935, Exell 1944).
The ideas in these early accounts can be organized
into a set of hypotheses concerning the causes of for-
est dwarfing (Leigh 1975, Grubb 1977). Elfin forests
might be dwarfed because of low productivity (low
photosynthetic performance per unit leaf area or mass)
due to low temperature, lack of sunlight, or nutrient
starvation. The latter could result from soils low in
mineral nutrients or from difficulties in acquiring or
transporting available mineral ions. Waterlogging and
consequent oxygen deprivation might inhibit root
metabolism and mineral uptake. Low rates of transpi-
ration in the elfin forest may hinder nutrient trans-304 Ecosystem Ecology and Forest Dynamics